▸ NVIDIA· PASCAL

NVIDIA Tesla P40

VRAM

24 GB

HIGH-END

BANDWIDTH

347

GB/S

MODELS Q4

316/428

74%

7B Q4 SPEED

~40

FAST

▸ MODEL COVERAGE @ Q474% OF ALL

▸ ESTIMATED SPEED· BY MODEL SIZE @ Q4

Average speeds at Q4 quantization. Actual performance varies by model architecture and context length.

3B

~93

TOK/S

7B

~40

TOK/S

14B

~20

TOK/S

32B

~9

TOK/S

70B

—

39.4GB NEEDED

▸ SPECIFICATIONS

VRAM

24 GB

BANDWIDTH

347 GB/s

FP16 COMPUTE

0.2 TFLOPS

TDP

250W

MEMORY

GDDR5

ARCHITECTURE

Pascal

CUDA CORES

3,840

PCIE

Gen 3 x16

226

FAST MODELS · >30 TOK/S

Real-time chat speed

281

USABLE · >10 TOK/S

Comfortable for all tasks

316

TOTAL COMPATIBLE

Fit in VRAM at Q4

▸ RENT IT IN THE CLOUD

Buying Tesla P40 costs $15–$40k and isn’t practical for most teams. Spin one up by the hour instead:

Rent on RunPod →Or Vast.ai →

Spin up in ~60s. Pay by the second. Cancel anytime.

Some links are affiliate links — we may earn a small commission at no extra cost to you. This helps keep FitMyLLM free and independent.

▸ COMPATIBLE MODELS· 316

Falcon-H1R Tiny 90M0.09B

FALCON·256K CTX· CHAT· REASONING

TOK/S · 2% VRAM

nomic-embed-text-v1.5 100M0.1B

EMBEDDING·8K CTX· CHAT

TOK/S · 2% VRAM

GPT-2 124M0.124B

GPT2·1K CTX· CHAT

TOK/S · 2% VRAM

SmolLM2 135M0.135B

SMOLLM·2K CTX· CHAT

TOK/S · 2% VRAM

SmolVLM 256M0.256B

SMOLLM·8K CTX· CHAT· VISION

TOK/S · 3% VRAM

Gemma 3 270M0.27B

GEMMA·32K CTX· CHAT

TOK/S · 3% VRAM

Snowflake Arctic Embed M v2.00.305B

EMBEDDING·8K CTX· EMBEDDING· MULTILINGUAL

TOK/S · 3% VRAM

bge-large-en-v1.5 335M0.335B

EMBEDDING·1K CTX· CHAT

TOK/S · 3% VRAM

mxbai-embed-large-v10.335B

EMBEDDING·1K CTX· EMBEDDING

TOK/S · 3% VRAM

Snowflake Arctic Embed L0.335B

EMBEDDING·1K CTX· EMBEDDING

TOK/S · 3% VRAM

GPT-2 Medium 345M0.345B

GPT2·1K CTX· CHAT

TOK/S · 3% VRAM

LFM·125K CTX· CHAT· TOOL_USE

TOK/S · 3% VRAM

SmolLM2 360M0.36B

SMOLLM·8K CTX· CHAT

TOK/S · 3% VRAM

Falcon-H1 0.5B0.5B

FALCON·128K CTX· CHAT

TOK/S · 3% VRAM

Qwen 1.5 0.5B0.5B

QWEN·32K CTX· CHAT

TOK/S · 3% VRAM

Qwen 2.5 0.5B0.5B

QWEN·32K CTX· CHAT

TOK/S · 3% VRAM

SmolVLM 500M0.5B

SMOLLM·8K CTX· CHAT· VISION

TOK/S · 3% VRAM

EMBEDDING·8K CTX· EMBEDDING

TOK/S · 3% VRAM

Snowflake Arctic Embed L v2.00.568B

EMBEDDING·8K CTX· EMBEDDING· MULTILINGUAL

TOK/S · 3% VRAM

QWEN·32K CTX· CHAT· REASONING

TOK/S · 4% VRAM

Qwen3-Embedding 0.6B0.6B

EMBEDDING·32K CTX· EMBEDDING· MULTILINGUAL

TOK/S · 4% VRAM

Falcon-H1R Tiny 0.6B0.6B

FALCON·32K CTX· CHAT· REASONING

TOK/S · 4% VRAM

Falcon Perception 0.6B0.6B

FALCON·4K CTX· VISION

TOK/S · 4% VRAM

LFM·125K CTX· CHAT· TOOL_USE

TOK/S · 4% VRAM

GPT-2 Large 774M0.774B

GPT2·1K CTX· CHAT

TOK/S · 4% VRAM

Qwen 3.5 0.8B0.8B

QWEN·256K CTX· CHAT· CODING· MULTILINGUAL

TOK/S · 4% VRAM

Qwen3.5-0.8B0.9B

QWEN·256K CTX· CHAT

TOK/S · 4% VRAM

FALCON·32K CTX· CHAT

TOK/S · 5% VRAM

INTERNLM·32K CTX· CHAT

TOK/S · 5% VRAM

OTHER·32K CTX· CHAT· VISION

TOK/S · 5% VRAM

TinyLlama 1.1B1.1B

LLAMA·2K CTX· CHAT

TOK/S · 5% VRAM

LFM2.5-1.2B-Thinking1.2B

LFM·122K CTX· CHAT· REASONING· TOOL_USE

TOK/S · 5% VRAM

Llama-3.2-1B1.2B

LLAMA·4K CTX· CHAT

TOK/S · 5% VRAM

LFM·125K CTX· CHAT· TOOL_USE· MULTILINGUAL

TOK/S · 5% VRAM

Zamba2 1.2B1.2B

OTHER·4K CTX· CHAT

TOK/S · 5% VRAM

DeepSeek Coder 1.3B1.3B

DEEPSEEK·16K CTX· CODING

TOK/S · 5% VRAM

EXAONE-4.0-1.2B1.3B

EXAONE·64K CTX· CHAT

TOK/S · 5% VRAM

OPT·2K CTX· CHAT

TOK/S · 5% VRAM

PHI·2K CTX· CODING

TOK/S · 5% VRAM

Phi-1.5 1.3B1.3B

PHI·2K CTX· CHAT· CODING

TOK/S · 5% VRAM

▸ NEXT STEP

Get personalized recommendations.

See ranked models with benchmark scores, run commands, and precise speed estimates for your Tesla P40.

▸Find models for Tesla P40 Compare with other GPUs

▸ DEVICE UNDER TEST

NVIDIA Tesla P40 — 24 GB VRAM.

▸ TESLA P40 SPEC

BRAND: NVIDIA
VRAM: 24 GB GDDR5
BANDWIDTH: 347 GB/s
FP16 COMPUTE: 0.2 TFLOPS
FP32 COMPUTE: 11.8 TFLOPS
CUDA CORES: 3,840
TDP: 250 W
ARCHITECTURE: Pascal

▸ AI CAPABILITY

316/ 428 models @ Q4

With 24 GB VRAM and 347 GB/s bandwidth, this GPU handles models up to 34.4B parameters.

Speed ≈ bandwidth / model_size × efficiency. A 7B model at Q4 runs at ~40 tok/s.

§ 01TOP MODELS FOR TESLA P40

316 FIT · SHOWING 20

MODEL	SIZE	VRAM Q4	TOK/S	AVG
Nous Capybara 34B	34.4B	21.5 GB	8	42.0
Yi-1.5 34B	34.4B	21.5 GB	8	45.3
Falcon-H1 34B	34B	21.3 GB	8	66.1
CodeLlama 34B	34B	21.3 GB	8	25.4
Nous Hermes 2 34B	34B	21.3 GB	8	47.0
Phind CodeLlama 34B	34B	21.3 GB	8	68.1
LLaVA-1.6 Yi 34B	34B	21.3 GB	8	47.4
WizardCoder Python 34B	34B	21.3 GB	8	73.2
Yi 34B	34B	21.3 GB	8	33.4
DeepSeek Coder 33B	33B	20.7 GB	8	26.0
Vicuna 33B	33B	20.7 GB	8	17.2
LLaMA 1 30B	33B	20.7 GB	8	17.8
DeepSeek-R1-Distill-Qwen-32B	32.8B	20.5 GB	8	46.9
Qwen3 32B	32.8B	20.5 GB	8	54.9
Qwen2.5-32B	32.5B	20.4 GB	9	54.3
Qwen 2.5 Coder 32B	32.5B	20.4 GB	9	48.0
QwQ-32B	32.5B	20.4 GB	9	43.7
OLMo-2-0325-32B	32.2B	20.2 GB	9	59.1
Aya Expanse 32B	32B	20.0 GB	9	35.9
Cogito 32B	32B	20.0 GB	9	39.4