“Our basic strategy is unchanged” – ABT’s exclusive interview with AMD

Our interview began pretty early right about 5:30 AM Pacific Time and this editor was just finishing up the GTX 580 evaluation.  Here was an impressive flagship video card that had answered every one of the criticisms of the GTX 480 and it looked as though Nvidia was making an aggressive comeback.  After exchanging pleasantries, I started off by asking Mr. Ossias how the success of the GTX 460 and the release of the GTX 580 have perhaps changed AMD’s strategy now that they are no longer the only DX11 game in town.

Mr. Ossias replied that AMD always anticipates competition and does not vary from their strategy of targeting individual segments by giving the best value in their respective price ranges.  “AMD anticipates competition from Nvidia and does not need or intend to vary from our existing strategy. We have a consistent strategy of targeting individual segments and offering the best combination of performance, features and value in each of the respective price ranges. As an example of such value, AMD offers Eyefinity technology which delivers a qualitative improvement in the gaming experience.  So, AMD’s basic strategy is unchanged.”

I next asked about the DX11 breakdown into segments in Mercury’s report regarding net and notebooks.

Mr Ossias said that AMD had a big proportion but no specific figures were available and he pointed out the high unit volume of mainstream products with mobility Radeon.  “Regarding the DX11 share in notebook segment, according to Mercury Research, AMD has sold over 82% of all DX11 GPUs thus far but they do not break the percentage down specifically in the notebook market. AMD is confident that its share is at least this high in the notebook market and probably more as it has sold many millions of DX11 notebook GPU’s in the past year more than the competition.”

Part of AMD’s success with early penetration and domination of the DX11 market may be attributed to AMD in getting DX11 HW to game devs long before the HD 58X0 was released; it had started in April with fairly early silicon.  According to Mr. Ossias, “This allowed us to sample the developers almost immediately after receiving our first silicon from the foundry.”  AMD has got very good at writing graphics drivers for future products using advanced simulators.  “Early seeding of developers allowed us to enable DX11 content to go into development well in advance so that DX11 titles were available when our Radeon HD 5xxx series products hit the market later in the year”.  And this has given AMD a tremendous momentum in capturing and keeping over 80% of the DX11 market.

I next asked for comment on DX11 and “extreme tessellation”.

Mr Ossias said that AMD had a “well-balanced product”.  Since the game devs have been using AMD HW to create DX11 games for about 20 months already, AMD’s upcoming HD 6000 series will be well-suited for these games and for rendering practical tessellation in game applications.

We next talked about mobile discreet graphics and I asked about the external GPU.

Mr. Ossias said that several products have it but the connector needs to become an industry standard.  Many times a product needs an installed base of users to drive it forward.   He then explained how AMD was looking to offer a good range of notebooks for all users with their Vision branding.

We moved on to Fusion and Mr. Ossias said that raising the graphics bar would make it more attractive to moving the industry forward.  “It will not only mean better graphics, but it will raise the level of expectation for consumers so that the devs rush in to satisfy it.”  It is important that the future of graphics will be an important part of the PC experience.  Eyefinity is an important part of it and “we simply haven’t been exploiting what graphics can do”.  As to Fusion, AMD is also going to embed a powerful graphics processor into the silicon.

We concluded our short interview by touching on the relativity of performance of integrated (IG) vs. discrete graphics.

“There remains a significant gap between internal graphics in existing platforms compared with the high performance expectations created by very popular (high volume) upper-end discrete products like the Radeon HD 5800, HD 5700 and new HD 6800 products.  As the graphics bar is raised, the apps will target faster graphics as their baseline and the overall user experience will benefit.”

At that point, a cell phone battery got dangerously low and we concluded the interview.  I wish to give my thanks to Mr. Ossias for allowing ABT to interview him for our readers.  Check our conclusion.