Nvidia’s New Graphics Card RTX A2000 Is Very Small

RTX A2000

We are now getting used to bulky GPUs with 3 or 4 fans in them along with massive heat sinks, but here we are the Nvidia RTX A2000. It’s quite small for a workstation graphics card, although one look under the hood and you’ll see it’s because this small card is somewhat closely related to the GeForce RTX 3060.

It doesn’t look that small without a point of reference for scale. That’s why you need to watch Richard Kerris, VP of Omniverse development at Nvidia, reveal this SIGGRAPH announcement.

This graphics card is actually only a bit behind an RTX 3060 in its specs, which isn’t a slouch in GPU terms.

Check out the specs:

  • CUDA Cores: 3,328
  • Tensor Cores: 104
  • RT Cores: 26
  • Memory: 6 GB GDDR6
  • Memory interface: 192-bit
  • Memory bandwidth: 288 GB/s
  • Interface: PCIe 4.0 x16
  • Total Board Power (TBP): 70W

That’s only a touch of the core count of the RTX 3060, although I’d assume the A2000 to work at significantly slower clock speeds to make its impressively minute 70W power envelope.

The clear distinction between the two is in VRAM. The A2000 comes with just 6GB of GDDR6, which also has ECC, or error-correcting code, due to it being a workstation Quadro card. While the RTX 3060 comes with 12GB of GDDR6, sans ECC.

A workstation graphics card with less memory than its gaming counterpart? That seems off, but But Nvidia has its reasons.

I think gamers are not quite so thirsty for a low-power graphics card with a small chassis, but there’s a part of me that would like to see a tiny GeForce card ready for a supremely tiny Mini-ITX build.

The new A2000 will be released in October if you are looking for a new mid-sized workstation. There are no pricing details available, but I expect it will cost you around $350.

Your Comments / What Do You Think ?

This site uses Akismet to reduce spam. Learn how your comment data is processed.