NV NVIDIA Consumer Discrete GPU

NVIDIA GeForce RTX 3080 10GB

Runs downloadable LLMs up to ~11B at Q4, ~123.8-147.6 tok/s on an 8B model.

NVIDIA GeForce RTX 3080 10GB
10GB
VRAM
760
GB/s bandwidth
11B
Max model (Q4)
$699
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 92.5-110.2 t/s
≤26k ctx · then ~4.9-5.8 (slow)
184.9-220.4 t/s
≤72k ctx · then ~9.7-11.6 (slow)
330.2-393.6 t/s
≤92k ctx · then ~17.4-20.7 (slow)
7-8B won't run won't run 123.8-147.6 t/s
≤25k ctx · then ~6.5-7.8 (slow)
13B won't run won't run 76.2-90.8 t/s
≤3k ctx · then ~4-4.8 (slow)
32B won't run won't run won't run
70B won't run won't run won't run
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 ~9.2 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 GeForce RTX 3080 10GB. Guidance — exact needs depend on your case, other parts and how much you offload to system RAM.

Power supply
≥ 750W · 80+ Gold
Headroom over the 320W card plus the rest of the build.
System RAM
≥ 16 GB · 32 GB recommended
At least the 10 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 GeForce RTX 3080 10GB

Videos about the NVIDIA GeForce RTX 3080 10GB

Should You Buy the nVidia RTX 3080 for Local AI? Gemma 4?

More hardware like the NVIDIA GeForce RTX 3080 10GB

About the NVIDIA GeForce RTX 3080 10GB

NVIDIA GeForce RTX 3080 10GB is a discrete GPU from NVIDIA on Ampere, 2020. For local genAI the numbers that matter are its 10 GB (how big a model + context fits) and 760 GB/s bandwidth (generation speed). Plan for a ~750W power supply.

Frequently asked questions

What size LLM can the NVIDIA GeForce RTX 3080 10GB run?

With ~9.2 GB usable it runs up to ~11B at Q4 (~3.5B at FP16) at 8k context. Bigger spills into system RAM (CPU/PCIe offload) and slows sharply.

How many tokens/sec does the NVIDIA GeForce RTX 3080 10GB generate?

Generation is bandwidth-bound. At 760 GB/s an 8B model at Q4 runs ~123.8-147.6 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 9.2 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 GeForce RTX 3080 10GB?

Besides the card: a ≥750W 80+ Gold power supply, at least 16GB of system RAM (2× the 10GB VRAM is ideal for CPU offload), a free PCIe x16 slot, any modern multi-core CPU, and a ≥1TB NVMe SSD for model weights.