It looks like we have some confirmation regarding the AMPERE GPU architecture from Nvidia. NVIDIA’s CEO” Jensen Huang” has just confirmed during a keynote press briefing that the upcoming Ampere architecture is going to launch for both “consumer” as well as the datacenter/HPC markets. This confirms all the previous rumors we have been hearing so far.
We will have Ampere lineup for the Gaming “Geforce cards” as well. The previous generation Volta and Turing series were actually “split” between professional and gaming segments. This changes with Ampere.
“Ampere will eventually replace Nvidia’s Turing and Volta chips with a single platform that streamlines Nvidia’s GPU lineup, Huang said in a pre-briefing with media members Wednesday. While consumers largely know Nvidia for its videogame hardware, the first launches with Ampere are aimed at AI needs in the cloud and for research.
“Unquestionably, it’s the first time that we’ve unified the acceleration workload of the entire data center into one single platform,” Huang said.
GA100 Specs (most likely the next TITAN as well)
8192 CUDA cores
Largest single GPU core count ever seen
8192 FP32 cores
4096 FP64 cores
512 tensor cores.
8 Graphics Processing Clusters on the GPU, each with 16 SM units and 8 TPCs.
826mm2 (even bigger than the Volta GV100 GPU [TITAN V] which was 815 mm2). The GPU also features more than twice the number of transistors at
54 Billion Transistors
Ampere GA100 GPU is single-handily the densest GPU ever built.
HUGE 6144-bit bus interface
Up to 48 GB HBM2e memory in six HBM2 stacks (TITAN may end up at 24-32GB, but the 48GB buffer is not out of the question)