NVIDIA has announced its two new GPUs: a 20GB RTX A4500, and a 12 GB RTX A2000.
NVIDIA RTX A4500, comprises a GA102 GPU with 7168 CUDA cores, which makes it the lowest core count from all already existing GA102 GPU graphics cards in the market.
The RTX A4500 has a TDP of 200W. This card offers 20GB GDDR6 memory across a 320-bit bus. The card has a maximum bandwidth of 640 GB/s.
The card is claimed to have 23.7 TFLOPS in FP32 compute power and 46.2 TFLOPs in double-precision.
NVIDIA also came forward with the slowest Ampere workstation card, RTX A2000, with 12GB GDDR6 memory support. This card is no different than the aforementioned product even though it comes with much shorter memory support. The model is GA106-based and possesses 3328 CUDA cores and up to 8 TFLOPS of single-precision compute power.
Both cards have been specially designed for desktop workstation systems. They offer support for quad DisplayPort 1.4(either full-sized DP or miniDP in case of the RTX A2000, a low-profile design).
Via PNY