Can DeepSeek R1 Distill 70B run on NVIDIA L20 48GB?

BARELY — Tight on Memory

B63Good

○Estimated from fit model

DeepSeek R1 Distill 70B needs ~53.6 GB VRAM. NVIDIA L20 48GB has 48.0 GB. With Q4_K_M quantization, expect ~10 tok/s.

Runtime: OllamaCapacity: OffloadBandwidth: HighStack: BasicBottleneck: Host offload

Operating mode

Choose the run profile you care about

Interactive favors responsiveness, while light API and scale-out lean harder on serving readiness. The fit stays the same, but the recommendation lens changes.

Current mode

Balanced

Balanced for general local use. Keeps the ranking neutral across personal and serving workflows.

Capabilities:

Select quantization to explore

Q4_K_M (Medium quality) — 53.6 GB, 9.6 tok/s, Very compromised (needs ~4.4 GB host RAM)

53.6 GB required48.0 GB available

112% VRAM needed

5.6 GB over capacity — needs offload or smaller quantization

Fit status

Very compromised (needs ~4.4 GB host RAM)

Decode

9.6 tok/s

TTFT

20258 ms

Safe context

Memory

53.6 GB / 48.0 GB

Offload

10%

Memory breakdown

Weights42.7 GB

KV Cache4.9 GB

Runtime1.2 GB

Headroom4.8 GB

See how fast it feels

See how fast it feelsDeepSeek R1 Distill 70B on NVIDIA L20 48GB
1st promptCold start — includes initialization
>Write a Python function to sort a list
Here's a Python function that sorts a list using quicksort:

def quicksort(arr):
 if len(arr) <= 1:
 return arr
 pivot = arr[len(arr) // 2]
 left = [x for x in arr if x < pivot]
 middle = [x for x in arr if x == pivot]
 right = [x for x in arr if x > pivot]
 return quicksort(left) + middle + quicksort(right)

# Example usage
numbers = [3, 6, 8, 10, 1, 2, 1]
print(quicksort(numbers)) # [1, 1, 2, 3, 6, 8, 10]

The algorithm picks a pivot, partitions the list into smaller, equal, and larger elements, then recursively sorts each partition and concatenates the results.
2nd promptWarm — model ready, no init overhead
>Now add type hints and a docstring
def quicksort(arr: list[int]) -> list[int]:
 """Sort a list using the quicksort algorithm.

 Args:
 arr: List of integers to sort.

 Returns:
 A new sorted list.
 """
 if len(arr) <= 1:
 return arr
 pivot = arr[len(arr) // 2]
 left = [x for x in arr if x < pivot]
 middle = [x for x in arr if x == pivot]
 right = [x for x in arr if x > pivot]
 return quicksort(left) + middle + quicksort(right)

Added type hints for the input and return type, plus a Google-style docstring.
Estimated: 9.6 tok/s decode · 20.3s TTFT (warm) · 24 tok/s prefill

What limits this setup

It fits through host-memory offload, and offload is the main reason performance drops.

CPU or host-memory offload is active

About 10% of the working set spills out of accelerator memory, which usually hurts latency and sustained decode throughput.

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.

Best improvement path

Remove offload with more accelerator memory

Prioritize a GPU or unified-memory tier that fits the whole model natively. Removing offload usually helps more than small compute gains.

Buy headroom, not only minimum fit

A slightly larger memory tier gives you safer context growth and makes the recommendation more future-proof.

Increase host RAM if you keep offloading

This setup may need roughly 4.4 GB of extra host RAM just for the offloaded portion, before OS and other tools.

Performance by workload

Workload	Grade	Fit	Decode	TTFT	Context
Chat	B	Runs with offload (needs ~2.6 GB host RAM)	10.5 tok/s	10017 ms	4K
Coding	B	Very compromised (needs ~4.4 GB host RAM)	9.6 tok/s	20258 ms	4K
Agentic Coding	F	Too heavy	8.0 tok/s	35404 ms	4K
Reasoning	B	Very compromised (needs ~4.4 GB host RAM)	9.6 tok/s	23941 ms	4K
RAG	F	Too heavy	8.0 tok/s	44254 ms	4K

Quantization options

How DeepSeek R1 Distill 70B (70B params) fits at each quantization level on NVIDIA L20 48GB (48.0 GB usable).

Quant	Bits	VRAM	Quality	Fit
Q2_K	2	27.3 GB	Low	A74
Q3_K_SBest for your GPU	3	34.3 GB	Low	A74
NVFP4	4	39.2 GB	Medium	F0
Q4_K_M	4	42.7 GB	Medium	F0
Q5_K_M	5	50.4 GB	High	F0
Q6_K	6	57.4 GB	High	F0
Q8_0	8	74.9 GB	Very High	F0
F16	16	143.5 GB	Maximum	F0

Get started

Copy-paste commands to run DeepSeek R1 Distill 70B on your machine.

Run

ollama run deepseek-r1:70b

Upgrade options

Hardware that runs DeepSeek R1 Distill 70B well

👁 NVIDIA
NVIDIA A16 64GBBudget pick

64 GB VRAM (+16)

Removes host-memory offload, which is usually the single biggest latency and throughput win.11.9 tok/s decode

Removes host-memory offload, which is usually the single biggest latency and throughput win.

Adds memory headroom for longer context windows and future model growth.

~$6,500 MSRP

👁 NVIDIA
RTX PRO 6000 Blackwell Workstation Edition 96GBBest value

96 GB VRAM (+48)1792 GB/s (+928)

Removes host-memory offload, which is usually the single biggest latency and throughput win.38.3 tok/s decode

Removes host-memory offload, which is usually the single biggest latency and throughput win.

Raises estimated decode speed by about 299%.

~$9,999 MSRP

👁 NVIDIA
RTX PRO 6000 Blackwell Server Edition 96GBNVIDIA upgrade

96 GB VRAM (+48)1597 GB/s (+733)

Removes host-memory offload, which is usually the single biggest latency and throughput win.34.2 tok/s decode

Removes host-memory offload, which is usually the single biggest latency and throughput win.

Raises estimated decode speed by about 256%.

~$9,999 MSRP

Frequently asked questions

See all results for NVIDIA L20 48GB See all hardware for DeepSeek R1 Distill 70B

URL: https://willitrunai.com/can-run/deepseek-r1-70b-on-l20-48gb

⇱ DeepSeek R1 Distill 70B on NVIDIA L20 48GB? YES

Can DeepSeek R1 Distill 70B run on NVIDIA L20 48GB?

BARELY — Tight on Memory

Choose the run profile you care about

Memory breakdown

See how fast it feels

What limits this setup

Best improvement path

Performance by workload

Quantization options

Get started

Hardware that runs DeepSeek R1 Distill 70B well

Frequently asked questions