Editor's Comment - June 27, 2011 at 10:55AM UTC:
Following the original article, we got contacted by multiple people involved in the recent BAPCo turmoil, who heatedly debated the claims set forth by controversial statements made by AMD-related sources. We are bringing a new article online which will detail "the other side of the coin." As a side note, we're preparing yet another interesting story closely tied to this article. Stay tuned.
The statements made in the article in no way reflect the opinions of Bright Side of News* nor the members of our publication.
Following our coverage on AMD's exit from BAPCo and blog post made by Nigel Dessau, we got a surprising call from the person at the heart of AMD which we had to check out. After the end of an eye opening conversation, we started calling our sources in order to confirm if the claims made by an obviously disappointed engineer hold any substance. We talked to our usual sources inside the company, as well as with a number of sources at their key partners and customers. The odd part was that all of our contacts said the same thing - the story checks out. Thus, we bring you the modestly edited version of our conversation.
AMD's BAPCo Exit is a Smokescreen
First and foremost, we started the discussion over the blog Nigel Dessau, AMD's Senior Vice President and Chief Marketing Officer wrote, stating clear reasons why AMD decided to leave the BAPCo and why AMD considers SYSmark 2012 an invalid benchmark.
"When I read Nigel's blog and saw the press release from BAPCo it made me sick because our CMO talks about transparency and honesty and it's all smoke and mirrors. At the end of the day, we actively had internal teams and external organizations hired to promote/discredit SYSmark. Not because it was inaccurate, but because it is accurate. Back in the original Athlon 64 and Opteron days, when we were winning in SYSmark we were heavily promoting it in the public sector, who in turn used it as a benchmark on which they based many of their purchases on. It was us who actually got BAPCo and SYSmark inside several government tenders to win orders measured in tens of thousands of systems. SYSmark was used to show how our K8 processors were beating Intel's NetBurst."
In the heat of the moment - BAPCo SYSmark 2012 GUI
As you can read here, the source alleges that AMD first had internal teams and organizations in order to promote SYSmark, and then turned those teams and organizations to discredit the benchmark when it didn’t work in their favor. The interesting part is if this holds true, it turns out that AMD was the architect of its own fall, since the company allegedly pushed BAPCo's SYSmark benchmark into many governmental tenders.The source moved on to say "The minute we started losing performance leadership, which was Woodcrest/Conroe/Merom timeframe was when we started to put together active groups that would discredit SYSmark."
Our source solely lays the blame on weak CPU performance set forth by Barcelona (K10) and Shanghai (K10.5) architectures. We all remember that a "cocktail from hell" happened in 2006, when Intel released Conroe (Core 2), Merom (Core 2 Mobile) and Woodcrest (Xeon) and only several months before that, AMD's executives Dirk Meyer and Hector J. Ruiz decided that 90nm K8 is selling like hotcakes and the company will keep the 90nm and ramp up the 65nm nine months after the original plan. This backfired spectacularly when Barcelona encountered a bug in silicon (the infamous Transition Lookaside Buffer bug) which sparked further delays.
Our source claims AMD had a solid position in BAPCo and that BAPCo's board accepted vast majority of suggestions for SYSmark and MobileMark benchmarks to the tune of 75-83%. But BAPCo could not fix the following: "the problem was that hound our processors performance was poor, the CPU performance wasn't there at the commercial level this significantly hurt our ability to win public sector and large enterprise business around the world."
Following the arrival of Conroe/Merom/Woodcrest, AMD regrouped and started to dedicate resources into internal and external teams that would actively work on discrediting SYSmark and MobileMark benchmarks similar to tactics that ATI and NVIDIA employed in the GPU wars.
Our source lays blame on former and present AMD leaders like Hector Jesus Ruiz, Derrick "Dirk" Meyer, Nigel Dessau and Patrick "Pat" Moorhead. According to the source, in one meeting with marketing and engineering teams present Nigel allegedly stated "how do we get rid of SYSmark." The gathered staff was surprised since the rationale was how do you go and tell the governments and large enterprises in Europe, Asia and North America that SYSmark doesn't matter. The valid question is "How can AMD tell the world governments and large enterprises that the tender for tens of thousands of computers needed to have SYSmark off the list, when we were one of the companies that lobbied for that benchmark to be in there in the beginning."
Over the past decade, we learned of multiple deals taking place where AMD contested that the metric used by some governments had a mandatory result which not even the fastest AMD commercial platforms could beat. All that time, AMD marketing and sales team were continuously pushing the engineers to focus on tactics to discredit benchmarks rather than creating new silicon to win. One of examples was government and large enterprise contracts where MobileMark is used. The source argues that "we could not compete with Kite Refresh in '07, Puma and Congo in '08... neither could we compete with Tigris in '09 and we're still challenged with Nile and Danube."
- 2007: Kite Refresh - Tyler CPU, M690T Chipset
- 2008: Puma - Griffin CPU, RS780M Chipset
- 2009: Congo - Conesus CPU, RS780M Chipset; Tigris - Caspian CPU, RS880M Chipset
- 2010: Nile - Geneva CPU, RS880 Chipset; Danube - Champlain CPU, RS880 Chipset
Our source is dismayed at the longevity of the chipsets and alleged lack of innovation as well as lack of will to invest in removing the bugs which plagued consumers such as long standing issues with USB 2.0 performance.
Corporate Culture, Lost
Given that the main source and other sources we used to confirm this story are still at or affiliated with the company, we then moved on to ask what happened to the corporate culture of AMD. After all, I professionally started to follow IT industry at the tail end of 1990s and had that luck to talk on numerous times with Jerry Sanders, company's founder and CEO. Former executive team featured prominent names such as Giuseppe Amato, Henri Richard and Pierre Brunswick from the sales and marketing side, and Dave Orton, Raja Koduri, Bob Drebin as brilliant engineers on another.
There were a lot of insiders that compared AMD to Ferrari, Ducati, Lance Armstrong (unsurprisingly, the company ended sponsoring all three of them) carefully nurturing "one against all" and "innovate to survive" mentalities.
While Intel struggled with NetBurst architecture and went on to create Yonah, a 32-bit architecture that was the foundation for Conroe/Merom/Woodcrest, while NVIDIA fall to their knees with NV30 i.e. GeForce FX and went to out-innovate the competition by creating the first GPGPU chip, the G80 (also known as NV50) - "AMD started to spend resources on how they could skew benchmarks in their favor."
This was a stark departure from the AMD mindset that saw that K5 cannot compete with Pentium, went on to buy NexGen and signed Derrick "Dirk" Meyer, young engineer credited as co-architect for brilliant DEC Alpha 21064 and 21264 processors. AMD K6 and K6-2 and K6-III served as stopgaps while Dirk's team was working on K7. When K7 or Athlon came out, it was the first x86 processor to beat Intel's Pentium II and Pentium III architecture, starting the famous "Clock Wars" race to 1GHz, won by AMD's "Magnolia", pushing Intel into releasing Pentium III 1.13GHz, which was infamously recalled after Tom's Hardware discovered a bug.
With K8, AMD introduced a 64-bit x86 instruction set and forced Intel to respond with Yamhill, a 64-bit set of extensions compatible with AMD's 64-bit set.
The author of these lines believes that the key reason that pushed AMD into regression was the failed merger between NVIDIA and AMD in 2005. This lead to a panic acquisition of ATI Technologies, for which AMD paid a grand total of 5.9 billion dollars (5.4B plus 350 plus 120 million).
This emptied the coffers and engineering team was left without resources to go and fight against the rising Intel threat. AMD's reaction after the departure of Henri Richard was appointing Nigel Dessau, seasoned executive with IBM and SUN Microsystems behind him. Nigel was faced with an impossible task of competing against Intel with blanks, and the only choice left was to deploy the Fear, Uncertainty, Doubt i.e. FUD strategy patented by his previous company, centennial Big Blue. A shift occurred and all engineers were left was not making a discussion that would go along the lines of "ok, how do we make our architecture better?", "how do we work to regain performance leadership?". According to one of our highly positioned sources, the culture switched to "how do we discredit benchmarks and skew the numbers?"
What happened disappointed our source "No, AMD has done nothing to actually innovate. Instead, what we are doing is discrediting the benchmarks in which we had a hand in creating. This is the saddest part." If you survived until this part, you have read what three different sources have said to us. We continue on the next page with the real reason for AMD's BAPCo Fallout.
The Beef: Why AMD Attacked BAPCo Now - Bulldozer CPU Performance
If you are wondering what lies in the timing of this announcement, our sources were pretty clear.
"Do you know why AMD chose now to go after SYSmark now?”
We expected that the source would say Llano APU performance etc., but no: "Because Bulldozer performance is sub-par. Because after such a long wait Bulldozer will not be a world beater. They knew that for another generation, we would come up way short on benchmarks like SYSmark. Removing ourselves from BapCo was the best way to make the benchmark look like it's no longer valid in the hope that prospective or existing customers stop using it."
Engineering Sample of AMD Bulldozer Processor Clocked at 2.8GHz. Picture credit: Pink Pig
Now, BSN* had received scores from B0 and B1 silicon and we decided not to run the story disclosing exact performance details since they plainly... were far below what we expected. After talking to journalists from other publications and developers that had access to engineering silicon, everyone was lead to believe that Bulldozer has a lot of "performance locks" which were set in place to prevent "premature benchmarculation" i.e. leaks.
"Bulldozer is going to disappoint people because we did not get the resources to build a great CPU, and it's not that we needed billions of dollars to make it a leader. We needed investment in people, tools and technology."
As every competitive spirit, one of our sources went on to confirm the quote above and expressed concern that "Bulldozer’s client performance is not going to be enough to deliver growth in the enterprise and in my view, that is the biggest driver behind our current strategy."
When asked about core performance, surprising information was that a Bulldozer core versus the existing cores in Llano will result in minimal improvements overall. Our sources went on to say that the launch of Llano clearly shows what is the current and future strategy - downplay CPU performance every chance you get. Everything has to revolve around the GPU.
Comments continued "As soon as you have analyzed the Bulldozer core, you're quickly going to find out that it disappoints at SYSmark. It won't matter, because by that time I am sure the current marketing strategy will be changed again when we have our new CEO in place."
"What We Want from AMD"
As the article draws to a close, we aren't going to comment on the following statement made by yet another source:
"There are more people like me at AMD that want to go and fight from an engineering perspective in the same spirit of the old Jerry Sanders days. We don't want an AMD that is going off and writing blogs about transparency and fluff, while engineers are frustrated. Engineers are not getting the resources they need, we're not getting the time and the investment we to build and optimize new architectures.
Instead, they feed marketing fluff and kool-aid, and all these self-centered blogs - instead of focusing on doing what we know how to do best. Fight, survive and challenge against all odds.
Thankfully, we'll have a new CEO soon and hopefully all this strategy nonsense will be out the door."
A Message to the new CEO
Furthermore, this is a personalized message to AMD's new CEO:
"Many of our best engineers are leaving to join companies like Apple, Freescale, Qualcomm, Samsung even NVIDIA and Intel - people that thought they would never work for another company.
Look at AMD's history - how we beat Intel Pentium III, which was based on one of the best CPU architectures of all time - the Pentium Pro - and we beat the them with the K7 (original Athlon). Look at K8, 64-bit innovation and the integrated memory controller. Look at how ATI came back from the brink and beat by NVIDIA soundly with R300 and took performance leadership across the board.
We created the netbook in OLPC project with Nicholas Negroponte when everybody was laughing at him and then gave the whole market to Intel. We were shipping millions of graphics chips for cellphones and were in TVs long before today’s leaders. We know how to win."
At the end of our conversation, we asked our main source what does he plan to do, was this rant his farewell with AMD or not. The answer was plain and simple: "I was here before they came, I am here with them now and I am going to be here well after they’re gone. My heart and soul, and more importantly my legacy is with this company and will continue to be."
Time will tell, but as the proverb from my former life says - all stories come true.