16 GB VRAM (+4)560 GB/s (+224)
C
Raises estimated decode speed by about 53%.34.4 tok/s decode
Raises estimated decode speed by about 53%.
Adds memory headroom for longer context windows and future model growth.
~$349 MSRP
Can gemma 3 12b it run on Intel Arc A730M 12GB?
YES — Tight Fit
C49Usable
○Estimated from fit model
gemma 3 12b it needs ~10.8 GB VRAM. Intel Arc A730M 12GB has 12.0 GB. With Q4_K_M quantization, expect ~23 tok/s.
Runtime: llama.cppCapacity: TightBandwidth: LowStack: StandardBottleneck: Balanced
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.
Select quantization to explore
Q4_K_M (Medium quality) — 10.8 GB, 22.5 tok/s, Tight fit
10.8 GB required12.0 GB available
90% VRAM used
Fit status
Tight fit
Decode
22.5 tok/s
TTFT
8608 ms
Safe context
29K
Memory
10.8 GB / 12.0 GB
Memory breakdown
Weights7.3 GB
KV Cache1.4 GB
Runtime0.9 GB
Headroom1.2 GB
See how fast it feels
See how fast it feelsgemma 3 12b it on Intel Arc A730M 12GB
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: 22.5 tok/s decode · 8.6s TTFT (warm) · 56 tok/s prefill
What limits this setup
The raw memory story may look fine, but the software ecosystem is still a constraint here.
Runtime ecosystem is narrower than CUDA
Intel GPUs can look attractive on memory per dollar, but local AI tooling, kernels, and model coverage are still broader and easier on CUDA today.
Best improvement path
Prefer CUDA if you want the path of least resistance
If your goal is maximum runtime coverage, easier troubleshooting, and better support for new local AI releases, CUDA is usually still the safer upgrade path.
Performance by workload
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Chat | C | Tight fit | 22.5 tok/s | 4695 ms | 29K |
| Coding | C | Tight fit | 22.5 tok/s | 8608 ms | 29K |
| Agentic Coding | C | Runs with offload | 16.2 tok/s | 17382 ms | 29K |
| Reasoning | C | Tight fit | 22.5 tok/s | 10173 ms | 29K |
| RAG | C | Runs with offload | 16.2 tok/s | 21727 ms | 29K |
Quantization options
How gemma 3 12b it (12B params) fits at each quantization level on Intel Arc A730M 12GB (12.0 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 4.7 GB | Low | C52 |
Q3_K_S | 3 | 5.9 GB | Low | C52 |
NVFP4 | 4 |
Get started
Copy-paste commands to run gemma 3 12b it on your machine.
Run
lms load hf-maziyarpanahi--gemma-3-12b-it-gguf && lms server startUpgrade options