Intel ARC and the Latest Drivers: Does this bode Well for ARC's future?

[D1craig]

After the uproar about the rx 580 not supporting dx 12_1 are we really giving praise to something for getting better dx 9 performance?

 

[Gerard]

After the uproar about the rx 580 not supporting dx 12_1 are we really giving praise to something for getting better dx 9 performance?

Seems that way, after all it was only launched 21 years ago so on the cutting edge of apis.

 

[humbug]

Yup, now card pricing seems to be determined by how much cocaine the ceo's want to snort.

 

[0h_no_not_again]

Intel have lost so much money. I honestly fear they will throw in the towel. People keep saying, myself included, that Nvidia/AMD are taking the ****. However, A770 is not selling is it? It seems people would just pay up or just keep sitting it out.

I doubt they will. GPU's are strategically important, and getting more so in the modern world in both enterprise and consumer markets. They have seen the huge slice of pie they're missing out on and realised they need to compete. Further, if their gpu's fall further behind, it will absolutely start to hurt their market offerings of CPUs, SoC's and all the other lovely devices that are ubiquitous in _everything_ today.

 

[random_matt]

So what is the ARC performance at currently (770)? I take it around 3060 levels? I had a look a while ago but I've not been keeping updated on the progression.

And "early access" didn't exist.

DX 11 games were not amazing, my last run on Odyssey matched a Vega 64 more or less.

 

[random_matt]

Remeber the days when £250 to £350 would by you the top tier card.

Walked out of PC World (Did not know any better) with a newly released 4870 at £199.

 

[J.D]

DX 11 games were not amazing, my last run on Odyssey matched a Vega 64 more or less.

Hopefully in the not so distant future, this will improve.

 

[Angilion]

Intel have lost so much money. I honestly fear they will throw in the towel. People keep saying, myself included, that Nvidia/AMD are taking the ****. However, A770 is not selling is it? It seems people would just pay up or just keep sitting it out.

I don't think Intel will throw in the towel because I think Intel are mainly interested in Arc as a specific type of processor rather than as a consumer graphics card product. And I mean Arc, not specifically Alchemist. I think Intel's in it for the long haul and is after the bigger markets for that sort of processor architecture, not really for the consumer graphics card market. That market is in rapid decline and the one and a bit companies running it as a not-quite-monopoly are running it into the ground while extracting maximum profit margins from minimum customers. That's an approach for a dying market and it will hurry its demise. Intel would be daft to consider that as the key market for the future because that market doesn't have much of a future.

Alchemist doesn't cut it but it is very valuable experience and the fiasco with drivers has shown that Intel sorely needs that experience. If Intel can work Alchemist up to an adequate budget product (even if they overprice it and don't sell many - the key thing here is perceived functionality) then that and the experience gained would put Intel in a better position for the next gen, Battlemage. Which might be a genuine contender.

 

[Angilion]

Remeber the days when £250 to £350 would by you the top tier card.

I can remember when it would buy you the most game-changing graphics card ever made. A card that went beyond being top tier by creating an entirely new market in which it was the only "tier". 3Dfx Voodoo. In my case, the Orchid Righteous 3D, which was the first (and when I bought it the only) graphics card using the 3Dfx Voodoo processor.

£250 in 1996 would be more like £450 today, but that's still relatively cheap compared to the current situation.

 

[Author_25]

I don't think Intel will throw in the towel because I think Intel are mainly interested in Arc as a specific type of processor rather than as a consumer graphics card product. And I mean Arc, not specifically Alchemist. I think Intel's in it for the long haul and is after the bigger markets for that sort of processor architecture, not really for the consumer graphics card market. That market is in rapid decline and the one and a bit companies running it as a not-quite-monopoly are running it into the ground while extracting maximum profit margins from minimum customers. That's an approach for a dying market and it will hurry its demise. Intel would be daft to consider that as the key market for the future because that market doesn't have much of a future.

Alchemist doesn't cut it but it is very valuable experience and the fiasco with drivers has shown that Intel sorely needs that experience. If Intel can work Alchemist up to an adequate budget product (even if they overprice it and don't sell many - the key thing here is perceived functionality) then that and the experience gained would put Intel in a better position for the next gen, Battlemage. Which might be a genuine contender.

THe A750 and A770 do feel more like a beta test than an actual launch, except getting people to pay to join it, as there has been no real marketing of the card, or push for sales, and the only area where there is any activity is on the driver side, and appears to be the main avenue for Intel, on Arc.

Its almost like the hardware is there to enable improvements to the software.

 

[eeii]

I can remember when it would buy you the most game-changing graphics card ever made. A card that went beyond being top tier by creating an entirely new market in which it was the only "tier". 3Dfx Voodoo. In my case, the Orchid Righteous 3D, which was the first (and when I bought it the only) graphics card using the 3Dfx Voodoo processor.

£250 in 1996 would be more like £450 today, but that's still relatively cheap compared to the current situation.

ahhhhh the Orchid Righteous 3D that takes me back.

 

[Cyber-Mav]

Intel making rapid progress on the driver side. May be worth getting an arc gpu seeing as how prices are dropping now too.
Got to remember it took amd/ati over 25 years to sort out opengl driver issues so intel is heading in the right direction

 

[KompuKare]

It seems that very fine isn't red wine but blue then

Got a bit excited about Skyrim:

As it's near double there but then I remembered, Skyrim SE is DX11 only the old binaries are DX9. And nobody much cares for those.

Plus, double of what?

 

[Marrond]

I’d bite at £260 at most for 770 8GB as a strictly gaming card. I’d pay the current asking price for 16GB ONLY after Blender 3.6 comes out and implements GPU rendering for Intel GPUs and they prove working and are competitive with Nvidia OptiX. I believe it’s achievable considering respectable RT performance on Intel cards and the fact that they actually do try to make it work and fix problems... Radeons are absolutely ******* for the purpose so supplementing my 6900XT with A770 would be a decent upgrade over my current 1080Ti that I keep around solely for that purpose.

 

[Author_25]

Latest Driver release has removed the overlay only application and you can know have the Desktop standalone application only.

Also great number of improvements as the graph above shows for the A750.

Also the A750 has had a price drop, $250, now.

 

[Quartz]

Plus, double of what?

The original launch driver, not the last driver.

 

[random_matt]

And the A770 is now £399, no chance.

 

[Cyber-Mav]

A770 at 300 would have me interested

 

[Quartz]

The new drivers do not appear to be on Intel's website right now.

 

[KompuKare]

The new drivers do not appear to be on Intel's website right now.

Look on the bright side. Previously their customers were beta testers, now the press gets to do the beta testing. Assuming the "press" didn't just all publish a press release, that is!