Notebookcheck Logo

NVIDIA unveils the RTX A6000 and A40: new Ampere-based cards for creative or data professionals

The new RTX A6000 card. (Source: NVIDIA)
The new RTX A6000 card. (Source: NVIDIA)
The RTX A6000 and A40 have been introduced as state-of-the-art GPUs for content creation, data-centers or engineering. They are based on the latest Ampere architecture, and have been equipped with next-gen tensor, RT and CUDA cores. These cards are therefore rated to outstrip their Quadro predecessors in terms of throughput and performance.

NVIDIA has unveiled its latest high-powered "professional" RTX cards. One of these is, as some leaks more or less predicted, the RTX A6000. It is accompanied by the somewhat less anticipated A40 GPU, which has 48GB of GDDR6 memory in common with its new sibling.

Both new products are of course based on the 7 nanometer (nm) Ampere architecture. NVIDIA has revealed that they are also set up with new CUDA cores, as well as second-gen dedicated ray-tracing (RT) cores and third-gen tensor cores. They are touted to deliver twice the FP32 performance, twice the throughput while completing RT-related tasks such as shading and five times the throughput compared to the previous generation.

On that note, avid pro-grade graphics fans may have noticed that these 2 new offerings appear to have dropped the usual Quadro branding. However, that name is still present in many of the new line's workflow-enhancing features and software enhancements. They include NVIDIA Quadro Virtual Data Center Workstation, NVIDIA Virtual Compute Server, Quadro Sync and NVIDIA GRID.

Their maker has not released the full gamut of their specs so far: for example, it has revealed that the A40 has a bandwidth of 696 gigabytes per second (GB/s), but has not done the same for the A6000. However, it does confirm that the A6000 and A40 are rated for 300W of power used at most, and have 4 and 3 DislayPorts respectively.

They are also fitted with the projected 8-pin power connector, and support the third-gen NVLINK necessary to scale a single rig up to 96GB, should that be necessary. The RTX A6000 is already in the hands of OEMs such as Dell, Lenovo and HP that will make new workstations based on this card, whereas the A40 will be incorporated into new servers made by partners such as Fujitsu, Dell and Cisco.

Those machines will be available in 2021; meanwhile, the A6000 card itself will become available from NVIDIA's website, Ingram Micro, PNY or Leadtek from mid-December 2020.

The new NVIDIA RTX A40. (Source: NVIDIA)
The new NVIDIA RTX A40. (Source: NVIDIA)

Source(s)

Read all 2 comments / answer
static version load dynamic
Loading Comments
Comment on this article
Please share our article, every link counts!
> Expert Reviews and News on Laptops, Smartphones and Tech Innovations > News > News Archive > Newsarchive 2020 10 > NVIDIA unveils the RTX A6000 and A40: new Ampere-based cards for creative or data professionals
Deirdre O'Donnell, 2020-10- 5 (Update: 2020-10- 6)