DELL Dimension E521 is a great machine
I finally received the two Dell Dimension E521s ordered. One has a 22inch LCD the other 19inch. I figured it is cheaper to buy from DELL--once you add Microsoft Windows and LCDs screens to the equation, a self-built machine is more expensive.
The ones I ordered came with the lowest AMD CPU these days, the X2 3600+, a dual core power house running at 1.9GHZ armed with AMD-V and AMD64 technologies (which Intel's Israeli team copied). Of course, I need to do some upgrades.
Openning the box was super easy, plugging in RAM is also trivial. Now, the interesting part: replacing the CPU with a faster chip. I prepared to do some major manual labor.
It turned out it's super easy to upgrade the CPU too. Just loosen the two screws, and you can rotate and lift the heatsink assembly, then you just lift the lever and replace the CPU. All done in two minutes.
The machines are for office work, but as all AMD solutions, they are Vista and Doom3 capable.
A case like the one used for the E521 would cost more than $50. Try add the other parts yourself.
posted by Sharikou, Ph. D. @ 6/05/2007 05:42:00 PM
120 Comments:
How about you give us some REAL news like Barcelona benchmarks. How about the potential Delays of Barcelona?
AMD not meeting deadline for cray..how about that?
no one wants an outdated E521.
AMD going down.
Dell uses AMD
Dell goes down...
hmmmmmmmmmmmmmmmmmmm I bet they're beating their heads over and over, WTH were we thinking?
hahahha vista and DOOM 3 capable...........gaming with a 3600X2 and AMD/ATI IGP......AMD fanboi's dream come true....only AMD fanbois are dumb enough to play games with IGPs when you grow up maybe you can afford a discrete graphics solution........
AMD-V copied? Intel had Virtualization well before AMD. Nice try though.
AMD not meeting deadline for cray..how about that?
Dell uses AMD
Dell goes down...
Indeed. It's not hard to see why Sun is starting to use Xeon CPUs in it's servers.
Oh Intel hate-peoples. The Sharikou just bought more computer-cpu-power than in all of Intel production plant for real-cheap.
Burninator is 100% super-objective so let us look at the facts about graphics. It is true that AMD only recent bought the ATI. But ask yourself, why would super-success company like AMD only buy out 1 graphic company? With super-cash from all Opteron sale (AMD actually MAKE money with each Opteron manufacture because water used has cance-curing properties after production!). AMD could buy BOTH ATI and Nvidia so why just one? Answer Obvious: Look at 'name' of Nvidia: See: Letter N and letter I, and letter 'D' which sometimes sounds like 'T': OBVIOUS!!! NVidia is simply a front for EVIL Intel. Many years ago Intel knew it would go out of business because of ferocious AMD saving-world. So it make a 'graphics' company called Nvidia, knowing that when AMD get big enough, it buy the graphics company, and then EVIL Intel Agents could go to work from inside to destroy good AMD and steal all technology!! AMD was too smart for silly Intel trick so only bought ATI.
This means that ATI is new, and therefore is only ahead by 3x generations of evil Nvidia/Intel monopoly. Even so, there is 1 super-detailed review that shows fair comparison of ATI cards It say: "New ATI X2900 XTX Shames Nvidia 8800 to Second Place" Now, this is definitely a pro-Nvidia statement since ATI is 3x generations ahead, so Nvidia is FARTHER BACK from 2nd place, but with strong evil bias on web today it still OK.
This lead to 7th point: DO NOT BELIEVE THE FRAMERATE MYTH! Oh, some fanboy site like Anand or HardOCP/HotHardware evil boys try to put up pictures showing ATI have 'lower' framerate. What they forget is that ATI technology is still 3x generations ahead of Nvidia! It only display the GOOD frames on purpose for 3x superior game play! The stupid-evil NVIDIA cards display crap-distortion that proven to lower scores! Since ATI GPU is 3x ahead of Nvidia, the evil Nvidia cards must get 3x greater framerates to keep up. (Oh Intel Fanboys, this must remind you of bad-stupid Prescott that is at 3 FAKE Ghz and still slower than K5!).
Now that ATI is part of Super-AMD, expect within mere 2 months updated GPUS that back to normal 10x generation lead expected of all AMD. I talk with Hector and he say that new technology will mean that graphics so advance that players of ATI cards will be able to frag just by looking at evil NVIDIA players in super-realtime! Also, super ATI GPUs will automatically detect aliens when not needed for super-high tech display. Intel need box of Crayons on primitive 32 bit bus! Haha!
Hahahha great view points the burninator you have almost convinced me.....i would tell you to open start typing a blog with your "interesting" take on things and then i remember this is what i come to sharikou's blog for...........
haha burninator that was awesome. I always love your insights. They are much more accurate than those of Sharikous.
you are so RIGHT. I had a question thought.
Since Intel started Nvidia long time ago (proven fact), do you think it's possible that Intel spread the cancer to Nvidia? and now Nvidia is stealing framerates from ATI?
Sharikou
Try add the other parts yourself.
This seems like it would have been a better deal, but it lacks Dell support...
Way to go Sharikou!
AMD based computers are in many ways superior to what intel has to offer. You did a great choice here by going with real brand (instead of hiked up on advertisement pseudo-brand like intel)
To all intel fangirls: learn from this man, he got PC that pawn what you got by a mile and all because he reads what people say (it is spoiled all over the internet: intel is inferior to AMD)
You're so right! Intel will frag himself like never before... I hate my C2D... I'm so sorry that I've bought it 2 months ago, right now I'm going to sell this stupid C2D = Crap to dump...
Hype. Thats all that this Core2Duo is. A way for intel to extract money from people's packets via 'shady' benchmarks and cool TV commercial.
Here's what real people are saying:
Posted by: Hammm Feb 26 2006, 11:37 PM
Unless you encode video as your life, AMD offers by far the better value for money and especially in games and dual core.
Athlons are currently the best in the buusiness except maybe the AMD Opterons
LinkToForum
lol where the hell all those hardcore amd fanbois came from?
a few months back it seems they all got extinct and now they are back. But i tihnk thats cool, more insight in the world of fanbois and there thoughts, means more fun for me. :)
If any of you could read Estonian I could link to hundreds of posts talking about AMD users who had switched over to Core2 and saw the superior performance. They don't believe hype, only performance they see in their applications and games.
Of cource that might be a bit biased since vast majority of those people OC their CPUs and get far more performance out of them compared to their stock speeds. As we know AMD OC ability sucks compared to C2
Of course, I need to do some upgrades.
On a brand new AMD system? LOL. That awful huh?
waxedass said: To all intel fangirls: learn from this man, he got PC that pawn what you got by a mile and all because he reads what people say (it is spoiled all over the internet: intel is inferior to AMD)
Sharikou just bought a PC that needed an immediate upgrade. Yep, i learned something alright. that all AMD PC's are obsolete right out of the box. No thanks.
waraxe
Hype. Thats all that this Core2Duo is. A way for intel to extract money from people's packets via 'shady' benchmarks and cool TV commercial.
Here's what real people are saying:
Posted by: Hammm Feb 26 2006, 11:37 PM
Unless you encode video as your life, AMD offers by far the better value for money and especially in games and dual core.
Athlons are currently the best in the buusiness except maybe the AMD Opterons
Wow. I'm truly stunned... by you utter stupidity. First of all, who the h*ll uses forum as a form of supporting evidence? Secondly, please double check the date of the post.
"Feb 2006".
It was at least 4 months prior to Core 2's launch, and the poster was comparing P4 with Athlon 64. Now you're using this to back your statement "Hype. That's what Core 2's all about".
Kid, this kind of stupidity is funny at first, but not anymore.
It was at least 4 months prior to Core 2's launch, and the poster was comparing P4 with Athlon 64. Now you're using this to back your statement
However uncle snack yumi yumi for USD2 only, although the A64 was born same generation with you and your beloved P4, but, as you can see until this days the A64 is still considered as a faster and competitive CPU compare to you and your beloved P4.. In fact, in some benchmarks, the old timer A64 still manage to surpass the youngest fastest QX of Intel 65nm tech.
May I ask you one question Uncle snack yumi yumi? Is there any single benchmark showing that Intel P4 surpassing the AMD current fastest CPU?
What has Pentium4 got to do with anything? Its production is (almost?) over and Core2 based CPUs are already everywhere with prices starting from $85 on newegg and another pricedrop coming in the end of July.
Btw, what if I would really find some benchmark where P4 is faster than AMD64?
For starters I'll give you this, this,
this, this, and this. What did I win?
Also, those cheap e2000 series CPUs OC rather well. I know several people running those <2GHz CPUs at >>3GHz with air cooling.
http://www.dailytech.com/Quick+and+Dirty+AMD+K10+Cinebench/article7574.htm
i realy hope for amd that this was still buged silicon...
No it is not, they say this is their latest silicon.
k10 1.6Gzh = 27sec
k10 1.8Gzh ~ 22 sec
k10 2.0Gzh ~ 17 sec
k10 2.2Gzh ~ 12 sec
K10 2.4Gzh ~ 7 sec
so, at the same clock speed, opteron 1268 SE will frag xeon 3220 by 59%~60% ;-)
Please learn your math skills.
2.4GHz=17s
1.6GHz=17*(2.4/1.6)=25.5s
I love it! Hey Sharikou, you keep buying your machines through Dell. Every AMD cpu you buy through Dell is one less that AMD can make decent money from selling somewhere else.
The rest of your comments are hogwash. You describe switching out a CPU with ease as if it were a surprise! "Plugging in RAM is also trivial," suggests you also expected something more difficult.
A four year old can put together a stock PC, but you describe buying and upgrading your PC's as if it were some sort of minor miracle!
I can put it other way also:
1.6=27s
2.4=27/(2.4/1.6)=18s
Or to put it this way, at same clock speed Xeon is 18/17=6% faster than Barcelona
Where you get k10 2.4GHz=17s??
I think my math skills is much better than yours.. ekeke
lol, now i know how all the "estimated" and "projected" performance of k10 got accomplish, if they use the same "math" as pezal...
This comment has been removed by the author.
I think my math skills is much better than yours.. ekeke
Are all AMD fanbois pulling numbers out of their ass?
http://tinyurl.com/36dvtw
K10 @ 1.6GHz v Kentsfield @ 2.4GHz
Kentsfield vs K10 Cinebench result difference,
(27-17)/17 = 58.8%
K10 Clock speed disadvantage,
(2.4-1.6)/1.6 = 50%
It's simple math you dummy. So even if K10 was @ 2.4GHz, Intel's old Kentsfield frags K10.
AMD BK Q2'08
performance of k10 got accomplish, if they use the same "math" as pezal...
Must have LOL. Pezal do you even know the percentage change forumla? LOL
I am telling you guys. AMD has shifted from competing with the best to getting beaten by technology which is 6 months to a year old.....this is the new AMD deal with it.
pezal
"Where you get k10 2.4GHz=17s??"
First, it was 18s, not 17. I got it by using basic math 6'th grade math. What did you use?
Must have LOL
Yea.. me too.. Wakakakaka..
Pezal do you even know the percentage change forumla?
% formula = 100 x (B - S)/B
Thus,
K10 2.4GZH frag Xeon 2.4GZH
= 100 x (17s - 7s)/17s
= 100 x 10s/17s
= 100 x 0.588
= 58.8%
K10 2.4GZH frag Xeon 2.4GZH
= 100 x (17s - 7s)/17s
= 100 x 10s/17s
= 100 x 0.588
= 58.8%
http://www.dailytech.com/article
.aspx?newsid=7574
WTF
K10 @ 1.6GHz scored 27s.
Kentsfield @ 2.4GHz scored 17s.
Do the MATH you moron. Kentsfield had a 50% higher clock frequency than K10. The difference between the two scores was 58.8%. Even if K10 was clocked@ 2.4GHz, it would still lose to Kentsfield. If you can't see that Pezal, you need to go back to 6th grade math.
Now I finally understand where AMD got the "40%+ performance advantage" claim over Kentsfield, they used Pezal's "math."
K10 @ 1.6GHz scored 27s.
Kentsfield @ 2.4GHz scored 17s.
k10 1.6Gzh = 27sec
k10 1.8Gzh ~ 22 sec
k10 2.0Gzh ~ 17 sec
k10 2.2Gzh ~ 12 sec
K10 2.4Gzh ~ 7 sec
Correction:
opteron 1268 SE will frags xeon 3220 by 100% - 58.8% = 41.2%
pezal, could you please post the formula you use to come up with those numbers? I'd really like to see how you conclude that increasing clock speed by 50% increases processing power by around 4 times or 400%
Bunch of idiots.
pezal said...
k10 1.6Gzh = 27sec
k10 1.8Gzh ~ 22 sec
k10 2.0Gzh ~ 17 sec
k10 2.2Gzh ~ 12 sec
K10 2.4Gzh ~ 7 sec
If you had a brain, you can get that 3.2GHz K10 would finish in half the time, 13.5s. Ho ho and others have shown that Kentsfield is faster.
hoho,
C'mon now, you know it's due to the superior AMD architechture :)
There is another possibility: he uses AMD CPU and for some weird reason it can't calculate correctly.
It seems very obviouse that these numbers don't show a "significant" improvement that had been claimed by AMD, but we know many things that may have caused this unsolved mystery:
1. It's probably a very low power rated AMD Opteron Cpu.
2. AMD based system uses 667 mgz ram vs 800 mgz on intel based system (CB is not memory intensive as far as I know. Is it?)
3. Intel's system uses the lates chipset x38 vs an older one from Nvidia.
4. Barcelona is a pre-release silicon vs 100% working silicon Kentsfield.
We can't conclude leadership just by looking at one aspect of benchmarking.
People we need competition...
Without AMD there would have never been a conroe. Copetetion drives new inovations, better and faster products.
I am sure you people don't want to wind up with a celeron for 300 some $'s or whatever...
This is all the proof we need. Kentsfield and Clovertown will frag Barcelona. Intel does not even need Penryn to frag AMD. It's scary to think that they have Nehalem coming in H2'08 as well.
AMD BK Q2'08.
tommy, did you read the article? Most of your questions are ansvered there already.
"1. It's probably a very low power rated AMD Opteron Cpu."
It is not the highest end but I doubt it is the lowest end. They said some have "bootup capable" 2GHz quadcores. That doesn't sound like too good to me. Having 2.3GHz on roadmap doesn't tell much also as Intel has 5GHz on its roadmap too, just check the uncencored "Tera tera tera" PDF.
Also the point still remains: clock to clock Barcelona is slower than Clovertown, at least in that (floating point intensive) benchmark.
"CB is not memory intensive as far as I know. Is it?"
It is not, that means memory speed difference makes next to no difference.
"3. Intel's system uses the lates chipset x38 vs an older one from Nvidia."
CB is almost entirely dependent on CPU, chipset has almost nothing to do with it, especially on AMD. Also at least comparing X38 vs NV Intel chipsets the difference isn't that much.
Another interesting question is where is AMD own chipset?
"4. Barcelona is a pre-release silicon vs 100% working silicon Kentsfield."
It's the latest silicon they've got. If it is pre-release then the rumours about it being late might become true. Also has there ever been such a thing that two spins of the same silicon show different performance results? So late in production they are only fixing bugs, not optimizing the chip.
"People we need competition...
Without AMD there would have never been a conroe. Copetetion drives new inovations, better and faster products."
I agree. Too bad AMD started behaving so sucidally lately.
tommy, did you read the article? Most of your questions are ansvered there already.
Yes, I skimmed it.
ho ho - I agree with you in most of the scenarios and I am also skeptical about the clock-clock comparison. Of course this just a one aspect of benchmark so we can't conlclude alot. We need more benchmarks in order to make a fair judgement between two proce
Another interesting question is where is AMD own chipset?
That is very interesting... I am sure they will develop a chipset for their own platform...
tommy
"I am also skeptical about the clock-clock comparison"
I wouldn't do any hard decisions based on that benchmark either. Though it does tell us that the claimed 40% performance over Clovertown are most probably either false or only thanks to superior memory bandwidth. That bandwidth is there on K8 based servers too so it isn't that big deal.
I guess they are counting the overall performance as a 40% over Clovertown, but God knows based on what benchmarks?
Another question. What about those Spec benchmarks claimmed by AMD based on simulations???
This comment has been removed by the author.
Those were simulated at 2.6GHz and they used the memory intensive benchmarks. Had they compared against K8 the difference would have been a lot less than difference between K10 vs Clovertown
You mean they used those Apps that most common in server aspects?
Had they compared against K8 the difference would have been a lot less than difference between K10 vs Clovertown
Didn't AMD include K8 in that chart also? I remmeber seeing it...
Ok everybody have a seat. Pezal, stop chewing on the crayon and sit down...
Ok children, pay attention Pezal.
If takes you 20 minutes to walk to Tommy's house at the rate of 100 steps per minute, how long will it take you if you walk twice as fast(200 steps a minute?
The correct answer is 10 minutes!!!! or half the time.
now if you take 1.6ghz barcelona and make it twice as fast to 3.2Ghz, assuming 100% linearity, you'll only get 13.5seconds...
How in the world did you get 400% linearity LOL..
Pezal you've outdone yourself.
tommy
"You mean they used those Apps that most common in server aspects?"
Which of these are commonly run on servers? Perhaps speech recognision or ray tracing? Perhaps even quantum mechanics!
Too bad they didn't use the JBB, that would have shown some interesting results.
"Didn't AMD include K8 in that chart also? I remmeber seeing it..."
Then perhaps you have seen some different charts.
I was talking about this chart right here
"I was talking about this chart right here"
2.6GHz K10's "simulated" performance.
Ok everybody have a seat. Pezal, stop chewing on the crayon and sit down...
Ok children, pay attention Pezal.
If takes you 20 minutes to walk to Tommy's house at the rate of 100 steps per minute, how long will it take you if you walk twice as fast(200 steps a minute?
The correct answer is 10 minutes!!!! or half the time.
now if you take 1.6ghz barcelona and make it twice as fast to 3.2Ghz, assuming 100% linearity, you'll only get 13.5seconds...
How in the world did you get 400% linearity LOL..
Pezal you've outdone yourself.
Pezal's AMD calculator must be malfunctioning.
2.6GHz K10's "simulated" performance.
I know it's based on simulation, but we were talking about the comparison between K10 and K8 and that's I provided the link.
If those numbers are correct on that chart, then it makes sense but at the same time looking at those CB benchmark numbers it's just not relevant amd makes no sense.
http://news.yahoo.com/s/infoworld/20070606/tc_infoworld/89156
Barcelona DELAYED till October. Nothing will be shipped out in July as said earlier.
HUGE HUGE HUGE blow to AMD.
Hahahahh
2900XT late and poor performer.....
2900XT not having UVD.......
AMD VP selling stocks.....
barcelona delayed.....
barcelona having yield issues.......
Barcelona's terrible performance in every benchmark shown so far.....
And the doctor is busy reporting about how great a Dell Dimension E521 is which apparently needs to be upgraded upon reciept........Pathetic just like AMD and their fanbois..........
Openning the box was super easy, plugging in RAM is also trivial. Now, the interesting part: replacing the CPU with a faster chip. I prepared to do some major manual labor.
Apparently placing in extra ram, taking off a heatsink and replacing the processor is a great accomplishment for sharikou.......excellent operation doctor keep up the good work.......this is where AMD fanbois get their info.....a man who thought replacing a processor will need some "major manual labor"........
tommy, this shows little improvement from K8.
now if you take 1.6ghz barcelona and make it twice as fast to 3.2Ghz, assuming 100% linearity, you'll only get 13.5seconds...
You moron! Didn't you know AMD CPUs scale super-lineraly? In fact, they scale quadratically. So, doubling the clock speed quadruples the performance of AMD CPU. Also, there are only so many clock cycles that an Intel CPU can steal from an AMD CPU. Once the AMD CPU is clocked above that threshold, you get super-linear, almost exponential performance.
Fanboy!
he - thanks for the link. I see there is very little improvement shown based on the simulated benchmarks of Barcenlona and the actual benchmarks of Opteron 3GHZ.
I still don't understand where that 40% increase over Clovertown comes from?
I still don't understand where that 40% increase over Clovertown comes from?
That would be in specfp_rate. Here, Clovertown is severely limited by the FSB bandwidth, and hence does not scale well. Additionally, Clovertown has a comparatively animic fp unit. So, Barcelona will dominate it in that benchmark.
However, with Penryn, Intel claims that it can outperform Clovertown by about 45% in specfp_rate. In that case, Penryn will be withing +/-5% of Barcelona, if Barcelona actually runs at 2.6 GHz.
The Doctor really "P"iled it "H"igh and "D"eep this time, LOL
"I ordered came with the lowest AMD CPU... Of course, I need to do some upgrades." ..
"replacing the CPU with a faster chip. I prepared to do some major manual labor." ...
"The machines are for office work,"
What a dell machine destined for "office" work and it needs a memory and CPU upgrade. Okay I'm done laughing.....
Thats right, buy a spanking new Dell that is customizable, what do you need to do to that crap system? UPDGRADE. What an endorsement by a Dell "Dude" and the minute you receive it it needs an UPGRADE? What a product, what a price, what a value. Must be that AMD chip. With endorsements like that its no wonder Dell and AMD have lost all the momentum and leadership!
Wow, with a "Ph"ony "D"octorate thinking like this AMD doesn't need any INTEL marketers they only need you to continue to tell the world how great the computers wiht their CPUs are. Its NO wonder they are cutting prices again so people who are stupid enough to buy can use the extra money to upgrade to another faster CPU just to do "office" work.
DUDE your fire articles and INTEL BKs are better then this sorry post!
arcebolgna is HERE!! Better yet call it Bologna or Phenom's father... okay done laughing.
Here we are a few months away from launch an all AMD can do is demo a sorry 1.6GHz. Then they give some special people a few minutes on a "early" stepping. Where is the latest fast speed up stepping.
http://dailytech.com/Quick+and+Dirty+AMD+K10+Cinebench/article7574.htm
Lets review 1.6GHz and 27 seconds. Lets scale that from 1.6 to 2.3 GHz that would imply 15.2 seconds for Cinebench compared to 17 seconds for a Kenstfield. What a dissapointment only 10% faster. AMD in for some real trouble when Penrym comes. I don't care about them fancy talk about bus limitations. 50% faster my Ass. All AMD has done for the past year is talk and dissapoint. Oh they bought ATI, that was a good story, they dropped prices, they lost billions, they talked about native quadcore they talked about fusion, they talked and talked but forgot to deliver. Got crappy finances, got failed schedules. They are on the fine road to bankruptcy.
Lets talke Barcebolgna, the difference between early steppings and later steppings is most likely around hard logic bugs and speed headroom issues, power problems. You simply don't fix fundamental architecture/benchmark performance with stepping revisions. You fix execution issues. Thus these benchmark results pretty much caps the K10 performance for one family of benchmarks. This is simply a disaster for AMD. It again confirms why they haven't been quick to demo any of their limited early stepping silicon. People in the know will get more and more info and come to the same conclusions. AMD needs to stop calling this the K10 and call it the K9 it is. Its a total dog for the timing and respin this product got. It should be much better then this on such a simple benchmark compared to their current K8 90nm products.
WHy does AMD talk about simluated numbers? Because their real silicon and technology can't produce any real numbers. Like their products can't produce real profits they are reducing attempting to fabricate money thru debt and lawsuites.
Sharikou you are right on BK in 2008 but it is AMD.. keep buying them Dell AMD products that will be a collectors item right up there with TRS80, commodore PET from other bankrupt computer/cpu manufactures.
Keep up the good "Ph"ony "D"ishing, it makes great entertainment for lunchtime reading!
Barcelona delayed due to yield issues:- http://news.yahoo.com/s/infoworld/20070606/tc_infoworld/89156
Everyone knows it is true. AMD is finished. Penryn will be here before Barcelona.
AMD BK Q2'08.
Giant said...
Barcelona delayed due to yield issues:- http://news.yahoo.com/s/infoworld/20070606/tc_infow
Barcelona is not delayed. The article you provided says so:
AMD stuck to its guns on Wednesday, insisting that it is still on schedule, and pointing as evidence to its demonstration on Tuesday of prototype Barcelona chips in servers from infrastructure partners Supermicro Computer, Tyan Computer, and Uniwide Technologies at the Computex trade show in Taipei.
"We remain on track for a summer 2007 launch with partners shipping platforms in the third quarter," said AMD spokeswoman Marian Kelley. She blamed the delay rumors on Cray's announcement Monday that it expects revenue to drop because of a delay in shipping its quad-core Cray XT4 systems. In fact, Cray was referring to AMD's midrange Budapest processor, not Barcelona, she said.
"The guidance that we've been giving and that we're sticking to is that Budapest will follow Barcelona in the second half of 2007," Kelley said.
That guy didn't even read the article he quoted - thats just funny and intel-fanboy-like-ish.
lex said...
Lets review 1.6GHz and 27 seconds. Lets scale that from 1.6 to 2.3 GHz that would imply 15.2 seconds for Cinebench compared to 17 seconds for a Kenstfield. What a dissapointment only 10% faster. AMD in for some real trouble when Penrym comes. I don't care about them fancy talk about bus limitations. 50% faster my Ass. All AMD has done for the past year is talk and dissapoint. Oh they bought ATI, that was a good story, they dropped prices, they lost billions, they talked about native quadcore they talked about fusion, they talked and talked but forgot to deliver. Got crappy finances, got failed schedules. They are on the fine road to bankruptcy.
And here i give you an opinion of real expect on this Cninebench test:
What I see is this: the rigs that were compared (Xeon and K10) were not similar enough. In fact, all we know about them is that they used different CPUs and RAM. Not a good start. Further, consider: possible different OS's, possible different OS settings, possible different drivers, possible different installed apps, no attention paid to BIOS settings, no attention paid to graphics hardware differences, no attention paid to different harddrives, anti-spyware running?, anti-virus running? This was very much a "quick and dirty" comparison, of which we know essentially nothing about.
In short, don't read anything in to this. The only thing this is good for is a reference as possibly the lowest Cinebench score K10 will get.
..(LINK)..
lex said...
The Doctor really "P"iled it "H"igh and "D"eep this time, LOL
Whats with the mentality of an 8-years old? I really don't get it, does intel hires school kids to promote itself?
core2dude said...
That would be in specfp_rate. Here, Clovertown is severely limited by the FSB bandwidth, and hence does not scale well. Additionally, Clovertown has a comparatively animic fp unit.
I have a deep respect to intel-fans who understand just how far their favorite company is behine the world leader is CPU disign - the AMD
Mo said...
How abotu you give us some REAL news Barcelona benchmarks. How about the potential Dalays of Barcelona?
AMD nt meeting deadline for cray..how about that?
no one wants an outdated E521.
AMD going down.
Dell uses AMD
Dell goes down...
hmmmmmmmmmmmmmmmmmmm I bit theiy're beating their heads over and over, WTH weer we thinkng?
Athlon FX-62 OWNS Core2Duo Extreme by as much as 23% Touch this you intel infidel.
waraxe
"And here i give you an opinion of real expect on this Cninebench test:"
You call some random guy on a forum an expert? What next, you call yourself the one who made Alpha?
Also most of his points are moot anyway.
1) K10 had vastly superior memory performance compared to Xeon
2) OS doesn't matter. Even running stuff through Wine on Linux gives me next to no speed difference on CPU limited scenarios
3) drivers doesn't matter, it is a CPU benchmark
4) installed apps do not matter
5) BIOS makes at most 3-5% difference and on both machines
6) graphics HW makes no difference in CPU benchmark
7) spy ware and anti virus doesn't matter as it is not CPU intensive and could be running on both
Only thing I can conclude from this that the guy knows next to nothing about the subject. As you quoted him as "expert" you must belong to the same league or lower.
That Xeon has around 6GiB/s real-world memory bandwidth. That Barcelona has at least as much thanks to better efficiency. If we would scale Xeon back to 1.6GHz we would have to cut down its memory bandwidth also. That would leave it to around 4GiB/s on Xeon against 6GiB/s on K10. That means even with vastly superior memory bandwidth (and latency) Barcelona still gets its ass kicked.
"I have a deep respect to intel-fans who understand just how far their favorite company is behine the world leader is CPU disign - the AMD"
Platform design and CPU design are two different things. Intel is behind in platform design, not in CPU design.
"Athlon FX-62 OWNS Core2Duo Extreme by as much as 23% Touch this you intel infidel."
I'm sorry but I can't see any Barcelonas on that picture. Also it is a known fact that AMD has better memory bandwidth. Though as majority of benchmarks show it doesn't help too much and in the end it still looses to Core2.
waraxe said...
Athlon FX-62 OWNS Core2Duo Extreme by as much as 23% Touch this you intel infidel.
the_ghost is that you?
HAHA, you give me memory benchmark to prove your point? LOL
Is FX-62 a Barcelona chip?
I can't stop laughing at your stupidity.
What's even more funny? Core 2 STILL 6% faster than the already dead Barcelona. Penryn will leave the barcelona in dust....
Do you guys remember what some people were saying?
Core 2 Duo will take back most of the market that it lost....
Penryn will at the least match barcelona's performance
Nehalem will be the final blow to AMD.
Does AMD even have anything after K10? lol
This is too fun.
Hehehe.. Now i realize all the intelers has a lower IQ in math.. ;-)
k10 2.4GZH will totally frogging Intel Xeon 2.4GZH by 40%~50% and that what will going to happen this Q3..
k10 (early stepping up) 1.6GZH = 27 sec
step up to 1.8GZH ~ 22 sec
step up to 2.0GZH ~ 17 sec
Step up to 2.2GZH ~ 12 sec
Step up to 2.4GZH - 7 sec
with 33% clock increment it will boost k10 performance up to 26%. Meaning to say, k10 2.4GZH will be faster up to 26% over k10 1.6GZH, and this is more than enough to frogging xeon 2.4GZH by 40%~45% at the same clock speed. ;-)
Umm, you do know that difference between 27s and 7s is around 385%, do you? How can 50%* clock speed increase give you 385% performance boost?
*) 1.6 is 100%, 2.4 is 150%, difference is 50%.
Also anyone who uses math can calculate that scaling linearly Xeon would beat K10 by around 8% in that benchmark. Though the difference might be even bigger as you can't scale memory bandwidth with clock speed.
Umm, you do know that difference between 27s and 7s is around 385%
= 100% - [100 x (27-7)/27]
= 100% - 74.07%
= 25.9%
~ 26%
And can you show me your calculation that make you get the 285%???
1.6 is 100%, 2.4 is 150%, difference is 50%.
1.6 is 100%
1.8 is 111%
2.0 is 120%
2.2 is 127%
2.4 is 133%
And can you show me as well how you get the 150%???
correction ;-)
with 33% clock increment it will boost k10 performance up to 74%. Meaning to say, k10 2.4GZH will be faster up to 74% over k10 1.6GZH, and this is more than enough to frogging xeon 2.4GZH by 40%~45% at the same clock speed. ;-)
Waraxe: Whats with the mentality of an 8-years old? I really don't get it, does intel hires school kids to promote itself?
Maybe Intel should hire you then since you have the English composition skills of an eight year old.....
How many alter egoes do you have sharikou everyone knows its you. Why go under all these names we all know you are dumb, a liar and a fake......
AMD needs to work harder.
pezal
"And can you show me your calculation that make you get the 285%???<"
If you insist
27s is X%
7s is 100 %
___
X=(27*100)/7
X=385%
That means if you have a job that takes 27s time to finish you have to increase performance by roughly 3.85 times to get the job done in 7s.
What do your calculations mean? They make no sense.
"And can you show me as well how you get the 150%???"
1.6 is 100%
2.4 is X%
X=(2.4*100)/1.6
X=150%
That means speed difference is 50%.
Your calculations are wrong. 1.8GHz is 12.5% faster than 1.6GHz, 2.0 is 25% faster etc.
"with 33% clock increment it will boost k10 performance up to 74%."
From what planet are you from exactly?
If your CPU is clocked at 1Ghz and does the job in 1 minute and you increase the clock speed to 2Ghz or 100% it will do the job in 30 seconds or 50% the time. It won't do it in 15s or 25% of the time. You claimed it would, just you used 33% instead of 100.
"Meaning to say, k10 2.4GZH will be faster up to 74% over k10 1.6GZH"
At 2.4GHz it has 50% higher clock speed than at 1.6GHz. There is no possible way for performance to increase >50%. Xeon at same clock speed was around 58% faster than K10 at 1.6GHz. That means at best K10 will be ~8% slower at same clock speed as Xeon
Is US school system really so bad that people have no ideas how to calculate with percentages?
'k10 (early stepping up) 1.6GZH = 27 sec
step up to 1.8GZH ~ 22 sec
step up to 2.0GZH ~ 17 sec
Step up to 2.2GZH ~ 12 sec
Step up to 2.4GZH - 7 sec'
Pezal, where did you get those numbers from, you seem to have just made them up ?
You cannot have a quadroupling of performance from a 50% increase in clock speed because thay means it is not linear.
with 33% clock increment it will boost k10 performance up to 74%. Meaning to say, k10 2.4GZH will be faster up to 74% over k10 1.6GZH, and this is more than enough to frogging xeon 2.4GZH by 40%~45% at the same clock speed. ;-)
I think you are a certifiable psychopath.
If you insist
27s is X%
7s is 100 %
___
X=(27*100)/7
X=385%
1.6 is 100%
2.4 is X%
X=(2.4*100)/1.6
X=150%
Wakakakaka... Damn you make me LOL.. Wakakakakaa..
K10 1.6Gzh = 27 sec
k10 2.4GZH = 7 sec
Diff = 27 sec - 7 sec
= 20 sec
so, performance in % = 100 X 20/27
= 74 %
Ho ho , why do you bother with such a retard as Pezal ?
And for you Pezal , an advice : seek professional help.
To Ho HO..
Here is a link for you to learn how to use the percentage calculation formulas.. take your time ;-)
http://www.cornerstoneresults.com/RefLib/KnlgeBk/f_fr_Percentage_Calculation_Formulas.pdf
pezal
"so, performance in % = 100 X 20/27
= 74 %"
Well, it all falls down as that 7 second number is taken randomly and it has no valid sources. That makes your entire "calculation" based on wrong data. Not to mention that even if those numbers would be valid your calculations would still be wrong.
My suggestion to you would be to move to country where they actually teach math. Either that or wait until you are old enough to go to elementary school. After you've got some math skills come back here and present your arguments.
I said:
"If your CPU is clocked at 1Ghz and does the job in 1 minute and you increase the clock speed to 2Ghz or 100% it will do the job in 30 seconds or 50% the time."
Please explain where is the math error in this sentence. One of us must be wrong and since you refuse to show and explain where and how you got that 7 seconds number I say it is you.
"Here is a link for you to learn how to use the percentage calculation formulas.."
How nice of you. Now please explain what formula you used to get that 7 seconds number?
savantu
"Ho ho , why do you bother with such a retard as Pezal ?"
In its weird kind way it is fun to prove other people wrong :)
Why bother with a person that claims 33% clock increase will yield a 74% performance increase? AMD fannies make me laugh.
I think this statement "My suggestion to you would be to move to country where they actually teach math. Either that or wait until you are old enough to go to elementary school. After you've got some math skills come back here and present your arguments." is much appropriate to dedicate to the person as below,
Ho Ho said
1.6 is 100%
2.4 is X%
X=(2.4*100)/1.6
X=150%
That means speed difference is 50%.
Your calculations are wrong. 1.8GHz is 12.5% faster than 1.6GHz, 2.0 is 25% faster etc.
27s is X%
7s is 100 %
___
X=(27*100)/7
X=385%
Ok, if you say so. It would be even better if you'd point out the mistakes in those following calculations.
Also we are still waiting for the formulae you used to get that 7 second time at 2.4GHz
PEZAL!!! You make for Evil Intel Fanboy with math-wrong calculating!
How can you say 2.4Ghz SUper-Quad Barcelona make for 7 second time? TOTAL LIES!!! YOU ARE NOW FULLY SEEN EXPOSED AS INTEL FANBOY!!
Here is REAL mathematics to show AMD in Correct Way!
First: 1.6 Ghz make for 27 seconds.
Step 1: Fanboy reviewers are total liars, the time was actually 20 seconds in fair review.
Step b: Oh Pezal-Intel fanboys, you forget Barcelona is 4-way QUAD CPU power! This means increasing clock make for 4 time improvement!! So with simple math, 100Mhz increase on Barcelona is 4x of 100Mhz!
Step 5: Do not forget 10x generation performance lead!! This make for 10x speed improvement too! So 4 x 10 x 100Mhz MAke for 4000Mhz fanboys!
Step d: Barcelona will easily clock at 4 Ghz (do not believe Gigahertz Myth from Evil Intel that clock speed is not important!) So make all steps in combine:
4Ghz - 1.6 Ghz = 2.4Ghz on REAL Barcelonas (AMD already made about 50 million of these, holding them back to fool silly Intel since they already BK)
2.4Ghz x 4 super-core x 10 genearation lead = 90Ghz Equivalent rate.
90 Ghz Equivalent / 1.6 Ghz = 56.25 time speedup!!
20 second / 56.26 = 0.3555 seconds for REAL Benchmark!!!
Do not believe Liar-Fanboy Intel Pumper Pezal!! Anyone who say Barcelona cannot benchmark in the 0.3555 seconds is total fanboy liar!! Learn basics of the mathematics fanboys!
LMAO!
The Burninator has proved that pezal is nothing more than a paid Intel pumper fanboy!
In other news after reading your post I was laughing so hard I almost woke the dead!
Sharikou, I hope you're taking notes here. The Burninator is providing a 100% guideline on how to run your little fantasy blog here!
Here are more likely correct values
K10 1.6GHz ~ 27 sec
K10 1.8GHz ~ 24 sec
K10 2.0GHz ~ 21.6 sec
K10 2.2GHz ~ 19.6 sec
K10 2.4GHz ~ 18 sec
50% increase in cpu speed gives 50% less time taken
The life of an AMD fanboi:
1.) Ignore any reviews that prove an AMD CPU is slower than an Intel CPU, even if the review is by an accurate source. Then automagically brand that reviewer as an Intel supported site/pumper.
PS: if the same reviewer says that an AMD CPU outperforms an Intel CPU even in one small useless benchmark, then it is automagically proof postitive that the AMD CPU is superior to the Intel CPU. If Intel wins a certain benchmark its because the Intel CPU was optimized for said benchmark. When AMD wins a certain benchmark, its simply because the AMD CPU is superior.
2.) Fuzzy math. Make up numbers. Sure they look good, but once proven they are bunk, stand by them anyways. Why not, the AMD fanboi leader lies about having a PhD, why not lie about a few numbers.
3.) Keep insisting on comparing AMD's future products to Intel's current products, and completely ignore the fact that Intel's new CPU's are hitting the market first.
4.) Insist that wattage counts when it comes to CPU's, but totally discount the fact that wattage doesn't matter at all when it comes to video cards. Even when the wattage for the video card is up to 10x delta than the wattage delta with the CPU's
5.) Continue to insist that the 2900 is superior to the 8800 family, even when proven by every site under the sun that it's simply not true.
6.) wakakaka
evil_merlin
"3.) Keep insisting on comparing AMD's future products to Intel's current products"
Did you miss the posts where he compared K8 against P4?
;)
ok, just add the point for the fun of it
7) DAMMIT era, Nvidia chipset is good, GPU is good ... post DAMIIT era, ATI fragged everything that NVIDIA has to offer :)
8) NVIDIA is evil (my own prediction on their future comments)
:)
AMD Barcelona can potentially smash Intel
http://blogs.zdnet.com/Ou/?p=492
Its crying time again.....
... and Intel dedicated GPUs will enter the market in less than a year :)
I wonder how much will they have in common wiht Larrabee ...
to little to late...for intel graphics cards or chips
"to little to late"
You mean too little as K10 or too late as R600?
Anyway, if that GPU has opensource drivers I'll probably be getting this for my Gentoo box, especially if they are similar to Larrabee.
pezal said...
k10 1.6Gzh = 27sec
k10 1.8Gzh ~ 22 sec
k10 2.0Gzh ~ 17 sec
k10 2.2Gzh ~ 12 sec
K10 2.4Gzh ~ 7 sec
So for every extra 0.2Ghz the k10 completes 5 seconds quicker, that would mean :-
k10 2.6Ghz ~ 2sec
k10 2.8Ghz ~ -3 sec
Now thats some scaling :) , ramp up k10 speed high enough and it's completing tasks before you start them.
I don't know about anyone else but that's some impressive maths... not
usually i just like calling you an AMD fanboy, but today i get to call you something else. a moron. only a moron considers replacing a CPU "major manual labor". remove heatsink, pull lever, swap cpus, lower lever, put heatsink back on. I would asume the thermal grease on their would be fine if they're for office use. jesus christ. die.
to little to late...for intel graphics cards or chips
Yeah, because the 60% market share they have already is too little.
I think your mommy is calling, it's past your bedtime.
This is to all those posting on here that AMD is uber superior and all benchmarks are evil Intel scandal lies of doom deathness evil.
It's okay to like Intel now. I write this on an AMD, and my AMD gaming rig sits upstairs waiting for some ownage on FEAR, but they're not delivering for us anymore. AMD has ALWAYS been better than Intel. ALWAYS. Faster, Cooler, Better Overclocking. But it's over guys.
I was in your position. I posted over and over while C2D benchmarks were coming out how much they blew, and how much AMD was going to "totally ass pwn3n4t3 them Intel (which is actually latin for homosexuality) nubz0rz3z." C2D rocks. Fast, Cool, and sweet Jesus can they overclock. They overclock higher than I've ever seen anything overclock. I know it will take you time like it did for me, but you'll come to accept it. We've been e-soldiers for AMD for so long, we've lost sight of what we were fighting for. It's been like David and Goliath between the two companies, but for the first time in a long time, Goliath is the good guy. I hope to god AMD doesn't go under, but I'm not going to buy, are argue for an inferior CPU.
PS
Learn Math Jackass
before anyone (Any AMD fanboys) goes looking it up, i am aware that Intel is not actually latin for homosexuality
I'm sure what it is but I do know that "hilaris" is latin for "gay" but it is meant as cheerful, or merry, not gay as in homosexual.
I figured it out, Intel runs this website to further discredit anyone who supports AMD.
http://www.extremetech.com/article2/0,1697,2143301,00.asp
Article states:
That 1.6-GHz figure is true; I received that number from one of AMD's partners, along with my contribution to the story: "disappointing". And that's news, if only because three of AMD's partners used that exact word in describing the chip. One partner used this term: "chaos". And these guys had AMD boards in their booth.
"Whats with the mentality of an 8-years old? I really don't get it, does intel hires school kids to promote itself?"
LOL, I won't comment about the silly masturbators arguing about benchmark scaling vs frequency...
Give me a break...all of you idiots wish you were even in this business. Too bad you are all outside looking in and beating your sorry organ
guess what PhD, your goddess Charlie from Inq posted this on the forum...
Charlie
The ones floating at the show are so broken they are not worth benchmarking. They are stable enough to finish up platforms and BIOSes, which is what this round of samples was meant to do.
They are far from full clocked, half the FP/SSE resources are broken, and the memory controller is barely functional.
On the up side, when they get debugged, things will get better/faster, presumably by a lot.
On the down side, they are not out, not imminent, and Intel has a new product coming out.
-Charlie
enough to wake you up from a wet dream with K10 huh?
AMD got the PhD, INTEL doesn't need any body on the WEB to push their products. They got a multi-hundred million dollar marketing budget that has TV and printmedia. You guys should see the adds every day in the WSJ. They won't bother pushing to idiots like you guys. ANybody stupid enough to buy a Dell and have to upgrade it to run office applications needs no money wasted on him because he is too stupid to have any real disposble income. The idots is living in his mommas basement
K10.. AMD really blew it with the naming..
K8 to K10. Skipping K9 was a huge mistake. It would have been so appropriately named. Phenom will have a new meaning too to geeks next year.. LOL
GMA X3500 frags AMD 690G:
http://www.fudzilla.com/index.php?option=com_content&task=view&id=1362&Itemid=1
HD2900 XT is the fastest ATI card this year. Nvidia doesn't even need anything faster than their current products, yet the Geforce 9 is coming this year. http://www.fudzilla.com/index.php?option=com_content&task=view&id=1340&Itemid=1
As has been widely reported, Barcelona has been delayed to Q4. They can't even clock it above 1.6Ghz. Pathetic. http://www.fudzilla.com/index.php?option=com_content&task=view&id=1370&Itemid=1
AMD BK Q2'08.
Also from FudZilla which seems to be the AMD fanboi's site to pull from:
Intel is working hard on Penryn 45 nanometre core and it looks like that they are doing fine. The initial samples are out now and they work around 3 GHz.
Our sources have confirmed that there will be a new revision coming this summer that will enable even higher clocks and Intel partners are confident about this one.
It looks we will see Penryn before the end of the year.
Penryn could have a 2x higher clock than Barcelona!
Putting on my AMD hat for the moment, although the 1.6GHz news is very disappointing coming as it does at the end of 2Q07 when it should have being released, it is not the first time that AMD have had clock problems and eventually overcome then, indeed it seems a bit of a tradition. No doubt they will hit around 2.5 or so on release I think, though you have to wonder where that 3Ghz story/screenshot ever came from ?? Certainly nearer to Oz than Kansas you have to feel.
Another bright spot is that this chip does seem a lot more power efficient than the current K8 which is good news; this will be a relief for people who bought QuadFX systems. Lets hope they do not have to push the voltage up to get the clock speeds and instead can do it with engineering aided by IBM.
I'm still hopeful K10 can just push it's head in front of Intel just so it keeps this price war up and gives me a choice between two and not one, speaking as a selfish consumer of course, we have been rather spoiled in the last year.
Taking off my AMD hat.
I'm very disappointed on what AMD have been leading us to believe for the last few months. With the talk of 40% faster than Intel etc etc I had assumed that they had chips running at 2.5GHz+ and these were actual test results. It seems now these were more likely to be theoretical figures for a chip that was not yet capable of that. If that is true then I think AMD misled by ommission, which is not too good. They also misled people about the graphics line as well with the delay due to launching all parts at once, which did not happen. And of course everyone thought it would match the top end nvidia part, or better, which it failed to do.
The problem AMD faces and that Intel and to some extent nvidia do not, is that of high xpectations. Intel came from the Prescott era and when they delivered it took people by surprise and they got the kudos for it, nvidia did the same thing coming back from the FX range which was dire. On the other hand AMD are coming from the K7 and K8 and the R300 way back in time and peoples expectations are a lot higher, something AMD pumps up as well, so they have to deliver.
It would be better for them to be quiet and then give us "surprise and delight" rather than pumping us up and then giving us "surprise and disappointment".
It looks like both R600 and K10 will come good, but people expected them to be good from the start and currently this does not seem the case.
Yomama...
Got a link to that?
Thx.
I am requesting that the good doctor post his academic transcript online. Just a simple request... that's all!
I am requesting that the good doctor post his academic transcript online. Just a simple request... that's all!
the chances of that are like the chances of Pezal or Penix renouncing the AMD fanboi life they live...
Or to put it another way, the chances of Ph(ake)d doing that are the chances of Intel going BK 2q08.
I really hope you're not questioning the good doctor's credentials. I can't imagine anyone faking something like that.
Intel's Rumored Graphics Chip No-Show
One of the biggest rumors to come out of Computex this week is that Intel Corp. is developing a new high-end graphics processor, which it will begin sampling to graphics cards makers in Taiwan in the next year or two. Graphics card makers at the show said they've seen no signs of it, however. Geez! Another Intel rumor....
It has been known for quite some time that Intel plans to enter graphics market. What we learnt from computex is that they are planning on doing it much sooner than first anticipated.
Also Intel has said it won't be targeting the high-end but up to $300 market where most revenue is made. I think that might actually be a quite good idea. You won't have to make huge investments to beat competitors high-end GPUs. Only downside is you don't have the bragging rights.
Also Intel has said it won't be targeting the high-end but up to $300 market where most revenue is made.
Did Intel say that? Is Intel even talking about this chip?
It could just be speculation...
Stupid AMD fanboy Sharikou. YOu buy DELL PCs? Gosh, you must be ftarded
core2dude
"Did Intel say that? Is Intel even talking about this chip?"
Well, it is not exactly Intel itself but its vendors:
Computex 2007: Intel planning to enter discrete graphics card market in 2008
Post a Comment
<< Home