What GPU do I need to run RedHatAI/Llama-3.2-1B-Instruct-FP8?
1.5B parameters, published in F8_E4M3. View on Hugging Face
Set up Llama-3.2-1B-Instruct-FP8Llama-3.2-1B-Instruct-FP8 is published by RedHatAI on Hugging Face, with 176,078 downloads and 4 likes to date. It's a LlamaForCausalLM model built for text-generation, published natively in F8_E4M3.
VRAM required & cheapest live GPU fit
Required VRAM = weight size at each precision, plus a fixed overhead for KV-cache, activations, and fragmentation. Full formula and assumptions: methodology.
A GPU is only matched to a row if its hardware supports that precision, and the primary recommendation is always a single-GPU fit when one exists.
INT4 requires a quantized checkpoint actually published for this model, check its Hugging Face page before relying on this row.
Cheapest way to run Llama-3.2-1B-Instruct-FP8 at its published (F8_E4M3) precision: 1× RTX 4070 Super on simplepod, at $0.100/hr per GPU ($0.100/hr total). Quantizing to FP8 or INT4 (rows above) can cost less, but requires a compatible quantized checkpoint to exist for this model.
Llama-3.2-1B-Instruct-FP8: common questions
How much VRAM does Llama-3.2-1B-Instruct-FP8 need?
1.7 GB at FP8 (native), 0.8 GB at INT4 (quantized). All of those sit under 6 GB, the smallest capacity this page reasons about, so VRAM is not what limits where this model runs. The 1.4 GB of weights plus inference overhead is the whole requirement.
Is Llama-3.2-1B-Instruct-FP8 already quantized?
Yes. It is published in FP8, one byte per parameter, so the 1.7 GB figure is already a quantized footprint rather than a full-precision one. Only INT4 goes below it, at 0.8 GB. A GPU without FP8 tensor cores cannot run it as published, which is why cards here are matched on precision support and not on VRAM alone.
How many copies of Llama-3.2-1B-Instruct-FP8 fit on one RTX 4070 Super?
7, by VRAM alone. That card carries 12.0 GB and one copy needs 1.7 GB at FP8 (native), on a live rate of $0.100/hr for the whole card. Throughput is not modelled here, so 7 copies is not 7 times the requests served.
Does quantizing Llama-3.2-1B-Instruct-FP8 lower the GPU bill?
Yes. At FP8 (native) the cheapest live fit is one RTX 4070 Super on simplepod at $0.100/hr. At INT4 (quantized) it drops to one A16 on vultr at $0.059/hr, provided a quantized checkpoint exists for it.
Weight-to-VRAM math, the fit rules, and how live prices are normalized: full methodology.
More RedHatAI models
- gemma-4-31B-it-FP8-block (31.3B, F8_E4M3)
- gemma-4-26B-A4B-it-FP8-dynamic (26.5B, F8_E4M3)
- gemma-4-12B-it-FP8-Dynamic (13.0B, F8_E4M3)
- Llama-3.2-1B-Instruct-FP8-dynamic (1.5B, F8_E4M3)
- Qwen3.5-9B-FP8-dynamic (9.4B, F8_E4M3)
- gemma-4-31B-it-FP8-dynamic (31.3B, F8_E4M3)
