cucker tarlsonWhat do you mean by that ?

StrayKATI wish I understood what you mean (sorry).. but that last sentence would be damn cool.

Every gpu purchase for me has been about remastering old titles in HD. Well, Directx 12 has no features. Everything is bog-standard.
Let's say, I think Intel could enable low-level-optimisation into the mainstream. Don't you think 'pc' is their closed ecosystem?

mtcn77Every gpu purchase for me has been about remastering old titles in HD. Well, Directx 12 has no features. Everything is bog-standard.
Let's say, I think Intel could enable low-level-optimisation into the mainstream. Don't you think 'pc' is their closed ecosystem?

Ah.. in HD. I know what you mean there. But I thought you meant more than that (as in, actual graphical remastering through some magic at the hardware level.. which would be awesome.. heh).

Really ?? Ohhhh im so excited to see all blue graphics and INTEL sticker on it !!

Intel can make efficient and powerful GPUs, ask any i7 5775C owner. That said, i'm not expecting the first product to be good.
They have to take care of software, very urgently.

oxidizedWell they might not be on track with 10nm, but their 14nm is actually even better than specifics, besides, their 10nm could be even on par with TSMC 7nm for all we know. Also, i'm not sure you understand Intel is worth more than AMD and nvidia put together + TSMC and GloFo as of now, they could open new fabs, or they might increase productivity of existent ones.

I understand that very well, but these are fabs we're talking about, not nuts and bolts factories. Fabs are simply not built within two years. They're not that easily extended either.

bugI understand that very well, but these are fabs we're talking about, not nuts and bolts factories. Fabs are simply not built within two years. They're not that easily extended either.

As long as there's money as driving force, nothing is a problem...Besides we can't know if they even need additional fabs, they handled their production quite good in all these years.

It will probably need a proprietary pci-e socket.

phanbueyIt will probably need a proprietary pci-e socket.

Knowing intel they will target pciex4 and 4 only to sell chipsets ,to be fair to intel , though very lately, they have changed gear in the driver's team, finally.

BlueberriesThere are a lot of brilliant minds at Intel and I strongly believe they wouldn't waste their time and money if they didn't think they could produce a competitive product.

That being said... I'm not expecting much. I don't think their first line of GPUs will have any substantial footing in a market against NVIDIA; but I wouldn't doubt it if their second or third iteration does. Intel has the power to build a small army, and given time, they will.

There have always been brilliant minds at intel. That didn't stop them from utterly failing both times they tried to do a dedicated GPU.

oxidizedAs long as there's money as driving force, nothing is a problem...Besides we can't know if they even need additional fabs, they handled their production quite good in all these years.

Money doesn't solve everything. See intel's woes with the 10nm process. Also, see how their first two attempts at dedicated GPUs turned out. They need people who can focus people and spending in the right direction.

This being said, something's really not right here. It really hasn't been that long since Raja defected from AMD and went on to intel. Building an architecture from the ground up takes a while. It seems that intel are in a hurry to ship this thing, and that's never a good sign.

Even if the hardware part will be well done, intel still has to invest a lot in drivers. Their IGP drivers are the worst thing since square wheels. And they'll have significantly less than two years to make the drivers work well (they have to have the hardware first). This is a big warning sign here.

The question is simply how fast it can manage - Intel already has GPU if you count the iGPU+onboard memory, except then it has hardly reached 1 TFLOPs FP32, where AMD is easily 2 TFLOPs north with their seriously nerfed 2400G.

Intel has understood the fanboism correctly.. Starting the hype train 2 years early..i wonder who will aboard to the Intel side in all the comment Section..

There must be some guyz telling that it will beat AMD & Nvidia already ..lol

Interesting choice of words, calling it a Graphics Processor, instead of a Graphics Processing Unit.
It brings up the possibility of modifications to existing compute cards to increase it's GPU capabilities.

For us who will be alive&kicking in 2020 things will be more cleal then..

scorpion_amd13This being said, something's really not right here. It really hasn't been that long since Raja defected from AMD and went on to intel. Building an architecture from the ground up takes a while. It seems that intel are in a hurry to ship this thing, and that's never a good sign.

Raja wasn’t brought in on the ground floor. This had already been under development for awhile before he left AMD.

The venue is telling: SIGRAPH over GamesCom. Intel's dedicated cards are going to be targeted at professionals over gamers. I would be shocked if Intel had any interest in maintaining gaming graphics drivers because it's a costly endeavor.

I’m wondering if the time until release (18+ months) is why it was at SIGRAPH. It’s always possible when they have a more finished product they may hit Gamescom too?

Maybe they will have multiple models?

Yes please. We need a 3rd competitor. I really miss the 3DFx, nVidia, ATI, Matrox, S3, etc, days.

rtwjunkieMaybe they will have multiple models?

It's possible but I'm still not convinced Intel's priority is the gaming segment. It's highly competitive and not their field of expertise. Intel, on the other hand, is very concerned about CPU orders falling in light of packing racks with SIMD compute cards. They can't stop that trend but they can offer a competing product to get some of that pie they're losing on the CPU side.

In fact, I wouldn't be surprised at all if Intel's card has no ROPs or TMUs (things mostly for gaming/rendering). Think Xeon Phi but instead of x86, it's a new SIMD architecture (a whole lot of shaders and not much else).

Intel did try making a deep learning/neural CPU and succeeded but in that process, they may have learned that the task is better suited for SIMD cards so they never released it to consumers knowing it can't compete.

Intel's discreet card may have tensor/neural features as well that wouldn't help with gaming.

RH92Hyping a product 2 years in advance ...what could possibly go wrong right ?

Two years or a year and 4 months aside, if this isn't an investor appeasement stunt, then I don't know what is. They lost some value on their stock, this is their attempt at recovering it.

FordGT90ConceptIt's possible but I'm still not convinced Intel's priority is the gaming segment. It's highly competitive and not their field of expertise. Intel, on the other hand, is very concerned about CPU orders falling in light of packing racks with SIMD compute cards. They can't stop that trend but they can offer a competing product to get some of that pie they're losing on the CPU side.

In fact, I wouldn't be surprised at all if Intel's card has no ROPs or TMUs (things mostly for gaming/rendering). Think Xeon Phi but instead of x86, it's a new SIMD architecture (a whole lot of shaders and not much else).

Intel did try making a deep learning/neural CPU and succeeded but in that process, they may have learned that the task is better suited for SIMD cards so they never released it to consumers knowing it can't compete.

Intel's discreet card may have tensor/neural features as well that wouldn't help with gaming.

I agree that it would make the most sense if Intel aimed for the professional market with their cards.
But then again, why mention DX12...?

Not getting onboard this hype train. They hyped 3DXpoint as well, and the reality hasnt thus far lived up to the promise. The densities are a bit underwhelming, we still dont have the 3DXpoint ram, and the price is too high. And then there's 10nm. And then there's basically no competition in terms of core count against Threadripper. instead we get a "we forgot to mention it was overclocked" and chilled to absolute zero high core count CPU marketing blitz.

What makes intel think it can compete in the dGPU space, when they cant deliver on their promises, and their on core GPU's are not that great?

mad1394Dedicated graphics market a 3 player game ? Best news I heard all year!

and on the same day we get news of 3 DRAM makers coming together to rake in profits by exploiting the consumer, :D

make it 5 competitors - for a truly healthy competition.

Well, it's better to have a backup plan in case your main and only product is no longer a monopoly.

GoldenXWell, it's better to have a backup plan in case your main and only product is no longer a monopoly.

also intel and nvidia hate eachother. so its nice to have bad blood for good competition. hehehe humans are tricky creatures

the DRAM makers are all friends ;)

Dr_b_Not getting onboard this hype train. They hyped 3DXpoint as well, and the reality hasnt thus far lived up to the promise. The densities are a bit underwhelming, we still dont have the 3DXpoint ram, and the price is too high. And then there's 10nm. And then there's basically no competition in terms of core count against Threadripper. instead we get a "we forgot to mention it was overclocked" and chilled to absolute zero high core count CPU marketing blitz.

What makes intel think it can compete in the dGPU space, when they cant deliver on their promises, and their on core GPU's are not that great?

Let them fail then, if you think they suck so much. :p

But they shouldn't stop before they even start. What if AMD had such a defeatist attitude 5 years ago?

KursahLarabee for Gaming Part Deux?

It'd be nice to see another player in the gaming graphics arena, please by all means! More competition is good! :)

Bring S3, PowerVR/STMicro/VideoLogic back along with Hitachi SuperH!