[adamc]

Speaking only as a gamer, I think graphics cards might eventually pay off, but it's a market that will require something like 10 years of runway. It's going to take a long time to both make competitive cards and convince gamers they have value. Lots of pretty good AMD GPUs have struggled for marketshare.

[dog436zkj3p7]

Yes, Intel should've never allowed itself to fall so far behind the competition in the GPU space. It's never too late, but at this point this would be financially risky and would require an enormous resource investment concentrated on both gaming and compute, while possibly never paying off. As you mention, AMD is struggling against Nvidia too, and their GPU tech is solid. A viable strategy for Intel would probably require selling at cost or even at a loss, while offering significantly better specs (more VRAM than competition seems like a viable path right now, but the AI bubble might not last).

However, I wouldn't be surprised to see the whole Arc effort killed off before another year passes, in line with Intel's absolute inability to stay committed to a product line. There's probably a reason why discrete Arc only appeared once Pat was in charge and I wouldn't be surprised to see it gone with his departure.

[throwaway48476]

Killing Arc would be very MBA brained. They already have to develop GPU IP for laptop chips. It's a volume game and AMD realized that developing one CPU chiplet for the whole portfolio.