NVIDIA’s Ampere RTX 3000 playing cards are about to release in a couple of months and the leaks have began pouring in already. The newest of which occurs to be a screenshot of the specs of the RTX 3080 Ti Founder Edition graphics card. This newest leak, whilst it’s not likely to be true at this juncture, states that the RTX 3080 Ti may have 5376 CUDA Cores with a Boost clock of 2200 MHz. It additionally signifies that the high-end RTX 3080 Ti may have 12GB of GDDR6 VRAM working at 18 Gbps giving it a complete bandwidth of 863 GB/s and will likely be designed with a TDP of 320W. The GPU powering the RTX 3080 Ti will supposedly be an NVIDIA Ampere GA102-400-A1. As according to some again of the hand calculations, those specifications point out that the RTX 3080 Ti may well be anyplace from 33% to 78% extra tough than the RTX 2080 Ti.
Rumoured specs of the NVIDIA RTX 3080 Ti Founders Edition
Twitter person CyberPunkCat (https://twitter.com/CyberCatPunk) posted a picture which seems to be a screenshot of the specs of the NVIDIA Ampere RTX 3080 Ti Founders Edition. It states that the GPU powering it is going to be a GA102-400-A1 which may have 5376 CUDA Cores with a spice up frequency of 2200 MHz. The Single Precision (FP32) efficiency is discussed to be 21 TFLOPS and at the reminiscence aspect of items, the RTX 3080 Ti will characteristic 12 GB of Samsung GDDR6 NAND chips working @ 18 Gbps. The overall bandwidth will likely be 836 GB/s and the TDP for the board will likely be 320W.
Previous rumours point out the presence of 3 GPUs, the GA102-200, GA102-300 and GA102-400. If we’re to take a look at the plausibility of those specs then a spice up clock of 2200MHz may be very a lot conceivable for the reason that RTX 2080 which is constructed on a 12nm procedure did arrange 2098 MHz spice up clock at the MSI RTX 2080 Ti Lightning Z version. And if we’re to imagine unique cooling, then this actual graphics card used to be even overclocked to 2505 MHz below LN2.
Coming to the core depend, with 5376 CUDA Cores @ a spice up clock of 2200 MHz, the FP32 efficiency comes out to be 23.6 TFLOPS. The leak states 21 TFLOPS of FP32 efficiency which issues to a base clock of roughly 1950 MHz. Compared to the 11.eight TFLOPS that the NVIDIA RTX 2080 Ti Founder Edition used to be in a position to at base clock, the RTX 3080 Ti finally ends up being 78% extra tough at 21 TFLOPS. NVIDIA has typically maintained a gen-on-gen efficiency increment of roughly 30% so this type of excessive efficiency delta turns into very tricky to consider. Even the Ampere A100, which is an undertaking grade GPU powering the DGX-A100 handiest supplies 19.Five TFLOPS of FP32 efficiency. So for NVIDIA to liberate a client card that’s extra tough than its undertaking card turns out extraordinarily not likely. Even the efficiency delta between the Ampere A100 and the Volta V100 is 24.5% so this actual leak with its supernormal efficiency delta is as just right as faux.
While no longer meting out deadly doses of sarcasm, this curious creature can frequently be discovered tinkering with tech, enjoying ‘vidya’ video games or exploring the darkest corners of the Internets. #PCMasterRace https://www.linkedin.com/in/mithunmohandas/