AMD Talked to NVIDIA Before Acquiring ATI: Report

[Wile E]

Yup Nvidia should of taken the merge deal, since Crossfire was auto supported by Intel from the get go. Now they are lacking in the Glue Logic dept, SLI/CUDA/Physx isnt doin so well like NV intended it to do.

CUDA and SLI are doing very well in their intended markets at their intended market share percentage. Just look at Tesla.

GPU powered Physx, not so much.

Nvidia is doing just fine, and is in no way hurting from passing on the deal.

Say what about multi gpu before AMD/ATI Uhmm Do us a favor look up ATI 128 Rage Fury Maxx

Heck your the only that has said he has had bad...yeah it was YOU not the company or product.....

3dfx still beat them both to the market with multi-card setups. Feb '98 they introduced SLI on VooDoo2 vs Oct '99 for the Maxx. but really, that's just picking nits. Who had it first is pretty much irrelevant in today's market.

And, actually, I had a bad time with Crossfire (and a handful of other driver bugs), too. That's why I went with single gpu nVidia this round. May try SLI at some point to get a feel for the other side of the coin if another cheap 580 comes my way, but a single powerful GPU will do for now.

But, would also happily go back to AMD if they offered what I want in a single gpu at my next upgrade point, and I knew the bugs were squashed.

@Wile E
I beg to differ. There are many examples how customer satisfaction is something obscure and hypothetical. Also the thing called marketing is something even more devious and suspicious. If you studied it or just read about it you should know that it is invented just to sell products to people that don't want them. Advertising is also another tool in this bag of "magics", but I agree that it is needed in some occasions - new company that nobody knows, or completely new product that people should be aware of. But nowhere they teach you to make products that actually people need(please note "need" not "want"). And all the other "deeds" companies do to make you "believe" and walk you to the store. Fanboism was born because of that. That's why I like when companies don't overuse these "magics" and let people think for themselves if they really need or not their products. This commercially driven society we live in is something very strange and I don't like the trend. So that's why I like to think for myself what I "need" not "want". Sure from time to time I oblige to my "urges" of new "toy" but try to keep it "real". This is my opinion and I don't expect you to think likewise, just wanted you to understand my point of view. Like you say "Keep it real bro"

There is no real need in the markets we are discussing. Almost nobody NEEDS powerful computers. We just want them. It's all based on consumer wants. If a company doesn't produce what the consumer wants, and doesn't make the majority of their customers happy, they cannot be successful, and cannot turn a profit. These companies are not forcing people to buy their goods, so the idea that they do nothing to satisfy their customer base is a false notion.

 

[BlackOmega]

CUDA and SLI are doing very well in their intended markets at their intended market share percentage. Just look at Tesla.

GPU powered Physx, not so much.

The reason PhysX is not doing so well is because they're trying to force anyone who uses it to buy an Nvidia card as their primary. In their arrogance, they're effectively killing off PhysX because who the hell is going to take the extra time to code for it when a relatively small percentage of people are going to use it? Probably not many that aren't reimbursed for doing so by Nvidia.

Aegia had a great idea with PhysX, Nvidia bought them and have now effectively screwed it up.

 

[TheoneandonlyMrK]

Aegia had a great idea with PhysX, Nvidia bought them and have now effectively screwed it up.

defo

you can buy a gpu off them to do compute but try to use its compute with an amd gpu and its not officially allowed, boll4x and pointless, their is allways going tobe a way so why bother and its an ocassional(due to few physx games) lost sale

 

[BlackOmega]

defo

you can buy a gpu off them to do compute but try to use its compute with an amd gpu and its not officially allowed, boll4x and pointless, their is allways going tobe a way so why bother and its an ocassional(due to few physx games) lost sale

Exactly. There's always workarounds to anything that's limited by a driver. Hell, back in the Nforce 4 days, you could even mod the hardware to run SLI even where Nvidia had laser cut it.

Also, with PhysX, when it was still allowed with an AMD GPU as the primary, I actually bought a decent midrange card specifically to run it. Ever since they disabled it through drivers, I haven't bought another Nvidia card for the sole purpose of running PhysX.

Not to mention, their advertising scheme is losing momentum since there's less and less games that have it. Another notch in their fail belt.

 

[eidairaman1]

Maybe in the Science world but the Consumer market SLI is slowly disappearing and NV cant make a Standard Glue Logic anymore

CUDA and SLI are doing very well in their intended markets at their intended market share percentage. Just look at Tesla.

GPU powered Physx, not so much.

Nvidia is doing just fine, and is in no way hurting from passing on the deal.

3dfx still beat them both to the market with multi-card setups. Feb '98 they introduced SLI on VooDoo2 vs Oct '99 for the Maxx. but really, that's just picking nits. Who had it first is pretty much irrelevant in today's market.

And, actually, I had a bad time with Crossfire (and a handful of other driver bugs), too. That's why I went with single gpu nVidia this round. May try SLI at some point to get a feel for the other side of the coin if another cheap 580 comes my way, but a single powerful GPU will do for now.

But, would also happily go back to AMD if they offered what I want in a single gpu at my next upgrade point, and I knew the bugs were squashed.



There is no real need in the markets we are discussing. Almost nobody NEEDS powerful computers. We just want them. It's all based on consumer wants. If a company doesn't produce what the consumer wants, and doesn't make the majority of their customers happy, they cannot be successful, and cannot turn a profit. These companies are not forcing people to buy their goods, so the idea that they do nothing to satisfy their customer base is a false notion.

 

[Prima.Vera]

LOL


I am very glad AMD didn't get in bed with NVIDIA...

ATI and AMD make a great couple, both are failing companies that I could careless about other then to create competition for NVIDIA and Intel.

I personally wouldn't/couldn't run a GPU without CUDA and can imagine running a CPU other than Intel

WOW! Fanbuoyisms at is finest!!

Classic!

 

[Wile E]

Maybe in the Science world but the Consumer market SLI is slowly disappearing and NV cant make a Standard Glue Logic anymore

I'm willing to bet that SLI and CUDA make more money in the science market anyway.

Either way, doesn't change the fact that they are successful. It's not like if the merger happened they would only go after teh consumer market. Mergers include all their markets.

 

[Totally]

Why is everyone surprised? It's common knowledge. I know I'm late in saying this but, this isn't exactly news, I pretty sure there was an article outlining the same exact things here posted a few years ago at the time of ATi aquisition. I can't find the article on here but did find an article where Tom's revisited the subject.

http://www.tomsguide.com/us/amd-nvidia-merger,review-1061-4.html

 

[largon]

AMD certainly got less than it paid for with that five-point-four billion dollars. Graphics division of AMD is pretty much insignificant to the company's financials. All AMD got was graphics chip design teams. They could've bought S3 instead, for a lot less and dump most of the cash on a new fab.

Ergo, ATi buy-out was a bad deal for AMD.