We come close to AMD's top brass, and key industry watchers, right at AMD's home ground in hot, dry Austin, Texas. What did they say?
With somewhat disappointing Bulldozer showing, and perceived wavering roadmap, there were comments in the industry circles how AMD will give up the main Intel competitor seat to someone like Qualcomm, or even cease to be in the CPU business altogether. This weeks' preview of the next-generation Fusion APU gave us a chance to mingle with AMD's top execs, see their new CEO and hear – unofficially, of course – what's the way forward.
The hot, dry Austin, in Texas, is the place where the top AMD management is, after all, so the chance of meeting high-level personnel at a press & analyst event is high. This time we're not talking more about Trinity yet, except that the thing does seem to perform well, and yes, there is a 17 W Ultrathin-class (read: UltraBook in Intel speak) Trinity with all the features enabled, i.e. two dual core modules and the full GPU – of course at a lower clock speed. That will make some interesting impact in that market later this spring…
Back to the main point: what are the problems facing AMD, where is the long term root of the problem, and how & when should we expect these problems solved?
Let's start from the GPUs – the are not a problem as of now. Put aside all the Nvidia gaming stuff in the GTX680, the HD7970, now in it's updated run, will keep its stand. As the second generation of these cards is expected to run at 1.1 GHz or so even at default from some vendors, it should keep up with the Nvidia offering well in games, while providing superior compute performance. According to the people around, the initial Sea Islands rollout is on target for late this year too, and AMD negates having 28 nm problems now.
The APUs aren't a problem either: the Trinity is expected to be sold out, whatever AMD can make – so, we come to the quantities question there. AMD gave assurances they can do many millions of Trinities this year, and, according to what was seen, the 'good enough' statement, as worn as it sounds, is right on target here.
Then we come to the CPUs, the real problem for years – I did overhear a discussion between the senior execs how did the whole thing turn out with the 'low IPC performance' Bulldozer instead of the other alternatives on the table, and how it will take some two years to completely turn around the switch to another much more IPC-efficient core architecture. The main culprit seems not to be the previous CEO, Dirk Meyer, but his predecessor Hector Ruiz, the 'Taco Bell' that replaced 'KFC goatee' Richard Sanders, the man who made AMD succesful.
They said that Sanders was so disappointed with Hector, saying once that 'he left Hector a roadmap to execute that anyone, absolutely anyone' would be able to get done. Of course, Hector did damage the company in many ways, which by itself is a separate story.
The new CEO, the Lenovo-experienced Rory Reed, is a tall jovial guy who – literally – jumps to the podium with a smiling face of a typical Hollywood actor, and seems to be fond of certain catchphrases during his speeches. However, he does bring energy to the rest of the team, and the will to move forward – far more than the staid engineer that Dirk Meyer was, or the cunning character of Hector Ruiz who was giving himself multimillion bonuses when the company was losing money big time. His experiences and relationship with the Chinese might be of help to the company's expansion in China, which is inevitable if AMD wants to stay relevant.
Talking about moving forward, the feel is that AMD is there to survive and thrive once again. Rather than repeating the bland statements of 'we don't want to compete in the high end, just on volume' which make no sense since a fabless company cannot compete on volumes against a competitor with seven large fabs at its disposal, this time there's clear indication that, after Piledriver, there will be substantial changes in both cores and system architecture from Steamroller onwards, that should help make AMD competitive closer to the top. I was told that delaying the socket migration beyond the AM3+, C32 and G34 to new socket is a good move,since AMD can design more aggressive, rather than stop gap, sockets for future platforms with better features like more memory and HyperTransport channels, as well as integrated PCIe v3, for greater future scalability. For the first time, some execs do acknowledge that Bulldozer approach may not have been the best one at the time, and things need to change. I was told that there is some good frequency scalability in the Piledriver core which should help gain some per-core performance ground.
So, in the near term, AMD will use APU to keep its presence in desktop and mobile market, and even low power 5 – 10 W part derivatives or Trinity may arrive for high-end HD++ tablets. The CPU core radical refresh is expected to complete within two years from now, along with brand new socket platforms, proving AMD a new base from which to attack the high end, again, just like in the good old Opteron/Althon 64 early times. The GPUs will continue to be the crown jewel of the company till then, though… expect new high end mobile HD7900 series later this month, and Sea Islands by yearend.