Skip to content

[Engine] Inkling-2-Mini SQ-10 segfaults in CPU expert computation on RTX 4070 SUPER #6569

Open

Description

@mskutin

Area

[Engine]

Desktop app, Terminal or source?

Terminal (the strata command)

What happened

Inkling-2-Mini SQ-10 loaded and completed the one-token startup check. On a 7,424-token request within the configured 8,192-token context, the engine exited with -11 (SIGSEGV) before generating a reply. The expected result is that the engine processes the prompt and returns a reply.

A standalone benchmark reproduced the segfault with a synthetic 128-token prompt, FP16 KV, and drafting disabled. It exited with status 139 before producing a result.

Smallest steps or command that reproduce it

With SQ-10 already installed, run the packaged engine directly:

strata_dir="$HOME/.local/share/strata"
engine="$strata_dir/components/shapes/inkling2/engine/0.1.65-linux-x86_64-cuda-cuda13/strata-inkling"
model="$strata_dir/Strata-data/models/inkling-2-mini/Inkling-2-Mini-SQ-10.strata"

(
  ulimit -c 0
  timeout --signal=TERM --kill-after=5s 45s "$engine" \
    --bench --native "$model" --max-context 8192 --parallel 1 \
    --kv fp16 --vram-reserve-mib 2048 --ram-experts-gb 3 \
    --draft off --prompt-len 128 --gen 8
)

Reproduction status

Reproduced locally

Strata version

Terminal 0.2.19, strata-inkling 0.1.65, Linux x86_64 CUDA 13 package, engine commit c6f6427d9de9.

Graphics card, memory and system

One NVIDIA GeForce RTX 4070 SUPER (12,282 MiB reported), driver 615.78.08, AMD Ryzen 7 7800X3D, 61 GiB usable RAM, Omarchy 4.5.0, Linux 7.2.15-1-omarchy x86_64.

Model, size and settings

Inkling-2-Mini SQ-10, 8K context, one request, vision off. The original serving run used FP8 KV, DSpark, a 2,048 MiB VRAM reserve, and Strata-generated expert budgets of 8.60 GB VRAM and 0.00 GB RAM. The engine reported 2,339 experts in VRAM and 733 on the host.

Logs

Standalone benchmark output:

[inkling2-ep] Loaded strata (SQ-10) | gpus=1 (p2p=pinned, ep_split=0.50) | experts: 2312 VRAM (GPU0=6.16 GiB, GPU1=0.00 GiB), 760 host (doorbell=off, numa=1, thr=0) | kv=fp16
timeout: the monitored command dumped core

The original serving crash recorded SIGSEGV (SEGV_MAPERR). Top frames from coredumpctl, with names demangled and shortened:

strata::quant::sk::gemm_sk_q8_1
strata::inkling2::eval_inkling2_expert_batch
EpDualGpuCudaBackend::forward_batch
ink2::run_serve_loop

The kernel logged segfault at 900 in strata-inkling. No OOM kill appeared in the kernel journal around the serving crash. strata doctor --live --profile launch-inkling-sq10-8k found the engine, model and tokenizer files and reported Live GPU probe: ok.

Related model support discussion: #6442.

Activity

  1. mskutin commented on Nov 22, 2026

    @mskutin
    Author

    Additional reproduction and serving timeline from the same machine and engine build.

    FP8 benchmark reproduction

    The engine also exited with status 139 on a synthetic 512-token prompt with FP8 KV, drafting disabled, and an explicit 3 GB RAM expert budget. The FP16 reproduction in the issue body used a separate benchmark run.

    strata_dir="$HOME/.local/share/strata"
    engine="$strata_dir/components/shapes/inkling2/engine/0.1.65-linux-x86_64-cuda-cuda13/strata-inkling"
    model="$strata_dir/Strata-data/models/inkling-2-mini/Inkling-2-Mini-SQ-10.strata"
    
    (
      ulimit -c 0
      timeout --signal=TERM --kill-after=5s 60s "$engine" \
        --bench --native "$model" --max-context 8192 --parallel 1 \
        --kv fp8 --vram-reserve-mib 2048 --vram-experts-gb 8.60 \
        --ram-experts-gb 3 --draft off --prompt-len 512 --gen 16
    )
    [inkling2-ep] Loaded strata (SQ-10) | gpus=1 (p2p=pinned, ep_split=0.50) | experts: 2312 VRAM (GPU0=6.16 GiB, GPU1=0.00 GiB), 760 host (doorbell=off, numa=1, thr=0) | kv=fp8
    timeout: the monitored command dumped core
    

    Original serving timeline

    The one-token startup check completed. The next request produced these server log entries.

    [strata] reading the prompt: 7,424 tokens, 10 s so far
    [strata] reading the prompt: 7,424 tokens, 20 s so far
    [strata] reading the prompt: 7,424 tokens, 30 s so far
    [strata] the engine stopped unexpectedly (exit code -11).
    [strata] done: 0 tokens in 39 s (no speed) (error, cancel=False)
    

    The engine-stop line is shortened. The backtrace in the issue body is from this serving crash.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions