Yes absolutely, but the markup on VRAM is already insane. We’re not talking about a huge price difference here. Even if their $600 card now costs $670, they’d get a lot of sales just by setting themselves apart from NVIDIA, especially for those interested in running AI models locally. I know plenty of people buying Macs now just cause of AI spending $3000+ or more. NVIDIA & AMD have a duopoly though, so if one says they’re not going to do something then the other doesn’t have a strong incentive to either.
if you’re interested in running local AI models, that market should be looking at the strix halo desktops. that was basically 80% of the marketing framework did for its framework desktop.
would you have prefered they paid more for bleeding edge 3gb gddr stacks and put 24gb when only a subset would actually need it?
Yes absolutely, but the markup on VRAM is already insane. We’re not talking about a huge price difference here. Even if their $600 card now costs $670, they’d get a lot of sales just by setting themselves apart from NVIDIA, especially for those interested in running AI models locally. I know plenty of people buying Macs now just cause of AI spending $3000+ or more. NVIDIA & AMD have a duopoly though, so if one says they’re not going to do something then the other doesn’t have a strong incentive to either.
if you’re interested in running local AI models, that market should be looking at the strix halo desktops. that was basically 80% of the marketing framework did for its framework desktop.
I’d prefer they offered it as an overpriced premium SKU for those of us who can’t say no to 4k.