Besides its new Jetson Nano 2GB SDK, NVIDIA also announced two new GPUs during its GTC 2020 keynote. Those cards are the RTX A6000 and A40 GPUs, and both cards are designed for the world of professionals and high-performance computing (HPC).
Like its consumer-ready GeForce RTX counterparts, both the A6000 and A40 are built with NVIDIA’s Ampere GPU architecture. This means that they also come with the brand’s new 2nd generation RT Cores, 3rd generation Tensor Cores, and CUDA cores to help significantly accelerate graphics, rendering, and AI-related tasks. Among other things.
Specs-wise, both cards feature 48GB of GDDR6 graphics memory and is scalable up to 96GB via a 3rd generation NVLink bridge. Additionally, both cards run on the PCIe Gen 4 interface and have a 300W TDP. Ports-wise, the A6000 features four DisplayPort 1.4 ports, while the A40 only comes with three DisplayPort 1.4 ports.
As for availability, the NVIDIA RTX A6000 will be available through channel partners starting from mid-December, while the A40 will only be available through OEM workstation and server vendors early next year. Also, NVIDIA’s Virtual Workstation will be made available next year. At the time of writing, pricing for either GPU wasn’t available.