NVIDIA
Operating mode
Use this to bias workload recommendations toward responsiveness, background autonomy, lighter serving, or multi-GPU scale-out.
Current mode
Balanced
Balanced for general local use. Keeps the ranking neutral across personal and serving workflows.
The Tesla P40 is a Pascal-generation datacenter GPU from 2016, built for inference workloads before the era of large language models. At 24 GB of GDDR5X, it was notable as one of the first high-VRAM inference accelerators, and it saw renewed interest from the LLM community when NVLink 3090-class cards were scarce. It can run 7B models at Q4 quantization but generation will be slow by current standards. Available on the used market for very low prices, it remains a viable ultra-budget option for hobbyists building an inference server, though modern alternatives are strongly preferred.
Beyond LLMs
What AI tasks this GPU can handle — from text generation to image and video creation.
| Capability | Status | Representative Model |
|---|---|---|
| LLM Chat (7B) | Runs natively | Llama 3.1 8B Q4 |
| LLM Coding (30B) | Runs natively | Qwen 3 30B Q4 |
| LLM Large (70B) | Won’t fit | Llama 3.1 70B Q4 |
| Image Gen (SDXL) | Runs natively | SDXL 1.0 FP16 |
| Image Gen (Flux) | Runs with offload | Flux.1 Dev FP16 |
| Image Gen (SD 3.5) | Runs natively | SD 3.5 Large FP16 |
| Video Short (25f) | Runs natively | LTX Video 2B |
| Video Long (100f) | Won't fit | Wan Video 14B |
Architecture
Pascal is NVIDIA's first 16nm FinFET GPU architecture, powering the GTX 10-series consumer cards and Tesla P100/P40 datacenter accelerators. It introduced unified memory architecture and NVLink interconnect for datacenter GPUs.
AI Relevance
No dedicated Tensor Cores — all AI inference runs on standard CUDA cores at FP16 or FP32 precision. Still usable for small models (7B Q4) on cards with sufficient VRAM like the GTX 1080 Ti (11 GB) or P40 (24 GB), but significantly slower than Turing and newer.
Kaufberatung
Ausgezeichnete Wahl für lokale KI
Führt 26 von 50 Top-Modellen gut aus — ein starker Allrounder für lokale Inferenz.
24.0 GB
VRAM
$5,699
UVP
$237/GB
Kosten pro GB VRAM
Beste Modelle für diese GPU
What will limit you first
This setup is broadly balanced for this model.
Very little memory headroom
You can run the model, but there is not much room left for longer context, bigger batches, extra apps, or future model updates.
Older PCIe generation
PCIe 3.0 is workable, but it compounds the penalty when you offload heavily or try to scale across multiple cards.
Best upgrade itinerary
Buy headroom, not only minimum fit
A slightly larger memory tier gives you safer context growth and makes the recommendation more future-proof.
Unlocks 1 additional models that do not fit on the current setup.
Mehr Spielraum gewünscht? MacBook Pro M4 Max 36GB (36.0 GB unified memory) ist die nächste Stufe.
Chat
SThis model is a direct match for chat. It belongs to a current frontier family for local AI. It fits natively with comfortable headroom. Known channels: huggingface, ollama, lm-studio.
Coding
SThis model is a direct match for coding. It belongs to a current frontier family for local AI. It should run, but memory headroom will be limited. Known channels: huggingface, ollama, lm-studio.
Agentic Coding
SThis model is still usable for agentic-coding, but it is not the most specialized pick. It belongs to a current frontier family for local AI. It should run, but memory headroom will be limited. Known channels: huggingface, lm-studio.
Reasoning
SThis model is a direct match for reasoning. It belongs to a current frontier family for local AI. It fits natively with comfortable headroom. Known channels: huggingface, ollama, lm-studio.
RAG
AThis model is a direct match for rag. It sits in the middle of the current model mix. It fits natively with comfortable headroom. Known channels: huggingface, ollama.
Qwen3-Coder 30B A3B InstructQwen3-VL 30B A3B Instruct
GPT-OSS 20BQwen 3 14B
Phi-4-reasoning-plus 14BQwen 3 30B A3BQwen 3.5 9BQwen 3.5 27B
Magistral Small 2507Devstral Small 2 24B InstructQwen 3.6 27BQwen 3 8BDevstral Small 1.1Qwen 3.5 4BNemotron Cascade 2 30B A3BNemotron 3 Nano 30B
Gemma 4 26B A4BMinistral 3 14BNemotron Nano 8BPhi-4 Mini Reasoning 4BQwen 3.5 35B A3B
Jina Embeddings v3Qwen 3 32B
BGE M3Qwen 3.6 35B A3B
EXAONE 4.0 32BQwen 3.5 397B A17BDevstral 2 123B Instruct
Kimi K2.5Kimi K2.6
DeepSeek V4 ProQwen 3.5 122B A10BDeepSeek V4 FlashMistral Small 4 119B
Command A 111BQwen 2.5 VL 72BGPT-OSS 120BQwen3-Coder-Next
GLM-5.1Pixtral Large 124BGLM-5DeepSeek V3.2Qwen 3 235B A22BQwen3-Coder 480B A35B InstructGemma 4 31BLeanstral 119B A6BDeepSeek Coder V2 236BDeepSeek R1 671BDeepSeek V3.1 671BFast erreichbar
Hochwertige Modelle, die etwas mehr Speicher benötigen
Image & Video Generation
41 of 52 models can generate images or video on your Tesla P40 24GB
| Model | Max Resolution | Gen Time | Grade |
|---|---|---|---|
| SD TurboImage | 512×512 | ~2.2s | S |
| Stable Diffusion 1.5Image | 512×768 | ~4.4s | S |
| Realistic Vision v5.1Image | 512×768 | ~4.4s | S |
| DreamShaper 8Image | 512×768 | ~4.4s | S |
| LCM DreamShaper v7Image | 512×768 | ~1.3s | S |
| PixArt-SigmaImage | 1024×1024 | ~17.6s | S |
| FramePack I2VVideo | 256×256 | ~32.3s/frame | S |
| SDXL TurboImage | 512×512 | ~2.2s | S |
| SDXL LightningImage | 1024×1024 | ~6.6s | S |
| Stable Diffusion XL 1.0Image | 1024×1024 | ~17.6s | S |
| Playground v2.5Image | 1024×1024 | ~26.4s | S |
| RealVisXL v5.0Image | 1024×1024 | ~19.8s | S |
| DreamShaper XLImage | 1024×1024 | ~19.8s | S |
| Juggernaut XL v9Image | 1024×1024 | ~19.8s | S |
| Animagine XL 3.1Image | 1024×1024 | ~19.8s | S |
| Pony Diffusion V6 XLImage | 1024×1024 | ~19.8s | S |
| Animagine XL 4.0Image | 1024×1024 | ~19.8s | S |
| Illustrious XLImage | 1024×1024 | ~19.8s | S |
| Wan Video 2.1 1.3BVideo | 256×256 | ~12.9s/frame | S |
| Stable Diffusion 3.5 MediumImage | 1024×1024 | ~30.8s | S |
| Flux.2 Klein 4BImage | 1024×1024 | ~5.3s | S |
| LTX Video 2BVideo | 768×512 | ~15.3s/frame | S |
| KolorsImage | 1024×1024 | ~35.2s | S |
| Stable CascadeImage | 1024×1024 | ~44s | S |
| AuraFlow v0.3Image | 1536×1536 | ~1m 19s | S |
| Stable Diffusion 3.5 LargeImage | 1024×1024 | ~1m 37s | S |
| Stable Diffusion 3.5 Large TurboImage | 1024×1024 | ~17.6s | S |
| CogVideoX 2BVideo | 720×480 | ~15.3s/frame | A |
| HunyuanVideoVideo | 256×256 | ~32.3s/frame | A |
| ChromaImage | 256×256 | ~32.3s | A |
| Z-Image TurboImage | 1536×1536 | ~18.2s | B |
| Flux.1 DevImage | 256×256 | ~1m 19s | B |
| Flux.1 SchnellImage | 256×256 | ~15.4s | B |
| LTX Video 13BVideo | 256×256 | ~32.3s/frame | B |
| Flux.1 Kontext DevImage | 256×256 | ~1m 28s | B |
| AnimateDiff v1.5.3Video | 512×768 | ~8s/frame | B |
| Cosmos Diffusion 7BVideo | 256×256 | ~48.7s/frame | B |
| CogVideoX 5BVideo | 256×256 | ~46.3s/frame | B |
| Wan2.2 TI2V 5BVideo | 256×256 | ~46.3s/frame | B |
| Flux.2 Klein 9BImage | 256×256 | ~16.1s | D |
| Flux.1 Fill DevImage | 256×256 | ~1m 15s | D |
| Mochi 1 PreviewVideo | 256×256 | ~29.1s/frame | F |
| HunyuanVideo 1.5Video | 256×256 | ~27s/frame | F |
| Helios 14BVideo | 256×256 | ~33.3s/frame | F |
| SkyReels V2 14BVideo | 256×256 | ~33.3s/frame | F |
| Wan Video 2.1 14BVideo | 256×256 | ~33.3s/frame | F |
| Wan Video 2.2 14BVideo | 256×256 | ~33.3s/frame | F |
| Qwen ImageImage | 256×256 | ~29.6s | F |
| Qwen Image EditImage | 256×256 | ~29.6s | F |
| Flux.2 DevImage | 256×256 | ~13m 53s | F |
| MAGI-1Video | 256×256 | ~41.3s/frame | F |
| HunyuanImage 3.0Image | 256×256 | ~52.2s | F |
Image models estimated at 1024×1024 (28 steps, FP16). Video models estimated at 768×512 (25 frames, 30 steps, FP16). Actual performance varies with runtime and system load.
Upgrade paths
See what you unlock with more powerful hardware
Upgrade-Optionen
Unlocks 1 additional models that do not fit on the current setup.
Lifts average decode speed across fitting models by about 33%.
ca. $2,499 MSRP
RTX 5000 Ada 32GBNVIDIA-UpgradeUnlocks 6 additional models that do not fit on the current setup.
Lifts average decode speed across fitting models by about 85%.
ca. $4,000 MSRP
Unlocks 17 additional models that do not fit on the current setup.
ca. $1,099 MSRP
Unlocks 45 additional models that do not fit on the current setup.
Lifts average decode speed across fitting models by about 372%.
ca. $8,000 MSRP
Tesla P40 24GB (24 GB VRAM) can run these top models: Qwen3-Coder 30B A3B Instruct (score: 93/100), Qwen3-VL 30B A3B Instruct (score: 93/100), GPT-OSS 20B (score: 93/100). See the full compatibility list above.
Tesla P40 24GB has 24 GB of VRAM available for AI model inference. This determines which models and quantization levels you can run locally.
Yes, Tesla P40 24GB is excellent for running LLMs locally with top compatibility scores above 80/100.
For coding on Tesla P40 24GB, we recommend Devstral Small 2 24B Instruct. It achieves 15.0 tokens per second with 40K context window. This model is a direct match for coding. It belongs to a current frontier family for local AI. It should run, but memory headroom will be limited. Known channels: huggingface, ollama, lm-studio.
There are 4 upgrade path(s) from Tesla P40 24GB: MacBook Pro M4 Max 36GB, RTX 5000 Ada 32GB. Upgrading would unlock larger models and faster inference speeds.
Yes, Tesla P40 24GB with 24 GB of usable memory can run Flux.1 Dev at FP16 natively. Flux is a 12B parameter diffusion transformer that produces high-quality images. You can also run the Schnell variant for faster generation.
Tesla P40 24GB (24 GB VRAM) can handle various AI generation tasks beyond LLMs. For image generation, SDXL and Stable Diffusion 3.5 run well. Flux.1 Dev also runs natively for state-of-the-art image quality. For video, LTX Video 2.3 can generate short clips. Check the AI Capability Matrix above for detailed compatibility.
Tesla P40 24GB is excellent for AI image generation. With 24 GB of usable memory, it runs all major diffusion models including Flux.1, SDXL, and Stable Diffusion 3.5 at full precision. You can generate high-resolution images quickly and even handle video generation models.
Yes, Tesla P40 24GB with 24 GB of usable memory can run Qwen 3.5 27B at Q4_K_M (~16.5 GB) with ~7 GB headroom for context and runtime. Quality at Q4 is very close to full precision for most tasks. Run it with: ollama run qwen3.5:27b
With 24 GB on Tesla P40 24GB, Q4_K_M is the sweet spot for 27B-35B models, Q6_K for 14B models, and Q8_0 for 8B-9B models. The general rule: use the highest quantization that fits with at least 2-3 GB headroom for KV cache and runtime.
Tesla P40 24GB has enough memory for many local LLMs, but bandwidth still matters a lot for real speed. Once a model fits, a faster-memory GPU can feel significantly better than a slower setup with similar capacity.
Compare with similar
