Other than 600W GPU: RDNA 3 is + 50% performance per watt

Other than 600W GPU: RDNA 3 is + 50% performance per watt

Other than 600W GPU

AMD recently released more details regarding the next generation of Radeon graphics cards. Based on the RDNA 3 architecture, the RX 7000s will offer a performance increase per Watt of over 50%. In particular, the new series of GPUs will boast the following features:

5nm process node Advanced chiplet packaging New computation unit Optimized graphics pipeline Next generation AMD Infinity Cache Over 50% performance / Watt more than RDNA 2 Speaking of the technical specifications, the Sunnyvale company will abandon the CU (Computer Unit) in favor of the WGP (Work Group Processor), each of which will be composed of two CUs, but with double the SIMD32 clusters compared to to only 2 of each RDNA2 CU. The flagship model, the Radeon RX 7900 XT, based on Navi 31, will have a single GCD (Graphics Compute Die) with 48 WGP, 12 SA and 6 SE, for a total of 12,288 Stream Processors. In addition to this, there should be 6 MCDs (Multi Cache Die) with 64MB of Infinity Cache per die and a 384-bit bus interface. As for the memory, the Navi 31 GPUs can be coupled to 24GB of VRAM, an amount equal to the flagships of the competitor NVIDIA.

Going down we then have the Navi 32 GPU (which should equip the Radeon RX 7800), which features a single GCD, four MCDs and 4 Shader Engine. Next, we have Navi 33, at the base of the Radeon RX 7700, which should offer superior performance to the RX 6800 and RX 6900, all with a nominal consumption of 230W, but which could reach 270-300W, thus outlining a decisive leap in performance compared to to the previous series. For your convenience, we present the summary table relating to the specifications of the Radeon RX 7000 series cards.

| ); }

NVIDIA's new AD102-based GeForce RTX 4090: 600W+ power usage CONFIRMED

NVIDIA's next-gen GeForce RTX 4090 graphics card, which will be powered by the AD102 'Ada Lovelace' GPU, will have a TBP (typical board power) of 600W... and that's in reference form, not in custom AIB form where 600W+ could be in the rear-view mirror.

The rumors are coming from Tom @ Moore's Law is Dead, who said his sources has confirmed the 600W power consumption news. NVIDIA is going to do a 'lot of weird things to not lose this generation' to AMD and its next-gen RDNA 3 architecture.

Tom's sources explained: 'It seems the top 6000/8000 Lovelace Professional card could use 320-375W. Right now they are testing prototypes at the limits of an EPS 8-pin (375W), or they are even testing some prototypes with a 16-pin connector. It's still 2-slot, and still a blower cooler. So there are limited to what they can cool'. The source added: 'But this is an unprecedented increase in power for a standard class of professional card'.

The leaks continue: 'In mid-2021, top Lovelace was expected to be a 400-500W card, but 500W+ was always an option... at this point 500-600W seems incredibly likely. The RTX 4090 will have a TBP of 600W. It also seems to be farther along than the other cards... so this is probably the first one to launch'.

What about system integrators (SIs) and OEMs and the 600W power consumption of the next-gen GeForce RTX 4090? Tom talked with some sources in the industry, who said: 'Honestly we (major OEM) don't see it as a huge challenge. Now we know we need to support 600W, and so we will'.

'If people will notice... a lot of OEMs are suddenly bringing out larger chassis that include much better PSUs for a reason. We're getting ready. 1000W+ PSUs will just be the standard soon'.