Rent a GPU

Can the NVIDIA RTX PRO 6000 Blackwell run Phi-3.5 Mini 3.8B?

RTX PRO 6000 Blackwell × Phi-3.5 Mini 3.8B8K context
YesRuns fully on the GPU
Memory at FP1611.5 GB of 96 GB
Weights 7.6 GBKV cache 3.2 GBRuntime 0.6 GB
QuantFP16Full quality
Speed~141tokens per second
Max context128Kthe model's full window
  1. FP16
  2. Q8_0
  3. Q6_K
  4. Q5_K_M
  5. Q4_K_M
  6. Q3_K_M
  7. Q2_K

Yes. The NVIDIA RTX PRO 6000 Blackwell can run Phi-3.5 Mini 3.8B fully in VRAM at FP16 (full quality), using about 11.5 GB of the 96 GB available, at roughly 141 tokens/second.

Try other context lengths

Run it with Ollama

With Ollama installed, one command downloads and starts Phi-3.5 Mini 3.8B:

ollama run phi3.5:3.8b

Ollama's default download is Q4_K_M (about 6.1 GB here). You have room for FP16, which is full quality: pick that GGUF in LM Studio or from Hugging Face.

New to this? Step-by-step setup for the RTX PRO 6000 Blackwell

Run it in the cloud

No NVIDIA RTX PRO 6000 Blackwell? Run Phi-3.5 Mini 3.8B in the cloud.

Rent a cloud GPU by the hour. No upfront cost, stop any time.

Rent a GPU on RunPodRent a GPU on Vast.ai

Memory needed by quantization

At 8,192 tokens of context. The black line marks the 96 GB the NVIDIA RTX PRO 6000 Blackwell can use.

QuantQualityMemoryFits?Speed
FP16Full quality11.5 GBFits~141 tok/s
Q8_0Near-lossless7.9 GBFits~265 tok/s
Q6_KExcellent7 GBFits~343 tok/s
Q5_K_MVery good6.5 GBFits~396 tok/s
Q4_K_MGood, most popular6.1 GBFits~464 tok/s
Q3_K_MNoticeably degraded5.7 GBFits~576 tok/s
Q2_KHeavily degraded5.3 GBFits~751 tok/s

Measured speed from real RTX PRO 6000 Blackwell owners

Tokens per second that people measured on their own NVIDIA RTX PRO 6000 Blackwell, next to our estimate.

No one has measured Phi-3.5 Mini 3.8B on the NVIDIA RTX PRO 6000 Blackwell yet. The speeds above are estimates. Run the benchmark and yours becomes the first real number on this page.

Have an RTX PRO 6000 Blackwell? Add your real speed

One command times Phi-3.5 Mini 3.8B in your own Ollama, shows you exactly what it would send, and asks before sending anything. No sign-up, no personal data.

macOS and Linux

curl -fsSL https://gpufit.dev/bench.sh | sh -s -- phi3.5:3.8b

Windows (PowerShell)

$env:GPUFIT_MODEL="phi3.5:3.8b"; irm https://gpufit.dev/bench.ps1 | iex

How the benchmark works

GPUs that run Phi-3.5 Mini 3.8B comfortably

Models that run well on the NVIDIA RTX PRO 6000 Blackwell

Frequently asked questions

Can NVIDIA RTX PRO 6000 Blackwell run Phi-3.5 Mini 3.8B?

Yes. The NVIDIA RTX PRO 6000 Blackwell can run Phi-3.5 Mini 3.8B fully in VRAM at FP16 (full quality), using about 11.5 GB of the 96 GB available, at roughly 141 tokens/second.

How much VRAM does Phi-3.5 Mini 3.8B need?

About 6.1 GB at Q4_K_M with 8,192 tokens of context, and about 11.5 GB at full FP16 precision.

How fast is Phi-3.5 Mini 3.8B on the NVIDIA RTX PRO 6000 Blackwell?

Roughly 141 tokens per second at FP16, estimated from the NVIDIA RTX PRO 6000 Blackwell's 1792 GB/s memory bandwidth.