AM2+ HT 3.0 motherboard now available
The killer is coming. K10 frags Core 2 more than K8 did Netburst.
Intelers are always in denial. In any case, Intel won't have similar technology until 2009 the earliest. It took the brilliant Israeli engineers three years to figure out Pentium 3. It has take them five years to grasp the essence HyperTransport 1.0 and direct connect architecture 1.0.
They plan to have a copy of HT (codenamed "quickpass") fully operational in 2009, which they call the greatest Intel innovation since 80286.
AMD's strategy is simple: market share at all costs. As long as Intel's market share is shrinking, its ability to wage war is severely hampered. As AMD takes more share, it only needs a small window of opportunity to finish off Intel due to the latter's large size and inefficiencies.
A 2.6GHZ K10 is sufficient to frag a 4GHZ Intel Quad -- recall Intel CPUs do not scale with frequency due to FSB bottleneck. Such a crushing performance lead will force Intel to cut prices, coupled with the loss of market share, Intel's demise is the inevitable result -- unless Intel quickly downsizes itself and seeks a level of survival that is about 1/3 of its current size.
But by 2009, Intel will have BKed, and won't be able to pay the anti-trust penalties.
PS: some readers seem surprised at those primitive attacks, get used to it -- Intelers have low IQ.
posted by Sharikou, Ph. D. @ 10/03/2007 09:16:00 AM
151 Comments:
Sharikou, so tell me again, when is Intel going to BK?
And explain to me just how this frags Intel's CURRENT C2D line up considering that the Barcelona can't even compete with intel's C2D based DESKTOP CPU's?
of course you can't
Sharikou. News for those that like to eat shit.
You are lucky you can hide behind a blog Sharikou. No one in the real world is stupid enough to hire you, let alone let you do freelance work for them.
Sharikou, news for those that avoid reality!
And meanwhile, Intel is getting ready to release 45nm processors with an evolutionary transistor redesign using high-k (hafnium-based) materials and metal gates --which greatly reduce electron leakage, allowing their chips to run cool while also running fast.
Poor AMD doesn't have this technology yet, and IBM can't give it to them becauset THEY don't have it yet, either!
Intel maintains the performance and efficiency lead through mid-2008, also enjoying manufacturing supremacy, allowing them to drop the prices as much as they like, undercutting AMD's profits.
Then Nehalem comes in mid-2008 -- wiping away any (perceived or real) advantages AMD has with Hypertransport and the integrated memory controller.
Sharikou, the idiot, wrote,
AM2+ HT 3.0 motherboard now available
It just so happens that when you go to the web site he put in his post, the only thing that appears is a 'coming soon' bitmap.
And how come HT 3.0 has been delayed on Server Barcelona SKUs until 2009+?
Here's what's 'coming soon': another round of debt issuance and negative earnings.
How do you spell 'loser'?
S-H-A-R-I-K-O-U
And how come HT 3.0 has been delayed on Server Barcelona SKUs until 2009+?
because AMD's biggest customers (HP and Dell) have told AMD to back off on changing the socket yet again.
so now its 2009 Sharidouche?
When it gets closer to 2009 are you going to push it back again till 2010?
Why don't you just say 2099 just to be safe.
Why 2009? What happened to Q2 2008?
AMD's strategy is simple: market share at all costs.
Including "bankruptcy", apparently.
Have you all noticed how Sharidouche had to expand on his original blog verbiage.
It's becoming increasingly apparent how desperate he finds himself these days, resorting to anti-semitic statements and intimations.
We'll wave a fond adieu to this tool as Intel closes in with Penryn.
I'm teary eyed just thinking about it.
Sniff, sniff
This is why AMD is dead meat. Hector and his retards could never change the fundamental design of the transistor.
http://www.spectrum.ieee.org/oct07/5553/1
Warning to oneretard and onlymoron - this is an article in an engineering publication. It uses words greater than one sylable. You will not understand it.
Anyone with an IQ over room temperature will find it very enjoyable to read. I can't imgae spending two years of my life trying to figure how to deposit a layer of material a single atom in thickness.
so now its 2009 Sharidouche?
When it gets closer to 2009 are you going to push it back again till 2010?
Well of course we all knew it was coming. Remember he already pushed it back once a year ago.
That, or there will be some excuse about Intel is lying, or cheating, or a monopoly or ......
This is why AMD is dead meat. Hector and his retards could never change the fundamental design of the transistor.
If it wasn't for the fact that IBM wants to keep Intel in check somewhat by giving AMD access to most of it's latest silicon technologies then it's clear what would have happened. AMD would be stuck with some old archaic process while Intel kept advancing technology and they would have died off years ago.
But now Intel will have the biggest edge ever: the high-k metal transistor. This is an Intel exclusive. AMD might admit how important it is (See 2006 December analyst day AMD presentations) but only Intel has it.
Now he makes a grudging compliment to the "Israili" engineers in a feeble attempt to make up for the fact he's an anti-semtic fucktard. We see through you sharibitch.
Market share at all costs must include driving your business into the ground.
2.6ghz K10? They don't have a K10 working past 2.0ghz right now, with 30% yeilds...Phenom is pushed back to next year...Nehalam will be out next year and unless AMD can fix thier processes, they're fucked....
BK by 2009 now...you're a cockknocker...
See this contract:
AMD/Intel x86 License Agreement Contract
http://contracts.corporate.findlaw.com/agreements/amd/intel.license.2001.01.01.html
It lists that the license will be terminated if:
(1) the filing by the other party of a petition in
bankruptcy or insolvency;
So when AMD goes BK in 2008 that's it for them. Finished. Even Samsung or IBM buying AMD cannot save them.
Of course, if you believe
Sharikou's ridiculous story that Intel will BK when they did they would take x86 with them, killing AMD in the process!
YES! I have been waiting a year for this day I knew Sharikou would reset Intel's BK once we approached Q2'08. So here is a trip down memory lane. Sharikou, prepare to get OWNED... by your own words.
All quotes below from Sharikou's blog:
Dell was excited, I heard, he is going to switch to near 100% AMD by 1Q08, just one quarter ahead of Intel's projected BK
I re-iterate my projection that Intel BK by 2Q08.
With the K10 onslaught, the overcapacity of the x86 market will lead to the BK of Intel by 2Q08.
The K10 will open a performance gap that will never be filled -- as Intel BKs in 2Q08.
Once K10 is out, all Core 2 will be sold at below $100. Intel's 2Q08 BK is coming
But, Intel will BK by 2Q08.
By 2Q08, IRS won't get a dime for its $2 billion invoice as Intel BKs.
AMD will have 45nm in 2Q08 -- but Intel will BK by then any way, as 65nm K10 will frag Intel's 45nm by at least 20%.
Intel BK by 2Q08, that what I said in 2006, and that is what will happen in 2Q08.
It's a cruel world for Intelers out there. Intel's BK in 2Q08 is unavoidable due to AMD's massive capacity ramp.
Intel BK by 2Q08 is set in the stone. 40% architectural advantage, that's K8 over Netburst all over again. Expect Patty to be begging on the streets soon.
Intel BK by 2Q08.
I have predicted Intel's BK time to be 2Q08, many don't believe me, some say it's crazy. But, folks, most of predictions came true and let's wait and see.
Therefore, Intel will BK by 2Q08
As I predicted long ago, AMD should exit 2006 with 40% market share (run rate). Intel's BK by 2Q08 is pretty much in the bag.
K8L demoed. Intel BK by 2Q08.
Intel BK by 2Q08.
Sharikou said:
"It took the brilliant Israeli engineers three years to figure out Pentium 3. It has take them five years to grasp the essence HyperTransport 1.0 and direct connect architecture 1.0."
"AMD stands for American innovation and American power."
"No wonder the Israeli engineers are so quiet these days, no more loud mouth bragging."
Three quotes in one day? Looks like someone's come out of the closet.
Intel fanboi,
I'm going to take some of the quotes you've posted below and put them on my latest blog entry.
I didn't have the energy to go search for the individual posts to link back to this tool's blog.
If you have some of them, please post them here and I'll snarf them up.
This comment has been removed by the author.
This one is my favorite:
Intel may suffer losses in 2Q06, 3Q06, 4Q06, 1Q07, 2Q07, 3Q07
For Intel, the common scene for the next 12 months will be impariment of goodwill (abandonment of acquired technologies and businesses), writing off inventory, writing off assets (old FABs, etc), one time charge for layoffs, price crash, unit share loss, etc, etc. Intel's stock holder equity will vaporize faster than that of Enron. GAAP loss is a certainty, but I also see operating loss from 3Q06 onward. I previously projected mass layoffs in 2007, but since Intel failed to execute a survival strategy, I projected a faster collapse in Feb 2006.
2Q06: 40% price drop, 20% volume drop, impairment on goodwill, one time charges for massive 16,000 layoff. GAAP loss in the bag. Expect Intel to warn soon. Currently, Wall Street expects Intel to earn $0.15 per share. Expect ($0.05) loss at minimum.
3Q06: Conroe ramps to 10%. 20% additional price drop of P4, mobile CPU price crash, 10% additional volume drop. Operating loss expected. Plus more impairment on goodwill, plus one time charge on inventory write off.
4Q06: Conroe ramps to 20%. Merom ramps to 10%. Further reduction of P4 and Core Duo price. Revenue flat from 3Q06 in a seasonally up quarter. Operating loss. AMD exit 2006 with 40% market share (run rate).
1Q07: Conroe ramps to 35% of desktop units. Revenue down more than seasonal as AMD's 65nm parts flood the market.
2Q07: Intel server market share drops to 40% as AMD ramps Rev H quadcore. Bulldozer hits hard.
3Q07: AMD FAB36 ramps to 20,000wspm at 65nm. FAB38 ramps. Chartered FAB7 ramps. AMD grabs >50% of market share.
4Q07: Go figure.
It's Q4'07 now. Intel will be going BK any day now folks!
Open mouth, insert foot. BK 05, 06, 07, 08... keep throwing out numbers and maybe you just might be right. Or not.
Anytime someone thinks you can't get any dumber, you vastly exceed our expectations. Congratulations. You're a moron.
For one, it's not "QuickPass" it's QuickPath. Secondly, I would really like to know if any other people PHDs actually use the word "frag". Third, Intel created a system like HT years ago, but because back then it was so expensive to implement, the industry didn't want it yet, so they stuck with the inexpensive FSB.
The ultimate irony would be an AMD BK in 2Q08.
So, in a nutshell, Sharidouche is hitting a big fat 0%.
Good going asshat!
No wonder this guy cant land a job.
What happened to Intel BK in 2Q08? I guess Sharikou is a lot like AMD... always postponing things.
Sharifarce will change it again and again, my guess is BK Q209 now, and he'll probabaly say that they'll never bring Nehalam to market...
AMD's problems aren't Intel's trying to drive them out of business, AMD's problem is Hector Ruiz driving them out of business...
Oh - and that's fucking ECS board - only posers like Sharifag would put an ECS board underneath a "killer" proc...given that, my guess is, like ECS, Phenom will be for the low end market only....
Wow! 27 comments and not even one by a Sharikou supporter/AMD fanboi.
Looks like the only ones posting here now are those of us who like to poke fun at Sharikou!
Where'd Shari's fanbois go?
Did they finally wise up?
Too embarressed?
What?
The world is fortunate the scientist developing those super computers, Torrenza co- processors and New Virtualization techniques are not consulting with the pro Intel crowd at this blog.
I’m equally thrilled to see Intel making great strides in the ever shrinking transistor.
Gentlemen, I would like to announce the winner of fragging this blog contest; the Intel group. Can we move on now?
What is even better, is while Shairdouche is frothing at the mouth how Intel won't have this or that technology till 2009, Intel DOESN'T NEED the technology right now, as its current generation of processors are just as good and in MANY instances better than AMD's newest stuff.
But of course, the fanbois fail to understand that simple fact that Intel IS firmly in the lead at this time and AMD's response has been sub-par. Barcelona was supposed to be the great equalizer. And it hasn't been.
But don't let Oneexpert know that. He may go and do something even more stupid than he usually does. But then again, since he can't cut and paste his suicide note, we should be OK.
And how come HT 3.0 has been delayed on Server Barcelona SKUs until 2009+?
because AMD's biggest customers (HP and Dell) have told AMD to back off on changing the socket yet again.
It's actually because HT3 support would need new mobo designs and HP, Dell didn't want that, at least according to the Inq a few days ago.
AMD Announces New Chip-Scale Packaging for Flash Memory
AMD Announces New Chip-Scale Packaging for Flash Memory
SUNNYVALE, CA -- December 10, 1997 --AMD today announced it will offer flash memory devices in chip-scale packages (CSPs). AMD has chosen the Fine-pitch Ball Grid Array (FBGA) because of customers' requirements for a small form-factor package that is independent of die shrinks. Occupying roughly one-third of the board space required by standard thin small outline packages (TSOP), AMD's new packaging provides system designers with an easy-to-use, solution that offers the same reliability as traditional TSOP packages.
"At only 30 percent the size of a standard TSOP flash chip, AMD's FBGA package provides portable system designers with the ability to reduce the size of their product or integrate more features," said Walid Maghribi, group vice president of AMD's Memory Group. "The 0.8mm ball pitch is easy for customers to use with today's personal computer board technology and allows AMD to continue to implement cost reductions without changing the electrical footprint of the package."
"Our new 1.8-volt-only device was developed in concert with our customers' future product plans and represents a milestone in AMD's drive to ever-lower voltages," Maghribi continued. "Eventually, digital cellular phones will be capable of providing one month of standby time as is common with pagers today. In addition, our higher density 1.8-volt-only devices will make possible entirely new types of cellular phone designs, further shrinking size and increasing functionality."
About FBGA
FBGA packages are constructed in a similar manner to conventional TSOP and BGA packages, using existing equipment and proven manufacturing processes. Because the die is attached to a flexible substrate, the package size is determined by the dimensions of the substrate rather than the dimensions of the die. Any die that is smaller than the package cavity can still be assembled in the FBGA package. This allows the package dimensions and electrical footprint to remain fixed as smaller, cost reduced die are put into production or as customers migrate between densities.
FBGA Advantages
The FBGA package has three major advantages over strictly die-sized packages:
* With FBGA packages, system manufacturers will never have the expense of re-laying out a board, retooling a programmer or revising an assembly line because of an AMD die revision. Manufacturing equipment, assembly processes, and documentation remain unchanged, guaranteeing OEMs an uninterrupted manufacturing ramp on existing production lines. Since FBGA products do not become obsolete due to die changes, replacement parts will be available for field service maintenance and repair.
* Die-sized packages lack the necessary die area to place a useable ball grid array for lower density flash memories. Today, FBGA is the only feasible CSP solution for flash memories below 8Mb. As 0.25-micron devices go into production, FBGA packaging is expected to be the only viable CSP alternative for 8Mb devices as well.
* Since die revisions do not require package requalifications or costly retooling of handling and testing equipment by the manufacturer, the FBGA is a more cost-effective CSP solution.
AMD's 8Mb devices are packaged in a 6x9mm FBGA with ball pitch of 0.8mm., while the 16Mb flash is slightly larger at 8x9mm. Since AMD's flash die are considerably smaller than competitive offerings, FBGA packages offer the smallest form factor CSP solution, even smaller than competitive devices in mBGA packaging. In addition, substrate-based packages allow all AMD flash devices, regardless of density, to have a common 0.8mm. pitch electrical footprint and therefore, to be interchangeable.
Availability
AMD will offer a full family of FBGA devices, ranging from 4Mb to 32Mb, in its low-voltage flash families. Production deliveries will begin in Q1 1998.
Package, Price, Availability
The Am29SL800 is available now in a 48-pin TSOP and a 48-ball Fine Pitch BGA (FBGA). FBGA provides the smallest, least expensive package solution for flash memory. The Am29SL800-170EI, industrial temperature range, is $8.70 in 10,000-piece quantities. The Am29SL800-170EC, commercial temperature range, is $8.35 in 10,000-piece quantities. The 1.8-volt family is offered at a 20 percent price premium over the 2.7-volt family.
Safe Harbor Statement
The forward-looking statements in this release with respect to new product offerings are subject to risks and uncertainties, and actual results may differ from those currently expected. Such risks and uncertainties include the impact of competitive products and pricing, the timely development of wafer fabrication process technologies, the effect of changing economic conditions, and such risks and uncertainties detailed from time-to-time in the company's SEC reports.
BUY AMD hi performance, energy saving, leading edge, cpus, platforms, and video solutions.
1997
r said:
"Gentlemen, I would like to announce the winner of fragging this blog contest; the Intel group. Can we move on now?"
Pezal, winner of the DUMBEST post of all time, has been owned. Christian Howell, who has yet to post a photo of his supposed Microsoft Blue Badge, has been owned. Sharikou, has been owned by his own words. I would love to move this gang over to Scientia's and own him (the grand prize shall we say), but he moderates his own blog. Abinstein was kind enough to stop by for a while and demonstrate his apologist attitude and twisted logic, but he also moderates his own blog. So here we stay...
Scientia or Abinstein wouldn't last two minutes with Pointer, Roborat, 13ring, gutterrat, merlin, etc. I wish they would post here or at Roborat's more often. They won't because they will lose.
Holy shit, can you read Howell? Apparently not, cause I said the same fucking thing you said.
Yep being a fanboi seems to damage your intelligence...
And look, more cutting and pasting from oneexpert that fucking moron.
Barondouchebag only reads what he wants to read into any given post. Cut him some slack...anyone who dresses like a silver penis obviously has issues....
Giant said...
This is why AMD is dead meat. Hector and his retards could never change the fundamental design of the transistor.
blah, blah, blah, blah, ...
If it wasn't for the fact that IBM wants to keep Intel in check somewhat by giving AMD access to most of it's latest silicon technologies then it's clear what would have happened.
You know nothing! Obviously AMD has licensed some IP from IBM. But the majority of it is owned by both parties in result to their co-development. Actually some of the IP is owned by Sony, FreeScale and a few more.
Learn the difference!
But now Intel will have the biggest edge ever: the high-k metal transistor. This is an Intel exclusive.
Right! And I guess you just missed the press release where IBM/AMD (and others) announced their own technological advancements. Actually they announced it before Intel did.
Face the facts! Stop lying to yourself! Your bias and you look like a fool!
Giant said...
(1) the filing by the other party of a petition in bankruptcy or insolvency;
So when AMD goes BK in 2008 that's it for them. Finished. Even Samsung or IBM buying AMD cannot save them.
What are you a law expert too now?
1. There are too many '*****' sections in this license to know that it doesn't add exceptions, conditions or worst, nullifies section 6.2.b.1
2. This is a 2001 agreement... how do you know another agreement hasn't voided or amended that clause? You simply don't!
3. Its clearly written that the licensing is terminated 'IF' the other party wishes so. Read it and you'll realize that Intel could have terminated it a while back! They have no reason to do so because:
a) they get royalties from AMD
b) they get to use AMD patents too
Look at 6.2.b.7.ii:
(ii) a party acquires, by merger, acquisition of assets or otherwise, all or any portion of another legal entity such that either the assets or market value of such party after the close of such transaction are greater than one and one third (1 1/3) of the assets or market value of such party prior to such transaction.
Here is a clause to prevent AMD from getting bigger... and you consider this "Free Market"?
Anyway, wouldn't the ATI acquisition nullify this license agreement?
if you believe Sharikou's ridiculous story that Intel will BK when they did they would take x86 with them, killing AMD in the process!
What are you saying? That if Intel was to go bankrupt Intel's IT just disappears and AMD looses too?
I've got news for you... it doesn't work quite that way. From what I've read AMD would only have everything to gain!
Look, its another clueless fanboi coming out of the woodwork to support the sinking ship.
Good job Jeach, you look like a total ass now.
Jeach! said ...
Right! And I guess you just missed the press release where IBM/AMD (and others) announced their own technological advancements. Actually they announced it before Intel did.
Face the facts! Stop lying to yourself! Your bias and you look like a fool!
I'm not sure who is lying here. Are you refering to the IBM-AMD's 'me-too' annoucement made at the same day of Intel announcement, which until now has no single working die shown?
Christian Howell, who has yet to post a photo of his supposed Microsoft Blue Badge, has been owned.
You punk bitch. I don't owe you anything. I told you to search on oms HW forum. I won't be and I won't be scanning or digitizing it again.
You'll just come up with something else just like the sorry verminous assholes over there.
At least I'm man enough to use my real name. COWARD!
Holy shit, can you read Howell? Apparently not, cause I said the same fucking thing you said.
Yep being a fanboi seems to damage your intelligence...
The socket wouldn't have to change stupid. Only the circuitry n the mobo to recognize the split power planes. 1207+ just means made for HT3, not won't fit into 1207.
So I guess you're the idiot little bitch fanboi.
That's funny Chris "BaronFudrix" Howell calling somebody a fanboi.
ROFL
So Baron, when is Intel going to BK? Are you going to push back your estimate since Intel is making money hand over fist?
Oh, I forgot, AMD is doing fine in your mind because now that they are slower you merely ask if the computer will POST. If so, it's fine.
Nice try Fudrix, you lose fanboy.
Yeah you use your real name and if you were one of my employees, you'd be out on the street...not for posting on a blog during working hours tho, more for being a numb-shit who thinks he knows more than he does and when he's called on it and proven wrong resorts to childish name calling and...you really should just go crawl into a corner with Sharidouche and his circle jerk buddies and leave the internet tech boards to real techs and engineers, you wanna be...
Not posting a real name has nothing to do with fear of saying who we are, it has everything to do with remaining safely anonymous when there are fucktards like you roaming the internet...
And I'll be timing how long it takes you to whine to sharibitch to get this post removed....
lolz...Chris Chris Chris..
Because you got banned for flamming others in the other forum, so you now come to this playground?
Pathetic Baron...Pathetic.
Baron, so when are you going to show us the picture of a sober woman that you nearly had *** with on the dance floor, for four times?
The HT3 CPU is not pin compatible to the HT2 CPU's, thus it needs a new socket. I don't care if its still 1207 pin, you can't just take one out of the existing sockets and put them in another.
thus a new socket. Play the game all you want, the two CPU's are not pin compatible, and the big guys told AMD to go pound sand until they made them all work on the same socket.
Whats next? Claiming you work for Microsoft again?
At least I'm man enough to use my real name. COWARD!
Just because you are unemployed doesn't mean the rest of us don't have jobs. While I do have a rather lax internet usage clause at my current office it still doesn't mean Hector would like all the stuff I say. Especially because its all facts.
evil_moron
But of course, the fanbois fail to understand that simple fact that Intel IS firmly in the lead at this time and AMD's response has been sub-par. Barcelona was supposed to be the great equalizer. And it hasn't been.
According to this chart from Anandtech's review, Barcelona beats Clovertown clock for clock in 5 out of 8 benchmarks with Intel winning only 2 (1 of which used Intel optimized binaries). As Barcelona scales to 2.6 Ghz in November and 3.0 Ghz by Q1 08 the performance per watt will easily swing back to AMD. Also if you look at Anand's performance per watt measurement you'll see it only uses 1 benchmark (Cinebench) so it is hardly all encompassing or conclusive. Pick another random benchmark and the results could be radically different. Try not to cry Intel Fannies... LOL.
This comment has been removed by the author.
So what barcey wins clock for clock. If I invent a cpu company and sell products that operate at 300mhz and are faster than both AMD and Intel clock for clock by 50% does that make my cpu better?
According to you it does...
And yes we're crying. Crying with laughter. Q3 earnings release soon. I will probably die of laughter when AMD release their "Profits"
TechWithNOLife
Clock for clock?
Holy shit boy your reality is FUCKED.
Who the FUCK cares about clock for clock. Intel's running a wee bit faster than AMD's fastest Barcelona now, and its going to get MUCH MUCH worse for AMD in November.
By the way fuckwit, like Sharidouche and the other fanbois, you leave out the rest of the story. From your link [b]When you are looking for the highest performance however, Intel has still a solid advantage with it's 3 GHz Xeon x5365[/b]
Whoops! SOLID AVANTAGE FOR INTEL.
Point Intel.
Game & match to Intel.
yeah lets all go out and buy Intel's 3.0 GHz chip, slow it way down, just so AMD can compete.
Get a fucking clue. Or is that part of the brain that was damaged by the fact you are a product of your mothers brother sleeping with her. I know its mean of me to make fun of your dad like that, but sleeping with his sister is bad m'kay? The results are persons like you, Sharidouche and oneexpert.
Incest is BAD. It leads to birth defects like incredible stupidity.
At least I'm man enough to use my real name. COWARD!
How are those emplyment offers going? Not too many since the recruiters are googling you the minute they get your resume.
That's also why you are never getting laid by a woman either.
AMD is superior because if you underclock Intel processors AMD beats them in SOME (but not all) benchmarks.
Ooookkk!
2.6ghz in November? LAWL! Don't expect anything new until December\January, and it won't be 2.6ghz.
The HT3 CPU is not pin compatible to the HT2 CPU's, thus it needs a new socket. I don't care if its still 1207 pin, you can't just take one out of the existing sockets and put them in another.
Do you mean the same compatibility problem that all Intel sockets 775 suffer? I don’t think so.
Scientia or Abinstein wouldn't last two minutes with Pointer, Roborat, 13ring, gutterrat, merlin, etc. I wish they would post here or at Roborat's more often. They won't because they will lose.
Your right, it’s impossible for them to win. They are no experts in insults/ abuse/ swearing/ name-calling/ foul language /invective.
Scientia or Abinstein wouldn't last two minutes with Pointer, Roborat, 13ring, gutterrat, merlin, etc. I wish they would post here or at Roborat's more often. They won't because they will lose.
Sci is too much of a pussy to leave his safe-heaven.
He hides behide his censorship and acts like a machoman... but if he ever leaves his territory, he will very much so get ripped a new A-hole. Remember the dedicated thread at Toms?
See how he refrains from posting on Greorge Ou's blog but attacks george every chance he gets on AMDZone....
If he is so damn smart like he claims, why doesn't he go spread his knowledge.
He thinks G3MX and SSE5 are top "developments" of 2007 when they are simply just ideas on paper LOL.
what a joke.
Aguia, are we talking about Intel here?
NO.
Stop moving the goal post around.
The reason HT3 isn't available is due to pin changes between HT3 and HT2 and Dell and HP don't want to deal with it.
But keep supporting the fuckwits here. It just makes you look like an ass.
Aguia, are we talking about Intel here?
NO.
Why not is this blog exclusive about AMD?
Stop moving the goal post around.
The reason HT3 isn't available is due to pin changes between HT3 and HT2 and Dell and HP don't want to deal with it.
So, talking about the pin changes Intel does every year in Intel socket 775 is not so interesting?
Do you know if HP and Dell like to deal with those?
But keep supporting the fuckwits here. It just makes you look like an ass.
But I was talking to you.
So Baron, when is Intel going to BK? Are you going to push back your estimate since Intel is making money hand over fist?
I have NEVER said that. I actually asked Sharikou to stop saying it. It just goes to show you're a Klingon, clinging to the words of those above you.
For shame.
lolz...Chris Chris Chris..
Because you got banned for flamming others in the other forum, so you now come to this playground?
Pathetic Baron...Pathetic.
Baron, so when are you going to show us the picture of a sober woman that you nearly had *** with on the dance floor, for four times?
Better than being a half gay little bitch. I'm not banned from anywhere except the Underachiever's Forum.
As far as what I do on the dance floor, I can only tell you what club I go to. Only little punks kiss and tell.
Just because you are unemployed doesn't mean the rest of us don't have jobs. While I do have a rather lax internet usage clause at my current office it still doesn't mean Hector would like all the stuff I say. Especially because its all facts.
Well, since I have three PCs at home at two at work I guess I'd take my status over yours. If you knew anything you'd know that as long as my work gets done no one can say anything.
A guy sued and won causing what's known as a PRECEDENT. That means if someone says I'm on the Internet too much they have to show that my work doesn't get done.
Well considering I have 4 PC's (two of which are servers), two Mac's, one Sun box and two laptops at home and well over 1000 computers at the office, I guess I win.
Do you enjoy wipping out your e-penis? Does having more computers than someone make you feel like a real man?
Chris "BaronFudrix" Howell doesn't have to worry about getting fired. He works for the government and they tolerate a large amount a laziness.
The people of New York City should be pissed at him for stealing their tax dollars.
"They are no experts in insults/ abuse/ swearing/ name-calling/ foul language /invective."
Aguia, you're absolutely correct. And besides, what's the point of winning when the other guys only spit out their opinions against your facts?
Fudrix,
Let me know when Cosgrave gets my email with subject "DOITT not DOING their job?". I can't wait for the results of the FOIA of your websurfing habits.
bubba -
"This is why AMD is dead meat. Hector and his retards could never change the fundamental design of the transistor."
You apparently do not know the high-K matter. If you read the article slightly more carefully, then you'd notice that it always write "we and others". In other words, the same research is done at IBM/AMD labs (and some others) as well.
What really makes Intel the first high-volume high-K maker is its finding the right metal gate composite to work with the gate-last approach that combined can go into volume production.
It's no small feat but for sure they're not the only ones like you suggested.
Aguia, you're absolutely correct. And besides, what's the point of winning when the other guys only spit out their opinions against your facts?
You seem to be confusing yourself with other people. You're just another fake. Their technical knowledge of process technologies far exceeds yours.
I have decided to create a Christian Howell OWNED Counter. Every time CH gets owned I will increment the counter. So what do we have today?
Christian Howell said today in regards to his Microsoft Blue Badge:
"I told you to search on Toms HW forum. I won't be and I won't be scanning or digitizing it again."
Ha ha ha. I went ahead and did that. Guess what I found! A post by Christian saying he worked for Microsoft, then a challenge by the members to product a Blue Badge photo, then a claim by Chris that the picture was posted at his own blog, then multiple posts by members saying that no such photo was ever posted! Great going Chris, telling us to find a link on another forum that would provide a link back to your blog that NEVER had the photo. You never worked for Microsoft. You just got OWNED.
Christian Howell OWNED Counter = 1
Intel fanboi -
"Christian Howell, who has yet to post a photo of his supposed Microsoft Blue Badge, has been owned."
Why would anyone post his badge for you to see? Oh you really want it but you can't get it, don't you? I think it's you who's owned by your unsatisfied desire, no?
No to mention your wrong logic has been owned by me soundly and squarely - we know you won't admit it, and I find not much honor in stating it, but before you call others owned, please look at yourself first. I know I will. :)
Thanks.
Christian Howell said:
"The socket wouldn't have to change stupid. Only the circuitry n the mobo to recognize the split power planes. 1207+ just means made for HT3, not won't fit into 1207."
To which Evil Merlin responded:
"The HT3 CPU is not pin compatible to the HT2 CPU's, thus it needs a new socket. I don't care if its still 1207 pin, you can't just take one out of the existing sockets and put them in another."
Christian Howell OWNED Counter = 2
This comment has been removed by the author.
"The HT3 CPU is not pin compatible to the HT2 CPU's, thus it needs a new socket."
I don't know what else to say but this is completely wrong. HT3 is 100% compatible with HT2. You can put a HT3-capable processor into HT2-enabled motherboard with no problem. However, unless the motherboard is redesigned, the system is not going to take any advantage of HT3.
Barcelona has HT3 built-in. As soon as any server board with HT3 capability exists, AMD can enable HT3 on the chips that it produces.
Again, Intel fanboi, in an attempt to FUD others as "owned", you are apparently "owned" by your own ignorance.
Abinstein the Apologist said:
"Why would anyone post his badge for you to see?"
Fantastic! Abinstein may be here to stay. I may have to start a Stupid Counter for him to go along with Baron's. Might be too much work though. So, why would anyone post his badge for us to see? Great question! You should ask Christian, not me, since HE was the one who said he posted it.
Abinstein Stupid Counter = 1
No to mention your wrong logic has been owned by me soundly and squarely.
Ha ha ha. Where did that happen? In Abinsteinland? Are you referring to the argument where you called me a racist for pointing out Sharikou was a racist? Let me quote you: "You claim that racism offends you, but when you single out Jews from an attack on Intel Israel your self is acting more like a racist." You are like a third rate politico on cable television.
Abinstein Stupid Counter = 2
Get with the picture asshole.
http://www.theinquirer.net/gb/inquirer/news/2007/09/26/amd-barcelona-ht3-turn
Looks like even Shanghai will have HT3 disabled.
Oh SNAP another AMD fanboi owned.
You wanna go another round?
Or are you going to try to move the goal posts again fanboi.
Do us all a favour and just STFU. You are useless, and ill informed at best. I think I have it figured out, Sharikou, onexpert, abinstein and the rest of the band of usual morons are all related.
Remember what I said boys, sex with mom and/or your sisters is not only illegal, it causes things like oneexpert.
Abinstein the Twister of Facts said:
"I don't know what else to say but this is completely wrong. HT3 is 100% compatible with HT2."
which was a response to what Evil Merlin said:
"The HT3 CPU is not pin compatible to the HT2 CPU's, thus it needs a new socket."
Ha ha ha. Third rate politico is an apt description for Abinstein! Hey Abinstein, stop your lies. You know exactly what EM meant. You left out the word "CPU". EM, was refering to the Inq article from last week, the same article CH refered to, concerning HT3 "enabled" CPUs which will be on a new socket. Where is that fourth HT3 channel on Barcelona going to go to without a new socket? Answer THAT!
Abinstein Stupid Counter = 3
You know nothing! Obviously AMD has licensed some IP from IBM. But the majority of it is owned by both parties in result to their co-development. Actually some of the IP is owned by Sony, FreeScale and a few more.
The R&D for new processes (45nm etc) is done at IBM's facilities in Fishkill, IBM contributes by far the most $$$ to R&D for the silicon technology. Conclusion? AMD needs IBM. Not the other way around.
Right! And I guess you just missed the press release where IBM/AMD (and others) announced their own technological advancements. Actually they announced it before Intel did.
I do remember exactly what happened. It was back in January. Intel organized an event with the press, they invited people from sites like Anandtech as well. They demonstrated working 45nm CPUs based on this revolutionary new high-k metal gate transistor. They showed games running, media encoding etc. across servers, desktops and notebooks. They also talked about the technology, the 45nm fabs etc. that they will have operational. IBM somehow got wind of this information, so they released a press release about their own high-k metal gate technology. But nothing was shown off, nothing whatsoever.
AMD has stated their 45nm process will just be the same old SOI stuff. No metal gate transistor until 32nm or late into the 45nm lifecycle.
Face the facts! Stop lying to yourself! Your bias and you look like a fool!
Why don't you face the facts? You are the one that claimed AMD would have a 4.7GHZ Barcelona CPU just because they use IBM's process technology and IBM has a 4.7GHZ POWER6 CPU. ;-0
Then you compare a press release to a live demo. That's laughable. Intel's 45nm CPUs based on this technology will launch in 38 days. When will your beloved AMD have this technology?
According to this chart from Anandtech's review, Barcelona beats Clovertown clock for clock in 5 out of 8 benchmarks with Intel winning only 2 (1 of which used Intel optimized binaries). As Barcelona scales to 2.6 Ghz in November and 3.0 Ghz by Q1 08 the performance per watt will easily swing back to AMD. Also if you look at Anand's performance per watt measurement you'll see it only uses 1 benchmark (Cinebench) so it is hardly all encompassing or conclusive. Pick another random benchmark and the results could be radically different. Try not to cry Intel Fannies... LOL.
TechReport conducted far more tests:
http://www.techreport.com/articles.x/13176
AMD gets fragged in most of them. Have you any proof that AMD will have a 3Ghz Barcelona by Q1'08? AMD's own Randy Allen said 2.5Ghz by December. While this is happening, Intel will launch Penryn . Are you so naive that you think AMD will increase speeds from 2Ghz to 3Ghz while Intel sits at 3Ghz? Ridiculous.
Try not to cry AMD Fannies... LOL.
Abinstein: Another opinion against my facts.
You Intel fanbois really have a great habit of proving me correct. :)
The dude has the balls to point to his biased blog full of misinformation as fact.
You are a bigger joke than sharikou if you think that BS you spew in your blog is anywhere close to being fact.
Great job intel fanboi. Your counter really helps put in perspective how retarded and ignorant these AMD fanbois really are.
Great job intel fanboi. Your counter really helps put in perspective how retarded and ignorant these AMD fanbois really are.
Indeed. I offer my congratulations to Intel Fanboi on his excellent counter. I only fear that if he is to track both Abinstein and Christian Howell it'll end up being a full time job for him!
Another opinion against my facts.
You Intel fanbois really have a great habit of proving me correct. :)
Your "fact" link takes you to
Not Found
Error 404
Just like your facts, nothing exists.
You read articles/documents and then paraphrase it in your blog and all of sudden you think you know the material? lol
You, Sci, and Sharikou should join the circus, it'll suit you better.
abinstein the tool, wrote
Barcelona has HT3 built-in. As soon as any server board with HT3 capability exists, AMD can enable HT3 on the chips that it produces.
Abinstein,
When will server boards with HT3 capability materialize for real?
Don't bother answering. The answer is on slide 32 in this presentation.
Don't try to talk about HT3 being enabled in Budapest either because it does not qualify as "Barcelona" which is a server CPU.
I think the Inquirer had it right in this article. Here's the quote:
Sun was the only one who wanted the new four-channel pinout and HT3 support - no wonder, as they were the only ones having guts to bring out 8S Opteron in the first place.HP, IBM and Dell might have been not too interested to bother about that - after all, they could be busy designing the new 2S, 4S and 8S Nehalems...
It must be nice having a guardian "Angel" like scientia to moderate, umm, shield you from all the hurtful feedback being delivered onto you in scientia's blog.
Abistein, if you would just admit to being an AMD fanboi outright then we might be able to engage in real conversation.
Did someone in Hollywood teach you how to deny? You seem to be pretty good at it.
And please, stop with the pathetic self-promotion of your blog. You'd be better off promoting AMDZone instead.
Thanks. I learn from the best. 13heat is the master of mixing logic with mockery. Pure entertainment. I wish I could be as funny as him. Giant is like a sledgehammer of facts. I'm way to lazy and uninformed to post like the Giant.
Still, it doesn't take much skill to counter an apologist like Abinstein, or the recently revealed to be a homophobe BaronMatrix.
A lot of the personal attacks are now starting to take away from discussing the actual blog or the facts. This has been getting worse since new people discovered this blog unfortunately and bring with them an aggressive attitude which is not called for. A lot of people here sound like a group of 7 year olds in the school playground and this is even rubbing off on some of the other posters who are now also being more aggressive.
A lot of the posts above are not only overly aggressive and rude but worse, they have very little substance also. So my respect for the posters is very little indeed. It's definitely a case of empty vessels make most noise.
Getting back to the latest blog from Sharikou, as well as the stretching of Intels demise into 2009 I was also interested to see the claim that a 2.6GHz K10 will be faster than a 4GHz Intel quad core. This claim means it is at least 54% faster. Even AMD only claimed 40% faster than Clovertown in some benchmarks and then subsequently a 2GHz K10 was faster than a 2.33GHz Clovertown in general (say 20% faster). So where does this 50% + increase come from?
As we know, 45nm Penryns are roughly 10% or so faster than Clovertown so this pushes the speed needed for an equivalent K10 even higher.
Of course this is all wishful thinking, Intel will not be releasing 4Ghz processors, much more liekly a max of 3.4-3.6GHz next year, that will still take a 3GHz or so K10 to match. It will be interesting to see who can win this new speed race. Currntly though Intel seems to have the edge.
Sorry Andy, you are 100% right, and a better man than me. But it just ain't gonna happen. Not here, not on this blog.
Mo said in regards to Abinstein:
Your "fact" link takes you to
Not Found
Error 404
Abinstein Stupid Counter = 4
Of course this is all wishful thinking, Intel will not be releasing 4Ghz processors, much more liekly a max of 3.4-3.6GHz next year
You have to consider the facts carefully in this matter. Intel has no trouble producing a 3Ghz quad core CPU. You can buy one for either desktops or workstations/servers now at Newegg.
That is Xeon X5365 with 120W TDP, 3GHz clockspeed produced on current 65nm process.
The 45nm process is not just another simple die shrink like 90nm -> 65nm. They have completely changed the transistor. This is the first major change to the transistor in over forty years. The high-k metal hafnium gate has allowed for a ten-fold reduction in leakage when compared with the 65nm process. Through billions of dollars in R&D Intel has finally solved the leakage issue that has plagued the whole industry since the 90nm process.
Consider the kind of increase in clockspeed that was obtained going from the 180nm process to the 130nm process for the Pentium 4 when leakage was not an issue. At 180nm Intel maxed out at 2Ghz. At 130nm they reached 3.46Ghz. But then at 90nm clockspeeds only reached 3.8Ghz for a Pentium 4. That's a rather pitiful 360mhz increase due to the extreme leakage that was present when when the CPUs were running at clockspeeds in excess of 3Ghz. (Obviously this wasn't nearly as much of a problem at lower clockspeeds, as with the Pentium M.)
Couple that data with the demonstrations from IDF. Intel has shown 3.4Ghz CPUs running on air cooling. On one of the enthusiast websites (I think it was XtremeSystems) a Yorkfield CPU was shown overlcocked to 4.1Ghz 100% stable with air cooling.
By the time Nehalem is launched in the second half of 2007 Intel will be at or be near clockspeeds of 4Ghz, mark my words.
Intel is on a roll. They have a clear lead in all three major CPU markets today and will be expanding upon that lead with Penryn in just 38 days. Meanwhile, AMD's new architecture is dead in the water. It's not as fast as Intel's existing Cloverotwn product. There's been no word out of AMD regarding their 45nm process, no announcement that their Shanghai CPU has taped out. Meanwhile Intel has already shown the next generation Nehalem CPU up and running. They're clearly on track for a release in H2'08.
@stupid Intel fanboys..
I dont know whether I should call you monkey or donkey?
Current AMD AM2 CPU has the same numbers of pins as the AMD AM2+ Barcelona or Phenom. Meaning to say, if you have no intention to upgrade your AM2 board to AM2+ such as the certain OEMS did except SUN, never mind, because you are still able to upgrade your CPU to phenom/Barcelona. The different between AM2 MB and AM2+ MB is the only AM2+ support DDR3 while the existing AM2 dont.. And of course you’ll have no problem as well upgrading your AM2 MB to AM2+ while maintaining the same AM2 CPU..
Intel Fanboi Stupid Counter = 1.5
Sorry Andy, you are 100% right, and a better man than me.
So you meant you are brainless than Andy? Ekeke..
Intel Fanboi stupidity counter=1.5+1.5 =3.0
I wish I could be as funny as him.
Funny as him? why not as funny as Mr bean? ain't mr bean is your grandmaster?
Intel fanboi stupidity counter = 1.5+1.5+1.5 =4.5
"You're just another fake. Their technical knowledge of process technologies far exceeds yours."
Another opinion against my facts. (Link updated)
You Intel fanbois really have a great habit of proving me correct. :)
This comment has been removed by the author.
Meanwhile Intel maintains the same socket with both support for DDR2 and DDR3 memory. I bought this motherboard for a dual core, now I have a quad core in it. I will be able to put a 45nm CPU in this board in 38 days. All I needed for quad core and 45nm CPU support was a simple BIOS update. If I want DDR3 memory later on I can go and put my CPU in a board with DDR3 memory.
This has left me with a lot of upgrade options, but I'm doubting now that I'll get a Yorkfield CPU. This Q6600 can last me until I build a Nehalem based computer next year. That will need a new motherboard and likely DDR3 memory anyway.
Oh, and before the AMD fanbois get started on 'You'll need a new motherboard for Nehalem!' AMD's next new architecture will need a new socket as well.
I bought a motherboard for dual core CPUs. It was compatible with quad core CPUs and the 45nm refresh of the quad core and dual core CPUs. I'll need a new motherboard for Nehalem.
AMD user buys AM2 motherboard. It's compatible with dual core, then later quad core (this is still to come for the AMD users though). Then next year they can upgrade the BIOS again and they should have 45nm Shanghai compatibility. AMD user will need a new motherboard for the bulldozer architecture.
Pezal said 'Ekeke'. For that he gets a counter and the count starts at 1000.
gutterrat the source of "tool" wrote:
"When will server boards with HT3 capability materialize for real?"
According to AMD's 2007 analyst day presentation, HT3 is planned for 2H07 Budapest. HT3 will also be available for desktops.
"Don't try to talk about HT3 being enabled in Budapest either because it does not qualify as "Barcelona" which is a server CPU."
Budapest is just the single-socket (AM2) version of Barcelona.
"HP, IBM and Dell might have been not too interested to bother about that - after all, they could be busy designing the new 2S, 4S and 8S Nehalems..."
This doesn't conflict with what I said at all - HT3 is 100% backward compatible. There is no need to design new motherboards for the chips.
"It must be nice having a guardian "Angel" like scientia to moderate, umm, shield you from all the hurtful feedback being delivered onto you in scientia's blog."
I don't know. Some people are probably too stupid to deliver personal attacks effectively.
"Abistein, if you would just admit to being an AMD fanboi outright then we might be able to engage in real conversation."
Gutterrat, if you would just admit being a FUDer outright then you might become a bit less despicable.
"Did someone in Hollywood teach you how to deny? You seem to be pretty good at it."
Maybe it's someone from Intel taught you how to FUD?
"And please, stop with the pathetic self-promotion of your blog. You'd be better off promoting AMDZone instead."
Unlike you who can do nothing but FUD, I was promoting truths and facts. Well, maybe in your FUDers mind making FUDs is not pathetic at all...
Mo -
"You read articles/documents and then paraphrase it in your blog and all of sudden you think you know the material?"
At least I read & understand the documents, and state them in factuality. Some people just push their heads into the sands and concurrently making FUDs out of their a**es - it's certainly no small feat.
Intel fanboi -
"Where is that fourth HT3 channel on Barcelona going to go to without a new socket? Answer THAT!"
I think you're being very ignorant here. Opteron with 3 HT links use far less pins of socket F; there is no need of a new socket, but since the pins are grounded/open anyway there's no point for AMD to enable the 4th link.
BTW, Intel fanboi, it appears you are not only fond of personal attacks but also low-class deception.
"Hey Abinstein, stop your lies. You know exactly what EM meant. You left out the word "CPU"."
Either you can't read or you purposely want to be a despicable lier like gutterrat, you somehow missed the very next sentence of mine right after your quote. I said: "You can put a HT3-capable processor into HT2-enabled motherboard with no problem."
Maybe you didn't learn this well from Intel, but let me teach you here: the processor in this context is the same as "CPU".
abinstein,
Every post you make further proves my point.
Budapest does not count. You think HT3 is needed on the desktop?
LOL
despicable lier you call me?
In addition to being a tool you now confirm that you are a Duck also.
"lier"
LOL
Maybe you didn't learn this well from Intel, but let me teach you here: the processor in this context is the same as "CPU".
Ekeke.. In long terms, CPU is stand for "Central Processing Unit".. Abinstein, I really respect your blog man.. keep it up.. ;-)
"Every post you make further proves my point."
That's precisely because the only point you've made is that you lie & FUD.
"Budapest does not count. You think HT3 is needed on the desktop?"
I know you're very ignorant of the matter but yes, HT3 is very needed on the desktop, for example, for IGP memory access.
BTW, although "lier" is a typo, but it would fit you as well especially with the "despicable" adjective in front of it.
HT3 is needed on the desktop? What a bunch of FUDs...
Abinstein, since you're so sure that bandwidth is dreadfully needed for desktop applications, would you mind showing me a benchmarks that sees tremendous performance boost due to increased bandwidth?
For server, HT3 is needed. For desktop and mobile? not really.
Actually, let's put it this way.
Can you show me a desktop benchmarks, or games, that see tremendous performance boost due to increased bandwidth?
IGP is limited by their architecture, not bandwidth. In fact, for desktop applications, you don't see a lot of programs that need that much bandwidth.
abistein,
I know you're very ignorant of the matter but...
Is that an 'opinion' or a conclusion based on 'facts' taht you fabricated?
You are telling me that you AMD fanbois need HT 3.0 on the desktop in 2007?
That's a real stretch.
Please tell me that you need HT 3.0 to play HD content on your PC.
Do you AMD fanbois really run your graphic apps using Integrated Graphics?
LOL
"despicable lier" who?
ROFLMAO
You must try a bit harder, I'm afraid you are just making me laugh.
yomamafor2,
abinstein needs to be careful for I see him walking right into a rathole from which he may not be able to climb out.
:)
Meanwhile Intel maintains the same socket with both support for DDR2 and DDR3 memory.
Same socket that isn’t pin compatible. Who cares that Intel socket 775 exists since 2004 if you can’t plug a new processor on it?
All I needed for quad core and 45nm CPU support was a simple BIOS update.
False. Go to any web site Asus, ASrock, … and see the CPU compatible table, even today selling motherboards most of them cant run quad core CPUs, and guess what? They are all socket 775!
Try to do that with one 2005 motherboard.
Just go back and list from 2004 which socket 775 can run the 2005 PentiumD.
Which Intel socket 775 motherboard from 2005 can run a 2006 Core 2 Duo.
…
Do that simple exercise and I will agree with you.
Intel is already in socket 775 5th (incompatible) revision. I will not even mention the four bus speed updates, wait I just did ;)
HT3 is needed on the desktop? What a bunch of FUDs...
Not really. It’s VERY needed because of PCIe 2.0.
Where server could need it for more RAID or Ethernet cards? I think the 6.4GB/s are enough for that.
Why do you think Intel increases their FSB speed?
Intel increased its bus speed every time there was a new AGP interface.
Now you are going to see the 1600Mhz FSB because of PCIe 2.0.
acquia wrote,
HT3 is needed on the desktop? What a bunch of FUDs...
Not really. It’s VERY needed because of PCIe 2.0.
Where server could need it for more RAID or Ethernet cards? I think the 6.4GB/s are enough for that.
You are not making sense.
Abinstein claimed that HT 3.0 was was needed for IGP.
Huh?
Abinstein claimed that HT 3.0 was was needed for IGP.
Huh?
Perfectly correct!
While Intel IGP traffic goes with in the chipset.
AMD IGP traffic goes through the CPU NB in order to use the built in IMC.
So with DDR2 memory speed at 1066MHz the HT bus would become the bottle neck in the communication.
HT 1000Mhz = 8GB/s
PC2-6400 DDR2-SDRAM (dual channel) = 12.8 GB/s
Intel doesn’t have FSB problems when using IGP because the IMC is in the chipset. So there is no traffic stressing the FSB.
AMD needs a 1600Mhz HT in order to achieve 12.8Gb/s.
Gutrat if you don’t understand this then you aren’t one expert. ;)
evil_moron
Who the FUCK cares about clock for clock.
That says it all right there. Thanks for reaffirming your lack of intelligence. Now grow some manners and see a shrink to deal with that anger built up inside. Might do you good.
pezal said...
@stupid Intel fanboys..
I dont know whether I should call you monkey or donkey?
One thing is for sure we all know what to call you Mr. Dumbest Post of all time award winner. I am glad to see you are back this board was missing some of your "mathematical" skills. :)
Intel Fanboi stupidity counter=1.5+1.5 =3.0
3 as in the number of cores you get when you can not yield quad cores....then again AMD knows there are plenty of gullible fanbois who will lap this up....
Abinstein says: I was promoting truths and facts. Well, maybe in your FUDers mind making FUDs is not pathetic at all...
Truth and fact coming from a dude who had to try three times before he could post his link correctly......i am better of getting "truth and fact" from a 5 year old...oh wait thats what your blog does. My apologies.
idiot4life says: That says it all right there. Thanks for reaffirming your lack of intelligence. Now grow some manners and see a shrink to deal with that anger built up inside. Might do you good.
Great point douche....Intel should just shut down for a year or two and wait till AMD reaches 3ghz so that we can have a true clock for clock match instead of underclocking intel to compete with AMD.
Glad you arent the decision maker of a company but there is always Hector who has the same mind set as you and we see how high AMD is soaring...
AMD's strategy of competing with the competition's year old products and underperforming is going as planned.
Barcelona has HT3 built-in. As soon as any server board with HT3 capability exists, AMD can enable HT3 on the chips that it produces.
Abinstein,
When will server boards with HT3 capability materialize for real?
Don't bother answering. The answer is on slide 32 in this presentation.
Don't try to talk about HT3 being enabled in Budapest either because it does not qualify as "Barcelona" which is a server CPU.
Wow, you have some real issues. You guys must all be like short bus graduates. Only the power planes need to be split, not the socket.
The boards with split power have been released. Tyan has one as does SuperMicro (Anand had one). Companies wanted the seamless plug-in I guess for the MP servers.
As far as Budapest, IT WILL HAVE HT3 as 1P is AM2+.
I resisted the temptation to say dumb ass. Well, maybe I didn't.
Anyway, Digitimes already reported on MSI, ASUS, and DFI with HT3 mobos.
There'll be more when nVidia drops their chipset next month. I would say that AMD will catch up in perf in a big way, even with Kuma which should sit halfway between K8 and Phenom, perf-wise.
And remember dual core is just becoming the mainstream and will be for a few years.
Meanwhile Intel maintains the same socket with both support for DDR2 and DDR3 memory. I bought this motherboard for a dual core, now I have a quad core in it. I will be able to put a 45nm CPU in this board in 38 days. All I needed for quad core and 45nm CPU support was a simple BIOS update. If I want DDR3 memory later on I can go and put my CPU in a board with DDR3 memory.
You mean like how XBit reported that K10 supports both DDR2 and DDR3?
All that's needed is the mobo and RAM slots. AMD is sticking with DDR2 for now because they can use 1066+ at better CAS ratings. And DDR3 is EXPENSIVE AS HELL right now.
Damn, so no more Intel BK 2Q08? :(
2009 now? Boo!
This comment has been removed by the author.
Andyw35 said..."A lot of the personal attacks are now starting to take away from discussing the actual blog or the facts. This has been getting worse since new people discovered this blog unfortunately and bring with them an aggressive attitude which is not called for. A lot of people here sound like a group of 7 year olds in the school playground and this is even rubbing off on some of the other posters who are now also being more aggressive."
You're right Andy, unfortunately - it's no more mature to claim you have a ph.d when you do not, claim you're an unbiased "journalist" when he's not. I have no problems being mature when the site/blog warrants it...this one does not..
Shardifreak is a liar, an maladjusted ex-intel employee who thinks he knows it all when in fact he probably can't wipe his own ass without help from his douchebag followers. Half the people who post in his defense are no better, the ones that used to post in his defense who at least had some intelligence left this place long ago. He's worth nothing more than ridicule and heckling...
This comment has been removed by the author.
abinstein
Gather around everybody It's fact time.
I will take The factmaker's latest entry on the die sizes of quad core cpus.
Fact: AMD has the largest monilithic die 294mm^2.
Factmakes assumption: Nehlem is roughly 270mm^2.... he got this from looking at a 300mm waffer. Assumption, not fact.
Intel can cram in a LOT MORE cache and still be below AMD's Barcelona, be more cost productive and Intel will probably get better yields than the massive barcy.
Sci and Abi continue to claim that barcy yields are "excellent" yet this is private data, thus it's not a fact because you dont know.
though you do acknowledge that nehlems got 2-way SMTs and TWICE the cache, you fail to fully understand Intels statement. such a chip is too large to manufacture on a 65nm process, Such a chip, that INtel has proposed or developed, is too large, what part dont you get? Not even AMD could do it, it's already reaching 300mm^2 with HALF the cache. Intel was able to implement 2-way SMTs, double the cache and STILL be below AMD's quadcore. Thisis on an advanced 45nm process that AMD is still "dreaming" about. AMD lacks Intel full 1-2 years on the manufacturing process.
What if AMD matched the cache of Intel.... can you imagine the die size? Would it be economical for AMD?
In terms of Arch. design, Intel lacks AMD by a year? Are you kidding me?
Since last year, Intel has shipped out over 1 million quad cores, how many has AMD shipped out?
Since last year, Intel has kept it's lead in the arch performance over AMD on outdated bottleneck FSB.
Since last year, INtel has gained market while amd has lost it.
Since last year, Intel has made B/Millions while AMD has been loosing MILLIONS.
All above facts.
All you posted was FUD. Now your fud is debunked.
If Intel can keep up with FSB, AMD can only begin to fathom what it will do with IMC, HT links, 2-way SMT and massive cache....
Dark days lay ahead of AMD.
Abinstein the Master of Deception said:
"Either you can't read or you purposely want to be a despicable lier like gutterrat, you somehow missed the very next sentence of mine right after your quote. I said: "You can put a HT3-capable processor into HT2-enabled motherboard with no problem.""
Ha ha ha. You really belong on Hannity and Colms. Your third rate tactics and ability to lie will make you a favorite on cable news. Just because you state a FACT after a lie in the same paragraph does not mean that your lie becomes true. Try again Abinstein.
Abinstein Stupid Counter = 5
AMD Opteron lands another supercomputing coup
Appro joins Tri-whizz tournament
By Theo Valich: Friday, 05 October 2007, 11:02 AM
SUPERCOMPUTER BUILDER, Appro is building a big Opteron (Barcelona)-based machine for the Tri-Laboratory project.
In order to achieve 438 TFLOPS, a total of 3024 nodes with 12096 processors will be used. Yes, this amounts to 48.384 cores combined with 96.8 TB of system memory. Cluster is created by using InfiniBand 20Gbps 4 X DDR fabrics with Voltaire Grid Detector 288 port core switches.
All of this is being built using hardware from Supermicro.
AMD has sold more barcelona quads in one month than intel(copytel)sold in 11 months.
Nobody wants reworked pentium 3 quad relics.
BUY AMD the only real future proof quad core, hi performance, energy saving, low cost, hi tech, cpus, platforms, and video solutions.
This comment has been removed by the author.
Abinstein already got Pre-owned by pariander in Abi's own blog.... lol.
Sorry Sci but pariander had some VERY good arguments against your flawed logic.
You should REALLY read up on Cost Effectiveness vs. Quality. Intel Quad is very Cost effective to produce and of very high quality. It's a money maker, and having 1 year lead over AMD was also a great contributor to quadcore success.
While the Barcelona is a massive expensive, low yielding not very cost effective product with questionable performance and a year too late (compared to intel).
if AMD wasn't so damn COCKY and full themselves, they could have been in a whole new league today. Poor management/purchasing decisions, Lackluster marketing have all contributed to it's downfall.
Stop blaming Intel's monopolistic ways of selling being the reason why AMD is a failure.
It's AMD choices that make it a money loosing company. Lacking in cost effective processes like die shrinks, no matter how you defend it, ATI purchase was made AT THE WRONG TIME.
Is it not true that to date, AMD has sold everything it has produced?
Constant delays and lackluster performace has doomed AMD.
YOu can tout technical advances and quality ALL YOU WANT, but if AMD can't make money, it's going no-where. Remember it's in the business to make moolah....not work for free.
False. Go to any web site Asus, ASrock, … and see the CPU compatible table, even today selling motherboards most of them cant run quad core CPUs, and guess what? They are all socket 775!
Try to do that with one 2005 motherboard.
Just go back and list from 2004 which socket 775 can run the 2005 PentiumD.
Which Intel socket 775 motherboard from 2005 can run a 2006 Core 2 Duo.
I stated that the board that I purchased was good for Core 2 Duo, then Core 2 Quad, then finally for the 45nm refresh of Core 2 (Yorkfield/Wolfdale). If your truly interested that board is the ASUS P5B Deluxe. That's a motherboard that was released in June 2006. I bought it in late July 2006 along with an E6600.
FWIW, I do agree that keeping the same socket but having so many CPUs incompatible with the older boards is a silly thing to do. This was especially true when the 915/925 chipsets turned out to be totally incompatible with the Pentium D.
Intel fanboi -
"Just because you state a FACT after a lie in the same paragraph does not mean that your lie becomes true."
Let me get this straight from you - what was the "lie" that I supposed have stated?
You don't even pass as a good FUDer.
Aguia said: "Gutrat if you don’t understand this then you aren’t one expert. ;)"
Gutterrat, maybe it's time for you to admit that you did't know what you (or we) were talking about. In other words, you (like other Intel fanboi's here - pun intended) are doing nothing but trolling with FUDs.
Aguia is also right in stating that PCIe 2.0 is one reason for desktop to have HT3. In any rate, AMD still holds the absolute supremacy in terms of multi-socket server even until the first Nehalem generation; there is really no rush for HT3 at the server space.
aguia
Perfectly correct!
While Intel IGP traffic goes with in the chipset.
AMD IGP traffic goes through the CPU NB in order to use the built in IMC.
So with DDR2 memory speed at 1066MHz the HT bus would become the bottle neck in the communication.
HT 1000Mhz = 8GB/s
PC2-6400 DDR2-SDRAM (dual channel) = 12.8 GB/s
Intel doesn’t have FSB problems when using IGP because the IMC is in the chipset. So there is no traffic stressing the FSB.
AMD needs a 1600Mhz HT in order to achieve 12.8Gb/s.
Gutrat if you don’t understand this then you aren’t one expert. ;)
you sure you did your math right? Because if I remembered correctly, HT 1.1 already provided bandwidth up to 12.8Gb/s, and 22.4Gb/s with HT2.0.
I don't see any bottlenecking if DDR2 indeed operates at 12.8Gb/s.
http://www.hypertransport.org/
tech/tech_httwo.cfm?m=4
Looks like you're no one expert either ;)
Abinstein
Aguia is also right in stating that PCIe 2.0 is one reason for desktop to have HT3. In any rate, AMD still holds the absolute supremacy in terms of multi-socket server even until the first Nehalem generation; there is really no rush for HT3 at the server space.
Here the FUDster goes again...
I agree that PCI-E 2.0 might be the reason why AMD wants to implement HT3 on desktop, but given the fact that most video cards only use 90% of the bandwidth provided by PCI x8, it would be a while before we really need x32, no?
Again, you won't see substantial performance improvement on HT3 implemented machine until a lot later.
Correction, PCI-E improved the bandwidth of each lane by 2x, not 2x the lanes.
As for server space, I believe HT3 is very crucial to AMD's continual success in the 4S server segments. As opposed to desktop segments, severs usually require large amount of transaction between each components on the board.
If HT3 needs to be implemented, it would be better to be implemented in server, rather than desktop.
Or maybe Dell and HP don't see the performance boost by switching from HT2.0 to HT3.0...because Barcelona doesn't lack bandwidth, but it lacks proper architecture design.
Budapest is just the single-socket (AM2) version of Barcelona.
Right Albenstein. You say core2 lacks bandwidth between cores. More cores, require more bandwidth.
So why on earth would you have single socket solutiosn with more bandwidth than 4 socket ones?
AMD is retarded
acquia,
Duh.
I know that AMD suffers from not having the IGP next to the memory controller.
The point here is that HT 3.0 does not matter because there's enough bandwidth available on PCIe to satisfy the needs of most graphics cards.
The fact is that not all of the existing PCIe bandwidth today is used. abinstetin, the tool, needs to get a new pacifier as it seems that he's chewed through his last one.
Intel does not have the graphics bandwidth problem on IGP that AMD has.
Meanwhile Intel maintains the same socket with both support for DDR2 and DDR3 memory.
Same socket that isn’t pin compatible. Who cares that Intel socket 775 exists since 2004 if you can’t plug a new processor on it?
All I needed for quad core and 45nm CPU support was a simple BIOS update.
False. Go to any web site Asus, ASrock, … and see the CPU compatible table, even today selling motherboards most of them cant run quad core CPUs, and guess what? They are all socket 775!
Try to do that with one 2005 motherboard.
Just go back and list from 2004 which socket 775 can run the 2005 PentiumD.
Which Intel socket 775 motherboard from 2005 can run a 2006 Core 2 Duo.
…
Do that simple exercise and I will agree with you.
Intel is already in socket 775 5th (incompatible) revision. I will not even mention the four bus speed updates, wait I just did ;)
WTF? Why would anyone have owned an Intel 775 board in 2005? AMD were killing them back then. AMD was obviously the best choice. For the past 14 months though Intel is the only choice.
AM2...no thanks.
Chistian Howell hollered incoherent stuff about HT3 being available.
Christian, I have contacts at Microsoft and it will be pretty easy for me to verify whether you are a current employee.
Are you or are you not a current employee of Microsoft? Please answer yes or no and I will take care of the rest...
With respect to all of the other nonsense you wrote it appears that you haven't gotten the message.
HT 3.0 really is of marginal use on the desktop. The value is really is in the server/HPC domain. Who gives a shit if it's available on the desktop on a dual core machine which you
AMD fanbois like yourselves who spend 95% of your time playing games would use an appropriate discrete graphic solution instead of IGP anyway.
Unless you can't afford a discrete graphics solution in the first place.
Also, I resisted the temptation to call you schmuck. Well, maybe not now.
[b]AMD has sold more barcelona quads in one month than intel(copytel)sold in 11 months.
Nobody wants reworked pentium 3 quad relics.[/b]
PROVE it.
of course you won't be able to, because just like Sharikou, you are talking out your abused ass.
Put up the facts, or shut the fucking pie hole you call a mouth.
AMD has sold more barcelona quads in one month than intel(copytel)sold in 11 months.
Nobody wants reworked pentium 3 quad relics.
Haha. good one!
I guess no one wants cars with round wheels either because that would just be copying the original car right?
I love the Internet. I have a theory that the world is always in perfect balance. It's quite reassuring to have people like you around. It means there is a highly intelligent opposite to you somewhere on this planet counteracting your imbecility.
Chistian Howell hollered incoherent stuff about HT3 being available.
Christian, I have contacts at Microsoft and it will be pretty easy for me to verify whether you are a current employee.
Are you or are you not a current employee of Microsoft? Please answer yes or no and I will take care of the rest...
I quit my job at Microsoft and moved from Redmond in 2002. Oooh, you have contacts at Microsoft. I'm scared of you.
And you're right I will never agree with any of you cause you are ignorantly biased.
HT3 has NOTHING to do with the split power planes of the cores and IMC. That's why there is a spit plane board without HT3.
I'm upset that Barcelona didn't get it as it would make 8P even better with 1 hop latency out to 32 cores.
Just agree that AMD is a professional CPU manufacturer in the Top 10 of semi-con manufs.
We should all expect the leapfrog in perf from these two. The difference is that AMD people didn't call Intel people names when HeatBurst was nearly melting cases and was nearly 1GHz slower perf-wise.
No, now we have the AMD jihadists claiming that the Barcelona is 146% faster than anything Intel offers and is selling three times what Intel's quad-cores are, that Intel is going BK in Q2 of 2008, that they have PhD's and what not.
Good going fanbois!
The desperation on AMD fanbois is very apparent no matter how they Hide.
Multiple quarters of losses.
Constant delays.
lackluster ATI performance and delays.
AMD delays.
Barcelona failing to scale.
Failing to perform.
Employees have their benefits cut.
Bad time to be an AMD fanboi.
AMD is a professional CPU company...one being driven out of business by incompetent managament. Ruiz fucked Motorola now he's doing a fine job on AMD.
Frankly I'm glad Baron isn't in Redmond...one less poser we have to deal with out here...
Mo
The desperation on AMD fanbois is very apparent no matter how they Hide.
Multiple quarters of losses.
Constant delays.
lackluster ATI performance and delays.
AMD delays.
Barcelona failing to scale.
Failing to perform.
Employees have their benefits cut.
Bad time to be an AMD fanboi
At least the AMD supporters seem to have a little bit of class. The same can't be said for the Intel trolls with their vulgar language and lame insults. I'm actually really surprised that such behavior can be exhibitited by people who one would otherwise consider to be reasonably intelligent.
BaronFucktard blathered..."And you're right I will never agree with any of you cause you are ignorantly biased."
This is like the pot calling the kettle black....hows that 4x4 running? You know, the one you were going to have by last Christmas that you made excuse after excuse for not purchasing...
tech4life mumbled...."At least the AMD supporters seem to have a little bit of class. The same can't be said for the Intel trolls with their vulgar language and lame insults. I'm actually really surprised that such behavior can be exhibitited by people who one would otherwise consider to be reasonably intelligent."
I could provide intelligent responses, but why?
Sharislut is a lying piece of shit, has no journalistic abilities whatsoever, is NOT a ph.d, and yet people continue to support him...so as long as people continue to support this fuckwit, it's not worth the effort of anything more than vulgar insults and bashing. It's fun and a way to insert some humor into an otherwise "corporate" day...
People like him deserve everything they get, and this, most deservedly, is what they get. If you blindly follow this retard, or support him in anyway, then you (this is general "you" tech4life, not a you reference) deserve the same treatment....
Class?
Class is not lying to the world on your blog about having a PhD.
Class is not posting fantasies as some type of AMD fanboi jerk-off material.
Class is not blindly accepting everything that AMD says as the truth.
Class is not claiming some web site to have the facts, only to back out and call that site a sell out when they do a negative review of one of AMD's products, or god help them, review an Intel product and find it superior to AMD's offerings.
Class is not having a bunch of doorknob fuckers following your every typed word like some type of prophet.
Ruiz fucked Motorola now he's doing a fine job on AMD.
I've heard this before. Moto was the biggest mobile company over here when the craze went off. All good phones were moto or a long time. Things went sour for them a good 5/6 years ago though.
This is like the pot calling the kettle black....hows that 4x4 running? You know, the one you were going to have by last Christmas that you made excuse after excuse for not purchasing...
This is classic. It does however prove Baron does have a some sort of neurological functioning. No one in their right mind would buy a 4x4. What a waste of money and electricity
yomamafor2
Looks like you're no one expert either ;)
Look like you are going to get owned too.
HT at 800Mhz has data transfers up to 6.4GB/s.
HT at 1000Mhz has data transfers up to 8GB/s.
HT at 1600Mhz has data transfers up to 12.8GB/s.
Since current socket amd2 motherboards have the HT clocked at 1000Mhz they are limited to 8GB/s.
And the 22.4Gb/s with HT2.0 is only possible with 32x links (dual 16x links) otherwise would be 11.2Gb/s. Current motherboards use 16x links.
You don’t have to thanks for the explanation ;)
yomamafor2
I agree that PCI-E 2.0 might be the reason why AMD wants to implement HT3 on desktop, but given the fact that most video cards only use 90% of the bandwidth provided by PCI x8, it would be a while before we really need x32, no?
Well in that case you are assuming that there is only one single slot, where Ati and Nvidia are already talking of three and four. And you are also assuming that Ati, Nvidia and Intel are all going to stand still in improving their products performance.
Also the primary objective in improving the AGP bus speed was not the lack of bandwidth for the cards, but to allow the video cards to use the slow computer memory for texturing cache. However we already know it’s a bad idea for high end products but for mainstream products like IGP, Nvidia Turbo memory, Ati Hypermemory they love it.
Again, you won't see substantial performance improvement on HT3 implemented machine until a lot later.
I bet lots of the guys that bought brand new system based on PCIe 1.0 are all hopping that you are right. And if the AGP to PCIe transaction told anything to us is that there is going to be a 0% performance improvement.
pointer said...
I'm not sure who is lying here.
Are you serious? Do you really think that Intel is SO smart that they invent everything? Went the High-K route all alone and 'own' all the credit? Please... come back to earth! There are more than two dozen companies out there, which combined equal 100 times Intel's size, which have a vested interest in developing, researching , tooling and testing High-K ideas!
Read the article again, even Intel admits it:
In the mid-1990s, we at Intel and other major chip makers recognized that we were fast approaching the day when we would no longer be able to keep squeezing atoms out of the SiO2 gate insulator. [...]
Of course, we weren't alone. And there were still plenty of unknowns. By 2003, researchers in university labs and other semiconductor firms around the world had zeroed in on hafnium-based materials as the gate dielectric. A variety of them were under earnest study: hafnium oxides, hafnium silicates, and hafnium oxides containing nitrogen. The method of forming the high-k film, too, was unsettled, with different groups trying sputtering, chemical vapor deposition, and atomic layer deposition, which we eventually settled on. [...]
However, many of our peers saw the gate-last process, which we ultimately adopted, as too much of a departure and too challenging. [...]
By late 2006, though, nearly everyone, including us, had given up on the fully silicided gates approach. No one could move the silicide's work function quite close enough to where it needed to be.
Nevertheless, the search goes on at other major chip makers to find the materials with the right work function that could survive high temperatures and enable the industry standard gate-first process flow.
Are you refering to the IBM-AMD's 'me-too' annoucement made at the same day of Intel announcement, which until now has no single working die shown?
No I'm not referring to that... I'm just referring to common sense!
Now I don't know where AMD/IBM/Sony/Toshiba/et Al is in regards to their development of High-K. But if they are expected to be 12 to 18 months behind on High-K, that's ok with me. If it turns out problematic or disastrous, Intel could survive it... AMD could not! Let Intel lead on this one (let AMD learn from it)!
It's rumored that a High-K processor will melt down if there is a full moon out on the 13th day of a leap year.
Frankly I'm glad Baron isn't in Redmond...one less poser we have to deal with out here...
People like you are a big reason why I quit. Just a bunch of warm-bodies in some weird love-fest.
If you work in Redmond, you know what NTSELF is. From 1999 - 2002, I was the most prolific troubleshooter.
My emails were a-chriho and chrishow.
Do a search on the Public Folder between those times, you'll see who's the smart one and whose a warm-body with an affinity towards brown-nosing.
This is like the pot calling the kettle black....hows that 4x4 running? You know, the one you were going to have by last Christmas that you made excuse after excuse for not purchasing...
How does me changing my mind have anything to do with name-calling? Why would me changing my mind be an excuse?
I'm glad I waited as I should be able to get an X4 using a lot less power. And it's easier to setup water-cooling with one CPU.
Hopefully, HP Blackbird will have them when I'm ready to buy. No more fans for me.
This comment has been removed by the author.
Where in my post did I say I worked for that company? Never have, never will, far better companies to work for in the Pacific Northwest. I do work *with* people at MS, people far above whatever level you think you were at or probably ever will be...(see, two can flap around epenis' only some of us can do it without being self-absorbed little bitches.)
I could care less what you did there, or what you do now. You're an egotistical pretentious poser who thinks he's more than he is, which when you compare nothing to nothing, is still, in the end, nothing.
Have a nice day, Baronretard.
Tick(Yonah) Tock(Merom) Tick(Penrym) Tock(Nehalem)
INTEL BK in 2007, 2008, now 2009. Sorry PhD "Ph"ony "D"octorate its AMD that is going BK
I told you AMD cum lappers last year that in two cycles of the clock AMD would be done for sure. Well we are are now 3/4 of the way there in a few more days. The Phoney Doctorate and his fluffers are all lying low. THe phony continues to spew more laughable posts and his most ferverant fluffers keep true to form.
The most notable technical development of 2007? That insures AMDs continued down fall was their reliance on IBM technology. Without leading edge performance silicon how can you build a vaible business. YOu start with one hand and foot missing in the race.
First and foremost the biggest technical development in 2007 was INTEL's annoucement of something they had been working on since 2003, HighK/MetalG. In 2007 they annouced two fully functional and yielding CPUs with this.
Read here of the story! http://www.spectrum.ieee.org/oct07/5553
And in a few days IBM/AMD can buy one and go jeez how come we can't do that. Then go scramble for a few years and maybe in 2009 IBM will produce it too.
Today AMD/IBM got nothing. Matter of fact they had to hastly call a press conference claiming me too, me too at the first annoucment earlier this year. But they have yet to show a fully functional test chip let along a CPU. They are years away from figuring out how to harvest the 10x improvement in leakge and performace that HighK/metal gate bring and will be saddled with leaky SiO2 gateoxide for 45nm. AMD will be BK before they get to 32nm HighK.
On another front, Barcelona will soon be irrelevant. Penrym is coming soon and then we have Nehalem on deck. Barcelona pricing in the high end with their current a "promised' future performance will NOT be enough to rescue AMD from financial ruin. I said "promise" because like their promises in 2006 and thru 2007 the talk and walk has already been demonstrated by the silly men in green. Promise and under deliver late. A huge integrated quadcore on 65nm with AMD "mature" yields of 0.5 defect/cm^2 is a recipe for bankruptcy.
An aside did you see the recent reverse engineering of the Barcelona? Its 11 metal layers. Yes you heard it right 11 metal layers. Can you imagine the cost and yield of doing all those extra metal layers!! INTEL is doing with consderably less which reduces cost, improves yield, improves factory productivity. AMD has to deal with these kind of cost, output penatlies because they continue to be saddled with inferior silicon technology. And like I said silicon technology is the basic competitive building block. Its like sending AMD out to war with bows and arrows while INTEL has rifles. Yeah them indians won few battles but the Cavalry wins the war. AMD is lost, alway was, and still is.
Lastly,you fanbois were all tripping over each others hardons before C2D cameout. I told you all INTEL was simply to big, had too much money to lose the war. Sure it lost the Opeteron battle competing with Netbust, but in the end you can't out engineer INTEL. Good engineers with what looks like infinite fab capacity, far superior technology, 2x transistor budget permits you to do wonders. C2D continues to clean AMD's clock and we haven't even seen Penrym ramp and stepping improvements yet. Then you got Nehalem coming with new architecture, faster interface, native quadcores and IMC. AMD has now way to compete with this with inferior 65nm, higher power, lower performance, higher cost 65nm. When they do migrate to 45nm in 2009 it will be with inferior silicon oxide/polysilicon gates. AMD will be gone before they transition to 32nm HighK/Metal Gate.
A sound bite from Anadtech on AMD's latest offering in mobile which represents all the growth and premium pricing
"While battery life and power requirements look to be reasonably competitive with Intel's offerings, in terms of raw performance AMD gets beat pretty easily. This occurs even when comparing AMD's latest $400 Turion X2 TL-66 to a system using Intel's $250 Core 2 Duo T7300. If you are concerned with mobile CPU performance, AMD is essentially unable to compete at all with Intel's current Core 2 Duo laptops. While you can typically find AMD-based solutions for less money than Intel-based notebooks, features often end up being cut in order to reach a lower price point, "
Summary you can buy AMD, you get lower performance and cut rate components... What a deal from the boys in Green.
You guys can continue to argue about Barcelona benchmarks vs. intels current plaform, but just like the Netbust days. Nehalem is coming, then Westmere, then Sandybridge. The cavalry will always win against indians armed with last generation weapons.
THis is a fucking business, you need to deliver value and still make money. AMD delivers value but loses billions how is that a sustainable business. Its fools who continue to support that broken business model, its the one going BK.
Christian whatever the hell, wrote,
I quit my job at Microsoft and moved from Redmond in 2002. Oooh, you have contacts at Microsoft. I'm scared of you.
Most prolific poster on NTSELF? The most prolific troubleshooter you say?
You mean: you were an SDET?
You were not an SDE?
That explains quite a lot!
Are you sure you 'quit' and weren't 'let go'?
ROFLMAO
Jeach! said ...
...
Are you serious? Do you really think that Intel is SO smart that they invent everything?
...
No I'm not referring to that... I'm just referring to common sense!
...
when did i say Intel invented everything? Now I see where your thought came from ... 'Giant said ... But now Intel will have the biggest edge ever: the high-k metal transistor. This is an Intel exclusive.'
First, what Giant said also didn't imply Intel invented everyhting too. It just mean that Intel has it, and is the only company has it ready for production right now, no other espeially AMD/IBM can do it 'now'
Secondly, your reply did imply IBM has it too ('me-too') ... and my reply to you was that IBM's annoucement is a face-saving, paper annoucement.
Surely other company/entity would surely got it eventually, and might actually contributed to it here and there through idea, tooling, etc. It still doesn't change the fact that intel has it and AMD doesn't, an unique advantage to Intel not only it has a smaller node, but also a much better transistor characteristic.
This comment has been removed by the author.
Post a Comment
<< Home