Nvidia introduces new ampere GPUs RTX A6000 and A40 per ampere
Monday, October 5, 2020 | Written by Jim Thacker
At its GTC 2020 online conference, Nvidia unveiled the RTX A6000 (see above) and the RTX A40, the company's first professional GPUs to use their new Amp architecture.
The RTX A600 – originally called the Quadro RTX A6000 and is expected to ship later this year – replaces the existing Turing-based Quadro RTX 6000 and Quadro RTX 8000 workstation cards from Nvidia.
Nvidia claims that compatible GPU-based renderers do 2-3x better than with the Turing cards.
The RTX A40 is a passively cooled version of the A600, which is intended for use in servers and is to be delivered in 2021. Nvidia has not yet announced any prices for either GPU.
The first workstation GPUs to use Nvidia's Ampere architecture
The RTX A6000 and A40 are Nvidia's first professional cards to use the Ampere architecture after announcing the GeForce RTX 3090, 3080 and 3070 at the consumer level last month.
Unlike the current Turing architecture and its dedicated RT ray tracing cores, Nvidia has not introduced any new types of hardware during this time, although the designs of the existing cores have been updated.
Nvidia describes the RT cores of the second generation of Ampere cards with 1.7 times the throughput of their predecessors. and their third generation machine learning tensor cores as 2.7x improvement.
Similar core specifications as the current top-of-the-range Quadro RTX 8000
In terms of other core specs, the RTX A6000 pretty much sticks to the formula of the Quadro RTX A8000.
Both have 48 GB of built-in GDDR6 memory, NVLink support and four DisplayPort 1.4 outputs and have similar maximum power consumption.
The RTX A40 (above) is a passively cooled version of the A6000 – ie without a built-in fan – for use in servers and digital signage.
According to Nvidia, it has three DisplayPort ports, not four, and the A6000's GPU processing power has “dropped slightly”, otherwise the technical data is identical.
Double the GPU rendering speed of the Turing cards?
Nvidia claims that compatible GPU renderers perform about 2-3x better with the RTX A6000 than with the Turing cards it replaces.
In its blog post announcing the new cards, the company says the RTX A6000 scores “almost three times” that of the Quadro RTX 6000 in the built-in benchmark of the upcoming KeyShot 10.
The Amp architecture also "doubles" the performance of the Chaos Group's V-Ray GPU and Project Lavina, although this is based on the GeForce RTX 3080, not the A6000 itself.
Nvidia also quotes Digital Domain CG supervisor Bob White as saying that the studio achieved "a two-fold increase in performance with Redshift," although it was not specified which GPUs were used.
Prices and release dates
The RTX A6000 will be available from mid-December. The A40 will be available through workstation and server manufacturers early next year. Nvidia has not announced any prices yet.
For more information on the RTX A6000, visit the Nvidia website
Tags: amps, CUDA core, DisplayPort, GDDR6, GPU, GPU computing, GPU rendering, graphics card, graphics memory, GTC 2020, memory, memory bandwidth, NVIDIA, NVLink, power consumption, price, professional GPU, Quadro, Quadro RTX 6000, Quadro RTX 8000, Raytracing, Release Date, RT Core, RTX, RTX A40, RTX A6000, Server GPU, Specifications, TDP, Tensor Core, Turing, Workstation GPU