I saw this move more as setting up a worthy competitor to Snapdragon X Elite, and it could also probably crush AMD APUs if these RTX things are powerful.
(And also, I'm pretending Macs don't exist for this statement. They aren't even PCs anymore anyway, just giant iPhones, from a silicon perspective.)
To anyone actually paying attention, igpus have come a long way. They are no longer an 'I can play minecraft' thing.
I hadn't realized that "Arc" and "Integrated" overlapped, I thought that brand and that level of power was only being used on discrete cards.
I do think that integrated Arc will probably be killed by this deal though, not for being bad as it's obviously great, rather for being a way for Intel to cut costs with no downsides for Intel. If they can make RTX iGPUs now, and the Nvidia and RTX brand being the strongest in the gaming space... Intel isn't going to invest the money in continuing to develop Arc, even if Nvidia made it clear that they don't care, it just doesn't make any business sense now.
That is a loss for the cause of gaming competition. Although having Nvidia prop up Intel may prove to be a win for competition in terms of silicon in general versus them being sold off in parts, which could be a real possibility it seems.
Wait, RDR2 is badly optimized? When I played it on my Intel Arc B580 and Ryzen 7 5800X, it seemed to work pretty well! Way better than almost any UE5 title, like The Forever Winter (really cool concept, but couldn't get past 20-30 FPS, even dropping down to 10% render scale on a 1080p monitor). Or with the Borderlands 4 controversy, I thought there'd be way bigger fish to fry.
It was definitely going to upset the market. Now i understand the radio silence on a card that was supposed to have been coming by Xmas.
I just meant their integrated GPUs are what's completely safe here.
Also the arc wasn't in jeopardy, the arc cards have been improving with every release and the latest one got pretty rave reviews.
Out of 12 mobile GT 7xx cards only 3 were Fermi (and 2 of those were M and not GT) rest were Kepler.
This has been somewhat improved-- some mainboards will have HDMI and DisplayPort plumbed to the iGPU, but the classic "trader desk" with 4-6 screens hardly needs a 5090.
They could theoretically sell the same 7xx and 1030 chips indefinitely. I figure it's a static market like those strange 8/16Mb VGA chipsets that you sometimes see on server mainboards, just enough hardware to run diagnostics on a normally headless box.
Even before the Pascal GTs, most of the GT 7xx cards, which you would assume were Maxwell or Kepler from the numbering, were rebadged Fermi cards (4xx and 5xx)! That generation was just a dumping ground for all the old chips they had laying about, and given the prominence of halfway decent iGPUs by that point, I can't say I blame them for investing so little in the lineup.
That said, the dGPUs are definitely somewhat at risk, but I think the risk is only slightly elevated by this investment, given that it isn't exactly a cash cow and Intel has been doing all sorts of cost-cutting lately.