Friday, March 25th 2022
NVIDIA GeForce RTX 4090/4080 to Feature up to 24 GB of GDDR6X Memory and 600 Watt Board Power
After the data center-oriented Hopper architecture launch, NVIDIA is slowly preparing to transition the consumer section to new, gaming-focused designs codenamed Ada Lovelace. For starters, the source claims that NVIDIA is using the upcoming GeForce RTX 3090 Ti GPU as a test run for the next-generation Ada Lovelace AD102 GPU. Thanks to the authorities over at Igor's Lab, we have some additional information about the upcoming lineup. We have a sneak peek of a few features regarding the top-end GeForce RTX 4080 and RTX 4090 GPU SKUs. According to Igor's claims, NVIDIA is testing the PCIe Gen5 power connector and wants to see how it fares with the biggest GA102 SKU - GeForce RTX 3090 Ti.
Additionally, we find that the AD102 GPU is supposed to be pin-compatible with GA102. This means that the number of pins located on GA102 is the same as what we are going to see on AD102. There are 12 places for memory modules on the AD102 reference design board, resulting in up to 24 GB of GDDR6X memory. As much as 24 voltage converters surround the GPU, NVIDIA will likely implement uP9512 SKU. It can drive eight phases, resulting in three voltage converters per phase, ensuring proper power delivery. The total board power (TBP) is likely rated at up to 600 Watts, meaning that the GPU, memory, and power delivery combined output 600 Watts of heat. Igor notes that board partners will bundle 12+4 (12VHPWR) to four 8-pin (PCIe old) converters to enable PSU compatibility.
Source:
Igor's Lab
Additionally, we find that the AD102 GPU is supposed to be pin-compatible with GA102. This means that the number of pins located on GA102 is the same as what we are going to see on AD102. There are 12 places for memory modules on the AD102 reference design board, resulting in up to 24 GB of GDDR6X memory. As much as 24 voltage converters surround the GPU, NVIDIA will likely implement uP9512 SKU. It can drive eight phases, resulting in three voltage converters per phase, ensuring proper power delivery. The total board power (TBP) is likely rated at up to 600 Watts, meaning that the GPU, memory, and power delivery combined output 600 Watts of heat. Igor notes that board partners will bundle 12+4 (12VHPWR) to four 8-pin (PCIe old) converters to enable PSU compatibility.
107 Comments on NVIDIA GeForce RTX 4090/4080 to Feature up to 24 GB of GDDR6X Memory and 600 Watt Board Power
tough call
Or if dual-PSU cases become a thing again, because dual-systems is so 2018 and it's cheaper to buy 2 800w Bronze PSUs than 1 1600w Gold PSU. Assuming it's dual-PSU cases again, Phanteks will need an updated version of their PSU Combo adapter with the new 12+4 pin GPU connector instead of the older 8-pins. Or go back to using slaved PSU adapters for delay-turning on the second PSU. At least we now have the smaller SFX PSUs that could be squeezed into cases next to the big ATX PSUs.
Oh and don't forget that's only with the top end sku's, the 4070 should be two 8pin and the 4080 three 8pin, and if you consinder undervolting (which is probably much more common on those cards) you probably wouldn't need a psu upgrade.
I do wonder though how the newer PSU's will look, say 1000w, will it keep having 4/5 8pin connectors or will it just be one or two with a 450/600w capable 16pin?
Ada is just a die shrink of Ampere, so RT/Tensor perf would scale linearly with raster compare to Ampere.
There are AI supercomputers doing all the planet saving stuffs so in the end capitalism save itself from the problem it created :D. This would be a bit surprising to you but Capitalist countries have healthier air quality than non capitalist countries.
The price of electricity is literally in every mining calculator.
As far as gamers go apparently UK electricity is up to 28p/kwh, which is insane, but even so at 600W, that's £20/month, gaming 4 hours a day with a £3000 (?) GPU.
Given that people give some stupid amount of money (north of £100/month) to watch Sky Sports, then even when you deliberately buy the 'massively overpriced halo edition' GPU, the electricity, even in post-Putin inflation hellscape, is not that expensive.
You are missing quite a bit of context to justify cognitive dissonance. Another glaring one is the idea processing power somehow helps climate? All it really serves is understanding it, while slowly ruining it.
Also, lower SKUs arent all that efficient at all, the 104 also got its share of TDP bumps. Its a strange reading of facts you have here, buddy...
You make a great point when you say ADA is just shrunk ampere, that does explain quite well why we get what we seem to be getting. Their tech isnt advancing a whole lot anymore, we are in a moar corezzz situation here like I alluded to earlier as well. We'll see where that goes... gonna be interesting what AMD will do with RDNA but they then have the opportunity to gain feature parity with NV..