Tom's Hardware Forums » CPU & Components » CPUs » Nvidia get into CPU's
 

Nvidia get into CPU's




Word :   Username :  
 
Bottom
Author
 Thread : Nvidia get into CPU's
 
Other Forum Pr!ck
Profile: nimble knuckle
More Information

http://www.tomshardware.com/2008/0 [...] u_market_/

Some of us predicted this would happen in a number of previous threads.

Looks like they are looking at the low end to start.

This is good news ... fabless no less !!

Since they have no scruples pinching technology from anywhere and anyone, and have an aggressive business model I hope they get what they deserve.

I imagine the first thing they will do is buyout a benchmarking outfit and make sure their new cpu's get the highest scores ... lol.

Then they will start e-mailing marketing firms in a week and claiming top spot.

Intel will need to take the physics company they bought off the ice and start them working on a decent GPU.

AMD will ... what will AMD do??

Comments?


---------------
Great spirits have always encountered violent opposition from mediocre minds.

Flood attempt detected. You cannot post this message.

Related Product

Register or log in to remove.

Profile: Ancient Poster
More Information

They are trying to get into the same market that ATI was in when AMD bought them. The mobile handheld market. Not a desktop. Dout they could penetrate that market easily.


---------------
http://valid.x86-secret.com/cache/banner/381014.png
Resident Jerk
Profile: Faithful Poster
More Information

Intel already has a GPU in the works.

It's GPGPU capable, the entire point in buying Havok was to screw over ATI/nVidia (who were using Havok to work on their GPGPU / Integrated PPUs on their GPUs).

It was supposedly going to debut Q408 at 45nm but speculation is now putting it at H209 at 32nm.


---------------
TeamBAG Member
Other Forum Pr!ck
Profile: nimble knuckle
More Information

I think you underestimate Intel cnu ... they know their current GPU is low end .. they realise the future is in platforms with multiple performance points.

They have tons of cash and could go from low to high end in a matter of months ... well a year or so they could come up with something.

I imagine nether Intel or AMD is happy and they lawyers are now going to work overtime at all of the cross-licensing agreements.

Could be be good for the market though.


---------------
Great spirits have always encountered violent opposition from mediocre minds.

Flood attempt detected. You cannot post this message.

Resident Jerk
Profile: Faithful Poster
More Information

Reynod wrote :

I think you underestimate Intel cnu ... they know their current GPU is low end .. they realise the future is in platforms with multiple performance points.

They have tons of cash and could go from low to high end in a matter of months ... well a year or so they could come up with something.

I imagine nether Intel or AMD is happy and they lawyers are now going to work overtime at all of the cross-licensing agreements.

Could be be good for the market though.





How was that underestimating them? I said they have a GPU in the market that was originally slated for Q408 and is now supposedly pushed back to H209 at 32nm because apparently the scope of it was insane and would cause issues at 45nm.

They have integrated solutions right now, they are working on a discrete solution.


---------------
TeamBAG Member
Other Forum Pr!ck
Profile: nimble knuckle
More Information

They can put a lot of energy into a 45nm solution cnu ... quickly. Why would you think they would ditch a process that quickly that is very successful? A0 booting silicon wasn't it?

They have the basics with the gpu tech already.

I never thought the Havock buy was an asset denial acquisition ... the timing suggests both they and AMD are well aware of the cpu move.

Can't believe I am talking up Intel ... slap me !


---------------
Great spirits have always encountered violent opposition from mediocre minds.

Flood attempt detected. You cannot post this message.

Resident Jerk
Profile: Faithful Poster
More Information

Reynod wrote :

They can put a lot of energy into a 45nm solution cnu ... quickly. Why would you think they would ditch a process that quickly that is very successful? A0 booting silicon wasn't it?

They have the basics with the gpu tech already.

I never thought the Havock buy was an asset denial acquisition ... the timing suggests both they and AMD are well aware of the cpu move.

Can't believe I am talking up Intel ... slap me !



I don't think it was just to keep nvidia/AMD away from physics either, I was just making a point.

They already have one failed video card, I doubt they want to have another failure and are putting everything into this one. It also could be that it will be GPGPU capable and they are holding off until some thing is out that supports that. I'm not speculating, I'm going off of news reports.

http://beyond3d.com/content/news/565

http://arstechnica.com/news.ars/po [...] oduct.html

http://arstechnica.com/news.ars/po [...] erbee.html


---------------
TeamBAG Member
"Waffle's Revenge"
Profile: enthusiast
More Information

Hopefully they do something more interesting than clone x86 chips. Like buy digital. Competing in the big iron market and creating a new platform for gaming: The VaX-Box

Profile: Honorary Poster
More Information

Nvidia won't venture into Intel's main territory any time soon. They simply don't have the manufacturing power, technological edge to make it a competition. Heck, I don't even think in 3 years time, Nvidia will be able to compete with AMD. Like BM said a couple months earlier, the barrier of entering CPU market is simply too great. (however, this does not mean monopolization)

To pitch Nvidia against Intel, I guess we won't see that at least 5~10 years later.


---------------
http://i62.photobucket.com/albums/h82/TXSuperFly03/478x88copy.png
Profile: stranger

Considering the raw power of nvidias tesla-system, a CPU from nvidia could make intel afraid of the dark... so to speak...

Profile: Honorary Poster
More Information

However, Tesla is not suitable for desktop use. Its RISC based architecture, and just like Intel's Itanium, incapable of running x86 efficiently.

Also, considering the heat Tesla system outputs, as well as the power it consumes..... :sarcastic:

It'll be a long long while before Nvidia can be even remotely considered as AMD's competitor, let alone Intel's.


---------------
http://i62.photobucket.com/albums/h82/TXSuperFly03/478x88copy.png
Profile: stranger

Underestimation, (if that is even a word :-p) is never good.

Are Intel and AMD sisc..? just wondering, think I heard that a while back... Everything starts somewhere.

I belive that Nvidia will, within a year, release a desktop processor, that will be quite competitive. :-)

Profile: Honorary Poster
More Information

Thomaseron wrote :

Underestimation, (if that is even a word :-p) is never good.

Are Intel and AMD sisc..? just wondering, think I heard that a while back... Everything starts somewhere.

I belive that Nvidia will, within a year, release a desktop processor, that will be quite competitive. :-)



Intel and AMD's products are hybrids of CISC and RISC, but primarily CISC.

In order for Nvidia to compete in this sector, they will have to:

1. Obtain a X86 license (which is the hardest part)
2. Obtain enough resources
3. Hire enough engineers
4. Develop enough technologies (patents) to catch up to AMD and Intel.

This is not underestimation. Speculating that Nvidia will be able compete with Intel and AMD within a year is overestimation. It will take them considerable time to even catch up with AMD.

I believe Nvidia will stick to mobile sector, at least for a while, before they accomplished enough to venture farther into the water.


---------------
http://i62.photobucket.com/albums/h82/TXSuperFly03/478x88copy.png
Resident Jerk
Profile: Faithful Poster
More Information

yomamafor1 wrote :

Intel and AMD's products are hybrids of CISC and RISC, but primarily CISC.

In order for Nvidia to compete in this sector, they will have to:

1. Obtain a X86 license (which is the hardest part)
2. Obtain enough resources
3. Hire enough engineers
4. Develop enough technologies (patents) to catch up to AMD and Intel.

This is not underestimation. Speculating that Nvidia will be able compete with Intel and AMD within a year is overestimation. It will take them considerable time to even catch up with AMD.

I believe Nvidia will stick to mobile sector, at least for a while, before they accomplished enough to venture farther into the water.



Pretty soon they can just buy AMD.. Once Intel has a graphics card nVidia wouldn't have a "monopoly." Buy AMD, sell off ATI and have a CPU division.. Just have to transfer the X86 liscense. Then just have TSMC produce 32nm K11s with Tesla die on the package for parallelism.. Hah!

Seriously someone needs to sue nVidia for faulty business practices considering SLI will ONLY work on an Nvidia Chipset. I know it's a driver issue, hacked drivers work fine. Someone email the EU, they won't listen to Americans.


---------------
TeamBAG Member
Profile: Honorary Poster
More Information

Well currently Nvidia is not jeopardizing any company in EU, so I guess they wouldn't care.

Seriously though, Nvidia have really faulty business practices. Its getting really ridiculous.


---------------
http://i62.photobucket.com/albums/h82/TXSuperFly03/478x88copy.png

Go to: