Llama 3.3 70B Instruct (FP16) — 47.4 GBon Scaleway L40S-1-48G
Overview
Llama 3.3 70B Instruct is a 70B parameter dense language model by Meta, with code, multilingual, tool-calls capabilities. It supports a context window of up to 131,072 tokens.
Llama 3.3 70B Instruct is a 70-billion-parameter dense transformer model from Meta, optimized for instruction following, code generation, and multilingual conversation. It delivers performance competitive with larger models in the Llama family while remaining practical for single-node GPU deployments. The model supports tool calling and eight languages including English, French, Spanish, and German. With a 128K context window and grouped-query attention, it quantizes efficiently down to Q4 levels for self-hosted inference on consumer hardware.
At FP16 quantization (full-precision quality tier), the model weighs 131.43 GB. This exceeds the 48 GB of VRAM on Scaleway L40S-1-48G. Inference is still possible via CPU offload or memory-mapped loading from disk, but expect significantly reduced performance.
The NVIDIA L40S is a datacenter GPU with 48 GB of GDDR6 VRAM and 864 GB/s memory bandwidth. It delivers 362 FP16 TFLOPS with Ada Lovelace architecture. A versatile GPU for AI inference, training, and graphics workloads. Handles quantized models up to 40B parameters comfortably.
Hardware Requirements
| Model size | 131.43 GB |
| VRAM available | 48 GB |
| VRAM used | 47.4 GB |
| System RAM | 96 GB |
| Min RAM required | 85.4 GB |
| GPU layers | 28 / 80 |
| Context size | 512 |
| Backend | cuda13 |
| Flash attention | Yes |
| Reading from disk | Yes |
Performance Notes
Deploy
Prerequisites
Ensure your GPU nodes are prepared with the NVIDIA container toolkit:
ansible-playbook prositronic.infra.nvidia_container_toolkitCommand
helmfile --state-values-file <(curl -s https://www.prositronic.eu/values/llama-3-3-70b-instruct/fp16/nvidia-l40s.yaml) applyGenerated values.yaml
/values/llama-3-3-70b-instruct/fp16/nvidia-l40s.yaml
Loading values…Frequently Asked Questions
How much VRAM does Llama 3.3 70B Instruct (FP16) need?
The FP16 quantization of Llama 3.3 70B Instruct requires 131.43 GB. 28 of 80 layers fit in the 48 GB of VRAM on Scaleway L40S-1-48G; remaining layers are offloaded to CPU.
Can I run Llama 3.3 70B Instruct on Scaleway L40S-1-48G?
Yes, with reduced performance. Scaleway L40S-1-48G can run Llama 3.3 70B Instruct (FP16), but only 28 of 80 layers fit in VRAM. The rest are offloaded to CPU.
What is quantization?
Quantization reduces a model's numerical precision from its original floating-point format to a more compact representation. This shrinks the file size and VRAM footprint, making it possible to run large models on consumer hardware. The trade-off is a small reduction in output quality. FP16 compresses Llama 3.3 70B Instruct from its original size down to 131.43 GB.
What quantization should I choose for Llama 3.3 70B Instruct?
FP16 is a full-precision quantization. Higher-quality quants (Q8, Q6) preserve more model accuracy but need more VRAM. Lower quants (Q4, Q3, Q2) reduce VRAM usage at the cost of some quality. Choose based on your available hardware and quality requirements.
What is flash attention and why is it enabled?
Flash attention is a memory-efficient algorithm that speeds up the attention mechanism in transformer models. It reduces VRAM usage during inference by avoiding the materialisation of the full attention matrix. For Llama 3.3 70B Instruct on Scaleway L40S-1-48G, flash attention is enabled to maximise context length and throughput within the available 48 GB of VRAM.
Why are some layers offloaded to CPU?
Scaleway L40S-1-48G has 48 GB of VRAM, but Llama 3.3 70B Instruct (FP16) requires approximately 131.43 GB. Only 28 of 80 layers fit in VRAM; the remaining layers run on CPU, which is slower but still functional.
How do I run Llama 3.3 70B Instruct (FP16) with Ollama?
Run ollama run llama3.3:70b-instruct-fp16 to start Llama 3.3 70B Instruct (FP16). Ollama handles downloading the model weights automatically on first run.