Recipe

qwen3-6-27b-q4-0-intel-arc-pro-b70-32gb-llama-cpp-tp4-wepbjtb7

qwen3-6-27b-q4-0-intel-arc-pro-b70-32gb-llama-cpp-tp4-wepbjtb7

Observed LocalMaxxing leaderboard run. Evidence for compatibility, not an executable launch contract.

Record

Status
candidate
Source
localmaxxing
Engine
llama.cpp
Engine version
db44417-sycl-q8-fuseadd-reshapeadd-postreboot
Accelerators
4
Tensor parallel
4
Context tokens
1,024
chat
unknown
reasoning
unknown
tools
unknown
vision
unknown

Hugging Face model card

Identity

https://huggingface.co/Qwen/Qwen3.6-27B
Repository
Qwen/Qwen3.6-27B
Status
known
Link type
Exact Hub repository

Public Hugging Face repository confirmed by the Hub API.

Candidate: useful compatibility or speed evidence. The registry does not offer Run until promotion requirements are met.

Observed configuration

llama.cpp

Evidence only · candidate · reference

Candidate evidence — not a Run contract

Source
https://www.localmaxxing.com/en/runs/cmota1fpx0001l404wepbjtb7

No tokenized launch fields. This is measured or documented compatibility, not a Docker launch.

Measured speed

ConcurrencyContextPrefillDecodeTTFT msStatusSweep
11,02434.4observedqwen3-6-27b-q4-0-intel-arc-pro-b70-32gb-llama-cpp-tp4-wepbjtb7-sweep

Remaining fields

Identity, launch, related records, and measured speed are shown above. This is the rest of the normalized record.

hardware count
4
hardware id
intel-arc-pro-b70-32gb
id
qwen3-6-27b-q4-0-intel-arc-pro-b70-32gb-llama-cpp-tp4-wepbjtb7
model instance id
qwen-qwen3-6-27b--q4-0
recipe source
localmaxxing
schema version
local-ai-registry/v1
speed sweep ids
qwen3-6-27b-q4-0-intel-arc-pro-b70-32gb-llama-cpp-tp4-wepbjtb7-sweep
status
candidate

capabilities

engine

name
llama.cpp
version
db44417-sycl-q8-fuseadd-reshapeadd-postreboot

serving

max context tokens
1,024
tensor parallel
4
Provenance & metadata (3)

facts

capabilities.chat · provenance · captured at
2026-08-30T09:10:02Z
capabilities.chat · reason capability-not-verified
capabilities.chat · state unknown
capabilities.reasoning · provenance · captured at
2026-08-30T09:10:02Z
capabilities.reasoning · reason capability-not-verified
capabilities.reasoning · state unknown
capabilities.tools · provenance · captured at
2026-08-30T09:10:02Z
capabilities.tools · reason capability-not-verified
capabilities.tools · state unknown
capabilities.vision · provenance · captured at
2026-08-30T09:10:02Z
capabilities.vision · reason capability-not-verified
capabilities.vision · state unknown
engine.graph mode · provenance · captured at
2026-08-30T09:10:02Z
engine.graph mode · reason runtime-detail-not-published
engine.graph mode · state unknown
serving.kv cache tokens · provenance · captured at
2026-08-30T09:10:02Z
serving.kv cache tokens · reason kv-cache-capacity-not-published
serving.kv cache tokens · state unknown
serving.max concurrency · provenance · captured at
2026-08-31T23:03:15Z
serving.max concurrency · reason server-capacity-not-evidenced
serving.max concurrency · state unknown

metadata

localmaxxing · batch size
1
localmaxxing · hardware label
Intel Arc Pro B70
localmaxxing · notes
Diagnostic negative-scaling result. Qwen3.6 27B Q4_0 GGUF on 4x Intel Arc Pro B70 using llama.cpp SYCL Level Zero selector 0,1,2,3. Same quality-preserving reshape-through-ADD software path as the 3x best: Q8 cache, async copy, single-kernel allreduce, event barrier, f16 KV, no speculative decode, no power-limit changes. Prompt 102.210613 tok/s; decode 34.375523 tok/s. Clean post-reboot run; dmesg after run showed no new reset/AER/errors. JSONL: /home/steve/bench-results/qwen36-q4_0-gguf/sycl-k617023-postreboot-reshapeadd-quad0123-p512n512-r1-20260505T234410Z.jsonl
localmaxxing · run id
cmota1fpx0001l404wepbjtb7

provenance

captured at
2026-08-30T09:10:02Z

sources

captured atkindurl
2026-08-30T09:10:02Znormalized-recipewww.localmaxxing.com/en/runs/cmota1fpx0001l404wepbjtb7