They don’t have one that can properly compete with nvidia, which means it’s a waste of resources to even target that area and instead focus on one where they can.
I thought bringing chiplets to GPUs, meant they’d be able just add as many CUs and cash dies as they needed to get on top. Even if it’s $3.5k and 1000W, they should be able to. They could sell 100K units as some limited edition special thing, and pull mind share away from nVidia by having the undisputed top card.
But they don’t. Which is why I think they undervalue having a halo product. They don’t think it’ll push units further down the product stack. I think they’re clearly wrong about that. People buy cards that fit their budgets. But they buy brands they know to be the best.
It’s well outlined in the article too. AMD is far behind in regards to raytracing and upscaling, both enthusiast level relevant technologies. AMD already struggles to compete with nvidia in sales and this is just the final nail in the coffin if they simply cannot offer a competitive product that’s actually going to get bought by enthusiasts.
If we look at the Steam hardware survey, we see that an RTX 4090 is at .96% shares, a 7900 XTX is not even reaching half of that with .40%. And this discrepancy is just going to grow more and more due to AMD being behind on those things. So until they can improve (if they even can), there’s no point in wasting their already lower resources on a market they cannot compete in.
The one issue I see here is that Intel is very much pushing into the budget and at least lower midrange market at the moment too, which could mean AMD loses shares to Intel as well.
I think they underestimate the marketing value of a halo product.
That, or it’s just spin to account for the fact that they don’t have one.
They don’t have one that can properly compete with nvidia, which means it’s a waste of resources to even target that area and instead focus on one where they can.
I’m not sure I understand why they don’t.
I thought bringing chiplets to GPUs, meant they’d be able just add as many CUs and cash dies as they needed to get on top. Even if it’s $3.5k and 1000W, they should be able to. They could sell 100K units as some limited edition special thing, and pull mind share away from nVidia by having the undisputed top card.
But they don’t. Which is why I think they undervalue having a halo product. They don’t think it’ll push units further down the product stack. I think they’re clearly wrong about that. People buy cards that fit their budgets. But they buy brands they know to be the best.
It’s well outlined in the article too. AMD is far behind in regards to raytracing and upscaling, both enthusiast level relevant technologies. AMD already struggles to compete with nvidia in sales and this is just the final nail in the coffin if they simply cannot offer a competitive product that’s actually going to get bought by enthusiasts.
If we look at the Steam hardware survey, we see that an RTX 4090 is at .96% shares, a 7900 XTX is not even reaching half of that with .40%. And this discrepancy is just going to grow more and more due to AMD being behind on those things. So until they can improve (if they even can), there’s no point in wasting their already lower resources on a market they cannot compete in.
The one issue I see here is that Intel is very much pushing into the budget and at least lower midrange market at the moment too, which could mean AMD loses shares to Intel as well.