Updated minutes ago
RTX 3080
nvidia · ampere · 10 GB GDDR6X · 320W TDP
VRAM
10 GB
BF16 TFLOPS
47
Bandwidth
760 GB/s
From
$0.14/hr
Spec Sheet
VRAM10 GB GDDR6X
Memory Bandwidth760 GB/s
BF16 TFLOPS47
FP16 TFLOPS47
FP8 TFLOPS47
INT8 TOPS47
TDP320W
InterconnectPCIE
Max per Node4
PCIe Gen4
CUDA Compute Capability8.6
Tensor CoresYes
Pricing by Provider
| Provider | On-Demand | Reserved | Spot | Badge |
|---|---|---|---|---|
| tensordock | $0.22/hr | - | $0.14/hr | Cheapest |
| vast_ai | $0.28/hr | - | $0.16/hr |
Compatible Models (182)
Single GPU (110 models)
CodeGemma 7B8.5B FP8Qwen 3 8B8.2B FP8Llama 3.1 8B8.03B FP8Hermes 3 8B8.03B FP8Aya 23 8B8B FP8DeepSeek R1 Distill 8B8B FP8Llama 3 8B8B FP8Llama Guard 3 8B8B FP8Ministral 8B8B FP8Minitron 8B8B FP8Llama 3.3 8B8B FP8NV Embed v27.85B FP8InternLM 2.5 7B7.74B FP8GTE Qwen2 7B7.6B FP8Marco O17.6B FP8Qwen 2 Audio 7B7.6B FP8Qwen 2.5 7B7.6B FP8Qwen 2.5 Coder 7B7.6B FP8Qwen 2.5 Math 7B7.6B FP8Qwen 2.5 VL 7B7.6B FP8+90 more
Multi-GPU (72 models)
DeepSeek MoE 16Bx2 FP8CodeGen2 16Bx2 FP8DeepSeek V2 Litex2 FP8OctoCoder 15Bx2 FP8StarCoder2 15Bx2 FP8Nemotron 15Bx2 FP8Qwen 2.5 14Bx2 FP8DeepSeek R1 Distill 14Bx2 FP8Phi-4x2 FP8Qwen 2.5 Coder 14Bx2 FP8Qwen 1.5 MoE A2.7Bx2 FP8RWKV-6 14Bx2 FP8Phi 3 Medium 14Bx2 FP8Nekomata 14Bx2 FP8OLMo 2 13Bx2 FP8+57 more
Training Capabilities
Estimated GPU count for full fine-tuning (AdamW, BF16) and QLoRA
| Model Size | Full Fine-Tune | QLoRA |
|---|---|---|
| 7B model | 14 GPUs | 1 GPU |
| 13B model | 25 GPUs | 1 GPU |
| 70B model | 132 GPUs | 5 GPUs |
Energy Efficiency
Estimated tokens/second per Watt for popular models
Mistral 7B
0.33 t/s/WFP8
Qwen 2.5 7B
0.31 t/s/WFP8
Llama 3.1 8B
0.30 t/s/WFP8