From ATI to AMD back to ATI? A Journey in Futility @ [H]

Not without an IP license they won't

this probably means the AMD graphics partnership may have more to it than what was announced so far

Indeed, I do feel like there is more going on than just supplying Intel gpu. If Intel is serious about competing with Nvidia, I do see an eventually RTG buyout from Intel.
 
It's clear they inked some serious licensing deal with AMD already. We don't know the details but imo it means that AMD and Intel have seriously joined forces.


nah I wouldn't think that far ahead. What Intel needs to do is get a proper GPU design. They already have AMD licensing for patents they need to create their own GPU. With Raja's help they can do that from the ground up. What they needed was an engineer that was capable for thinking 2 steps ahead. They didn't have that before for the GPU side of things. Its going to take them time? Yeah, proabably 2 generations or 3 but this would be the ultimate goal. They don't want to be tied down to AMD when AMD graphics are looking the way they are against nV's. Its not a competition in the graphics realm. There is only 1 company right now and that is nV.
 
Now this should make Nvidia shake in its boots.


Why? Nvidia is 2 gens in from of AMD, and light years ahead of Intel. Whats there to be scared of. They have a virtual monopoly in DL. Where Intel and Google are much bigger companies but can't seem to do much. Shit google is forced too use nV products anyways even though they have their own products, nV's are better.

The fruits that Raja and Intel bare, will be seen many years down the road. Its not intimidate. By then we have no idea of what the land scape will be in the GPU markets. If anything, nV will continue to be stronger then now, if things go the way they have been.
 
nah I wouldn't think that far ahead. What Intel needs to do is get a proper GPU design. They already have AMD licensing for patents they need to create their own GPU. With Raja's help they can do that from the ground up. What they needed was an engineer that was capable for thinking 2 steps ahead. They didn't have that before for the GPU side of things. Its going to take them time? Yeah, proabably 2 generations or 3 but this would be the ultimate goal. They don't want to be tied down to AMD when AMD graphics are looking the way they are against nV's. Its not a competition in the graphics realm. There is only 1 company right now and that is nV.

They have no agreements for licensing GPU tech from AMD, cite a source if you say otherwise. CPU side, yeah they have cross licensing agreements. Raja can give them a direction but they are so far behind it's not funny so were talking years of development and what could become another failed try at the discrete market by Intel.
 
Yeah, Intel's integrated graphics sucks, aside from being very good in low-power modes.

Licensing just tells you HOW to do something conceptually. It doesn't tell you in detail, which is why just Licensing stuff is meaningless. It's why Intel did jack with their licensed Nvidia patents.

So let's assume we take Intel seriously the third time around? Intel is five years away from challenging AMD, and ten years from challenging Nvidia. Patents won't help with that, because they lack the experience.

We'll see if they have that kind of patience this time round.
 
Last edited:
They have no agreements for licensing GPU tech from AMD, cite a source if you say otherwise. CPU side, yeah they have cross licensing agreements. Raja can give them a direction but they are so far behind it's not funny so were talking years of development and what could become another failed try at the discrete market by Intel.


They have IP agreements from before, the same agreements they had with nV were replaced by AMD, Either they go with AMD or Stick with nV, they have to have one of them, otherwise they CAN'T even make an IGP. Simple as that.

Not when Intel can push 1 billion a quarter into R&D for graphics. nV can't even do that. How many billions was their contra fund? just need a fraction of that lol.

Don't forget, Phi even though not great in graphics IS capable in DL (the only thing holding it back is software), its pretty much just as good as anything AMD has made so far.

So they have the capability to do compute workloads very well, just have to figure out the graphics side of it out.
 
Last edited:
Based on Kyle's article, I feel there is already bad blood between RTG and AMD when Raja still helming RTG. With how Vega launch and his so called break and then exit of RTG, one cannot help to think that is he force out of his position.


If that's the case amd has to worry about poaching of their current engineers to this new intel gpu arm, that will no doubt be flooded with money.
 
If that's the case amd has to worry about poaching of their current engineers to this new intel gpu arm, that will no doubt be flooded with money.

AMD already signaled that their GPU division is kaput with firing Raja in the first place. (After failing to fund it properly for 6+ years)
 
Do we know if it's a firing? Maybe Raja himself decided it's time to go.
Either way there was extremely bad blood between Raja and Lisa Su (most likely over funding priorities), it might indeed be mutual.

The fact is, without Raja, AMD's GPU division is basically a goner, the entire reason why Intel has failed for so long in GPUs is due to not having a lead engineer like Raja to make a GPU worth a damn (despite having basically unlimited resources).
 
If that's the case amd has to worry about poaching of their current engineers to this new intel gpu arm, that will no doubt be flooded with money.
Indeed, with any bad blood, usually it involves more than 1 person. I won't be surprise if there are people who are loyal to Raja decides to join him at Intel.
 
Indeed, with any bad blood, usually it involves more than 1 person. I won't be surprise if there are people who are loyal to Raja decides to join him at Intel.


Yeah this is a problem for AMD if there is.

On the other side of things though, Intel is probably pretty happy this happened.
 
They have IP agreements from before, the same agreements they had with nV were replaced by AMD, Either they go with AMD or Stick with nV, they have to have one of them, otherwise they CAN'T even make an IGP. Simple as that.

Not when Intel can push 1 billion a quarter into R&D for graphics. nV can't even do that. How many billions was their contra fund? just need a fraction of that lol.

Don't forget, Phi even though not great in graphics IS capable in DL (the only thing holding it back is software), its pretty much just as good as anything AMD has made so far.

So they have the capability to do compute workloads very well, just have to figure out the graphics side of it out.

They can manufacture that same IGP as long as they make no changes to it, make any change and they can no longer sell it without some kind of a agreement. Intel has also stated and so has AMD that no IP agreements have been made between them. Billions in funds wont generate you a superior product just cause you push tons of cash into it. Takes many trial and errors before you actually get a product and usually the first one is still pretty rough around the edges that you sell to the public. Phi is a different animal and it's possible Raja can help them with that pretty quick. But making it a gaming card, I just dont see it.
 
They can manufacture that same IGP as long as they make no changes to it, make any change and they can no longer sell it without some kind of a agreement. Intel has also stated and so has AMD that no IP agreements have been made between them. Billions in funds wont generate you a superior product just cause you push tons of cash into it. Takes many trial and errors before you actually get a product and usually the first one is still pretty rough around the edges that you sell to the public. Phi is a different animal and it's possible Raja can help them with that pretty quick. But making it a gaming card, I just dont see it.


Raja already has the trial and errors under his belt, experience, stated that already, they need the experience at the top level, a leader that knows what is going on. Others just have to follow. Those billions will hire other experienced people fast, in a couple generations we might not even be talking about AMD in the GPU wars anymore because Intel can serious put a dent in their ability to get more talented engineers.

You think Intel is satisfied with their current IGP? Oh and they can use current and future nV patents too for many more years which are needed for their GPU's, So you are right they don't need jake from AMD.

Money is needed to fuel the development process, to get the right people in the right places. AMD doesn't have that, that is what Raja was complaining about 6 months to a year ago, and trying to get funding for RTG properly, but AMD had to focus on CPU's only. Hence why we see the mediocre GPU product lines for the past 2 generations. This will continue with Navi most likely. What did I say we hear all this about AMD hiring PR/Marketing, PM's, and Product Managers, that isn't what they need, they need engineers with the proper skills to be competitive, they lost A LOT of their senior engineers when Raja left the first time, when AMD screwed the pouch by cutting GPU's, going with smaller designs, it worked for them because well nV stood still with their Tesla architecture. That will not happen anytime soon now. nV knows Intel wants to get into HPC/DL badly, Phi is about a generation behind in these compared to nV's GPU's, if they slow down the performance advantage will be lost, this is what is keeping developers on nV's side, because they are familiar with CUDA and they can get more performance out of it with out too much work. With Intel its common programming retinues for them too, as its similar to programming on a CPU, so no performance advantage for nV means now developers can go to Intel without much worry about getting the most for their time.

Intel has been afraid of nV ever since the AI markets started to expand. 2 generations more wait, will give nV a cushion that will be hard for even Intel to break if things continue the way they are. its hard for them right now, 2 more generations might be the like what Qualcomm and Samsung did to nV in the cell phone market. nV had a better product for a generation in cell phones, but they were out classed by Samsung and Qualcomm when it came to corporate capabilities due to their market share and relationships. Better to start now then later right?

Can't come into the market late without spending more money then the competition did and the products must do better then the competition too. nV had better products then Qualcomm and Samsung, but they didn't have the cash to go up against them.
 
Raja already has the trial and errors under his belt, experience, stated that already, they need the experience at the top level, a leader that knows what is going on. Others just have to follow. Those billions will hire other experienced people fast, in a couple generations we might not even be talking about AMD in the GPU wars anymore because Intel can serious put a dent in their ability to get more talented engineers.

You think Intel is satisfied with their current IGP? Oh and they can use current and future nV patents too for many more years which are needed for their GPU's, So you are right they don't need jake from AMD.

Money is needed to fuel the development process, to get the right people in the right places. AMD doesn't have that, that is what Raja was complaining about 6 months to a year ago, and trying to get funding for RTG properly, but AMD had to focus on CPU's only. Hence why we see the mediocre GPU product lines for the past 2 generations. This will continue with Navi most likely. What did I say we hear all this about AMD hiring PR/Marketing, PM's, and Product Managers, that isn't what they need, they need engineers with the proper skills to be competitive, they lost A LOT of their senior engineers when Raja left the first time, when AMD screwed the pouch by cutting GPU's, going with smaller designs, it worked for them because well nV stood still with their Tesla architecture. That will not happen anytime soon now. nV knows Intel wants to get into HPC/DL badly, Phi is about a generation behind in these compared to nV's GPU's, if they slow down the performance advantage will be lost, this is what is keeping developers on nV's side, because they are familiar with CUDA and they can get more performance out of it with out too much work. With Intel its common programming retinues for them too, as its similar to programming on a CPU, so no performance advantage for nV means now developers can go to Intel without much worry about getting the most for their time.

Intel has been afraid of nV ever since the AI markets started to expand. 2 generations more wait, will give nV a cushion that will be hard for even Intel to break if things continue the way they are. its hard for them right now, 2 more generations might be the like what Qualcomm and Samsung did to nV in the cell phone market. nV had a better product for a generation in cell phones, but they were out classed by Samsung and Qualcomm when it came to corporate capabilities due to their market share and relationships. Better to start now then later right?

Can't come into the market late without spending more money then the competition did and the products must do better then the competition too. nV had better products then Qualcomm and Samsung, but they didn't have the cash to go up against them.


Sometimes you crack me up, you have been bashing Raja since Vega benchmarks were known along with all the RTG engineers and now AMD is screwed cause they might lose the engineers you say are no good. Raja is a good engineer no doubt but no one engineer makes or breaks anyone but I think he may help focus Intel instead of their usual all over the place approach they try. I also think Intel is quite aware their IGP sucks. AMD does in fact lack the same resources as Intel, wont argue that but they gave Raja a budget and he knew it was going to be less then what he wanted. AMD has done fine on the gpu side they just have been unable to make a halo product like the 1080ti but it's more of a niche market. Their bigger issue has been being late to the market on their products and allowing Nvidia to have months of sales without having a competing product. That has been the true curse of using HBM, it took much longer to ramp up production then they thought. We know too little about Navi to make any judgments on it's success or failure to compete. I actually expect them to approach their gpu issues very similar to how they did Ryzen, might work great or not, will be a new approach to GPU design I think. AI development is a new field, tho one I think will eventually face severe government regulation as the fear of Skynet starts to settle in with the masses. I agree Intel will be a bigger foe then AMD in that market. Intel has wanted their finger in every market, however they have lacked the focus to get that done and often have delivered less then stellar results, very hard to be nimble as a big corporation. It will be interesting to see what Intel does but I think were several years from anything meaningful being released.
 
Sometimes you crack me up, you have been bashing Raja since Vega benchmarks were known along with all the RTG engineers and now AMD is screwed cause they might lose the engineers you say are no good. Raja is a good engineer no doubt but no one engineer makes or breaks anyone but I think he may help focus Intel instead of their usual all over the place approach they try. I also think Intel is quite aware their IGP sucks. AMD does in fact lack the same resources as Intel, wont argue that but they gave Raja a budget and he knew it was going to be less then what he wanted. AMD has done fine on the gpu side they just have been unable to make a halo product like the 1080ti but it's more of a niche market. Their bigger issue has been being late to the market on their products and allowing Nvidia to have months of sales without having a competing product. That has been the true curse of using HBM, it took much longer to ramp up production then they thought. We know too little about Navi to make any judgments on it's success or failure to compete. I actually expect them to approach their gpu issues very similar to how they did Ryzen, might work great or not, will be a new approach to GPU design I think. AI development is a new field, tho one I think will eventually face severe government regulation as the fear of Skynet starts to settle in with the masses. I agree Intel will be a bigger foe then AMD in that market. Intel has wanted their finger in every market, however they have lacked the focus to get that done and often have delivered less then stellar results, very hard to be nimble as a big corporation. It will be interesting to see what Intel does but I think were several years from anything meaningful being released.


I never bashed Raja, said that some things he did were wrong lol. Actually I complemented him quite a few times, he is from my home time. Might want to look back again who was bashing whom. I wasn't one of the guys that bashed Raja.

Yeah again, just lumping every one together as saying something because you believe they are the same thinking people, not a good way to make a point.

AMD doesn't have the engineering man power to do anything much right now in graphics, this is well know. This is why they only have ONE GPU team that did both Polaris and Vega at the same time. You want me to do a google search on how many of them left AMD 7 years ago? You can do it and find out the AMD GPU crew is a skeleton screw just to get things by.

AMD's budget for GPU's is crap you know it, I know it, everyone knows it. 500 million a quarter for both CPU and GPU piled up for 3 + years most likely 5 years, and they were barely able to catch up to Intel, barely, because they haven't caught up. Delays to products, where does that come from? Hmm not enough money to do things faster? Having more teams to have development timelines overlapping each other.

Please you think Polaris looks good against Pascal midrange? Its crap, it sucks down how much more power? Its uses as much power as a gtx 1070, but performs like a gtx 1060! Yeah generation behind. Vega looks worse because GCN doesn't scale well.

Please AI, skynet, true consciences in computers is decades away, maybe even a century or more. The computational power to do what our minds do and the speed we do it is just not possible any time in the near or mid future, long long future. Need to achieve singularity before machine consciousness and even that optimistically 50 years away.

You make assumptions about things like Skynet, this isn't the movies man, it takes a lot more work then just putting chips together and tada you have self evolving computers. If I was on crack I still wouldn't believe the stuff you are talking about lol.
 
I never bashed Raja, said that some things he did were wrong lol. Actually I complemented him quite a few times, he is from my home time. Might want to look back again who was bashing whom. I wasn't one of the guys that bashed Raja.

Yeah again, just lumping every one together as saying something because you believe they are the same thinking people, not a good way to make a point.

AMD doesn't have the engineering man power to do anything much right now in graphics, this is well know. This is why they only have ONE GPU team that did both Polaris and Vega at the same time. You want me to do a google search on how many of them left AMD 7 years ago? You can do it and find out the AMD GPU crew is a skeleton screw just to get things by.

AMD's budget for GPU's is crap you know it, I know it, everyone knows it. 500 million a quarter for both CPU and GPU piled up for 3 + years most likely 5 years, and they were barely able to catch up to Intel, barely, because they haven't caught up. Delays to products, where does that come from? Hmm not enough money to do things faster? Having more teams to have development timelines overlapping each other.

Please you think Polaris looks good against Pascal midrange? Its crap, it sucks down how much more power? Its uses as much power as a gtx 1070, but performs like a gtx 1060! Yeah generation behind. Vega looks worse because GCN doesn't scale well.

Please AI, skynet, true consciences in computers is decades away, maybe even a century or more. The computational power to do what our minds do and the speed we do it is just not possible any time in the near or mid future, long long future. Need to achieve singularity before machine consciousness and even that optimistically 50 years away.

You make assumptions about things like Skynet, this isn't the movies man, it takes a lot more work then just putting chips together and tada you have self evolving computers. If I was on crack I still wouldn't believe the stuff you are talking about lol.


You attacked AMD engineers and everyone knows that, not new, you might have been slightly less harsh on Raja but you still harped on him as well. AMD did in fact shrink their engineering teams a well known fact but i would not call them a skeleton crew, marketing took a much bigger cut. When your a small company you get used to doing more with less, not always easy and a product failure can really cripple you like we saw with Bulldozer. They have done quite well with a situation that has left plenty of companies in the history books. You know the biggest delay was HBM production and Polaris is fine, most people dont give a shit about wattage use of a video card an OEM company will but not the average consumer. Vega 56 is a great card and why Nvidia launched the 1070ti, while the Vega uses more power it's just as fast and once again the difference on power is just not something most anyone will care about except miners. Vega 64 I agree is just not that great, eats far too much power to at best = a 1080. As for your thoughts on AI the government doesn't give a shit and many people do in fact fear a Skynet like issue. Otherwise you wouldn't have Tesla, Facebook and others talking about the impacts of AI. We have no clue how close or far away we are from true artificial intelligence until a program starts modifying it's own code in a unintended way. Likely it's quite a few years off but mankind has been known to have a breakthrough or two over the years. I expect regulation to appear as this becomes a larger topic and the more fear the general population has for it. Hollywood has nothing to do with the tech but has tons to do with stoking the public's fears.
 
We'll have to see what comes out of this but I will give credit where it's due. Considering what Raja had to work with during his time with AMD, he pushed out some decent products. Obviously it wasn't on the level of Nvidia based hardware but Nvidia has a lot more manpower & resources to work with. Raja will now be in a position where he'll have a lot more breathing room & resources to work with himself. He should have very little trouble surrounding himself with what he needs. Including a team of people of his choosing if Intel allows him to do so. Also, AMD removed his profile page from their site lol - http://www.amd.com/Pages/PageNotFou.../corporate-information/leadership/raja-koduri
 

Oh, my...

ATI guys must be even more pissed now.
Why would they be pissed? He's integrating AMD GPUs into an Intel platform and bulky cases with discrete graphics are going the way of the dinosaur. As Raja stated, people don't seem to have grasped the magnitude of this yet and AMD and Intel appear to just be getting started defining the PC market.
 
Why would they be pissed? He's integrating AMD GPUs into an Intel platform and bulky cases with discrete graphics are going the way of the dinosaur. As Raja stated, people don't seem to have grasped the magnitude of this yet and AMD and Intel appear to just be getting started defining the PC market.

We agree on something!!
 
...bulky cases with discrete graphics are going the way of the dinosaur. As Raja stated, people don't seem to have grasped the magnitude of this yet and AMD and Intel appear to just be getting started defining the PC market.

Hooray for console-like performance out of an extremely small form factor PC?

Meh.
 
Sometimes you crack me up, you have been bashing Raja since Vega benchmarks were known along with all the RTG engineers and now AMD is screwed cause they might lose the engineers you say are no good. Raja is a good engineer no doubt but no one engineer makes or breaks anyone but I think he may help focus Intel instead of their usual all over the place approach they try.

LOL! It just goes in the opposite direction. I remember all those fancy posts about the "dream team" (Keller+Koduri+Papermaster) and how Intel and Nvidia would be scared. It was all hype and BS of course. Then Keller left abruptly (due to discrepancies with Su) and the same posters that created the hype and BS pretended that Keller left AMD friendly because had finished his work... Now Koduri leaves (again due to discrepancies with Su) and the same posters are trying to dismiss again this departure.

We get yours is another instance of the typical double-standard attitude from the ADF. Koduri was considered a semi-God when he was inside AMD, now he has moved to Intel, then he is little more than a third-class manager.

AMD is a shrinking boat and smart people is moving away. Some of you can continue pretending that all is ok inside AMD and pretend that Intel/Nvidia are scared, but numbers show otherwise.

Vega fiasco is not Koduri's mistake. It is Lisa Su's mistake, who redirected resources from GPU division to CPU division, among other weird decisions. Koduri and his team cannot do miracles without the needed "time" and "money". And he will get plenty of time and money now at Intel.

I said it years ago, Lisa Su will kill the company. It is happening slooowly.... ;)
 
Last edited:
You attacked AMD engineers and everyone knows that, not new, you might have been slightly less harsh on Raja but you still harped on him as well. AMD did in fact shrink their engineering teams a well known fact but i would not call them a skeleton crew, marketing took a much bigger cut. When your a small company you get used to doing more with less, not always easy and a product failure can really cripple you like we saw with Bulldozer. They have done quite well with a situation that has left plenty of companies in the history books. You know the biggest delay was HBM production and Polaris is fine, most people dont give a shit about wattage use of a video card an OEM company will but not the average consumer. Vega 56 is a great card and why Nvidia launched the 1070ti, while the Vega uses more power it's just as fast and once again the difference on power is just not something most anyone will care about except miners. Vega 64 I agree is just not that great, eats far too much power to at best = a 1080. As for your thoughts on AI the government doesn't give a shit and many people do in fact fear a Skynet like issue. Otherwise you wouldn't have Tesla, Facebook and others talking about the impacts of AI. We have no clue how close or far away we are from true artificial intelligence until a program starts modifying it's own code in a unintended way. Likely it's quite a few years off but mankind has been known to have a breakthrough or two over the years. I expect regulation to appear as this becomes a larger topic and the more fear the general population has for it. Hollywood has nothing to do with the tech but has tons to do with stoking the public's fears.


Sorry man the entire Vega Line up is crap, we all know it, if you are trying to say anything of Vega can hold a candle to Pascal, well that is truly delusional.

I have stated this over a year ago, Raja DID NOT MAKE VEGA! the design for Vega started well BEFORE Raja came back! GCN was NOT RAJA's design! Raja was not at AMD when GCN was made and released. Not only that I even stated Raja will be fired if Vega fails, but it won't be his fault! Go look back and see what I posted, instead of making wild accusations!

https://hardforum.com/threads/vega-rumors.1933112/page-74#post-1043135536

From July of this year

Raja is really good, if I remember correctly he was one of the principle engineers for the r300, plus he is from my home town in India, so kinda favor him for that ;) He left AMD due to the budget cuts as many of the other sr. engineers did too. Can't do anything without a proper budget. And I think he was promised a lot when he went back to AMD. So its not something they can probably easily torch him because of Vega. Shit any one in his position knows HBM was a waste of resources, but once its there, he can't do anything, his hands were tied. Everything so far points to one thing, not enough money. Ryzen might turn that around but its going to take time.

https://hardforum.com/threads/vega-rumors.1933112/page-62#post-1043122002

Another same time frame

https://hardforum.com/threads/vega-rumors.1933112/page-31#post-1043082352

another

https://hardforum.com/threads/i-don...accountable-just-yet.1935933/#post-1043035045

another


Should I keep posting older posts?

Or do you just want to keep making things up!

Wtf, keep babbling about shit that I have never stated. Just shows you don't even know who said what on this very forum.

And the AI stuff, please tell me more lol, like how computers can create consciousness, when we have no fuckin clue why we have consciousness lol. Like a computer can just have feelings of life and death, without us knowing anything about these things to program a computer to understand these things. This isn't a yes no problem, its feelings. If we can't understand it why we have feelings, which we really don't know why, how the fuck are we going to make a computer to have them? You are talking like the first AI researchers from the 50's, most AI researchers now, know its not a possibility. AI can be programmed to do one or many different individual tasks better then humans, but they can't do is truly think and evolve, because why would a computer need to do that? It has to understand and have feelings to push itself. Why do humans get better at doing things? Because we want to do better, for certain reasons, the feelings of wanting to know more, the exhilaration of achievement. Many other things, power, money, you name it. Why would a piece of silicon crave for these things? It needs to have feelings. Skynet is BS, just because it can correlate between different systems doesn't mean its got feelings lol.
 
Last edited:
You guys are priceless. This is like when the View had John McCain on. 8 years ago he was the great satan but now that he's bad mouthed Trump they love him.

Raja joining Intel is not going to do a darn thing to AMDs graphics efforts. By the time you reach the top you are contributing very little to technical innovation. On the other hand you need to be a good leader, public speaker and champion of your team. Yet this sub forum has been flooded with the usual suspects railing on him for having terrible presentations and drastically overhyping late and disappointing products.
 
LOL! It just goes in the opposite direction. I remember all those fancy posts about the "dream team" (Keller+Koduri+Papermaster) and how Intel and Nvidia would be scared. It was all hype and BS of course. Then Keller left abruptly (due to discrepancies with Su) and the same posters pretended that Keller left AMD friendly because finished his work... Now Koduri leaves (again due to discrepancies with Su) and you are trying to dismiss again this departure.

AMD is a shrinking boat. Some fo you can continue pretending that all is ok inside AMD and pretend that Intel/Nvidia are scared, but numbers show otherwise.

Vega fiasco is not Koduri's mistake. It is Lisa Su's mistake, who redirected resources from GPU division to CPU division, among other weird decisions. Koduri and his team cannot do miracles without the needed "time" and "money". And he will get plenty of both now at Intel.

I said it years ago, Lisa Su will kill the company. It is happening slooowly.... ;)

Not like she walked into a well greased machine. I expected AMD to die off a long time ago but here they are still kicking.
 
Not like she walked into a well greased machine. I expected AMD to die off a long time ago but here they are still kicking.


They were damn close to it 2 years ago, still close not as bad though lol.
 
Did you mean to leave out #wait4navi and #RajaNaviFTW ?
:p


I think at this point we can say #navicoffinnails

I think they are going to go back to small die strategy and not compete with nV in the high end not till they get enough resources to do both GPU and CPU R&D properly.
 
Why would they be pissed? He's integrating AMD GPUs into an Intel platform and bulky cases with discrete graphics are going the way of the dinosaur. As Raja stated, people don't seem to have grasped the magnitude of this yet and AMD and Intel appear to just be getting started defining the PC market.
Because they already feel they fell behind because of AMD and now Intel will confront them on the whole GPU spectrum as well, with an R&D budget for each segment that dwarfs AMD's total.
 
Last edited:
Because they already feel they fell behind because od AMD and now Intel will confront them on the whole GPU spectrum as well, with an R&D budget for each segment that dwarfs AMD's total.

If this is true - I'm friggin excited. Some high powered competition!!!

I've been waiting for this for years, actually, if you're right. I'm only afraid it'd be aimed at distributed compute. Although building their own Xbox units (CPU+GPU at least) and being vertically integrated makes a ton of sense. Higher usage of their fabs as well resulting in lower overhead. Overall more competitive.
 
If this is true - I'm friggin excited. Some high powered competition!!!

I've been waiting for this for years, actually, if you're right. I'm only afraid it'd be aimed at distributed compute. Although building their own Xbox units (CPU+GPU at least) and being vertically integrated makes a ton of sense. Higher usage of their fabs as well resulting in lower overhead. Overall more competitive.
Going forward under Koduri’s leadership, the company will unify and expand differentiated IP across computing, graphics, media, imaging and machine intelligence capabilities for the client and data center segments, artificial intelligence, and emerging opportunities like edge computing.
Don't know, but this sounds pretty broad to me and hints at gaming too. Maybe I'm just reading too much into it.
 
Its going to take time, at least 2 gens 3-5 years, from now before we see or know anything about whats Intel's take. In the meantime though I expect them to focus on compute.
 
Back
Top