# llama_qwen3.6_A3B

llama.cpp CUDA runner for **Qwen3.6-35B-A3B** (UD-Q4_K_M) at **128k context** on 4n4rch02.

Port **18008**, OpenAI-compatible API at `http://192.168.3.189:18008/v1`.

## Host

- 4n4rch02 (192.168.3.189), Arch Linux, driver 610.43.03
- AMD Ryzen 7 3700X — **8 physical cores** / 16 SMT threads
- 62 GiB RAM, 15 GiB swap
- CUDA0: RTX 3060 12 GB — **shared with the KDE/Wayland desktop** (~1.1 GiB at idle)
- CUDA1: RTX 3080 10 GB — dedicated

## Modell

`/mnt/2TSAM990nvme/docker-volume-outsource/llm-models/hf/hub/models--unsloth--Qwen3.6-35B-A3B-GGUF/snapshots/a483e9e6cbd595906af30beda3187c2663a1118c/Qwen3.6-35B-A3B-UD-Q4_K_M.gguf`

20.6 GiB, unsloth Dynamic Q4_K_M (imatrix). The compose mounts the **repo dir**, not
the snapshot dir — the snapshot entry is a relative symlink into `../../blobs/`, so
both have to be inside the mount.

### Architektur (`qwen35moe`) — warum 128k hier billig ist

This is a **hybrid SSM + attention MoE**, not a dense-attention model:

| | |
|---|---|
| Layers | 40 |
| `full_attention_interval` | 4 → **only 10 layers use full attention** |
| Remaining 30 layers | gated-delta SSM, constant-size recurrent state |
| Attention heads | 16 Q / 2 KV, `key_length` = `value_length` = 256 |
| Experts | 256 total, 8 active, expert FFN 512, shared expert 512 |
| Native context | 262144 |

Only the 10 full-attention layers grow a KV cache, so 128k of KV costs just

```
10 layers x 2 kv-heads x 256 dim x 2 (K+V) x 1.0625 B/elem (q8_0) x 131072 tok = 1.33 GiB
```

Measured: the whole runtime footprint with *all* experts on CPU is 4.65 GB at 128k ctx.
**The context is not the constraint here — the 20.6 GiB of weights against 22 GiB of
VRAM is.** That is what the tuning below is about.

The model is multimodal-capable (the chat template emits `<|vision_start|>` /
`<|image_pad|>` tokens), but **no mmproj file is present**, so this deployment is
text-only. It is a reasoning model with `<think>` tags and XML-style tool calls, so
`--jinja` is mandatory.

## Tuning

### 1. `--n-cpu-moe` collides with `-sm layer` — `-ts` is not optional

`--n-cpu-moe N` moves the expert tensors of the **first N layers** to host RAM.
`-sm layer` assigns the **first** layers to CUDA0. Those are the same layers, so
CUDA0 gets the lightweight ones and CUDA1 ends up holding every heavy expert layer.

With the default split, **every** value of `--n-cpu-moe` from 12 down to 4 died the
same way — CUDA1 out of memory while allocating the KV cache:

```
ggml_backend_cuda_buffer_type_alloc_buffer: allocating 680.00 MiB on device 1: cudaMalloc failed: out of memory
alloc_tensor_range: failed to allocate CUDA1 buffer of size 713031680
llama_init_from_model: failed to initialize the context: failed to allocate buffer for kv cache
```

`-ts 24,16` moves the layer boundary back toward CUDA0 and fixes it. A heavy
(expert-bearing) layer is ~469 MiB; a layer whose experts are on CPU is ~59 MiB.

### 2. `--threads 8`, not 16

The CPU-side expert matmuls are memory-bandwidth-bound, so the 8 SMT siblings only
add contention. Measured at `ncmoe=8`, 8k depth:

| threads | prefill | generation |
|---|---|---|
| 8 | 556 t/s | **57.3 t/s** |
| 16 | 556 t/s | 39.8 t/s |

**+44% generation.** Prefill is unaffected because it runs on the GPU.

### 3. Expert offload vs. VRAM headroom

All at 128k ctx, q8_0 KV, `--threads 8`, `-ts 24,16`:

| `--n-cpu-moe` | gen @512 | gen @8k | CUDA0 free | CUDA1 free |
|---|---|---|---|---|
| 8 | 59.9 t/s | 57.3 t/s | 388 MiB | 531 MiB |
| **10 (default)** | **55.6 t/s** | **53.3 t/s** | **1312 MiB** | **529 MiB** |
| 12 | 51.3 t/s | 49.1 t/s | 2236 MiB | 527 MiB |
| 40 (all experts on CPU) | 15.3 t/s | 16.1 t/s | — | — |

`ncmoe=10` is the shipped default: it gives up 7% throughput for **~1.3 GiB of spare
VRAM on CUDA0**, which the desktop shares. At `ncmoe=8` only 388 MiB is left there,
and a browser opening a few video tabs is enough to OOM a `restart: unless-stopped`
service into a crash loop. If the desktop is idle, `ncmoe=8` is the faster setting.

CUDA1 is the binding constraint in every case — 529 MiB free is not enough for another
469 MiB expert layer, which is why `ncmoe` cannot go below 8 at this context size.

### 4. Throughput over depth (shipped config)

`ncmoe=10`, `-ts 24,16`, `--threads 8`, 128k ctx allocated, q8_0 KV:

| prompt depth | prefill | generation |
|---|---|---|
| 512 | 420 t/s | 56.3 t/s |
| 32k (27169 tok) | 486 t/s | 47.8 t/s |
| ~97k (99109 tok) | 425 t/s | 34.1 t/s |

Generation falls off ~40% between empty and ~97k, which is the attention cost on the
10 full-attention layers. Prefill stays flat around 420–490 t/s, so filling the whole
128k window takes roughly 4–5 minutes.

### 5. Speculative decoding makes it slower — don't enable it

Counter-intuitive but consistent, `--n-predict 400`, thinking disabled:

| workload | none | ngram-mod | ngram-simple | ngram-cache |
|---|---|---|---|---|
| rewrite (output ≈ copy of prompt) | **55.6** | 40.6 | 39.7 | 47.0 |
| extend (structured code) | **55.3** | 32.7 | 27.2 | 38.1 |
| prose (novel text, control) | **56.1** | 54.7 | 52.9 | 48.4 |

Every mode loses, and it loses *most* on exactly the structured workloads
speculation is supposed to win. The reason is the CPU-side experts: verifying a
K-token draft costs about K times the CPU expert work, because each token routes to
its own subset of 8 out of 256 experts, so there is no weight reuse to amortize.
On a fully GPU-resident model batch verification is nearly free; with
`--n-cpu-moe` it is not. Reconsider only if the model ever fits entirely in VRAM.

### 6. Rejected

- `-ub 256`: frees only ~80 MiB per GPU (not the 469 MiB an extra layer needs) and
  costs **38% prefill** (556 → 347 t/s). Keep the default `-ub 512`.
- Lower `-ts` toward CUDA1 (e.g. `23,17`): would leave CUDA1 at ~60 MiB free.

## Vergleich: A3B (MoE) vs. Qwen3.6-27B (dense)

Same box, same llama.cpp image, same `-ctk/-ctv q8_0 -fa on --threads 8`, same
benchmark, measured at identical prompt depths:

| Prompt-Tiefe | A3B gen | 27B gen | A3B prefill | 27B prefill |
|---|---|---|---|---|
| 512 | **56,3 t/s** | 24,1 t/s | 420 t/s | **539 t/s** |
| 8k | **53,3 t/s** | 23,2 t/s | 497 t/s | **817 t/s** |
| 32k | **47,8 t/s** | 20,5 t/s | 486 t/s | **767 t/s** |

**Generierung: A3B ist durchgehend ~2,3x schneller.** Nur 3B der 35B Parameter sind
pro Token aktiv, das dichte 27B muss alle 27B lesen — auch wenn es komplett im VRAM
liegt und das A3B ein Drittel seiner Experten im RAM hat.

**Prefill: das 27B ist ~1,6x schneller**, weil es vollständig GPU-resident ist,
während beim A3B jeder Prefill-Batch durch die CPU-Experten muss.

**Kontext: das 27B schafft die 128k hier gar nicht.** Es hat 65 Layer mit 4 KV-Heads
gegen 40 Layer mit 2 — bei `full_attention_interval=4` also 16 statt 10
Full-Attention-Layer und damit **4,25 GiB KV bei 128k statt 1,33 GiB**. Zusammen mit
~15 GiB Gewichten reicht das nicht: schon das kleinere Q4_K_S scheitert beim
Compute-Buffer:

```
allocating 1145.13 MiB on device 0: cudaMalloc failed: out of memory
graph_reserve: failed to allocate compute buffers
```

Gemessen wurde es deshalb bei 64k ctx.

### Caveats zu diesem Vergleich

- Gemessen wurde **Q4_K_S (16,1 GB)**, nicht Q4_K_M (17 GB) — die einzige lokal
  vorhandene, mit upstream llama.cpp ladbare 27B-Datei dieser Klasse. Q4_K_M ist ~5%
  größer und wäre entsprechend etwas langsamer, der Abstand also eher noch größer.
- Der ollama-Blob von `qwen3.6:27b-q4_K_M` lässt sich mit upstream llama.cpp **nicht**
  laden: `key qwen35.rope.dimension_sections has wrong array length; expected 4, got 3`.
  Ollama fährt einen eigenen Fork; ein Vergleich exakt dieser Datei ist im selben
  Runner nicht möglich.
- Ältere 27B-Zahlen unter `PrismQuant/benchmarks/` (34,3 t/s) stammen von einer
  **RX 7900 XTX, gedrosselt** — nicht mit dieser Maschine vergleichbar.

### A note on the CUDA0 headroom

Free VRAM on CUDA0 was observed drifting between ~1100 and ~1300 MiB across restarts
purely from desktop compositor churn — a ~200 MiB swing with nothing else changing.
That is why the default keeps >1 GiB spare there rather than the 388 MiB that
`ncmoe=8` leaves.

## `--predict` ist ein Default, kein Deckel

Despite the name, `--predict` does **not** cap what a client can ask for. Measured
against `--predict 8192`:

| Request | Result |
|---|---|
| `max_tokens: 9000` | **8893** tokens, `finish_reason: stop` — the flag is ignored |
| no `max_tokens` | **8192** tokens, `finish_reason: length` — the flag applies |

So it only binds clients that send no `max_tokens` at all. That is still worth having
here: without it such a client generates until the 128k window is full, which at
~40 t/s is **roughly 50 minutes**, and with `-np 1` the single slot is blocked for all
of it. Set to `32768` — a runaway is bounded to ~12 minutes, while no realistic long
answer gets truncated. Many OpenAI clients do not send `max_tokens`, and at 8192 they
were silently cut off with `finish_reason: length`.

Note the sibling `gemma4-26b-llama-runner` README describes `--predict` as a "harte
Cap fuer Gesamt-Generierung" — by this measurement that is wrong there too.

## `--reasoning-budget` dagegen ist ein echter Deckel

Same flag family, opposite semantics — measured with `bench/reasoning_budget_probe.py`,
which counts the thinking block exactly via the server's `/tokenize` endpoint:

| Request | Reasoning tokens | Answer |
|---|---|---|
| server default | **4095** | 1227 tok, complete |
| `reasoning_budget: 16384` in the request body | **4095** | 531 tok, complete |

A client **cannot** raise it. Where `--predict` was a default anyone could override,
this one is a hard server-side ceiling, so changing it means editing the compose file
and restarting.

### More budget did not buy better answers

Same number-theory problem (correct answer: empty solution set, verified by brute
force), one run per setting at `temperature 0.3`:

| budget | thinking | total tokens | wall clock | answer |
|---|---|---|---|---|
| 4096 | 4095, cut mid-derivation | 5325 | ~2 min | **correct** |
| 16384 | 16383, cut mid-**repetition-loop** | 17417 | ~6 min | **correct** |

At 4096 the thinking is guillotined mid-derivation (`...Since $x^2 \equiv`) and the
model recovers, rebuilding the parity argument cleanly in the visible answer. At
16384 it never converges — it spends the extra ~12k tokens looping on
`Maybe it's $a^2 + b^2 = c^2 + 7$?` over and over, then answers correctly anyway.

So the extra budget cost **3.3x the tokens and wall clock for an identical answer**,
and the degenerate loop is exactly the rambling the brake exists to cut off. The
intuition that "it was truncated, so it needs more room" did not survive measurement:
this model does not use more thinking budget productively, it fills it.

Caveat: one problem, one sample per setting. Treat the direction as indicative, not
as a tuned value. Raising the budget does not weaken runaway protection either way,
since total generation is bounded by `--predict` / the client's `max_tokens`.

## Deploy

```bash
cd ~/projects/llama_qwen3.6_A3B
docker compose up -d
```

Loads in ~15 s (`--no-mmap`, weights read from NVMe).

## Benchmarks

```bash
python3 bench/bench.py --port 18008 --depths 512,8192,32768 --n-predict 128
./bench/sweep.sh 10:24,16:8 8:24,16:8      # <ncmoe>:<tensor-split>:<threads>
python3 bench/spec_bench.py --no-think     # copy/code/prose generation workloads
```

`bench.py` issues a warmup request first — without it the first measurement reads
~30% low because of one-off CUDA graph setup.

## Image

`ghcr.io/ggml-org/llama.cpp:server-cuda`, tested at build **b10121**
(commit `555881ebc8b0`, 2026-07-25). Updated from b10068 during this deployment;
b10121 is the first build here that supports the `qwen35moe` architecture end to end.

## VRAM-Exklusivitaet

Uses ~20 GB of the 22 GB total. It **cannot** run alongside `llama-gemma4` (18006) or
qwen-prism (18004). Ollama on 11434 loads models on demand and will fight for VRAM —
stop it or let its `keep_alive` expire before starting this runner.