Can I Run Ternary Bonsai 27B on Intel Arc B580?

Written by Jakub Rusinowski · Last updated July 15, 2026

Yes. Ternary Bonsai 27B needs 8.3 GB at 8K context (5.8 GB weights + 2.5 GB KV cache/overhead) at Ternary (1.58-bit, ~1.71 bpw) and the Intel Arc B580 has 12 GB, leaving about 3.7 GB of headroom, at roughly 35 tok/s (estimated) with room for up to 16K of context.

See what else this hardware can run →

The numbers

Ternary Bonsai 27B on the Intel Arc B580: VRAM by quantization

Assumes an 8K-token context with an f16 KV cache, on an estimated architecture — this model publishes no config we can read. A longer window needs more; a quantized KV cache needs less. “Max context” is the largest window that still fits in 12 GB. Figures are estimates from parameter count, quantization and memory bandwidth — the analyzer lets you tune KV-cache quant and context.

Context costs VRAM too. At Ternary (1.58-bit, ~1.71 bpw) on the Intel Arc B580: 8K 8.3 GB · 32K 13.6 GB ✗ · 128K 34.5 GB ✗. Past 32K it no longer fits 12 GB — a q8_0 KV cache buys roughly half of that back, which the calculator will price for you.

What to do next

Nothing to change — run it at Ternary (1.58-bit, ~1.71 bpw). If you want more capability, check whether the next size up also fits.

No compatible GPU? Bonsai 27B on 16 GB of system RAM, CPU only: runs comfortably.

Running Ternary Bonsai 27B on this card:

Other Bonsai 27B sizes on the Intel Arc B580

• 1-bit Bonsai 27B — runs (6.3 GB)

Ternary Bonsai 27B on nearby hardware

• Intel Arc B570 (10 GB) — runs

All Bonsai 27B sizes on the Intel Arc B580 | Ternary Bonsai 27B model page | Intel Arc B580 GPU page | VRAM calculator | Check your hardware