The future of multi-core computing
See this, 32 way, 256 K10 cores.
....
The mere mentioning of K10 makes Intelers over react in fury.
Hector is finally ready to pull the trigger on Intel.
posted by Sharikou, Ph. D. @ 8/19/2007 09:09:00 PM
Journal of Pervasive 64 bit Computing
Main Blog Page
Site Feed:
sharikou.blogspot.com/atom.xml
Blog Roll
Hot pages
sharikou.blogspot.com/atom.xml
Blog Roll
Hot pages
- Price War Strategic Analysis
- AMD and Intel Performance
- Intel's false performance claims
- AMD's Architecture Lead Persists
- AMD and Intel Performance
- AMD and Intel Architecture
- Bensley/Woodcrest and Athlon MP
- Bensley/Woodcrest is 2P only
- 64bit CORE Performance!
- Conroe Busted
- Capacity: AMD vs Intel
- Conroe's Impact
- AMD and Intel in China
- A Beautiful Picture!
- Temperature and System Reliability
- Google and AMD
- AMD's Grand Masters
- Intel's Predicament
- AMD and Dell
- NGMA Core and Colwell 6th Gen Arch
- AMD and Intel History
- The Tale of Intel and Anand
Analysis on IT trends and competitive strategies, with emphasis on micro processors, computer systems and networks. Based on latest news, backed up with real data, this site intends to provide a true and realtime picture of the fast changing IT landscape. This journal strives to be accurate on facts and sharp on criticisms. You may email your opinion to sharikou@yahoo.com or post comments here, be cool and intelligent.
About Me
- Name: Sharikou, Ph. D.
Freelance journalist on IT matters. Some of my writings have been published on online IT journals. Any original content on this journal is Copyrighted, but it's free for non-commercial use. Any Trademarks used on this site belong to their respective owners. Some of the pictures are links. If there is any issue with the content of this site, please email sharikou@yahoo.com .
Previous Posts
- X2 6000+ selling like hot cakes
- A web site for free market
- AMD ready to pull the trigger on INTEL
- INTEL guilty
- AMD grows units by 38% in 2Q07
- Intel Core 2 servers are literally hot
- Core 2 Duo thermal throttles
- The only thing that is keeping Intel alive
- Intel out of top 5
- Core 2 Quad price starts the free fall
Sunday, August 19, 2007
....
The mere mentioning of K10 makes Intelers over react in fury.
Hector is finally ready to pull the trigger on Intel.
132 Comments:
I think you should fix your link. No one cares about an orchestra.
$5 says the video actually does not exist. i.e. The orchestra is the video Sharikou intended to link to the whole time.
AMD is for low end crap only. Only real CPU makers like Intel, Sun and IBM have CPUs that can go to 16P and beyond.
I figured as much. I tried to look for such a video (knowing it doesn't exist) and found none.
This video isn't even of a true multi-core orchestra Sharikou! All the players are NOT on the same die, they are just 'glued' together on the stage!
Even worse, they are all dependent on that single front-side-bus 'conductor' that is a total bottleneck! Conductors have been in use for hundreds of years! This must mean that all AMD chips are based on obsolete technology and are hundreds of years old!
Oh and at 5:46 the piano clearly shows that Intel is superior... because I said so, and me making up shit from nothing is (sadly) more realistic than you doing it.
Go back to making Kim Jong Il porno tapes Sharikou, you've had it trying to fool anyone else.
This comment has been removed by the author.
This video isn't even of a true multi-core orchestra Sharikou! All the players are NOT on the same die, they are just 'glued' together on the stage!
Even worse, they are all dependent on that single front-side-bus 'conductor' that is a total bottleneck! Conductors have been in use for hundreds of years! This must mean that all AMD chips are based on obsolete technology and are hundreds of years old!
That is utterly hilarious! You have my thanks for the entertainment!
Oh... by the way, that piano player guy is really good, and so is the orchestra. While normally an orchestra would have nothing to do with AMD, Intel, or chips in general, there could be a lesson in here for AMD:
1. The orchestra actually knew what the hell it was doing and worked together.
AMD could try to have its management actually try to make half-intelligent decisions so the engineers can actually make real chips.
2. There was not a single power-point presentation displayed of how great the orchestra was, and no grandiose promises about how great the orchestra would be at some indeterminate time in the distant future. Instead, the orchestra actually did something and actually played music!
AMD could definitely take a dose of vitamin shut the fuck up & actually deliver to be more like the orchestra.
3. The orchestra concentrated on trying to be good and didn't sit around whining about the other orchestras out there and bitching that the other orchestras are unfair because they work harder and practice more.
AMD could learn to not bitch about Intel and instead focus on actually making good products people want to buy.
4. The orchestra had all of the essential instruments and musicians it needed to make great music and didn't try to force in a Mariachi/Polka band playing covers of Twisted Sister in the middle of the performance.
AMD could learn that instead of just going out an blowing billions of dollars on ATI just to 'look cool' that it should instead focus on doing things it actually has a clue about and actually make decent products for a change.
There ya go Sharikou, even your non-sensical posts can still show AMD how they ought to be doing things.
This comment has been removed by the author.
AMD X2-6400+ outperforms C2Ds
Have you any proof of this? We've all heard that BS story over and over.
In testing, performance of the 6400+ is 10-15% below the E6850, and the CPU has limited overclock capability, said the sources, who also pointed out that the price gap compared to the E6850 is not huge.
http://digitimes.com/mobos/a20070723PD205.html
Giant said...
In testing, performance of the 6400+ is 10-15% below the E6850, and the CPU has limited overclock ...
Looks pretty good to me ;)
http://www.xtremesystems.org/forums/showthread.php?t=151746&page=6
Michael said: Looks pretty good to me ;)
http://www.xtremesystems.org/forums/showthread.php?t=151
Yes thank you for proving giants point by the link the 6400+ IS slower than E6850 AND does have a limited overclock.........
Here is a lil info about the 6400+ from a review site. 6400+ quickly turning into the new prescott of the semi conductor industry.....
http://www.extremetech.com/article2/0,1697,2173093,00.asp
So you're AMD, and you need to move some processors. No one really wants your 7000 line of CPUs, which requires dual socketed motherboards. And the Athlon 64 6000+ can't keep up with the equivalent Intel Core 2 Duo E6750. What do you do?
You sort through the CPUs that are coming off the fab, cherry pick a few that can run at 3.2GHz, pack it in a cool looking black box—without a fan, no less.
Then you don't send out samples to the usual range of enthusiast web sites (including ExtremeTech), so that no one really knows how it will perform.
But this "black edition" approach being launched in almost a stealth fashion, with little fanfare to the media and no reviews going up on the day of launch, also smells like desperation on the part of AMD. With Intel riding fat in the catbird seat, AMD needs to keep users interested in its product, at least until viable AMD quad core CPUs for the desktop ship. The company's financial situation is precarious right now, and it needs all the revenue it can muster.
AMD desperate did he say naw that cant be with their yard sales with 50 dollar dual cores that they cant even give away to customers, their record breaking lows in the stock market, their 2.0 ghz snails pace processing technology and their hotter more inefficient GPU line i am surprised people are labeling them as desperate..........
AMD's dual-core Athlon 64 line hits 3.2GHz and still barely manages to keep up with Intel's line running at 2.6Ghz
I very much like Oneexpert's added commentary to quotes.
And of course, they are always wrong.
If there is one (ok two) things we can count on is:
1.) Shaikou posting stuff like a total and utter moron, claiming Intel will be BK Q208, so some other idiocy in his "posts".
2.) Oneexpert in a need to coddle up to his mis-informed lover posts utter crap.
Good going! You really do look like a moron!
Here is a link about what people have to say about the 6400+ launch.......ala sharikou's new egg reviews..........
http://www.dailytech.com/article.aspx?newsid=8505
"Seems like a waste to spin a new SKU"
"Also, who is going to be stupid enough to buy one? "
"I currently own an AMD CPU, it is my third AMD CPU, and I've never been disappointed, but their business sense of late, release schedules, SKU's, and lack of information concerning Barcelona have me worried."
"Gonna steal a line from one of the guys on TR and say that it's only black because AMD ran out of red ink :)"
"In any case, I expect some reviewers to get their hands on this and show (as expected) that it gets beaten senseless by Core 2 offerings. "
I think Sharikou has finally lost it.
First quad-core laptop hits U.S.
Pricing starts at $3,359.
Never mind that there's little-to-no software that can take advantage of four processing cores.
This comment has been removed by the author.
$251 for 6400+ vs. $266 for E6850. However, 6400+ does not include a cooling solution, while the E6850 does indeed come with the standard Intel fan. Add to the fact that the 6400+ uses 92% more power and is slower and you can see Intel is clearly the "smarter choice"!
AMD's dual-core Athlon 64 line hits 3.2GHz and still barely manages to keep up with Intel's line running at 2.6Ghz
Indeed. The tests at HKEPC have shown in most circumstances the E6750 is faster than the 6400+. The only processors that 6400+ are clearly faster than is the E6350 (2.33Ghz) and below.
From my previous post since oneexpert is an expert at copying and pasting kinda like how AMD is an expert at power points instead of actual CPUs anymore.........
ALL AMD FANBOIS ARE LIARS ESPECIALLY ONE EXPERT
No wonder they call you one retard you cant even read benchmarks correctly
http://www.hkepc.com/bbs/hwdb.php?tid=842214&tp=AMD-A64X2-6400&rid=842214
Here are the reviews from HKEPC that you are talking about.
E6750 running at 2.66ghz beats 6400+ running at 3.2ghz
Total benchmarks: 43
Benchmarks won by E6750: 22
Benchmarks won by 6400+: 21
----Winner E6750----
Average increase in performance in benchmarks won by E6750: 18.666%
Average increase in performance in benchmarks won by 6400+: 8.056%
----Winner E6750------
Clockspeed difference between E6750 and 6400+: 540Mhz
-----Winner E6750-----
Price for E6750: $212
http://www.pcrush.com/prodspec.asp?ln=1&itemno=108123&refid=1238
Price for 6400+: $220 - $240 WITHOUT HEATSINK!!!
-------Winner E6750-----
Winner of performance, for price, for power consumption E6750
You are a lying sack of shit One expert.
Says a lot when your top-of-the-line, special edition cpu is only worth $241.
Poor AMD.
Pathetic. OneRetard keeps posting the same nonsense over and over and over (which just goes to show you he is truly retarded) and the rest of the AMD zealots just hang off of sharidouche's ball sack. (or should I say her vagina?)...this whack job is a delusional lying freak who can barely type a coherent sentence half the time. When Intel doesn't bankrupt Q2 of next year, what's he going to do then?
Maybe he'll reapply for the position he was fired from at Intel....
Aug 20th 2007 , Memo Update :- Paul O to his Lieutenants
From the dark recesses in my broken mind, I have finally found real fear. I am anxious to escape the madness of this monopoly institution and once again enter the sanctuary of my excellent stock options. They call it the Phenom or Barcelona or K10 and it runs a next generation CPU Quad core natively at insane speeds. Some of its unlisted features are even more terrifying. If the K8 brought us to our feet, this new beast will trigger our
downward death spiral. I can't even begin to talk about the limited options we have..But I must go! I hear them coming!
....annnd when the fanpukes have nothing they look to try and make up witty corporate notes that in reality just make the author (See above) look more retarded...
chuckula & giant -
"Even worse, they are all dependent on that single front-side-bus 'conductor' that is a total bottleneck!"
The right analogy is that the ochestra is the computer, whereas the conductor is the user. The music is a massively multithreaded program.
Sadly you couldn't better hide your lack of knowledge in both music and computer.
From the dark recesses in my broken mind, I have finally found real fear.
yes those AMD powerpoints can be very scary especially with the lack of information and delusional promises they contain.
I am anxious to escape the madness of this monopoly institution and once again enter the sanctuary of my excellent stock options.
Yes i am sure the CEO of a 80% market share holding company would trade places with a financially broke AMD.
They call it the Phenom or Barcelona or K10 and it runs a next generation CPU Quad core natively at insane speeds.
Yes insane speeds of upto 2.0Ghz though 1.9Ghz would be the one most abundant in the channel.
Some of its unlisted features are even more terrifying.
yes features that are UNLISTED and not KNOWN to anyone would be very scary kinda like reverse hyperthreading with AM2 the performance benefits of that implementation were amazing AMD
If the K8 brought us to our feet, this new beast will trigger our
downward death spiral.
K8 brought netburst down but C2D just whooped k8 like a pinata (waiting for retard to tell me to put that tilda over the N)
I can't even begin to talk about the limited options we have..But I must go! I hear them coming!
Yes i have to go laugh at AMD yard sale and hand out books about manufacturing processors for dummies
Hector is so desparate, he sold the field behind the office.
http://www.theinquirer.net/?article=41782
This comment has been removed by the author.
AMD Up 0.21 (1.77%)
Don't forget the tilde!!
Smile...
Abinstein: It's called a joke you useless fanboy retard. Please show me the actual University you graduated from with an actual degree in computer engineering. Mine is from Purdue and I could probably design a better CPU after knocking back a fifth of tequila than you ever could (and I already have designed a pipelined CPU with exception handling in VHDL you useless tool).
Do you have a swastika carved in your forehead because the way you blindly follow anything Sharikou says reminds me of goose stepping Nazis. The Nazis had an old expression that if Hitler said 2 + 2 =5 then that's what 2 + 2 is. You seem to agree with that idea since you seem to have no problems when your fuhrer Sharikou puts up a post saying that a 32 way K10 system that only exists in his fantasies has 256 cores.... 32 x 4 is 128 you useless fucker. I'm sure you know a whole lot more about me when it comes to microarchitecture design when you can't even do 4th grade math.
Having been in the computer field since 1973 I’m still amazed at the religious passion some people have over what ever the latest and greatest piece of silicon Intel or AMD puts out (or announces). My watch has more CPU power than the first computer system I was introduced to in school back in 1970. I helped configure and build the first IBM PCs the City Government I work for back in 1981 and we paid a fortune for them ($3200.00 for each of 21 machines) by today’s standard. Your average computer Geek today wouldn’t stay in the same room with these for fear of catching something. I still remember the passion of those that got the real 5 mhz Intel 8086 16 bit CPUs in their IBM compatible vs. the inferior 16/8 bit 8088 running at 4.88 mhz in the IBM thing. The first real work computers I worked on were considered “super” computers for the day and they maxed out at about 1.3 meg of memory and cost millions each in 1973 dollars. This nation’s entire space program was built upon computer systems in the 60-70s that couldn’t be programmed to run your typical First Person Shooter game today if your life depended on it. I have such fond memories of my first 8 mhz Z80B CPU running the CPM OS. Those were the days.
What most followers of the Intel “religion” don’t seem to grasp is that the price they are paying for what ever Intel “heater” they are using today is the function of effective competition between AMD, which is 1/10th the size of Intel and Intel which not only has more of everything, including inventory it can’t sell at the price it wants to but that it also has laid off 6000 workers because it had to. I have very fond memories of all the marketing gimmicks Intel has pulled in the past 20+ years such as “SXing” older versions of its mainline chip in order to sell it at a lower price, introducing completely new CPU types, pin outs or memory type so that you had to buy a complete motherboard which also contained a new type of Intel chip set, etc. Intel is famous for this and that is one of the things that have driven customers away from Intel into the arms of AMD. They generally don’t do that. If you look at the ASP of what sells to the largest number of customers, a similarly configured Intel and AMD system built by the same system vendor will cost a lot more for the Intel system than the AMD model all else being equal. It is not unheard of to see $100-200.00 difference on the ASP systems today and the average customer could care less about the Call of Duty frame rate difference between a 3.0 Ghz Athlon 6000 X2 and the same clock rate Core Duo 51xx series. Neither do I and there is no real world difference between these two systems all else being equal. Highly specialized benchmarks can show me the difference but my computer powered wrist watch can’t measure the difference for most things.
My corporation has thousands of 2.2 – 2.6 Ghz Intel Netburst based PCs on lease and my 2003 AMD 64 3200 is faster than all of them (still). You can’t buy a system today as slow as my 2003 systems but neither can I find anything, other than some specific kinds of games that benefit from junking my 2003 systems to buy the latest systems being offered. I could double the CPU performance of my systems tomorrow but it wouldn’t do a thing for my Internet experience, speed up my hard drive performance or improve the graphics on my 8X AGP 256 MB 3D card. My two drive RAID setup, even being limited by going through the PCI bus will still beat the current crop of single drive machines regardless of the CPU power installed. Same for the latest, greatest graphic card(s). A $500.00 Graphic card does nothing for MS Office, Internet use or any of my 2D games which play just as well on my 2003 systems today as they did on my 550 mhz Pent III back in 1999. Putting a Group 5 Porsche engine in a 1968 VW does not make it a 24 hours Le Mans racer. That’s what buying the latest CPU every few months and upgrading your system typically tries to do.
Have I made the point by now? Without effective competition, those few that have large discretionary budgets and time on your hands would find it real expensive to completely rebuild your super systems every few months because the other 99% will not buy enough systems to ever lower the price. You have the Intel systems you have today because there is an AMD and you might want to reflect upon that. I spent more on Microsoft software upgrades for my existing systems last year than on the inflated gas prices we are paying today (an extra $1.50 a gallon x the number of gallons I use every year). There is no AMD equivalent for Microsoft and we all pay for that.
My Data Center where I work has close to 400 Intel servers. The newest 3.0 Ghz 4P (8 Core) ones are no faster than the same 3.0 Ghz 8P ones we bought in 2005. All are Netburst based. We are getting our first 3.0 4P (8 Core) AMD this month. It is up to 70-80% faster than the same 4P (8 Core) Netburst model. The Intel 51xx and 53xx Core Duo models aren’t offered in 4P (8 Core) systems by our System Vendor who happens to be No 1. There is a reason for that and is reflected in the reason Intel had to lay off 6000 loyal employees in order to lower its cost. The Intel CPU designs were always and remain desktop chips with the limitations that brings. The Opterons rule where 4 sockets and above are required. The Intels don’t scale well under load with their current designs using off Core memory controller. Funky memory designs that use more power than conventional DDR types makes their lower power claims moot when you have servers with 32-128 Gig of this stuff. There is a bigger picture here than just the CPU/Memory/Graphic Card frame rate performance you see so many people’s passions expended on.
It is alright to have favorites but many of you should really keep religious passion out of your computer system logic thinking. From one quarter to the next, one year to the next a healthy Intel and AMD will trade positions back and forth and we will all benefit from that. The minute there is only one, we all will lose. I have Intel and AMD systems and I don’t pray to either one. My AMD 386-40 is still faster than Intel’s 486-25 SX and Volkswriter running under Dos on this AMD 386-40 is still faster than MS Word for most things running on anything. The sum of the parts makes the product that keeps the lights on at Intel and AMD. 99% of the computer buying public isn’t going to junk their current systems to simply buy the latest Intel wonder thing. Most of us have been there and done that before. Take away AMD and Intel will ask $1000.00 for its next “SX” creation because it can. Take away AMD and you would be looking at junking your existing 32 bit only systems to move to Intel’s 64 bit Itanium 2. That thought should even scare the most passionate followers of the Intel Religion.
What I can't believe is the gent above posted all that on this joke of a blog...
This blog seems to be a front for intel Guerilla Marketing. Could Sharikou be this dumb? Sharikou sets all of you guys up with rediculous rants and articles and the paid pipers pounce on it. Ridiculous, this site is a waste.
Could Sharikou be that dumb?
No, he's even dumber.
He lies about being a Ph.d.
He can't type coherent sentences.
Anytime a site, ANY site, shows Intel's products in better light, the site is being paid by Intel, but if they show AMD in a better light, that's ok.
He make ludicrous "predictions" in the face of every bit of data that goes against what he's saying.
He has several log ins here to make it look like he has more support than he has. Him or not they're about as bright as a box of rocks.
I'll come up with some more later...
Even the leakiest dingy will go up in a rising tide, right before it fills completly with water and drowns the rats and silly AMD fanbois with it.
AMD BK in 2008
Do you have a swastika carved in your forehead because the way you blindly follow anything Sharikou says reminds me of goose stepping Nazis. The Nazis had an old expression that if Hitler said 2 + 2 =5 then that's what 2 + 2 is. You seem to agree with that idea since you seem to have no problems when your fuhrer Sharikou puts up a post saying that a 32 way K10 system that only exists in his fantasies has 256 cores.... 32 x 4 is 128 you useless fucker. I'm sure you know a whole lot more about me when it comes to microarchitecture design when you can't even do 4th grade math.
Chuckula, ignore abinstein. He is probably Sharikou in disguise.
This is funny. Intelers don't get it. :-)
"The future of multi-core computing"!
Tigerton launching September 2nd at speeds of up to 2.93Ghz. This will scale to 32P; 128 cores. 128 2.93Ghz processing cores. What will AMD have when barcelona finally launches? Only 8P, so 32 2GHz cores.
AMD's server performance will be less than 25% of Intel's performance in the highest end servers possible.
AMD is finished. BK in Q2'08
This comment has been removed by the author.
We finally have a new tactic by the pro-AMD faction; Thom wrote a post (above) where we get the usual pro-AMD talking points, but this time wrapped up in a sentimental story. His writing is actually quite good, but his facts and point of view make him useful only for fiction.
Anyone else notice the banner ads from GoogleSense on the front page of the Journal is for toilets?
Makes perfect sense, as all that Sharikou posts is crap.
Thom, you are a complete hypocrite. You should not be preaching about the goodness of competition and not to take sides whilst then failing to be neutral yourself.
You're as biased as most other folk on here but at least they have the honnesty to admit it, whether they are AMD or Intel.
Hector is finally ready to pull the trigger on Intel.
Indeed. Paul Otelini and Pat Gelsinger must be cowering in fear under their desks at the thought of a 2Ghz Barcelona CPU; or the aptly named 6400+ Black Edition (Because they ran out of red ink!)
Hey oneexpert, how do you like this?
E6850 nearly 3x as fast as 6400+!
http://xtreview.com/images/amd-athlon%2064%206400-officeexcel.PNG
6400+ fragged all over by the E6850.
AMD is finished. BK in Q2'08.
And the 2 shitcups goes to...
one goes to:
giant = idiot & crap eater, IQ < 50 (20% chance of recovery, but I'm skeptical)
& the second one
13ringinheat = biggest idiot of all times, braindead mofo, IQ < 20 (0% chance of recovery)
With a screen name like Anti Intel guy i cant imagine why our comments would chap your ass after all you being the beacon of unbiased factual info.
I guess you prefer living in delusions of a blazing fast turd brown edition 2.0ghz quad core,
A "top of the line" K8 running at 3.2ghz that competes with the competitors 5th best processor
A GPU that is hotter, slower, six months late and competes with the competition's third best card.
AMD and their fanbois.......when just being decent will suffice
Yep name calling is much easier than actual data, benchmarks and links........
@13ringinheat: Yeah right crapburger, your IQ just dropped to 0, congrats ! Now you're a vegetable!
@anti common sense.....
Hahahahahh a vegetable with more sense than you....
I like your name calling very entertaining....shitcup, crapburger.....some one is overly obsessed with shit......i guess its easier to relate to it when you are made up of said substance..........
I guess i will take the roll of vegetable in this blog and you can be the shit.....
@13ringinheat: You must be a Dingbat. That explains everything...
I'm still not convinced that he's serious... maybe we're just being played for fools. Anyone that thinks that AMD is in a favorable business position, is obviously lacking even a single brain cell. Is Sharikou really THAT stupid? I'm thinking maybe he's just just a pro-Intel guy jerking our chains around a bit for laughs. I just refuse to believe anyone could be THAT stupid.
giant,13ringinheat,Chuckula & Co
Would you care to explain your hatred for AMD ??
we will have some silence now :-))
Rodney = uselessjunk ! Look who's talking: the BIG jerker!
This comment has been removed by the author.
Q: What's blue and crawls up slow a woman's leg?
A: An aborted inteler...
anti-intel guy = onlyamd or oneretard
netrama - I use both companies products almost equally in my endeavours so I favor neither - I use what fits the application at hand best, so in terms of hating AMD (or Intel) I hate neither because both are necessary in this market and both have decent products.
What I hate is blatant ignorance, stupidity, and, most of all, liars. Sharikou is a liar and a fraud, the guys that kiss his ass here (most of which are probably himself anyway) are the worst sort, and deserve any foul treatment given them....
AMW, I’ve owned 6 Intel based machines and 4 AMDs in the last 20+ years. That doesn’t include upgrades to most of these. The AMD models were generally more upgradeable than the Intel ones because Intel designs their designs to have shorter life cycles. I didn’t say I didn’t have a bias but my bias comes from two decades of working with Intel products. I still own an Intel system and use it every day. It doesn’t do 64 bit stuff however.
Intel FanBoi, I’ve never equated history with sentimental things. The facts are the facts. AMD got its start because of a Government requirement for a second source before the Government would buy millions of what was then new Intel 8086/8088 class single chip CPUs. Perhaps all this occurred before you were born. For a company 10 times the size of AMD to only have 80% of the market and having to lay off 6000 employees says something but perhaps you don’t understand what it means to lose your job while all you’ve been told for years of service is how superior your products are over the competition. What you and your mindset don’t grasp is that there is a glut of CPU power in the market place that serves no useful function for 99% of the world’s PC users. The 2002 Intel I’ m typing this into is a whole lot slower in all ways than my 2003 AMD systems but none of that matters on the Internet or any normal MS Office use. What most people use a PC for has no value added to it by having the top of the line CPU. You understand the term “value” right? How about “utility”, that’s an economic term for the same thing and relates to what incentive potential customers have to purchase one product over another. Intel could pull a “Sun” tomorrow and put a whole bunch of cores in a new larger socket and some of you would act as if the “second coming” was upon us. How may cores does the dominate PC OS in the world support? As for new pro-AMD tactic, you couldn’t be further out in left field. I have to work with and depend on hundreds of Intel servers every day. My salient point was don’t let this become your “religion” because unlike a true religion, things change and some times quickly in this passion play and some people can’t handle that too well.
That’s all there was to the history lesson.
This comment has been removed by the author.
@Thom: Pick another flock!
@Spaztic Pizza: There's nothing left of you... only pizza.
Thom asked...."How may cores does the dominate PC OS in the world support?"
Replace the word "cores" with "sockets" and you'd be asking this question correctly.
@Spaztic Pizza: Just replace "spaztic" with "vomit" and see what you get.
Shouldnt you be off to elementary school now?
@Spaztic Pizza: Well, when your balls are slapping against the back of your ass, I'd say
you're in definitely.
Congratulations Sharidouche - I see that you're now letting people of "quality" post comments on your blog...
Thom,
Do you expect us to believe that your "facts are facts" just so happen to be the same disproven talking points we have heard over and over again from the likes of Sharikou, etc? "Intel HAD TO lay off 6000 workers, Intel sells heaters, Intel is 10X the size, nobody needs the extra processing power, blah blah blah." Next you'll tell us that Intel chips cause laptops to explode. You are like a repeat of Sharikou from 2006.
@Spaztic Pizza: You're right pizza delivery boy.... but also, you're pathetic. Top quality defines AMD not Intel technology.
http://www.youtube.com/watch?v=SYvupXNmaKQ
Geforce 9 and Harpertown coming in November.
http://www.theinquirer.net/?article=41836
AMD is finished. BK in Q2'08.
@Giant: What else? Intel Pee7, larrapeepee or funkySB? Boring, boring, boring...
Oh wow. A dual core to quad core CPU upgrade using the same motherboard and fan! Doesn't this seem familiar? Oh yes, it does. I did that same upgrade myself a few months ago. First update BIOS, then remove the computer side panel. Take off the fan, remove the CPU, insert new CPU, clean thermal paste off fan, apply new layer of thermal paste and put the fan back on, replace the tower side and power the system bac on. Done. E6600 to Q6600 upgrade in under 15 minutes.
What is so revolutionary about that? You can do the same thing with the socket 771 server CPUs. Replace any dual core CPU with a quad core CPU; just update the BIOS first.
Remember: Intel is 10 months ahead in quad core server CPUs; over one year ahead in desktop quad core CPUs.
Boring, boring, boring...
What does AMD have? 2GHz quad core CPUs? *snore* 3.2Ghz dual core CPUs? *snore*
@giant: Remember: Intel is 10 months ahead in quad core server CPUs; over one year ahead in desktop quad core CPUs.
:) :) :) LOL !
Intel sucks so much, they might as well drop a cock down their throat and write a plan for world domination.
LOL !
This comment has been removed by the author.
I think anti-Intel guy has even less brain cells than oneexpert. Stupidity here has reached new heights!
Inside news..
http://news.yahoo.com/s/pcworld/20070821/tc_pcworld/136214;_ylt=AoQKtuNy2LPYoiu4Qha9D5EE1vAI
Thom said...
Having been in the computer field since 1973 ...
I have been in this field for 10 years although first computing experience was 20 years ago on Apple II.
What most followers of the Intel “religion” don’t seem to grasp is that the price they are paying for what ever Intel “heater” they are using today is the function of effective competition between AMD
I totally agree with you on this one, although the tide has turned just a year back, and AMD is selling heater right now with all its highend desktop series, such as the FX and the 6000+/6400+. If not intel, we all have to pay few hundreds just for 3800 ..
it also has laid off 6000 workers because it had to.
Yeah, it is kinda sad in the business world. Those big corporation including AMD has to lay off its people when thing goes wrong.
I have very fond memories of all the marketing gimmicks Intel has pulled in the past 20+ years ...
yeah me too .. on the AMD end, it has been acting very hypocrate ... at one end it said the customer doesn't what the process nanometer is, and then at the other end it said about its native solution, going to be used immersion tools ... for us the user, what we really care is the absolute performance, performance per watt, etc. native quad core can be a way to have better efficiency, so do the better process technology too. I still remembered once Henry said intel has been pushing technology that is not ready for the market and he quote DDRx example, which take a few year for full adoption, and then on the same artcle he boast AMD first embrace the 64 bit in IA, and he failed to realized it also take years (much longer) to take off and i bet most of us are not using any 64 bit OS most of the time. it is good that AMD pushing the 64 bit technology in IA, but do not act so hypocrate on technology that Intel is pushing.
, etc. Intel is famous for this and that is one of the things that have driven customers away from Intel into the arms of AMD. They generally don’t do that.
yeah, this indeed is a good point that AMD has been doing. May be my brother is kind bad luck, he bought an AMD 939 system just 2 years back ... he is not able to do any upgrade to his system now ...
average customer could care less about the Call of Duty frame rate difference between a 3.0 Ghz Athlon 6000 X2 and the same clock rate Core Duo 51xx series.
yeah, i agree that there is no much differnt to own a 6000+ system to a celeron system for normal user who just wanna browse web and do some minor documentation. he/she can save the power bill by using the celeron system too as compared to the power hungry AMD 6000+ system
I totally agree that users should judge from what they need and from the maintenance (such as power) point of view too.
You have the Intel systems you have today because there is an AMD and you might want to reflect upon that
yeah competition is a good thing to have, or we will have to pay hundreds just for the AMD 3800 and vice versa.
server will not be that low in price before intel compete in it with its IA. Then AMD join in the competition and further stir up the innovation. Intel fight back with it Core design and it has been very good in the UP and DP end. Site like lowyat.net, which used to run with the AMD server just a year back, it is able to cope with increase server load just by switching to the latest intel solution and still with more headroom to grow.
AMD had to lay off its loyal employees as well as those former ATI in order to lower its cost. The AMD CPU designs few years always and remain server chips with the limitations that brings in the mobile end. That's why AMD has to split the design in the future to cater for the mobile to compete with Intel as its CPU is not as competitive / power efficient as intel's in the mobile segment.
Take away AMD and Intel will ask $1000.00 for its next “SX” creation because it can.
yes, i totally agree on this one and the reverse is true too. Quite hard to image that a AMD CPU that costs hundreds just a year back and now cost tens. Thanks to intel for putting up such a nice competition.
I see you're all still jumping through Sharikou's hoops. I just crack up.
Anti intel said: Rodney = uselessjunk ! Look who's talking: the BIG jerker!
Hahahha great comebacks there BIG JERKER....even retards can put up a better insult than that.....
Anti Intel guy said: they might as well drop a cock down their throat
Seems like you speak from experience.......
Anti Intel guy said: @Spaztic Pizza: Well, when your balls are slapping against the back of your ass, I'd say
you're in definitely.
Or in your case have another dude's balls slap against your ass........
Anti Intel guy said: Q: What's green and crawls up slow on a man's leg?
A: Its anti intel guy!!!
Now please go do what you do best and suck a dick and leave the computer talk to adults
Thank you
In the real world, where opinion, action and consequence aren’t separate things like they are in the Blob-o-Sphere there is an economic rational for everything that has any longevity to it. The first generation of the Apple IPhone will be a good example of something that has no economic rational and short life.
One anonymous poster said they were surprised that I’d spend the time to post my comment and that this site was a joke. Well, if this site is a joke it is because of the quality of the postings. I fail to see what terms, “fucker”, “Vagina”, “Cock”, “Ass”, etc have to do with 64 bit computing. Speaking of 64 bit computing, something AMD designed into their Opteron and Athlon lines not added on, the state of 64 bit computing today can be laid at the door steps of Microsoft. AMD laid the foundation; Intel followed, it’s up to the software industry to port their work. I went through the 8 to 16 and then the 16 to 32 bit conversions efforts so don’t bash AMD for the state of the software industry and Microsoft in particular. Microsoft is more interested in billions of dollars invested in eye candy these days. If there was effective (any) competition for MS’s OS line this might not be the case.
Another anonymous poster tries to defend Intel’s marketing gimmicks by accusing AMD of the same. Well, find me a corporation that does not try to defend its revenue stream and I show you one headed for bankruptcy. The difference here is a) Intel has done this since the 80826 line in the mid 1980s and b} they’ve done this every couple of years resulting in completely new motherboards, memory types, etc solely for the purpose of heading off competition not bring a better product to the table. When Intel introduced the Pent 4 line they came in at about 1.3 Ghz vs. the Pent III at 1 Ghz (more or less). Some observant people quickly pointed out that the Pent III and Pent 4 at those speeds were the same performance and Intel wanted a lot more for the Pent 4. Hence, the initial Pent 4s disappeared quickly as Intel ramped up the Netburst “heater” design to kill off the Pent III where AMD was competing effectively. If AMD did this to the extent Intel did over the last two decades they would only have 10% of the market instead of 20-24%.
Another anonymous poster, an echo here, is all worked up about the apparent performance of an unannounced product from AMD vs. the top end of the current Intel product line. I can’t help you if you don’t understand anything about the FAB process and initial yields. If AMD hits their performance goal of 20%/50% better Integer/Floating point performance at the same clock rate as the Intel 51xx/53xx series it won’t take much of a clock rate jump at 65 micron level to push way past the current series. Since they are already running the dual cores at 3 Ghz, I wouldn’t put much faith in the Quads running at 2.0 Ghz when delivered in mass. Have you noticed that the Intel Quad cores aren’t showing up in 4P+ servers. Do you know why?
Another anonymous poster, a trend here, offered up all sorts of performance differences between the top of the line Core Duo and the same AMD 6400 model. Strange, when I look up industry standard benchmarks for what Intel puts out as it Server chips (7xxx, 51xx, 53xx series) and the same for AMD, I just don’t see these kinds of differences. I take it as matter of faith that everyone can produce at least a 100 examples of data to support their position but in the real world, where tens of thousands of Servers are sold every week and have to stay in continuous 24/7service for at least three years, the workload benchmarks don’t agree with the popular cache based CPU results that favor the chip with the huge L1, L2 or L3 cache on board. That might have something to do with why Cray Research chose the Opteron to base their super scalar computers on or why IBM helped with some of the low level FAB processes in the Barcelona design. You think these corporations did this out of the goodness of their hearts? You think AMD bought ATI to get a piece of the Graphic card business? You might want to look at the CPU benchmarks for the Power6 CPU to get a better glimpse of where 65 or 45 micron level AMD CPUs are headed. I’m certain Intel will follow as soon as they get the memory controller on chip.
What seems to get lost amid all the crude, rude and vulgar one up comments is that this is a horse race and always has been. Just because you win one race doesn’t mean you get to be King of the World for ever simply because. If you Intel followers grasp the larger concept here, each of you would buy an AMD system with your next purchase. If enough of you did this, next year you would be able to buy a much better Intel system at a cheaper price and so on. It is no more complex than that and it only works when you don’t have a monopoly like we have with Microsoft and for the most part Intel too. I’ve never met anyone in my business that didn’t have horror stories about multiple Computer Vendor, etc. I’ve never met a saint in this business either. Intel has profited for years off the gimmicks surrounding the Pent 4 design. How many Pent 4 versions above 3 Ghz were there but never actually offering a 4 Ghz version? We had to turn off the Hperthreading on all our data base servers to keep it from freezing up our servers from time to time under load and get reliable performance data. At best HT added 10-15% to the potential performance under certain conditions. In practice it added nothing but headaches in a server environment and produced totally unrealistic performance metrics. That’s a marketing gimmick, no more or less.
The larger point is getting the best systems for the lowest price. You aren’t going to get that with a one horse race regardless of who that is. I don’t hide behind a cute screen name in order to not be recognized by one of my friends who might not approve of my language or tone. I am who my screen name says I am. The quality of this site will improve when those that spend time here engage their brains a little more and perhaps act their age, or a little more. There is no value added to this space by the childish rubbish being spewed here by a fast set of fingers attached to an otherwise vacant brain pan. There are plenty of sites tailored just for the sort of childish rubbish displayed here but this one should concentrate on 64 bit computing. It will arrive at some point and it won’t be because Intel makes it happen.
The latest rumor I heard is that Dr. Sharikou actually worked for Intel designing the competitor to the Pent 4. It had a lower clock rate but ran more instructions per cycle. Intel liked the sexy high clock rate model better. Better market appeal. One never really knows for sure.
Thom stated..."The latest rumor I heard is that Dr. Sharikou actually worked for Intel designing the competitor to the Pent 4. It had a lower clock rate but ran more instructions per cycle. Intel liked the sexy high clock rate model better. Better market appeal."
Somewhere, at some point, it was proven (and admitted to by this blogs loser creator) that he is NOT a doctor, does not have a ph.d and was fired from Intel for systems abuse. I'll see if I can find the links at some point if I feel up to putting in the effort.
That makes Sharidouche a liar, and his fanboi's are worse than he is...which is why I personally enjoy causing nothing but chaos here...I can't abide by lying pieces of shit, of which Sharidung and his various user ID's he uses to bolster his support here, fall squarely into...
So if you believe he's a ph.d outside of his own little mind, you're very very gullible.
AMD News...
Last Trade: 12.17
Change: Up 0.07 (0.58%)
Intel News...Ooops forgot this one..Sorry.
Last Trade: 23.89
Change: Down 0.22 (0.91%)
Thom wrote another post with even more AMD talking points! Thom, we have heard ALL of this before. You appear to be the smartest AMD Fanboi yet, so PLEASE come up with something new! And just in case you claim you are unbiased, I summarized your post below. You wear green tinted glasses my friend...
Thom wrote:
Speaking of 64 bit computing, something AMD designed into their Opteron and Athlon lines not added on
Thom wrote:
Intel followed
Thom wrote:
Intel ramped up the Netburst “heater”
Thom wrote:
Have you noticed that the Intel Quad cores aren’t showing up in 4P+ servers.
Thom wrote:
When I look up industry standard benchmarks for what Intel puts out as it Server chips and the same for AMD, I just don’t see these kinds of differences.
Thom wrote:
The workload benchmarks don’t agree with the popular cache based CPU results that favor the chip with the huge L1, L2 or L3 cache on board.
Thom wrote:
You might want to look at the CPU benchmarks for the Power6 CPU to get a better glimpse of where 65 or 45 micron level AMD CPUs are headed.
Thom wrote:
I’m certain Intel will follow as soon as they get the memory controller on chip.
Thom wrote:
Just because you win one race doesn’t mean you get to be King of the World for ever simply because.
Thom wrote:
It only works when you don’t have a monopoly.
Thom wrote:
Intel has profited for years off the gimmicks surrounding the Pent 4 design.
Thom wrote:
HT added nothing but headaches in a server environment and produced totally unrealistic performance metrics. That’s a marketing gimmick, no more or less.
Spaztic Pizza, simple test question, no tricks. Go to this website, www.spec.org. Look up the fastest intel 51xx, 53xx CPU SpecInt Rate 2006 results for 2P 8 core and 2P 4 core systems vs. the same AMD 2P 4 core and 4P 8 core Opterons, all running at 3 Ghz. You will find the fastest configurations by looking for Peak Spec Int Rate 2006 ratings above 50 for 2 P 4 core systems and above 100 for 2P 8 Core or 4P 8 Core systems. The top ratings will pretty much paint the same picture. Please tell me why the 51xx Duo Cores and 53xx Quad Cores running at 3 Ghz aren’t running away from the 3 Ghz Opterons? Even allowing for the differences between 32 bit vs. 64 bit tests, compiler versions and a host of other things there isn’t enough difference between these systems to make even a fine academic argument over. These are industry published results by Vendors that have millions on the line and it cost a bit to set up, run and get them certified. Who’s who of computing submits systems and results to this site. There is a difference between cache based CPU results that favor Intel’s super large caches vs. what actually happens when you throw real work at a “system” with multiple CPUs (sockets) that have to all synch and maintain peak performance. That’s why you don’t see 4P and above Intel based 51xx, 53xx servers in these results so far. The “Cores” don’t work well above 2P and while that might not matter a wit to you it matters a lot to Intel since servers is where they make the most profit per socket. That’s why Intel has said AMD is not the focus of their efforts. Look at the performance of the 4.7 Ghz Power6 on the same website. IBM helped with what you guys call the K10. What happens if the Power6 running virtualization of an x86 out performs your beloved Intel silicon? IBM is not short of cash and they’ve put their money on what you call the K10. You might want to open your eyes a little wider and see the whole picture.
Intel Fanboi it not my function or desire in life to feed you, cloth you, pick up after you, amuse you or entertain you. That is the function and responsibility of your parents. If everything I’ve said has been said before why did you repeat it, again? If you have a beef with something in that long list, spell it out. As to bias, there are no unbiased people on this planet but having bought and used both Intel and AMD products for nearly 20 years now I do have an insight many posting on this site do not. How many years have you bought and used both products? Answer the same question I posted to Spaztic Pizza.
Everything I said about Dr. Sharikou I made up just to see who would get spun up over it. Like I said in my original post, don’t let this become your religion. Things change.
More crap about an integrated memory controller? Honestly, who cares? The 486 had an integrated memory controller. Clearly, that must be superior to my Q6600 because my Q6600 does not feature an integrated memory controller!
could care less about the Call of Duty frame rate difference between a 3.0 Ghz Athlon 6000 X2 and the same clock rate Core Duo 51xx series. Neither do I and there is no real world difference between these two systems all else being equal.
So when AMD was faster than Intel, did you just tell people the difference doesn't matter, that no one would notice the difference?
I also laugh at the suggestions of AMD being placed up with other server CPU makers like IBM. AMD is for low end crap; they don't even scale past 8 sockets! The current Xeon MP does 32 sockets; though CPUs like the Itanium 2, POWER6 and SPARC can go far, far higher than that.
Intel's 45nm CPUs are churning out of it's fabs now. They're in no hurry to launch them though; not when they can wait for AMD to launch Barcelona then suddenly drop Penryn down on AMD to ensure that AMD has zero time to enjoy their launch. Even without Penryn; a 2Ghz Barcelona is going to be slow and pokey. It'll probably be beaten by the fastest 3.2Ghz dual core Opterons in single threaded applications!
AMD is finished. BK in Q2'08.
http://www.fudzilla.com/index.php?option=com_content&task=view&id=2557&Itemid=35
Thom, no tricks.
Go to this website that has certified benchmarks of a real busniess application, not synthetic tests.
Find an AMD system in the top 50.
http://www.sap.com/global/scripts/jump_frame.epx?content=http%3A//www50.sap.com/benchmarkdata/sd2tier.asp&CloseLabel=
Thom wrote:
Speaking of 64 bit computing, something AMD designed into their Opteron and Athlon lines not added on
They added a bunch of 64bit instructions to the CPU. That's it. Early Sempron CPUs for Socket 754 had the 64bit extensions disabled. Intel added these extensions to the Pentium 4 in the same manner.
Intel ramped up the Netburst “heater”
AMD ramped up the 4x4 heater and the HD 2900 XT heater.
Have you noticed that the Intel Quad cores aren’t showing up in 4P+ servers.
Have you noticed they're launching September 2nd, up to 32P? Where's your AMD server that can do that?
When I look up industry standard benchmarks for what Intel puts out as it Server chips and the same for AMD, I just don’t see these kinds of differences.
There are plenty of benchmarks. You just need to look harder. http://techreport.com/reviews/2006q4/xeon-vs-opteron/index.x?pg=1
(Note: Before some silly AMD fanboy says 'These don't use the fastest Opteron CPUs!' they don't use the fastest quad core Xeon either. 2.8Ghz Opteron vs. 2.66Ghz quad core Xeon. Since then clockspeed for AMD has raised 14% to 3.2Ghz; 12% for Intel to 3Ghz.
You might want to look at the CPU benchmarks for the Power6 CPU to get a better glimpse of where 65 or 45 micron level AMD CPUs are headed.
What kind of nosense is this? AMD's current 65nm CPUs are nothing but die shrinks of it's 90nm CPUs with higher latency L2 cache. It's upcoming quad core CPUs are a new architecture yes, but still x86. Since when is AMD in the business of making CPUs based on the POWER instruction set?
I’m certain Intel will follow as soon as they get the memory controller on chip.
Have you seen the poor results for AMD CPUs in video encoding? I'm sure they'll catch up when they finally add 128bit SSE support to their CPUs. Or design a core that can issue 4 instructions at once rather than just three. Or integrate a shared cache design. The list goes on and on. Read: An integrated memory controller is not the holy grail of CPU performance.
Just because you win one race doesn’t mean you get to be King of the World for ever simply because.
You might want to mention those words to AMD. Apparently, AMD thought they could live off K8 forever; and that Intel would keep up with Netburst forever.
With the revolutionary 45nm high-k process, the new 45nm refresh of the Core Architecture coming later this year (Penryn) and the new Nehalem architecture due out in the second half of 2008, Intel is not going to lose it's leadership position. Not while AMD is still working on Barcelona; trying to scale it above 2Ghz for mass production. According to the rumor sites, even the 2Ghz version is going to be quite scarce at launch.
It only works when you don’t have a monopoly.
Lets make a quick list of companies who make CPUs; just off the top of my head. Intel, AMD, IBM, VIA, Sun, Freescale and Texas Instruments. There are more, but those will prove my point nicely. How can one have a monopoly when there all these CPU manufacturers?
Intel has profited for years off the gimmicks surrounding the Pent 4 design.
AMD has attempted to profit for years off the gimmicks surrounding the K8 design and the misleading PR ratings.
HT added nothing but headaches in a server environment and produced totally unrealistic performance metrics. That’s a marketing gimmick, no more or less.
I do a lot of video encoding. So when the P4 line supporting HT was released, I bought one. the 2.8Ghz version. With HT enabled video encoding was 10 -> 20% faster than with it disabled. For me that was a tremendous gain. Granted, it didn't work well with some applications, but it wasn't a "gimmick" at all. If it didn't work well with your server applications there was/is nothing to stop you from turning it off in the BIOS.rv
Thom
Even allowing for the differences between 32 bit vs. 64 bit tests, compiler versions and a host of other things there isn’t enough difference between these systems to make even a fine academic argument over.
Just to point it out at 8 Cores (Intel 2P Quad-Cores vs AMD 4P Dual-Cores) price would be a valid point for an argument.
Just to point it out at 8 Cores (Intel 2P Quad-Cores vs AMD 4P Dual-Cores) price would be a valid point for an argument.
Indeed! Four eight cores:
2x Xeon 5365 - $1172 each, total - $2344 - Total TDP of 240W.
4x Opteron 8224SE - $2149 each, total - $8596 - Total TDP of 500W
Opteron system would use far more power, and costs much much more. You still need motherboard, memory etc. which would all cost about the same for each system.
HECTOR RUIZ: "In spite of all the hype and hoopla, there is no really such thing that Intel has leapfrogged AMD. It's quite the contrary. As a matter of fact, despite all the perceptions of Intel closing the gap, half the time they do a little bit better and the other half we do. And all of that will end with the introduction of Barcelona because it's such a significant jump in performance and quality."
Apparently all the benchmarks disagree with Hector Ruiz. Well, that's either because those sites are paid Intel pumpers or because Core 2 was designed for benchmarks!
I can't wait for the first Barcelona brown turd edition 2Ghz benchmarks.
This comment has been removed by the author.
Note to Giant: Thanks for the follow up post. I was too lazy to refute Thom's assertions.
Note to Thom: It is not your duty to amuse me, but you do anyway. You are a typical hypocritical, lying, distorting, fact-challenged AMD fanboi. You are a cut above because you can string together a coherent sentence and don't use cus words, but beyond that you are as pointless as the rest of them.
Yes great link there one expert.......now lets compare processors and GPUs that are supposed to directly compete with each other in performance and price segments using the same link.....
Since i know you are retarded and that might be too much for you to comprehend i have made this lil summary for you..........
CPU
AMD X2 6000+ at load 119.5 W
Intel E6700 at load 66.0 W (beats 6000+ in 99% of benchmarks)
Winner Intel
GPU
AMD 2900 XT at 3d load 205 W (highest on the chart despite being 5th highest in performance)
8800 GTS 640mb) at 3d load 128 W. (beats 2900XT in most benchmarks)
Winner Nvidia
Power used if you go with Intel and Nvidia at load 194W
Power used at load if you go with AMD 324.50W
Winner Intel and Nvidia
And you get better performance with the energy efficiency.
Power savings if you go with Intel and Nvidia 130.5 W
Go AMD if you want to destroy the planet while using slower outdated technology
Cant wait for that even more energy efficient 6400+ black edition and ofcourse barcelona turd brown edition.............
Thom asked me to answer the following question:
Please tell me why the 51xx Duo Cores and 53xx Quad Cores running at 3 Ghz aren’t running away from the 3 Ghz Opterons?
OK Thom. Lets resolve this. You can search the web all day looking for technical data points to support your argument, and I can do the same. You can word questions to make a success look like a failure (which you did above). So lets just drop all the spec arguments, the integrated memory controller arguments, the marketing gimmicks argument, etc., and look at THE BOTTOM LINE; How much can Intel charge for its product line vs. how much can AMD charge for its product line? I'll even give you a few dollars to account for Intel's name brand recognition.
Intel Desktop: $999
AMD Desktop: $251 ($300 for FX)
Intel laptop: $851
AMD laptop: $354
Intel Server: $1172
AMD Server: $360
Intel MP: $2,622
AMD MP: $2,149
So Thom, if AMD is so technically superior to Intel, how do YOU explain the above?
Go ahead Thom, AMUSE me with your response.
Reference:
http://www.amd.com/us-en/Processors/ProductInformation/0,,30_118_609,00.html?redir=CPT301
http://www.intel.com/intel/finance/pricelist/processor_price_list.pdf?iid=InvRel+pricelist_pdf
At last, intelligent life found on the blog net. Thom, get your blog page together. I’d like to email you. You are such a breath of fresh air.
Sharikou, AMD news is disappearing fast on the net. http://www.top500.org/lists/2007/06 no longer mentions processors in the statistics data, just vender's like Intel, but no mention of AMD at all. It really pisses me off how naive AMD is. Dell buys Alien ware and trashes it by only offering two-year-old AMD systems at twice the price that I can buy them else ware. Why? Because Alienware was AMD's best testimony to their superior performance and Dell is still getting those really big checks in the mail from their asshole friends. When I search Google (No longer a real search engine, Just sending you the “Political Correct” stuff so you can be the perfect “Google Bots’” they want you to be. Come on guy’s, most of my searches return ZERO RESULTS, instead of several million, like they did a couple of months ago.) Therefore, when I search Google for anything AMD, almost every result returned is a bunch of negative crap written by Inteler’s. Actually, this is cool, Intel is spending Billions trying to hose AMD’s Barcelona launch in the media Instead of trying to build a working product. Excellent I say, this seals their fate for certain. (A good thing!) Intel sounds so cool doesn’t it? Intel = Intelligence? How sad this is not the case.
Sharikou, AMD is obviously reading your blog. Have they offered you a job yet? They have followed
You’re every suggestion. The last being “Smash Intel with 90nm technology.” Athlon64 x2 6400 has just been released doing just that, and saving us cash too! Thanks AMD! Rock On!
Intel will fail as the 8-track tape did when cassettes were introduced. Barcelona is a holographic DVD by comparison. Do you understand the significance here? No more checks in the mail for Giant’s posts.
Yep, he will finally need to get a life. Good luck Giant it will all be over soon, cause you know talk is cheap buddy, unless Intel is sporting the bill, and all good things gotta come to an end, like the wild wood wee…. or whatever Intel is smoking these days! Famous quote from “Life of Brian” for Intelers; Just kidding, Crucifixion, uh, I mean Barcelonia!
The Doc
Excellent performance Sharikou. Definitely Direct Connect Architecture stuff. The “Front Side Bus” guys will never get it. Yellow River, wasn’t that written by I. P. Freely? I Couldn’t help that one, still very impressed though.
The Doc
Dual core CPUs at up to 3.06Ghz for mobile next year:
http://www.fudzilla.com/index.php?option=com_content&task=view&id=2579&Itemid=35
AMD Puma mobile platform pre-fragged by Intel.
AMD BK Q2'08.
P.S. It's blatantly obvious to all here that Dr Blog is a paid AMD pumper spreading FUD about Intel. What will people like him do when Nehalem comes out next year with CSI/QuickPath? They won't be able to whine about the FSB 'bottleneck' anymore!
They added a bunch of 64bit instructions to the CPU. That's it. Early Sempron CPUs for Socket 754 had the 64bit extensions disabled. Intel added these extensions to the Pentium 4 in the same manner.
You should to compare the Pentium 4 with the amd64 athlon 754 not the sempron one..
AMD ramped up the 4x4 heater and the HD 2900 XT heater.
And according to this site, the latest 65nm V8 of intel is sucking much power in idle compare to the OLD AMD 4X4.
http://techreport.com/reviews/2006q4/xeon-vs-opteron/index.x?pg=11
Have you noticed they're launching September 2nd, up to 32P? Where's your AMD server that can do that?
Don’t you remember that amd already reach the 32P level long long time ago?
http://findarticles.com/p/articles/mi_zdpcm/is_200305/ai_ziff41366
There are plenty of benchmarks. You just need to look harder. http://techreport.com/reviews/2006q4/xeon-vs-opteron/index.x?pg=1
Wow.. One of the AMD slower server processors get frogged by the intel fastest server processor.. In fact, both of the Intel Quad processor sucking more electricity power compare to the amd 4X4 processors..
http://techreport.com/reviews/2006q4/xeon-vs-opteron/index.x?pg=11
What kind of nosense is this? AMD's current 65nm CPUs are nothing but die shrinks of it's 90nm CPUs with higher latency L2 cache. It's upcoming quad core CPUs are a new architecture yes, but still x86. Since when is AMD in the business of making CPUs based on the POWER instruction set?
Don’t you know that the K10 design is almost similar with the Power6 design?
Have you seen the poor results for AMD CPUs in video encoding? I'm sure they'll catch up when they finally add 128bit SSE support to their CPUs. Or design a core that can issue 4 instructions at once rather than just three. Or integrate a shared cache design. The list goes on and on. Read: An integrated memory controller is not the holy grail of CPU performance.
Have you seen how poor are the results of the Intel CPUs scaling when it goes to the multi cores session? However, Im sure they’ll catch up to improve their Quad processors to be the actual native 4 cores rather than perform just like a 3.5 cores when they finally found out how to copy the AMD native Quad core design when Barcelona is launched this 10th September. Read: The FSB is not the future of the Intel processors.
You might want to mention those words to AMD. Apparently, AMD thought they could live off K8 forever; and that Intel would keep up with Netburst forever.
AMD will stand and lives forever if Intel cannot catch up to be a real native multi cores in the future. Intel thought they could live by keep gluing and gluing their antique processors?; and the AMD would keep up with the better design multi cores processors.
Not while AMD is still working on Barcelona; trying to scale it above 2Ghz for mass production. According to the rumor sites, even the 2Ghz version is going to be quite scarce at launch.
Yes.. AMD will only introducing a 2GHz server processor this 10th September. However, don’t forget that although it is only clocked at 2 Ghz, but it is 20% faster compare to the Intel 2.33Ghz.. ;-).. In desktop competition, AMD already shown their Phenom clocked up to 3Ghz and it could be clocked higher than that who knows?
This comment has been removed by the author.
Pezal, you're back! Still unable to do basic math like when you said a mythical 2.4Ghz K10 would be 450% faster than the one that was actually shown at Computex at 1.6Ghz?
You're long rant had one choice line:
Don’t you know that the K10 design is almost similar with the Power6 design?
You have absolutely no idea what the hell you are talking about and that line is the perfect proof. The K10 is about as similar to the Power 6 as any Intel CPU is... which is to say not at all. In fact, if anything the old Pentium 4's share more in common with the Power 6 than anything being currently produced by Intel or AMD and even that is a stretch but I'll explain below.
The Power 6 was designed from word 1 to have incredibly high clock speeds, while at the same time sacrificing some IPC (just like the P4 did). Look at the benchmarks comparing Power6 to Power5... the Power6 is definitely faster, but the clock speed difference is MUCH higher than the actual software execution speed difference. (See here note that this is an actual article, not a link to someone playing the trombone).
Power 6 relies on an older in-order architecture which means that it must execute instructions as they come to the chip, even if that would cause pipeline stalls while waiting on data from cache or memory. Intel brought out of order execution to the mass market with the Pentium Pro and AMD copied the idea in their chips, all chips from both companies have been out of order for years.
IBM decided to go with the simpler architecture since it wanted much higher clockspeeds. While OOE is useful, the logic necessary to implement it is quite complex and often will limit the speed of the overall CPU. So... IBM made a design decision to sacrifice IPC in favor of clockspeed. While the details are different than what Intel did with the Netburst architecture, the underlying tradeoff is exactly the same.
And once again, you are completely wrong in assuming that just because IBM throws AMD some scraps from its manufacturing process (which is not as good as Intel's) that somehow AMD's chips are all magically just like IBM's. Try taking some classes at a community college or getting your GED instead of posting here.
This comment has been removed by the author.
Giant, Chuckula, 13ringinheat & Co
Why exactly the hatred for AMD?? A few AMD fans here have out and explained why they like, what they like and also spoken from their obvious years of experience using both products. You guys show up here and make rubbish statements, bash Sharikou.
Here is an example: - 'AMD BK in Q2 '08'
Another one: - "Go AMD if you want to destroy the planet while using slower outdated technology"
What is the reason for all this love for Intel?? Care to explain...
Otherwise logic says you guys are out to grab AMD stock cheap or you are the marketing (mafia) money speak of Intel. A lot of you profiles are hidden and most of you have been sighted here ever since Conroe came out!!
BUY ATI/AMD energy saving safe video solutions...
Onemoron, its already been clearly shown that ATI/AMD's video products at the enthusiast end use more than 100 watts more than the Nvidia product.
Are you just fucking retarded or it it the case of serious stupidity?
Here is an example: - 'AMD BK in Q2 '08'
how is it any different than Sharidouche, the Ph(ake)d claims of Intel BK in Q2 08?
Lemme see one is supposedly from a self professed expert, the other is from a random reply in that self professed expert's blog.
I wonder who is really the moron...
Evil_Merlin: Although both comments are stupid... one is a little closer to reality than the other (hint: look at AMD's cash flow)
oneexpert
"Nvidia has Linux security hole"
If AMD would have drivers one could install and that would actually work they will have holes in them too. Luckily for us they don't have such drivers, problem solved.
Blah! Blah! Blah!
This comment has been removed by the author.
Netrama:
I do NOT hate AMD. Or more accurately, I don't hate the engineers and designers at AMD who actually work hard under difficult situations. Some of those guys went to school with me and I feel sorry that they are getting screwed over. I do have absolute contempt for Hector Ruiz, Randy Allen, the board of AMD, and the other higher ups who took AMD's success and drove it into the ground to make a quick buck and to get themselves on the covers of some business magazines.
I am probably the only person on this site who actually has given AMD advice that won't end in its bankruptcy. You need to learn the difference between a 'friend' and a 'sycophant'. In fact, just look at the Michael Vick case for the perfect explanation.
Michael Vick never had any true friends, instead he had a bunch of sick-ass sycophants who sucked up to him and always agreed with him and influenced his decisions in really bad ways to kill innocent dogs. They never said anything bad about him, and never ever questioned any of his actions or said what he was doing was wrong. Look where that ended him up, he's going to jail because he never had a single person come into his life and say that what he was doing was completely wrong and that he had to stop it. That person might not have sounded 'nice' to Michael Vick, but that person would have been the only friend he ever had.
Pezal, Sharikou, one_molester, etc. etc. are all just like Michael Vick's sycophants. They don't actually give a shit about AMD, they instead 1. hate Intel for some reason I can't understand, and 2. blindly agree like brain-dead lemmings with every buzzword that AMD marketing comes up with. They are literally incapable of rational thought. No matter what Intel does it's always 'evil' or 'stealing' from AMD. Here's an even better example: Intel implemented a shared L2 cache on the core 2 CPUs, and all the AMD fanboys started screaming about how stupid and wrong that is merely because AMD didn't do that with the K8. Guess what? The K10 has a shared cache just like what Intel did! The new party line is that Intel must have stolen the idea from AMD and that Intel's designs are just ripoffs of AMD. You see? No thinking, no analysis, no thought, just blind emotional hatred. That same kind of hatred is what got 6 million Jews gassed and that same hatred is why Muslims are blowing up other Muslims in Iraq right now.
I am the only true friend that AMD has on this website because I want the asshats who are destroying the company fired and I want them to start telling the truth about what they are making. You can't begin to heal a sick patient until you admit that something is wrong and go try to fix it.
When Intel does something smart I say it's smart. When Intel does something dumb, I say it's dumb.
When AMD does something smart, I say it's smart.
When AMD does something dum I say it's dumb.
That's called the truth, and AMD needs a lot more of it an a lot less of the Sharikou crowd.
Oh pezal, this from the very article you linked to Re: 32P Opterons, perhaps you should have read it first?
Also, the IA-64 has strength in server designs that use more than four processors (4P), scaling effectively to 32P and showing prowess in floating-point–intensive applications.
IA-64. Read: The ITANIUM 2 can scale to 32P and beyond. AMD's 64bit solution is called AMD64 or x86-64, not IA64.
If AMD would have drivers one could install and that would actually work they will have holes in them too. Luckily for us they don't have such drivers, problem solved.
Indeed! Under Linux, there are STILL no drivers for the HD2x00 series. Months after launch, that's just pathetic.
Speaking of graphics, G92 is still set for a Novemember launch, Nvidia style. Nvidia style meaning full product availability from retailers on the day of the launch. I think I'm more excited about this than any other upcoming 2007 product (Penryn included). Being an avid gamer, seeing Nvidia double performance once again just one year after the Geforce 8 launched will just be incredible. I can't wait to grab one of these on launch day.
http://www.fudzilla.com/index.php?option=com_content&task=view&id=2586&Itemid=51
So oneexpert, when you quoted AMD's massive 2006 growth (Buying Ati accounted for most of that growth) why did you forget this line from the very same article?
We know it won't be looking that way in 2007
It also seems that Henri Richard may well be resigning from AMD:
http://www.hexus.net/content/item.php?item=9646
Now we just need Hector Ruiz and Randy Allen to follow suit and then maybe AMD can get back on track!
AMD is finished. BK in Q2'08.
For those who can't tell, my postings here are a direct parody of Sharikou's fanatical pro-AMD stance, claiming Intel's bankruptcy in 2008 etc.
I owned an Athlon 64 X2 4200+ system, bought and built it the month the CPUs became available. It was a great system. Now I own a Q6600 system, Intel has the better CPUs now, simple.
I owned multiple Ati video cards in the past, a 9700 Pro and an X800 XT . Since then I've owned a 7600 GT and an 8800 GTS.
I just buy whoever offers the better technology at the time.
Chuckula has the right idea, Hector Ruiz, Randy Allen, Henri Richard and all the rest of their cronies have to go. Dirk Meyer seems like he knows what he's doing. He would make a fine CEO, or they could ask Jerry Sanders to return.
Giant, Intel Fanboi, ( and others ) I never said any AMD was faster then the top of the line Intel 51xx, 53xx series. I said there is no real world difference for the bulk of the people who buy PCs (desktops, laptops). I also said everyone could produce 100 examples to support their conclusions about who has the fastest doing this but it does not matter to the bottom line of either Intel or AMD. Still no one has answered my question about the relative sameness of the SPECINT Rate 2006 results between the 3 Ghz Opterons and the same in the 53xx series Intels. Why?
Giant, I’m glad HT worked for your single application. It cost tens of thousands of dollars in labor charges to research where it was causing problems and turn it off on thousands of servers my corporation has. You understand the magnitude difference here? The bios settings had to be changed and every server rebooted. Think in terms of the number of employees that would have had to be let go to pay for this. That’s not trivial pursuit in a 24/7 environment.
As for all those other CPU manufacturers that would take AMD’s place, really? What IBM, VIA, SUN, Freescale and Texas Instruments x86 CPU have you bought lately?
Chuckula, you make a few good points about IBM being able to speed up their entry level Power6 CPUs into a range Intel never could achieve and the relative clock rate vs. performance gap but what’s the point? If Intel could have gotten the Pent 4 to that level would we be having this conversation about Duo Core stuff? What happens when the Power6 goes to 5.5 Ghz? Nor did I ever say that the AMD product was the least bit like the Power6 (and the Power5 produced better results than the both Intel and AMD at a clock rate basis but it was still less powerful than the upper end Intel and AMD stuff). What I said was that IBM, a relatively conservative company when it comes to performance claims seeded some money to AMD to improve their NM switching process speeds and reduce power and heat output. Said another way, some of the base technology that allowed the Power6 to go into clock ranges that Intel only dreams of and produce a rather measurably higher industry standard CPU benchmark over either Intel or AMD (also found at www.spec.org) as a result might lead to a similar advance in the Athlon/Opteron line where its performance does scale well with clock increases. Since I work on dozens of Sun SPARCs, IBM Power5, soon to be Power6, Intel and AMD powered servers every day, I believe I know what I’m talking about.
Dr Blog, I have to work for a living. Don’t have time to blog all day/night as some on this site do. All it takes to put info on this sight is a fast set of fingers, access to Goggle and unlimited free time. Thanks for the offer.
Henri Richard is RESIGNING!
The movement has begun.
Also, Hector admits that there were indeed problems with Barcelona and thats why they had to keep delaying it.
oh how it feels to good to say I told you so!
Bubba, the problem with most application benchmarks is there are numerous other factors (variables involved beyond just a different CPU running the test) and most end up being Apples to Orange comparisons. I could point you to both Intel and AMD based results for Sales and Distribution at HP’s website. I deal with this everyday trying to get the most out of Sun SPARC, IBM Power5, Intel and AMD servers running some of the same applications but usually being too different to directly compare at the application level across platforms. A SAP Sales and Distribution test would be the last place I’d go to see relative CPU performance. We’ve doubled and even tripled the relative CPU power on our SAP systems and the per unit transaction processing speed does not speed up enough to talk about. We have plenty of CPU power in reserve even though our SAP systems are mounted on the world’s slowest 64 processor. CPU power is often not the determining factor in real world application performance.
enumae, the point of my qualifications (2P, 4P, etc) was simply to help narrow down the horde of configurations found at www.spec.org. The common denominator is two parts, one 8 cores and two, a system that can feed those 8 cores with work, in this case Data Base engine work and very large memory models (32 -128 Gig memory). The 4Ps are typically 4U+ chassis that can handle larger numbers of RAID controller cards while the 2P are typically 1U or 2U and can’t. The 1U and 2Us tend to be Blade or Application servers surrounding the 4P and above Data Base Servers. The heavy lifting is done on the DB server not the 2P types. Said another way, a 2P 8 Core 3.0 Ghz 53xx Quad can not serve the same I/O bus structure that the 4P and above can so while it looks like the 2P Quad Core Intel is a better deal and could do the same work as the 4P 8 Core Opteron, it can’t. That is why we are still getting 3-3.6 Ghz Netburst 7xxx Intel 4P 8 core Servers at work even now which are measurably slower than the same speed Opteron and a lot more expensive than any equal core 5xxx system. There is more to this than just the CPU/cache based results you get from optimized benchmark products. Think of it this way. You can put a Group 5 Porsche engine in a 1968 VW Beetle but will it perform the same in a 24 hour Le Mans as the Group 5 Porsche? A long time ago, Chevy thought they could beat the Group 5 Porches with hopped up Cavilers. On paper the Cavilers had more of everything including horsepower. The Porsches finished 1, 2, 3. The Cavilers didn’t win because the sum of the parts wasn’t up to the task of a 24 hour race. Every time the Porsches went off the track (often), them came back on and kept going. Most of the Cavilers didn’t finish. I’m not suggesting Intel’s 5xxx series are Chevy Cavilers but for reasons that most Intel fans don’t want to accept, the 5xxx series aren’t being installed into the Porsche level platforms and that is continuing to cost Intel market share and profit in its most profitable sector. AMD has talked about getting out of the desktop line. It might happen. They will still have their Server line; IBM’s research help and Cray Research choosing the Opteron over any Intel CPU. Intel would win. Who would ultimately lose? I’ve seen this cycle repeat for over 35 years starting when IBM was King of the Hill. The faces change but it still the same old story and one that’s I’ve simply tried to relate to based on my 35 years of slaying this dragon. Don’t make this your religion and I do buy and own both Intel and AMD systems. I reward the one that has my best interest in mind when I choose to buy another system.
t IBM being able to speed up their entry level Power6 CPUs into a range Intel never could achieve and the relative clock rate vs. performance gap but what’s the point? I never said that, please go back up to my previous post about sycophants and seek professional help.
IBM's 'entry level' Power 6 is massively more expensive than any Xeon you can buy and should be compared to the Itanium instead. In fact, it's 'entry level' Power 6 is only clocked at 3.5Ghz (see here) which clockspeed wise is done every day by regular Intel chips with some overclocking. Only the extremely expensive highest-end chips are at 4.7 Ghz.
You also fail to realize that the Power6 needs 790 million transistors to get things done, which is more in 2 cores than a present Quad Core Xeon uses. That means Power needs lots of power (electricity). Which is fine since it is a very high-end chip that often will have customized cooling solutions. The last thing you've forgotten to mention is that IBM is maybe going to be selling 20-30 thousand of these things a quarter (that may even be an optimistic number). Intel shipped over 1 million quad core chips in about 3 qaurters, so even what AMD fanboys call a 'paper launch' by Intel is over 10 times more than what IBM sells volume wise.
Don't make the mistake of assuming that just because Intel makes mass market chips that are inexpensive and run well that they are somehow 'stupid' because IBM can design extremely expensive chips for a limited production run. IBM does not actually make any money from those chips themselves... it makes money from the massively expensive machines that IBM sells and even more money from the multi-year service contracts that IBM gets from its customers. If IBM tried to simply sell these chips it would never ever make a profit, while Intel is able to just that.
If you'd actually been in IT for as long as you claim you have been, you'd know that it is all about engineering tradeoffs. Any undergrad could design some massive processor that would beat the crap out of anything currently in production on paper. The difference is in actually being able to produce the chips in an economical fashion. This is something that many people on this website cannot even comprehend, they are too busy reading powerpoint slides and empty promises to actually think about the real problems involved.
One final point, I just love how you are screaming about how great Power6 will be at 5.5Ghz. The funny thing is that IBM originlly promised Power6 at way above 5 Ghz (they even said 6Ghz) but was not able to actually deliver! Sounds a whole lot like another 3 letter company I know. At least IBM was still able to get something that worked. And to answer your dumb rhetorical question, at 5.5Ghz the chip will be at most 17% faster than what IBM sells now, and that is assuming perfect scaling which never happens in the real world. 4th grade math strikes back.
Why exactly the hatred for AMD??
I have no hatred for the company i just have a hatred for liars which turns out to be most amd fanbois so i guess amd is guilty by association.
A few AMD fans here have out and explained why they like, what they like and also spoken from their obvious years of experience using both products.
yes alot of AMD fans come up with things that happened in the past and fud that AMD promises with their slide shows. Currently Intel processors are unmatched and their products are actually on the shelf instead of slideshows
You guys show up here and make rubbish statements, bash Sharikou.
There is not one statement besides that obvious mockery of sharikous bankrupt statement that any intel person has said on here that is not validated by numerous benchmarks and links around the net. The same can not be said by AMD fanbois here so you tell me who has more rubbish statements......
Here is an example: - 'AMD BK in Q2 '08'
obvious mockery of sharikou's retard prediction
Another one: - "Go AMD if you want to destroy the planet while using slower outdated technology"
Obvious mockery of one experts retarded statements though this one is kinda true because AMD's technology currently is outdated and their products do use more power compared to the competitor that gives better performance which the links have proven......
What is the reason for all this love for Intel??
I love performance and they have the better product RITE NOW
Otherwise logic says you guys are out to grab AMD stock cheap or you are the marketing (mafia) money speak of Intel.
Logic also says that AMD is in big financial trouble or we are just refuting the BS AMD fanbois here put out which is almost always all lies
A lot of you profiles are hidden and most of you have been sighted here ever since Conroe came out!!
What do u need my profile for i am not looking to be your friend. I have been reading this blog as a joke ever since conroe's benchmarks were released and loved how all the AMD fanbois kept denying it till conroe was released knocked some sense into their delusional world.
Henri Richards leaving AMD in first week of September?
http://www.theinquirer.net/?article=41864
Thom said:
The problem with most application benchmarks is there are numerous other factors (variables involved beyond just a different CPU running the test) and most end up being Apples to Orange comparisons.
OK Thom, here we go with the AMD fanboi talking point bashing benchmarks.
1. When AMD was winning the benchmarks, there was not one AMD faboi like you making the above statement. You just don't like what the benchmarks are telling you now.
2. The point of a benchmark is to measure performance in a manner that is repeatable, portable, verifiable by second parties, and timely. Do you have an alternative way for multiple interested parties to determine performance on multiple configurations in a timely, verifiable, repeatable manner? Since you have been in the business for "35" years, your insight would be enlightening.
3. Are all those gaming companies and software vendors just unaware idiots when they create benchmark programs so that the user can test out different configurations? Or is the benchmark a pretty damn good reflection of what you can expect for your system?
4. Sharikou was the first with this argument. He claimed that all benchmark tests fit into the Core 2's L2 cache and thus were not accurate real world reflections. So Thom, like I said, you are a repeat of Sharikou from 2006.
You are a liar and a fraud. And worse, you are probably aware of it.
intel fanboi said: You are a liar and a fraud. And worse, you are probably aware of it.
Another AMD fanboi who is a liar and a fraud just like the company.......oh the surprise!!!
I also like the arguments these fanbois make about how AMD is "good enough" for apps. If we start following their advice we would never progress........oh that explains why the 2900XT , 4x4, AMD 65nm, AM2 and 6400+ X2 was released..........
AMD....when being just good enough will suffice
CPU power is often not the determining factor in real world application performance.
Then why do you post spec benchmarks showing AMD cpus are faster than Intel, if it doesn't matter?
Methinks you just don't like to admit that when comes to the real world, AMD systems are slower than Intel systems in the vast majority of the cases.
WOW
AMD's top salesman leaving the company two days before the barcelona launch.
That CAN'T be good news. What is it he doesn't want to have to live with?
Here's the email.
TO: All Employees
FROM: Hector Ruiz, AMD Chairman and CEO - Dirk Meyer, AMD President and COO
SUBJECT: Henri Richard
We are sad to announce that Henri Richard has made the decision to leave AMD.
When Henri joined AMD in 2002, his primary mission was to establish a world-class global Sales and Marketing organization. It is safe to say that he has accomplished that mission, and he is now ready for a new challenge in his career. While we will certainly miss him, it’s a measure of his success that the organization he leaves behind is fully poised to succeed and capable of maintaining and building the momentum that we have built.
In the last five years, we have increased exponentially our global account footprint, acquiring customers of every caliber, including the top PC and server OEMs around the world. In fact, from Toshiba and Acer to Lenovo, Dell, Sun and HP, we have become a critical strategic partner to our customers and a key component of enterprise solutions and consumer products worldwide.
Commensurate with this growth in our business, and the strength of our corporate and product brands, we have a Sales and Marketing organization that can support and continue to grow our strategic importance to our customers. Nevertheless, we know this was a difficult decision for Henri, especially as we are poised to enjoy the successes of our acquisition of ATI, our upcoming quad-core Barcelona product and our strong product and technology roadmaps.
Henri’s official departure date is still pending, and we will communicate a leadership plan shortly.
Look for a Q&A with Henri to appear on AMD Online in the next day or two. Henri will discuss AMD’s accomplishments over the last five years, and our phenomenal opportunities moving forward.
Please join us in thanking Henri for his leadership and wishing him well in his post-AMD career.
I'm sure the Sharidouche will say that Henri feels no need to hang out now that Hector's ready to pull the trigger on Intel.
Intel FanBoi, Budda, and who ever else feels up to the challenge, it is generally considered rude, crude, and immature behavior to not answer a question when asked first and then go off on a diatribe about all sorts of things not related to the question asked. Adding a little name calling, vulgarity on top of trying to win an argument with “weight of paper” (aka filling the space with as many facts and figures as your fast moving fingers can Google) method completes the picture of a very insecure, immature individual. Be that as it may and I concede you aren’t going to answer my first question or are incapable of answering it so I’ll give you one more chance with a simpler question. Let us see if you can keep childish emotion out of the arena for the 10 minutes it will take you to research this and make a recommendation.
You have a real job, one in which your decisions will impact thousands of co-workers and could cost millions to undo. Your IT Dept gives you a request for 1000 identical x86 servers with the following specs. 8 Cores, 32 bit and 64 bit x86 compatible, initial memory requirements of 64 Gig; growth to 128 Gig, 7 SAS PCI-E RAID 5 controllers, each with 32 15K rpm SAS drives (and spares for each controller online). Corporate policy limits your system Vendor to one choice, in this case HP and your Server models to any one of the DL/ML 300 or 500 series or the Intel® Itanium® 2 Integrity systems with 8 cores. For the sake of this problem, everything cost exactly the same regardless of the CPU model you chose. The memory, RAID controllers, drives are all the same, only the CPU selection varies. Each system fully equipped, installed and running is going to cost about $175,000. 1000 times $175,000 is real money. These servers will be in service until this time in 2010. They are replacing 3 year old 8P 2.0 Ghz Xeons with 32 Gig of memory running at 80% capacity. Load Growth is expected to more than triple in the next three years. While the servers themselves will cost the same, any cost outside this will be evaluated and added to the total system cost. (hint: any issues that have to be worked around or incompatibilities will be added to the price of each server) All these server models can run 64 bit software but the Corporation has mandated that all software on them be 32 bit only initially. They will run a RDBMS that takes all but 4 Gig of memory. The OS has no bearing on this problem. Make a chose and justify it based on what you can find related to these server models.
The wrong answer will cost you your job, reputation and make your life a little harder until you pick yourself back up and learn from this. The right answer will allow you to keep your job and reputation and another chance to fail later on. This word problem is played out a 1000 times a week across the world, each with a slightly different flavor that adds weight here and takes some away there. Any of the major system Vendors could be used for this. I chose the No 1 Vendor because they bend over backwards to provide as much flexibility in their systems as possible. They don’t care what your personal preferences are. It’s your choice (and consequence).
Chuckula you are absolutely correct. IBM machines have been massively expensive systems for going on 46 years now…and they still make and sell Mainframes despite no Mainframe being able to come close to the CPU power of my Intel based laptop. Funny about that but the 4.7 Ghz Power6 is still more powerful than anything Intel has to offer and they are running it without having the chassis melt. There is more to a computer system purchase at this level than just the raw CPU power as expressed by a CPU/Memory loop benchmark. If that was all that mattered, all but one CPU vendor would have died out 2 decades ago. If that was all the mattered, IBM would have gone BK in the eighties. Neither Intel or IBM is going BK and I strongly suspect neither is AMD. Bought by someone with deep pockets maybe but not BK.
By bye Henri...
Why would the head of marketing choose to leave right before the biggest launch for AMD
Is it because there are NO benchmarks on the K10 while Penrym is already widely published..
The rats are leaving and soon Hector will get fired.
AMD is finished
I’ve made three salient points on this blog that simply go over or through the heads of those that favor Intel.
The first, don’t let this become your religion because things change in this business, is simply common sense advice coming from years of experience dealing with an ever changing landscape. Those that do let these matters become their religion typically end up rude, ugly, crude people filled with irrational hatred that ultimately will cost them a lot. The behavior displayed by many Intel fans on this blog while hiding behind anonymous screen names will not be tolerated in the real world where you will have to make face to face communications with people who will simply dismiss you as a child. I’m the same on the blog as in real life and I’m taken seriously at work where millions are on line. If your focus is always simply to “win” an argument rather than carry on a conversation, be careful for what you wish for.
The second point was simply Econ 101 with a little real world marketing thrown in. If we end up with one X86 Vendor we won’t be taking about the latest CPU benchmark but the outrageous price Vendor X is asking (and getting) because we live in a captive market based around CPU type now. Some of you don’t appreciate that very much. The same is true for OS (Windows) now and that should be all the warning you need to not wish it to happen.
The third point, as represented by the two questions I ask, Why are the 5xxx series and the Opterons running at 3.0 Ghz producing the same certified SPEC INT 2006 RATE throughput numbers for 8 cores and the very much real world IT decision process problem with having to pick the best system to meet stated requirements was simply to get you to spend 10-15 minutes actually looking below the surface to see that all is not what it always seems or as some of you would wish it to be. Been there done that but I never let my personal preferences over ride my business logic where it counts. Many of you I think would and suffer eventually for that.
At no time have I said that the Intel 5xxx series was slower than the matching clock rate Opteron but some of you keep trying to prove that I said the Opteron was faster. By the same token some of you think power per watt and alike means anything in a Server situation with the boxes running full out 24/7 along with hundreds of 15 k rpm hard drives. It doesn’t. The biggest heat producers in my personal desktop systems, regardless of Vendor are the hard drives and power supply and then the memory. In my laptop it is the CPU however. What that should tell you is my desktops are used very much differently than my laptop and other things besides CPU power are important.
When I pointed out that what drives most of you is moot to the bulk of the people who buy PCs today, most of you simply dismissed that. When I bought my 2002 Intel Laptop with a Pent M in it, I purposely bought the lowest price 1.6 Ghz version because Intel wanted a considerable price premium for the next two levels up. The money I didn’t spend on the higher speed CPUs I put into more memory and faster hard drive. Which do you think provides a better performance feel in normal everyday use, the 2.0 Ghz Pent M laptop with the 30 Gig 4200 rpm drive and 256 meg of memory or the 1.6 with the 5400 rpm drive and 512 meg of memory? They cost the same. It even has a 7200 rpm drive today and the difference is noticeable for most things. I’ve put my 1.6 Pent M up against a lot of 2.0 Ghz Pent Ms with their standard drives and memory and beat them in everyday use factors. There is no doubt that my 1.6 Ghz CPU is measurably slower than the 1.8, 2.0 and 2.2s that followed it but I got a better system for the same money by understanding the sum of the parts for a given price has value. Your typical AMD based product of comparable performance cost enough less than the Intel product to add more value to the AMD system in other areas and come away with a better system. Most PC purchases are interested in value for their dollar and $100-200 difference in identical systems except for the CPU brand (and raw performance) still means something with those that work for a living and have other priorities in their life outside of “whos on first” today.
If the focus of your life it to turn every conversation into a verbal brawl rather than to gleen something that might be useful down the road in your life then many of you are probably going to find out the hard way that most people will not tolerate this kind of behavior. As you get older this will be made clearer to you.
Thom,
you said:
You have a real job, one in which your decisions will impact thousands of co-workers and could cost millions to undo. Your IT Dept gives you a request for 1000 identical x86 servers with the following specs. 8 Cores, 32 bit and 64 bit x86 compatible, initial memory requirements of 64 Gig; growth to 128 Gig, 7 SAS PCI-E RAID 5 controllers, each with 32 15K rpm SAS drives (and spares for each controller online)
Seven RAID controllers supporting 32 drives with spare is 256 drives per server. Is that correct?
Dr Blog (Roger)
Thom,
You have attempted to portray yourself as a reasonable individual, but your pro-AMD agenda is as obvious as Sharikou's. You are a liar.
And as far as AMD goes, I have never criticized that good company in any post. All I have ever done is defend Intel from liars and frauds such as yourself.
This comment has been removed by the author.
Thom said
"AMW, I’ve owned 6 Intel based machines and 4 AMDs in the last 20+ years. That doesn’t include upgrades to most of these. The AMD models were generally more upgradeable than the Intel ones because Intel designs their designs to have shorter life cycles. I didn’t say I didn’t have a bias but my bias comes from two decades of working with Intel products. "
What machines you have owned has nothing to do with my assertion that you were being hypocritical. Your main point in your first message was that you should not follow the cpu manufacturers, which is fine, but you addressed it to Intel favouring people only when you should have addressed it to both sides.
On a different matter I'm not sure how you can claim the AMD's are generally more upgradable, not when my 939 system based system was passed passed over for AM2 and my older Intel Northwood 478 system having been upgraded since to a Dothan !
I think your admitted AMD bias is blinding you as much as most other people on this blog, but you are not prepared to admit it and generally just sound like you think you are above the herd and end up being patronising. Sorry, but you do.
Well, i visit this blog usually for some laugh and occasionally post some comment,
especially for post that mix some real statement and false statement, making them sounds
real, or some hypocrate comments poster, some one clearly a fanboi and claiming that he/she
is not or pretending to be neutral.
Hypocrate, has been a trait in AMD, especially recent years. Bashed Intel on benchmark
pratice while itself not only doing the same, to some extent, worse. Basked intel on tying
its marketing agenda with the process technology while itself talks about the native design
and first to use immersion tools, etc
For the posters here, hypocrate is when you raised question why some hated AMD so much in such putting a date on its bankruptcy, while knowingly it was the blog owner that started all this more than a year ago and people that can't stand it and started to post parody comments.
hypocrate also shown by asking people not to make technology favorite into a religion while clearly himslef shown signs of a believer.
hypocrate is when keep claiming Intel CPU as heater, knowingly this thing has passed and AMD has taken the heater throne lately wit its 6000+ series, FX series, and 4x4 platform.
hypocrate is also in such effort of image bashing portraying intel laying off its 'loyal' employee for keeping cost low, without mention its religion (AMD) also did the same.
hypocrate in such that mixing a plain statement within other bashing statements making the plain statement sounds bad. there is nothing wrong in SXing the product in order to sell at lower price. this protect your higher end margin, target other crowd that has different budget. And remember, AMD has been doing this as well (back then it copied everything that Intel had, and now it has its Sempron line).
hypocrate in such protraying himslef of no bias, but putting effort in portraying AMD CPU as euqal if not better. A true statement that average user doesn't care about the frame rate differences with both company high end CPU, but enthusiast care! Average users should not buy high end model anyway. And when you do not have competitive product today, then quote old models to imply the superiority by saying my 3200+ is faster that all of the 2.6 Ghz Intel Netburst PC .... and the funniest thing, with a bracket (still). And also even go to the extent to compare AMD 386-40 to Intel 486-25SX, knowing in last time, the frequency matter and there is whooping 60% freq gap! Enough effort in setting up the psychology trap.
hypocrate in such talk long and lenghty on why he need to use 4P (where indeed, AMD still has some advantage over intel offereng under memory intensive workload), just to protray the superioiry. Remember, while some server computing needs are indeed memory intensive, there are some CPU processing intensive too. Effort also been made not to mention anything on UP and DP.
hypocrate in such keey saying the existence of AMD keep intel's price in check. Well this is true, but he failed to realized the reverse is true too. Without intel, we still need to pay hundreds just for the 3800, instead of tens.
Clearly, some have chosen a technology religion.
Pointer,
That may be my favorite post on this blog in a while. Sharikou is a liar, but at least he is honest about it :-) unlike Thom.
Post a Comment
<< Home