How much VRAM does the NVIDIA RTX 6000 Ada have?

The NVIDIA RTX 6000 Ada has 48 GB of GDDR6 with 960 GB/s memory bandwidth.

What is the NVIDIA RTX 6000 Ada best for?

With 48 GB of VRAM, the NVIDIA RTX 6000 Ada is ideal for running 70B-class models at Q4 quantization and large MoE models — a workstation sweet spot for local inference.

What LLMs can the NVIDIA RTX 6000 Ada run locally?

The NVIDIA RTX 6000 Ada can run 52 of the 70 open-weight models tracked by CanItRun natively in VRAM at 8k context. Top options include: Llama 3.3 70B Instruct at NVFP4, Llama 3.1 8B Instruct at FP32, Llama 3.2 3B Instruct at FP32.

Can the NVIDIA RTX 6000 Ada run Llama 3.3 70B Instruct?

Yes. The NVIDIA RTX 6000 Ada runs Llama 3.3 70B Instruct natively in VRAM at NVFP4 quantization, achieving approximately 27.4 tokens per second.

Can the NVIDIA RTX 6000 Ada run Qwen 3.6 27B?

Yes. The NVIDIA RTX 6000 Ada runs Qwen 3.6 27B natively in VRAM at NVFP4 quantization, achieving approximately 71.1 tokens per second.

Can the NVIDIA RTX 6000 Ada run Llama 3.1 8B Instruct?

Yes. The NVIDIA RTX 6000 Ada runs Llama 3.1 8B Instruct natively in VRAM at FP32 quantization, achieving approximately 30 tokens per second.

NVIDIA RTX 6000 Ada

The NVIDIA RTX 6000 Ada has 48 GB VRAM and 960 GB/s memory bandwidth. It can run 52 of our 70 tracked models natively in VRAM at 8k context.

With 48 GB GDDR6, the NVIDIA RTX 6000 Ada is a workstation-tier GPU that can run 52 models natively. It handles 70B-class models at Q4 quantization.

The NVIDIA RTX 6000 Ada Generation is the Ada Lovelace successor to the RTX A6000, upgrading memory bandwidth from 768 to 960 GB/s while keeping the same 48GB GDDR6 VRAM and workstation form factor. The jump in bandwidth meaningfully improves inference tokens-per-second on larger models. Like its predecessor, it supports NVLink, ECC memory, and fits in standard workstations, making it the top-tier single-GPU option for on-prem LLM workloads that need professional reliability.

NVIDIA RTX 6000 Ada: October 2022 Ada workstation with 48GB GDDR6 at 960 GB/s — A6000 successor.

70B at Q4 native with ~20% higher tokens/sec than A6000. ~30-45 t/s for 7B, ~12-18 t/s for 70B.

Full CUDA with ECC. NVLink support. Top single-GPU option for on-prem LLM needing professional reliability.

Vendor	NVIDIA
Architecture	Ada Lovelace
VRAM	48 GB
Memory type	GDDR6
Memory bandwidth	960 GB/s
Compute backend	CUDA
Tier	Workstation
Released	2022
Models (native)	52 / 70
Models (offload)	2 / 70

Software: Full llama.cpp and Ollama support out of the box. CUDA 12.x recommended; driver ≥ 525 required.

Popular models for this GPU

Nemotron 3 Super 120B GPT-OSS 120B Llama 4 Scout 109B GLM-4.5 Air 106B GLM-4.6V 106B

Models this GPU runs natively in VRAM (52)

Models that fit with CPU offload (2)

These use system RAM for layers that don't fit in VRAM — expect much slower inference.

Too large for this GPU (16)

Compare NVIDIA RTX 6000 Ada with other GPUs

Frequently asked questions

How much VRAM does the NVIDIA RTX 6000 Ada have?: The NVIDIA RTX 6000 Ada has 48 GB of GDDR6 with 960 GB/s memory bandwidth.
What is the NVIDIA RTX 6000 Ada best for?: With 48 GB of VRAM, the NVIDIA RTX 6000 Ada is ideal for running 70B-class models at Q4 quantization and large MoE models — a workstation sweet spot for local inference.
What LLMs can the NVIDIA RTX 6000 Ada run locally?: The NVIDIA RTX 6000 Ada can run 52 of the 70 open-weight models tracked by CanItRun natively in VRAM at 8k context. Top options include: Llama 3.3 70B Instruct at NVFP4, Llama 3.1 8B Instruct at FP32, Llama 3.2 3B Instruct at FP32.
Can the NVIDIA RTX 6000 Ada run Llama 3.3 70B Instruct?: Yes. The NVIDIA RTX 6000 Ada runs Llama 3.3 70B Instruct natively in VRAM at NVFP4 quantization, achieving approximately 27.4 tokens per second.
Can the NVIDIA RTX 6000 Ada run Qwen 3.6 27B?: Yes. The NVIDIA RTX 6000 Ada runs Qwen 3.6 27B natively in VRAM at NVFP4 quantization, achieving approximately 71.1 tokens per second.
Can the NVIDIA RTX 6000 Ada run Llama 3.1 8B Instruct?: Yes. The NVIDIA RTX 6000 Ada runs Llama 3.1 8B Instruct natively in VRAM at FP32 quantization, achieving approximately 30 tokens per second.