• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

AMD Radeon RX 7600 Final Specs and Power Figures Leaked, Uses 6 nm

IIRC the 7900s have zero gain from the dual-pumped FP32 units because the compiler doesn't really support the functionality yet.
What compiler? Do you have any source for that?
 
Hm, so now it's not certain if it uses the full N33?
 
I'm looking forward to reviewers benchmarking this alongside the 6600XT and noting what a nothing-burger RDNA3 is.

At $249 for a brand new 6650XT and absolutely tons of unsold stock, the MSRP of this needs to be basically $249 or lower.
 
I'm looking forward to reviewers benchmarking this alongside the 6600XT and noting what a nothing-burger RDNA3 is.
RX 7600 is RDNA 2.5
It doesn't use the same node and doesn't have the same L3 cache improvements the 7900XT(X) have, just by looking at it superficially. I'm sure reviewers will dig in to the architecture even more.
 
This card/chip is waaaay out of his comfort zone, too damn hot clocked to make up the small footprint and grunt. It will be okay for cheaper SKUs with a tamed power (115w, 150w) and cards between 100-250 USD, 300 is a robbery.
 
$269/€299

...ooof
 
What compiler? Do you have any source for that?
Yup


I’m guessing RDNA 3’s dual issue mode will have limited impact. It relies heavily on the compiler to find VOPD possibilities, and compilers are frustratingly stupid at seeing very simple optimizations. For example, the FMA test above uses one variable for two of the inputs, which should make it possible for the compiler to meet dual issue constraints. But obviously, the compiler didn’t make it happen. We also tested with clpeak, and see similar behavior there. Even when the compiler is able to emit VOPD instructions, performance will only improve if compute throughput is a bottleneck, rather than memory performance.

On the other hand, VOPD does leave potential for improvement. AMD can optimize games by replacing known shaders with hand-optimized assembly instead of relying on compiler code generation. Humans will be much better at seeing dual issue opportunities than a compiler can ever hope to. Wave64 mode is another opportunity. On RDNA 2, AMD seems to compile a lot of pixel shaders down to wave64 mode, where dual issue can happen without any scheduling or register allocation smarts from the compiler..




$269/€299

...ooof
Yeah they can go f___ themselves with that price.
 
I wonder if it will have an updated media/codec engine?
 
Considering RTX 3060 Ti sells for $200 used nowadays, not sure if there are buyers for this card for anything more than $200.
 
Back
Top