NV NVIDIA Prosumer Discrete GPU

NVIDIA RTX 6000 Ada 48GB

Runs downloadable LLMs up to ~71B at Q4, ~156.4-186.4 tok/s on an 8B model.

NVIDIA RTX 6000 Ada 48GB
48GB
VRAM
960
GB/s bandwidth
71B
Max model (Q4)
$6,800
MSRP

What you can run

At 8k context, FP16 KV. Full-speed tok/s holds up to "max ctx"; beyond it the context spills to system RAM (CPU/PCIe offload) and drops to the degraded figure.

Model sizeFP16Q8Q4
3B 116.8-139.2 t/s
≤559k ctx · then ~4.9-5.8 (slow)
233.6-278.4 t/s
≤605k ctx · then ~9.7-11.6 (slow)
417.1-497.1 t/s
≤625k ctx · then ~17.4-20.7 (slow)
7-8B 43.8-52.2 t/s
≤203k ctx · then ~1.8-2.2 (slow)
87.6-104.4 t/s
≤264k ctx · then ~3.7-4.4 (slow)
156.4-186.4 t/s
≤291k ctx · then ~6.5-7.8 (slow)
13B 27-32.1 t/s
≤102k ctx · then ~1.1-1.3 (slow)
53.9-64.2 t/s
≤181k ctx · then ~2.2-2.7 (slow)
96.3-114.7 t/s
≤216k ctx · then ~4-4.8 (slow)
32B won't run 21.9-26.1 t/s
≤41k ctx · then ~0.9-1.1 (slow)
39.1-46.6 t/s
≤94k ctx · then ~1.6-1.9 (slow)
70B won't run won't run 17.9-21.3 t/s
≤11k ctx · then ~0.7-0.9 (slow)
120B won't run won't run won't run
400B won't run won't run won't run

Estimate: tok/s ≈ bandwidth ÷ (params × bytes/param) × efficiency (NVIDIA). Cyan = fits ~44.2 GB usable. Beyond "ctx", KV spills to system RAM (CPU/PCIe offload) (rough degraded speed shown).

What else you need for a full build

Minimum supporting components for a desktop built around the NVIDIA RTX 6000 Ada 48GB. Guidance — exact needs depend on your case, other parts and how much you offload to system RAM.

Power supply
≥ 700W · 80+ Gold
Headroom over the 300W card plus the rest of the build.
System RAM
≥ 64 GB · 128 GB recommended
At least the 48 GB of VRAM; 2× lets you CPU-offload models larger than VRAM.
Motherboard
1× free PCIe x16 slot (4.0/5.0)
A full-length x16 electrical slot; PCIe 4.0 is ample, 5.0 future-proofs.
CPU
Any modern multi-core
The GPU runs inference; more cores and memory channels only matter when offloading to RAM.
Storage
≥ 1 TB NVMe SSD
Weights are large — a 70B model at Q4 is ~40 GB; budget for a few.
PSU connector
12V-2×6 / 12VHPWR (or 3× 8-pin)
High-power NVIDIA cards use the 16-pin 12V-2×6 power connector.

Run popular models on the NVIDIA RTX 6000 Ada 48GB

Videos about the NVIDIA RTX 6000 Ada 48GB

Ollama NVidia GPU Speed Test Comparison of RTX 4090, Tesla P40, A100 SXM 80GB, RTX 6000 Ada 48GB

More hardware like the NVIDIA RTX 6000 Ada 48GB

About the NVIDIA RTX 6000 Ada 48GB

NVIDIA RTX 6000 Ada 48GB is a discrete GPU from NVIDIA on Ada Lovelace, 2022. For local genAI the numbers that matter are its 48 GB (how big a model + context fits) and 960 GB/s bandwidth (generation speed). Plan for a ~700W power supply.

Frequently asked questions

What size LLM can the NVIDIA RTX 6000 Ada 48GB run?

With ~44.2 GB usable it runs up to ~71B at Q4 (~20B at FP16) at 8k context. Bigger spills into system RAM (CPU/PCIe offload) and slows sharply.

How many tokens/sec does the NVIDIA RTX 6000 Ada 48GB generate?

Generation is bandwidth-bound. At 960 GB/s an 8B model at Q4 runs ~156.4-186.4 tok/s (estimate).

Why does long context slow it down?

The KV cache for the context must also fit fast memory; once weights+KV exceed 44.2 GB it spills to system RAM (CPU/PCIe offload) and throughput collapses. Each row shows the max usable context.

What else do I need to run the NVIDIA RTX 6000 Ada 48GB?

Besides the card: a ≥700W 80+ Gold power supply, at least 48GB of system RAM (2× the 48GB VRAM is ideal for CPU offload), a free PCIe x16 slot, any modern multi-core CPU, and a ≥1TB NVMe SSD for model weights.