This was posted 2 years 3 months 8 days ago, and might be an out-dated deal.

Related
  • expired

Leadtek Quadro Nvidia RTX A2000 Workstation Graphic Card $821.88 Delivered @ Australian Warehouses

40

The NVIDIA RTX A2000 introduces NVIDIA RTX technology to professional workstations with a powerful, low-profile design. Transform your workflows with real-time ray tracing and accelerated AI to create photorealistic concepts, run AI-augmented applications, or review within compelling VR environments. Supported by major creative and design applications, the power of RTX is yours.

GPU Memory: 6 GB GDDR6 with error-correction code (ECC)
Display Ports: 4x mini DisplayPort 1.4
Max Power Consumption: 70 W
Graphics Bus: PCI Express Gen 4 x 16
Form Factor: 2.7 (H) x 6.6Ā¯ (L) Dual Slot
Thermal: Active
VR Ready: Yes

Related Stores

Australian Warehouses
Australian Warehouses

closed Comments

  • +1

    can it run crysis 3

  • what a cute gpu

  • +30

    I know this is NOT the purpose of this card and the comparison is rather dumb, but I had to figure it out so FYI: it delivers RTX 3050 level of performance.

    • +2

      My people.

    • +2

      Considering it runs on PCIe power (up to 75W), it's actually very impressive performance.
      Would be great in one of those optiplex SFF systems if it didn't cost so much

    • Nvidia also locks a few features behind the quadro tier too, which you pay for. Gpu virtualisation being one I think?

  • Quadro! Now that's a name I've not heard in a long time.

  • I think it's equivalent to GTX1660 super but has ray tracing capability according to Techpowerup

  • +4

    RTXA2000: 3328 CUDA Cores + 192bit GDDR6, 8.0 TFLOPS.
    RTX3050: 2560 CUDA Cores + 128bit GDDR6, 9.1 TFLOPS
    RTX3060: 4864 CUDA Cores + 256bit GDDR6, 12.8 TFLOPS

    TFLOPS cares more about calculation performance than gaming performance, as gaming performance is hugely video memory speed related.

    that being said A2000 with more CUDA cores with lower TFLOPS than RTX3050 is because less power requirement (which means less core frequency)

    hence gaming performance is more similar to RTX3050 (slower core speed + more memory bandwidth)

    • The 3060 Cuda cores and memory bus specs you listed are incorrect

      • +1

        my bad, it is 3584 / 192bit.

        Copied from the 3060ti side without notice

        Thank you for pointing out.

  • +2

    6 GB VRAM for AI Workstation, hmm

    • +2

      It does not make sense for AI tasks

      • +1

        yep, RTX 3060 12 GB is a better choice if you don't need ECC VRAM.
        RTX A4000 16 GB at this price would have been a great deal :D

        • +1

          You don't really need ECC for model training, maybe for inference in some cases, but not very often.

          The current A4000 price is not far from 3090 deals, which is much better for deep learning / training tasks (unless you have to put GPUs to datacenter, NVidia prohibit RTX cards via drivers licensing to be used in datacenters, with the exception of crypto mining).

  • +2

    ill go $200 otherwise tell him he's dreaming

Login or Join to leave a comment