NVIDIA unveils RTX A5000 and RTX A4000 – Ampere-based graphics cards for professional graphics work

0
308

NVIDIA unveils RTX A5000 and RTX A4000 – Ampere-based graphics cards for professional graphics work

NVIDIA has unveiled two new professional graphics cards at today’s GTC 2021 event – the RTX A5000 and RTX A4000. Both new products are built on the fresh architecture of Ampere and are aimed at professional users in the field of digital information visualization. Recall that last fall NVIDIA introduced the older model of the series, the RTX A6000.

NVIDIA
NVIDIA

The RTX A5000 uses 8192 CUDA cores, as well as 256 tensor and 64 RT cores. The video card received 24 GB of GDDR6 video memory with ECC error correction and support for the 384-bit bus. The theoretical peak memory bandwidth is declared at 768 GB / s.

According to NVIDIA, the RTX A5000 is ready to deliver 27.8 teraflops of single-precision performance. The performance of RT cores in ray tracing tasks can reach 54.2 teraflops, and the performance of tensor cores can reach 222.2 teraflops.

The RTX A5000 has a claimed 230W power rating. The video card is equipped with one 8-pin auxiliary power connector. The cooling system uses a tangential fan. The external interfaces include four DisplayPort 1.4 connectors. The manufacturer indicates that the card will occupy two expansion slots in the system.

For the novelty, support for GPU virtualization is announced through tools such as NVIDIA Virtual Workstation and NVIDIA Virtual Compute Server. The card offers support for NVLink technology, allowing you to organize workstations with two similar professional graphics accelerators. The cooling system of the novelty is of a tangential type. In a PC case, the video card will occupy two expansion slots.

Also Read:  NVIDIA Unveils Jetson Nano 2GB - A System For AI Developers Powered By Maxwell Graphics Accelerator For $ 59

The RTX A4000 model received 6144 CUDA cores, 192 tensors, and 48 RT cores. The peak performance of the RTX A4000 model in single-precision operations reaches 19.2 teraflops, in ray tracing tasks – up to 37.4 teraflops, and the performance of tensor cores can reach 153.4 teraflops.

The video card is equipped with 16 GB of GDDR6 video memory with a 256-bit bus and ECC error correction function. The theoretical memory bandwidth is 448 GB / s.

The declared power consumption level of the novelty is 140 watts. The video card is equipped with one 6-pin auxiliary power connector. The external graphics accelerator interfaces include four DisplayPort 1.4 connectors. The video card is compact and requires one free expansion slot for installation. For the younger model, support for NVLink technology is not announced.