Skip to content
local-ai

NVIDIA RTX 6000 Ada Generation

NVIDIA · Enthusiast

The Ada Lovelace successor to the A6000, with 48GB of ECC memory, faster tensor cores, and FP8 support, aimed at professional and prosumer local AI.

Specifications

VRAM 48GB
Memory bandwidth 960 GB/s
Power draw 300W
Type gpu

A premium new-buy workstation card, into several thousand pounds. Verify current pricing.

NVIDIA RTX 6000 Ada Generation: common questions

What models can the NVIDIA RTX 6000 Ada Generation run?
With 48GB of VRAM it can run models up to roughly 77B parameters at a 4-bit quantisation, or smaller models with more context. Use the hardware matrix for specifics; these figures are approximate.
How much power does the NVIDIA RTX 6000 Ada Generation draw?
About 300W under load, so pair it with a power supply that has real headroom.

Strengths

  • 48GB plus Ada FP8 tensor cores give strong single-card inference
  • 960 GB/s bandwidth is a meaningful step up over the Ampere A6000
  • A 300W draw keeps it workstation-friendly despite the performance

Weaknesses

  • Very expensive new, with far worse cost-per-GB than consumer 24GB cards
  • GDDR6 bandwidth still trails HBM data-centre GPUs for high-throughput serving
  • No NVLink, so scaling beyond one card relies on slower PCIe
See what models fit in 48GB →

Featured in builds

Entry last verified 18 August 2026. Specifications and especially prices change; verify before buying.