The issue is that the marked is not looking good. Amd have to sell any card they have. Why not selling now the card with high power and later you bring a little bit faster and less consuming variant.
Bondrewd
Veteran
Housefire perception no gucci.
How would that change things? It's going to perform somewhere between a 4080 and 4090 in either scenario, but if they drastically increase the power they lose all their marketing on perf/W, card size, not needing a PSU replacement etc and it's going to get memed as another vega
The investors.
AMD is playing their reliable execution card. The roadmap put RDNA3 to 2022 and that's it. It doesn't matter they are barely hitting it with a half-baked chip. It's still 2022.
The question is if even hiting those high speed will let them compete at higher price points. IF the cards are competing well at the $900-1k price range and making a profit why push a card to far to get to the next bracket.
a strong refresh in mid 2023 will still be a good stock bump for them
Bondrewd
Veteran
Eh it's just one part 2 SKUs really.
Frenetic Pony
Veteran
Could see up to a 40% clock increase with N31 respin, and a move to stacked cache should see high resolution and raytracing improvements beyond that. Even if the memory can only be improved by 20% (24gbps) the resulting card should be cost competitive with a 4080 at $1200, even if the 4080 moved to a $1k price point.
Bondrewd
Veteran
Counterproductive for what N3x parts actually are.
We've barely got to 20Gbps bins rofl, 24 is far and away.
My fire extinguisher needs a job to do, he es waiting for something like this
Investors first, screw the consumers. And everybody's cheering AMD as the "necessary competitor". No thanks.
At least Nvidia doesn't ship broken hardware.
define broken here ?
Gigabyte announces custom Radeon RX 7900 XTX and RX 7900 XT graphics cards, clocks up to 2680 MHz
https://videocardz.com/press-releas...-7900-xt-graphics-cards-clocks-up-to-2680-mhzNitro is looking fine too:
https://videocardz.com/newz/sapphire-radeon-rx-7900-xtx-nitro-graphics-card-has-been-pictured
Last edited:
Should be easy for them to add to the top of the stack then right. Put out a 7950 or something. Either it will launch at $1k cause Nvidia has already dropped the 4080 price or it will drop closer to $1200. Either way its better that than to have a crazy power hungry card just to hit a price point esp if we are talking 6 months of time
I am sure that would help them out sure even with out stacked cache. Although the stacked cache could end up making each chiplet smaller and less expensive. Dunno the costs there
Bondrewd
Veteran
ohhhhhhhhhh boy lmao you better not look at what bugs H100 ships with
What's "broken" hardware? It's functional, the performance just isn't where it was meant to be and it's priced accordingly. Don't like it, don't buy it – but they aren't "screwing" anyone. You don't think NV ships hardware without all sorts of errata, do you?
That being said, Navi31 seemingly being so busted that the performance is nowhere near where the competitor is and where it was projected to be is a huge failure for AMD, not something NV has done in recent years. Made worse by the fact that it's at a time when AMD keeps losing MS and relevance in dGPUs...they really didn't need this.
Bondrewd
Veteran
Very bad for messaging yes, but a $1k dorra GPU isn't a particularly high volume segment per se, and the relevant N3x offerings aren't busto.
No, they have deadlines by which they have to ship any given product.
Fo sho.
Speaking purely for myself (and yeah I am not The Market) if AMD can come out with a remotely sensible 7800XT I will very likely buy it. My last half-dozen GPUs have been NVIDIA but they are taking the piss now on prices and they can piss off.
AMD losing MS where? Regardless, the performance was aimed higher, but it would have never reached 4090 anyway. It's around 4080/16 speeds (excluding RT) at cheaper price, the chip is bigger when counting all the dies, but chiplets add some overhead and they're using (relatively) cheap older "7nm" node (N7x/N6x) on MCDs, GCD is same "5nm" node (N5x/N4x) but it's possible 4N is customized from denser N4x rather than N5x used by AMD. Not quite sure how that can be seen as a huge failure?
Similar threads
