Mark Roulo (nothanks@xxx.com) on 8/4/10 wrote:
---------------------------
>Richard Cownie (tich@pobox.com) on 8/4/10 wrote:
>---------------------------
>>But being a pioneer doesn't always lead to a sustainable
>>business. Right now NVidia just announced an
>>unexpected $150M drop in revenue for the quarter - that's
>>20% of their business that vanished overnight. And by
>>mid-2011 we'll have adequate CPU+GPU chips from both
>>Intel and AMD which will simply obliterate NVidia's
>>revenue from sub-$100 discrete GPUs. And in the $100-$200
>>range NVidia still has nothing at all that matches
>>the Evergreen series on features (DX11, Eyefinity) or
>>price-performance. A situation that's probably only
>>going to get worse with AMD's Southern Islands arriving
>>in Q4 2010 and Q1 2011.
>
>Without arguing whether or not the sub-$100 GPU market can vanish and be replaced
>by on-chip GPUs, does anyone have a feel for this question:
>
>

>If the sub-$100 discrete GPU market goes away, can nVidia survive on the high
>end (and, I suppose, embedded) chips alone? Or do they need the volume of the low
>end to justify the ongoing development?
>

>
>-Mark Roulo

I think that's the correct question to ask, along with one other that David Hess mentioned:

How much are OEMs willing to spend for incremental (2X, 3X, 4X, 8X) graphics performance over an IGP? And how much is the BOM for a low-end discrete solution including the chip cost, board cost, DRAM cost, design cost, etc.?

Since that question will literally define the low-end over the next few years. If the answer is that the minimum cost of a discrete solution is $20, and OEMs demand 4X performance for $20 of cost...then things could turn quite ugly.

There's a forum post at Beyond3D that I think is somewhat insightful to answering your question:

http://forum.beyond3d.com/showthread.php?t=48843&page=94

At the bottom of the page, NV's units by price bin are listed for desktop discrete:

<100]: 5.7M, 66%
[100,200]: 2M, 23%
[200,300]: 0.32M, 3.7%
>300: 0.47M, 5.5%

NV has roughly equal volume in non-desktops and desktops. However, I expect the notebook price stack to be heavily compressed, since nobody buys $300 GPUs for a notebook. I would expect even more of the revenue to be concentrated in 0-200. And the [0,60] range (or perhaps it's [0,40]) is the one that's vulnerable to sandy bridge and fusion.

David