Skip to content
Run LLM Locally

Phi-4 14B on the RTX 6000 Ada

With 48 GB of VRAM, the RTX 6000 Ada runs Phi-4 14B at Q4_K_M entirely on the GPU with an 8K context (about 11.2 GB needed). At Q8_0 it fits entirely.

ShareFacebookXWhatsAppLinkedInEmail

Phi-4 14B

Q4_K_M · 8K context

Excellent

The model needs about 11.2 GB and fits in the 46.6 GB available, with 35.3 GB to spare.

Your hardware
RTX 6000 Ada 48 GB · 64 GB RAM
Estimated memory
~11.2 GB estimate
VRAM (usable memory)
46.6 GB
GPU only
Yes
CPU/GPU offload
Not required
Generation speed
64–89 tok/s estimate
Largest context in fast memory
16K tokens
Confidence: low

Based on: memory bandwidth (960 GB/s) and model size

Memory breakdown

~11.2 GB (10.8 GB–11.8 GB) needed · estimate

Fits entirely in VRAM · 35.3 GB left

  • Model weights8.4 GB
  • KV cache1.6 GB
  • Runtime overhead0.7 GB
  • Safety margin0.5 GB
Assumptions: context 8K tokens · KV cache f16 · llama.cpp (CUDA) · flash attention on · weights = real file size

EstimateData updated 24 days agoSource: Run LLM Locally engine, Hugging Face config.json + model card

What makes this estimate less reliable

  • no benchmark for this hardware and model yet

Other quantizations on the RTX 6000 Ada

Phi-4 14B on RTX 6000 Ada + 64 GB RAM · 8K context
Phi-4 14B Q8_08K~17.6 GB · 41–55 tok/s estimateExcellent
Phi-4 14B Q6_K8K~14.2 GB · 51–70 tok/s estimateExcellent
Phi-4 14B Q5_K_M8K~12.8 GB · 56–78 tok/s estimateExcellent
Phi-4 14B Q4_K_M8K~11.2 GB · 64–89 tok/s estimateExcellent
Phi-4 14B IQ4_XS8K~10.2 GB · 70–99 tok/s estimateExcellent
Phi-4 14B Q3_K_M8K~9.6 GB · 74–106 tok/s estimateExcellent

Alternatives

GPUs that run Phi-4 14B Q4_K_M entirely in VRAM

Other models that fit the RTX 6000 Ada

Where to buy a RTX 6000 Ada

Prices and links follow the market you choose.

RTX 6000 Ada Generation

48 GB GDDR6 ECC · 960 GB/s · 300 W

ExcellentRuns Phi-4 14B Q4_K_M entirely in memory

Price not available

Availability unknown

View on amazon.com

Affiliate link

No current offer imported for this market: the link opens a marketplace search.

Frequently asked questions

Can the RTX 6000 Ada run Phi-4 14B?
The model needs about 11.2 GB and fits in the 46.6 GB available, with 35.3 GB to spare.
Which quantization should I use on the RTX 6000 Ada?
Q8_0 is the highest-precision quantization that fits entirely in VRAM at 8K context. ~17.6 GB / 46.6 GB.
What is the largest context on the RTX 6000 Ada?
At Q4_K_M, about 16K tokens fit entirely in VRAM with an f16 KV cache. A q8_0 KV cache roughly doubles that.
How fast is Phi-4 14B (RTX 6000 Ada)?
We estimate 64–89 tok/s (estimate, confidence low). Based on: memory bandwidth (960 GB/s) and model size.