Can embeddinggemma 300M run on AMD Instinct MI100 32GB?

YES — Runs Great

D35Poor

○Estimated from fit model

embeddinggemma 300M needs ~4.4 GB VRAM. AMD Instinct MI100 32GB has 32.0 GB. With Q6_K quantization, expect ~4 tok/s.

Runtime: llama.cppCapacity: RoomyBandwidth: HighStack: StandardBottleneck: Memory bandwidth

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

Q6_K (High quality) — 4.4 GB, 4.2 tok/s, Runs well

4.4 GB required32.0 GB available

14% VRAM used

Fit status

Runs well

Decode

4.2 tok/s

TTFT

46095 ms

Safe context

4.4M

Memory

4.4 GB / 32.0 GB

Memory breakdown

Weights0.2 GB

KV Cache0.1 GB

Runtime0.9 GB

Headroom3.2 GB

See how fast it feels

See how fast it feelsembeddinggemma 300M on AMD Instinct MI100 32GB
1st promptCold start — includes initialization
>What is local AI inference?
Local AI inference means running an AI language model directly on your own hardware — your laptop, desktop, or server — instead of sending requests to a remote cloud API.

When you run inference locally the model weights are loaded into your GPU or unified memory. Each token you generate requires reading those weights from memory, so memory bandwidth is the main bottleneck for decode speed.

Key benefits of running locally:
 - Full privacy: your prompts never leave your machine
 - No per-token cost or rate limits
 - Works offline once the model is downloaded
 - Latency depends only on your hardware
2nd promptWarm — model ready, no init overhead
>How much VRAM do I need?
It depends on the model size and quantization level. A rough rule of thumb:

 Model size Q4 (4-bit) Q8 (8-bit) FP16
 7B params ~4.3 GB ~7.5 GB ~14 GB
 13B params ~7.9 GB ~13.9 GB ~26 GB
 70B params ~42.7 GB ~74.9 GB ~140 GB

Most people use 4-bit quantization (Q4_K_M) which gives 90-95% of full quality at a fraction of the memory. A 24 GB GPU can comfortably run most 7B-13B models.
Estimated: 4.2 tok/s decode · 46.1s TTFT (warm) · 11 tok/s prefill

What limits this setup

This model fits, but memory bandwidth is the part holding decode speed back.

Throughput will feel slow

Estimated decode speed is only 4.2 tok/s, so this is more of a technical fit than a comfortable daily-driver setup.

Best improvement path

Prioritize bandwidth, not only capacity

If this workload feels slow, the next useful step is often a GPU tier with materially faster memory bandwidth rather than only a small bump in capacity.

Performance by workload

Workload	Grade	Fit	Decode	TTFT	Context
Chat	D	Runs well	4.2 tok/s	25143 ms	2.2M
Coding	D	Runs well	4.2 tok/s	46095 ms	4.4M
Agentic Coding	D	Runs well	4.2 tok/s	67048 ms	8.8M
Reasoning	D	Runs well	4.2 tok/s	54476 ms	4.4M
RAG	D	Runs well	4.2 tok/s	83810 ms	8.8M

Quantization options

How embeddinggemma 300M (0.30000001192092896B params) fits at each quantization level on AMD Instinct MI100 32GB (32.0 GB usable).

Quant	Bits	VRAM	Quality	Fit
Q2_K	2	0.1 GB	Low	C43
Q3_K_S	3	0.1 GB	Low	C43
NVFP4	4	0.2 GB	Medium	C43
Q4_K_M	4	0.2 GB	Medium	C43
Q5_K_M	5	0.2 GB	High	C43
Q6_K	6	0.2 GB	High	C43
Q8_0	8	0.3 GB	Very High	C43
F16Best for your GPU	16	0.6 GB	Maximum	C43

Get started

Copy-paste commands to run embeddinggemma 300M on your machine.

Run

docker run --rm -it ghcr.io/ggerganov/llama.cpp:full \
 --hf-repo "ggml-org/embeddinggemma-300M-GGUF" \
 --hf-file "embeddinggemma-300M-GGUF-Q6_K.gguf" \
 -c 4096 -ngl 99

Upgrade options