Does it really look like a competition for Intel?... When I read the article, my first reaction was: now they will control the whole "processing power". That's what people really spend the money on right now (in home desktops at least).
nforce + cpu + gfx == thin-client made by nvidia?
nforce + cpu + gfx + some cheap stuff (storage is cheap and home users don't really care about the sound) == PC based gaming platform?
Or if someone likes conspiracy theories: while working on their gfx chip they discovered something that's easily portable to x86 and will make a massive difference. Now they're secretly producing a new x86 chip that will change the cpu market as we know it...
The defiant tone of the author suggests to me that this was a bit of a screed piece by someone who feels threatened by any move that could change the balance of the wintel duopoly. I agree, Intel needs challengers.
Hmmm. I don't think the problem is just with design. The problem probably lays with fabrication.
Intel is large enough to both upgrade their fabs and CPU design alternate years. Fabs are becoming more capital intensive to build each year. Furthermore, Intel can afford to make mistakes. This is the reason why AMD is being killed - its 45nm process was way behind Intel's. AMD is now being split up into a foundry and CPU design company (but the future is bleak).
nVidia will never be a serious competitor – it is a fabless semiconductor company (its stuff gets manufactured by TSMC).
The looming Intel monopoly will be worse than Microsofts monopoly. IMHO the only hope for competition is for for companies such as Via to manufacture low-end CPUs at foundries.
Competition is great.. but if they are planning to go head to head with Intel, that is a recipe for failure. Hopefully that have something novel in mind.
Nvidia has a ton of great chipset makers whom have routinely produced stellar, incredible hardware. Intel has far more x86 experience, but Nvidia is no slouch. I don't see how they could possibly fail to produce a chip worth selling at SOME price point. It is only a matter of time before they develop the expertise to sell a chip at ANY price point.
I've wondered how one (company) gets into a really high tech industry. Wouldn't they have to do an awful lot of research - basically reinventing most of Intel's wheels for the past decade?
I guess hiring an x86 processor development team would give them a jump in, but without lots of existing work, documents, schematics, etc. would a team of people know enough to rebuild such complex work from scratch?
10 comments
[ 4.1 ms ] story [ 16.2 ms ] threadnforce + cpu + gfx == thin-client made by nvidia?
nforce + cpu + gfx + some cheap stuff (storage is cheap and home users don't really care about the sound) == PC based gaming platform?
Or if someone likes conspiracy theories: while working on their gfx chip they discovered something that's easily portable to x86 and will make a massive difference. Now they're secretly producing a new x86 chip that will change the cpu market as we know it...
Intel is large enough to both upgrade their fabs and CPU design alternate years. Fabs are becoming more capital intensive to build each year. Furthermore, Intel can afford to make mistakes. This is the reason why AMD is being killed - its 45nm process was way behind Intel's. AMD is now being split up into a foundry and CPU design company (but the future is bleak).
nVidia will never be a serious competitor – it is a fabless semiconductor company (its stuff gets manufactured by TSMC).
The looming Intel monopoly will be worse than Microsofts monopoly. IMHO the only hope for competition is for for companies such as Via to manufacture low-end CPUs at foundries.
Nvidia has a ton of great chipset makers whom have routinely produced stellar, incredible hardware. Intel has far more x86 experience, but Nvidia is no slouch. I don't see how they could possibly fail to produce a chip worth selling at SOME price point. It is only a matter of time before they develop the expertise to sell a chip at ANY price point.
http://www.theinquirer.net/inquirer/news/162/1017162/nvidia-...
http://www.theinquirer.net/inquirer/news/103/1031103/nvidia-...
I guess hiring an x86 processor development team would give them a jump in, but without lots of existing work, documents, schematics, etc. would a team of people know enough to rebuild such complex work from scratch?