As PC builders wait for the now delayed launch on the Nvidia GeForce RTX 3070 later this month, Nvidia has thrown the covers off its new GPUs for professionals.
These new GPUs are simply named the RTX A6000 and A40, with Nvidia seemingly abandoning the Quadro and Tesla monikers it usually paired with these sorts of cards.
The RTX A6000 is Nvidia’s new workstation GPU so let’s start there.
The GPU will sport 48GB of GDDR6 memory with error-correcting code(ECC), it supports PCIe Gen 4 and it’s actively cooled.
All of that memory will run at a clock of 16Gbps, with a memory bus of 384-bit according to a report by AnandTech. Unfortunately there is no word on clock speeds at this time.
The A40 is Nvidia’s new Ampere based data centre GPU, and as such this will be a passively cooled GPU. That thick cooler (pictured below) looks rather sexy we must admit. Pity it’s destined for a rack somewhere never to be seen again.
The A40 also features 48GB of GDDR6 memory with ECC though it will run slightly slower at 14.5Gbps on a 384-bit bus.
Both cards sport the GA102 GPU which is based on the Samsung 8nm process, the same we saw in the RTX 3080 and RTX 3090.
Nvidia has also shared some initial thoughts from folks who might use GPUs like this such as Blackmagic Design.
“We were impressed with the significantly better performance we measured on NVIDIA RTX A6000 than the prior generation of GPUs, both with general grading functions as well as our recent AI-based tools. Its 48 gigabytes of main memory greatly enhances Blackmagic RAW 8K and 12K camera workflows, making editing and grading performance fluid,” said president at Blackmagic Design, Dan May.
Nvidia expects the RTX A6000 to be available from it and partners such as PNY, Leadtek, Ingram Micro and Ryoyo from mid December with the A40 expected to form part of servers early next year.