July 4th, 2007
AMD posts blatantly deceptive benchmarks on Barcelona
Updated 8/16/2007 - AMD caught red handed
[UPDATE 7/6/2007 - AMD intends to drag their feet on removing the deceptive benchmarks]
After AMD admitted the bad news last week that their Barcelona product would be late and underwhelming on clock speed, the AMD propaganda machine has gone in to hyper drive with the latest salvo of blatantly deceptive benchmarks. After claiming to have the high-road on ethics, AMD showed hypocrisy on three separate occasions (one, two, and three). But this latest round of deceptive benchmarks is so outrageous that it’s criminal [UPDATE 7/5/2006 - Criminal might be too strong of a word. This is blatant deceptive advertising at the very least.]
[UPDATE 7/5/2006 - Despite the fact that AMD has plausible deniability that they created these simulated benchmarks much earlier in the year, the benchmarks were confidential earlier in the year and they're now making its rounds in the blogsphere and press as if it were news of a miraculous triumph by AMD. The AMD Barcelona spin campaign is in full throttle and this bogus information is posted prominently on AMD's Barcelona Product page. Fortunately, all this pressure has forced AMD to promise that they will remove the bogus benchmarks on their Website.]
On AMD’s “Barcelona” performance page, AMD shows the following fictitious and outdated information. Apparently some of these misleading numbers are even showing up on Wall Street Journal advertisements.
- It’s fictitious since AMD doesn’t have a 2.6 GHz Barcelona quad-core CPU and they won’t even have it in September which is already late by AMD’s original timeline. The fastest Barcelona processor coming out in September is 2.0 GHz. It isn’t really clear when AMD will be able to ramp up the clock speed an extra 30% to get to 2.6 GHz but it most likely won’t be any time soon because processors don’t just ramp 30% over night.
- The numbers AMD posted for Intel’s XEON X5355 and X5160 have been outdated since April 2007 and you need a magnifying glass to see that disclaimer in the fine print on the bottom. The actual up-to-date SPEC.org certified scores for the two Intel products listed are significantly higher. It not like AMD can claim that they forgot to include the very latest scores which were just posted days ago, we’re talking months here so it’s a blatant omission.
- Intel’s XEON X5365 3.0 GHz quad-core CPU which shipped back in April was deliberately omitted from these results even though AMD showed off numbers for a 2.6 GHz Barcelona chip which doesn’t even have a launch date yet. Putting in 2.0 GHz Barcelona scores would be shady enough since the part hasn’t officially launched yet but including 2.6 GHz Barcelona scores is just outrageous.
Full picture on quad-core SPECint_rate2006 performance:
* Not real product. Fastest Barcelona being released in September is 2.0 GHz
As you can see from above, AMD’s claim that they have a 20% clock-for-clock advantage with Barcelona is simply wrong. Based on the latest certified SPEC.org results, AMD has a little more than a 1% clock-for-clock performance advantage in a dual-socket 8-core Server configuration but they have 50% clock speed deficit when the Barcelona finally launches in September. That means Barcelona will not be the Intel quad-core killer that AMD has been promising for most of this year and it won’t even be close.
The deception doesn’t end with the quad-cores; AMD is also claiming to have an advantage on dual-core processors when in fact they have a major performance deficit. AMD claims to have a 2.5% advantage when Intel actually has a 14.7% advantage when you’re looking at the certified SPEC.org scores.
Full picture on dual-core SPECint_rate2006 performance:**
** UPDATE 7/5/2007 - Hans de Vries pointed out in the talkback that the AMD 2222SE has newer scores. The newer results have been added to this chart.
I’ve seen benchmarks get cherry picked and twisted before but this is just outrageous. AMD is deliberately leaving out Intel’s best scores, leaving out Intel’s best products that shipped months ago, and putting in theoretical Barcelona scores for products that don’t even have a ship date. After Henri Richard (AMD executive) came in front of our ZDNet cameras to slam Intel for “un-ethical behavior” and promising not to do the same, we have caught them on four separate occasions behaving unethically. After this latest incident, it’s clear that AMD has no intention of behaving honestly or ethically.
Updated 8/16/2007 - AMD caught red handed
Since there are a few websites and some individuals that continue to accuse me of wrongly accusing AMD of leaking the bogus simulated Barcelona 2.6 GHz benchmarks, I need to set the record straight. AMD in mid-July last month was still giving out those benchmarks to the press weeks after they had already removed them from their own website. Ironically that same week, I had an HD Video conference on July 20th with AMD’s public relations people and they slammed me for implying that AMD had leaked it to the press and that it was a Russian partner of AMD that distributed the benchmarks. But when TechArp’s Adrian Wong wrote a story using the same benchmarks, I informed Wong that he was using discredited benchmarks that AMD was forced to remove and Wong informed me that AMD had given it to him. Wong was not happy that he was given junked benchmarks from AMD and posted the following note in his Barcelona article:
Adrian Wong: George also pointed out that AMD removed the “simulated” benchmark results from their website on July 6th, and promised to post proper benchmark results of their Barcelona processor. However, we received the very same benchmark results just last week. If these results have already been officially junked by AMD, why are we still being served the same poo?
AMD initially denied they had given it to TechArp and told me that their European PR manager stated that they stopped distributing those benchmarks in April 2007. Later in the week on July 25th an anonymous source of mine sent me an AMD PowerPoint presentation that was being given out to partners and I confronted AMD with it. Four hours later AMD informed me that there was an AMD press event in Kuala Lampur on July 16 and “the simulated benchmark slide was included in the deck that was presented”.
I had originally decided to let this slide and not make another posting of it but it seems that a few individuals still want to smear my reputation for implying that AMD was responsible for leaking the slides to the press. Because of this I have no choice but to set the record straight that AMD was in fact responsible for distributing these discredited benchmarks even after they had been officially junked.
George Ou is Technical Director of ZDNet. See his full profile and disclosure of his industry affiliations.
