VOOZH about

URL: https://willitrunai.com/can-run/command-r-plus-104b-on-instinct-mi210-64gb

⇱

Can Command R+ 104B run on AMD Instinct MI210 64GB?

BARELY — Tight on Memory

C54Usable
○Estimated from fit model

Command R+ 104B needs ~74.2 GB VRAM. AMD Instinct MI210 64GB has 64.0 GB. With Q4_K_M quantization, expect ~11 tok/s.

Runtime: llama.cppCapacity: OffloadBandwidth: HighStack: StandardBottleneck: Host offload
Share:

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) — 74.2 GB, 10.5 tok/s, Very compromised (needs ~8.7 GB host RAM)
74.2 GB required64.0 GB available
116% VRAM needed

10.2 GB over capacity — needs offload or smaller quantization

Fit status

Very compromised (needs ~8.7 GB host RAM)

Decode

10.5 tok/s

TTFT

18434 ms

Safe context

4K

Memory

74.2 GB / 64.0 GB

Offload

10%

Memory breakdown

Weights63.4 GB
KV Cache3.4 GB
Runtime0.9 GB
Headroom6.4 GB

See how fast it feels

See how fast it feelsCommand R+ 104B on AMD Instinct MI210 64GB
1st promptCold start — includes initialization
>Solve: If a train travels at 60 mph for 2.5 hours, how far does it go?
Let me work through this step by step. Given information: - Speed = 60 mph (miles per hour) - Time = 2.5 hours Formula: Distance = Speed x Time Calculation: Distance = 60 mph x 2.5 h = 150 miles The train travels 150 miles. Double-check: at 60 mph the train covers 60 miles each hour. In half an hour it covers 30 miles. So in 2.5 hours: 60 + 60 + 30 = 150 miles. Confirmed.
2nd promptWarm — model ready, no init overhead
>What if the train stops for 20 minutes halfway?
The stop doesn't change the total distance covered while moving, but it affects total trip time. Breakdown: - First half: 1.25 hours at 60 mph = 75 miles - Stop: 20 minutes (no distance) - Second half: 1.25 hours at 60 mph = 75 miles Total distance: 150 miles (same) Total time: 2.5 h + 0.33 h = 2.83 hours Average speed: 150 / 2.83 = 53 mph The distance stays the same but average speed drops to 53 mph because of the stop.
Estimated: 10.5 tok/s decode · 18.4s TTFT (warm) · 26 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 8.7 GB of extra host RAM just for the offloaded portion, before OS and other tools.

Performance by workload

WorkloadGradeFitDecodeTTFTContext
ChatCVery compromised (needs ~7.4 GB host RAM)11.0 tok/s9574 ms4K
CodingCVery compromised (needs ~8.7 GB host RAM)10.5 tok/s18434 ms4K
Agentic CodingFToo heavy9.6 tok/s29481 ms4K
ReasoningCVery compromised (needs ~8.7 GB host RAM)10.5 tok/s21786 ms4K
RAGFToo heavy9.6 tok/s36852 ms

Quantization options

How Command R+ 104B (104B params) fits at each quantization level on AMD Instinct MI210 64GB (64.0 GB usable).

QuantBitsVRAMQualityFit
Q2_KBest for your GPU
2
40.6 GB
LowB65
Q3_K_S
3
51.0 GB
LowF0

Get started

Copy-paste commands to run Command R+ 104B on your machine.

Run

ollama run command-r-plus

Upgrade options

Hardware that runs Command R+ 104B well

AMD Instinct MI350X 288GBBudget pick
288 GB VRAM (+224)8000 GB/s (+6362)
B
Removes host-memory offload, which is usually the single biggest latency and throughput win.100.1 tok/s decode

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

Raises estimated decode speed by about 853%.

~$8,000 MSRP

AMD Instinct MI300A 128GBBest value
128 GB VRAM (+64)5300 GB/s (+3662)
A
Removes host-memory offload, which is usually the single biggest latency and throughput win.63.6 tok/s decode

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

Raises estimated decode speed by about 506%.

~$12,000 MSRP

AMD Instinct MI250X 128GBAMD upgrade
128 GB VRAM (+64)3200 GB/s (+1562)
A
Removes host-memory offload, which is usually the single biggest latency and throughput win.42.8 tok/s decode

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

Raises estimated decode speed by about 308%.

~$15,000 MSRP

Frequently asked questions

See all results for AMD Instinct MI210 64GBSee all hardware for Command R+ 104B
4K
NVFP4
4
58.2 GB
Medium
F0
Q4_K_M
4
63.4 GB
MediumF0
Q5_K_M
5
74.9 GB
HighF0
Q6_K
6
85.3 GB
HighF0
Q8_0
8
111.3 GB
Very HighF0
F16
16
213.2 GB
MaximumF0

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.