MetaLlama 3.2 1B Instruct on AMD Radeon RX 9070
Yes — AMD Radeon RX 9070 handles Llama 3.2 1B Instruct well at Q8_0 — 212 tok/s. Solid daily-driver performance on 16 GB VRAM.
Model Size
1.24B
Device VRAM
16 GB
Bandwidth
640 GB/s
Quantization
Q8_0
Performance by Quantization
OwnRig currently has one published compatibility entry for Llama 3.2 1B Instruct on AMD Radeon RX 9070 at Q8_0. This is the best supported pairing we can stand behind today.
| Quantization | Speed | TTFT | Fits in VRAM | Rating | Confidence |
|---|---|---|---|---|---|
| Q8_0 | 212 tok/s | 20ms | ✓ Yes | Good | estimated |
Notes
Q8_0
RDNA 4 card with improving ROCm support; Vulkan remains the safer default in llama.cpp. Q8_0 fits in VRAM on this card.
About Llama 3.2 1B Instruct
Llama 3.2 1B Instruct (1.24B) is a chat, coding model. The smallest Llama model. Runs on integrated GPUs and even CPUs. Useful for basic classification, simple Q&A, and as a draft model for speculative decoding. Limited reasoning capability.
View all Llama 3.2 1B Instruct hardware options →About AMD Radeon RX 9070
AMD Radeon RX 9070 has 16 GB at 640 GB/s. Street price: $549.
See all models AMD Radeon RX 9070 can run →Estimate method: Estimated from rtx-4060-ti-16gb compatibility patterns plus ROCm/Vulkan adjustments. Reference hardware source: github.com (2026-04-18)
Performance varies by driver version, inference engine, quantization method, context length, and system configuration. Figures shown are estimates based on community benchmarks and may not reflect your exact setup. Product names are trademarks of their respective owners. OwnRig is independent and not affiliated with any hardware or AI model provider.