Combining --enable-prefix-caching with the DFlash drafter crashed engine init with "block_size must be divisible by hash_block_size" in HybridKVCacheCoordinator. The drafter's attention KV page is ~2x the target's, so vLLM's page-size unification scales the target's mamba+attn block 2240->4480 to match; the (align-mode) mamba block then differs from cache_config.block_size and resolve_kv_cache_block_sizes backs off to hash_block_size = LCM (4480), which the drafter group (still 2240) is not divisible by. But the GCD (2240) divides every group and is the correct finer hash granularity; the back-off's `block_size != cache_block` test is a buggy proxy for "non-align mamba". patch_prefix_align.py makes the back-off align-aware (only back off when mamba_cache_mode != "align"), so resolve uses the GCD. Prefix caching is on by default (a win for multi-turn / long-context, neutral for single-turn c=1); set PREFIX_CACHE=0 to disable. Validated on GB10: READY, DFlash accept ~7.7 tok/step on code, ~13x warm-prefix TTFT (2.30s->0.18s), KV pool ~422k tokens (no regression). See docs/FINDINGS.md section 1a.
11 KiB
FINDINGS — DFlash + dense levers for Qwen3.5-122B-A10B on DGX Spark
Single-stream (c=1) decode of Qwen3.5-122B-A10B (hybrid GDN + mamba + 128-expert
MoE, ~10B active) on GB10 / SM121, 128 GB / 119 GiB unified, ~273 GB/s. The agent this
backs (Hermes) is ~73 % tool-calls. All numbers temperature 0.
Credit. This builds on albond's recipe (hybrid INT4+FP8 + INT8 lm-head + MTP-2 on vLLM 0.19, and the e2e benchmark method). The new work here is forward-porting the dense levers to vLLM 0.23, swapping MTP for DFlash, and composing them — landing at 59.0 e2e tok/s vs albond's 51.58 on his own harness (+14%), and ~2× on real agent traffic. See README → Building on the albond recipe for the side-by-side.
1. Getting DFlash to run on the hybrid 122B in vLLM
The DFlash drafter is non-causal (it block-drafts 16 tokens in one parallel
forward) — only the FLASH_ATTN (FA2) backend supports non-causal attention.
But the hybrid GDN+mamba+MoE target's KV-cache page geometry won't unify with
the drafter's attention spec:
- vLLM auto-aligns the hybrid (attention block 2240, mamba page padded +0.54 % to
match), so
max_page_sizeis a padded value.unify_kv_cache_spec_page_sizescales the drafter's attention block byratioand then assertspage == max— which fails, becausepage_size_bytesignoresblock_sizeoncepage_size_paddedis set. - Fix (
patch_unify2.py): keep the scaledblock_sizeand pad the <1 % remainder (mirrors vLLM's ownHiddenStateCacheSpechandling). The earlier "pad but keep block_size=16" patch mis-strided the drafter KV → acceptance collapsed to 1.47 (a real bug, not a quant mismatch). --mamba-block-size 256(from other Spark recipes) breaks the 122B: it makes the mamba group block ≠ cache block, tripping the coordinator hash assert. Omit it.--no-enable-prefix-cachingroutes toKVCacheCoordinatorNoPrefixCache, which has no hash assert. It is the default; prefix caching is irrelevant at c=1 single-turn but a large win for agentic multi-turn / long-context re-reads — see §1a for enabling it with DFlash.
Working stack: patch_unify2 + prefix-off + INT4 (bf16 KV) target + drafter
pinned to FLASH_ATTN. No FA4 shim needed — vLLM gates FA4 to cap families
90/100/110 (excludes 120), so the drafter runs FA2. (The whole fa4-sm120 saga is
SGLang-only; SGLang's DFlash works too but its sm121 base decode is ~2× slower
than vLLM's, so it loses on absolute throughput.)
1a. Prefix caching + DFlash together (PREFIX_CACHE=1)
Enabling --enable-prefix-caching with the DFlash drafter crashed engine init with
AssertionError: block_size must be divisible by hash_block_size
(HybridKVCacheCoordinator.__init__). Prefix caching alone (no DFlash) and DFlash alone
both work; only the combination broke. Diagnosed on the GB10 (2026-06-28):
- The DFlash drafter's attention layers carry a 2× larger KV page (9 175 040 vs the
target's 4 587 520 B — the drafter has ~2× the KV heads). So the drafter is
max_page. unify_kv_cache_spec_page_sizetherefore scales the target's mamba + attention block 2240 → 4480 (ratio 2) to match the drafter page; the drafter group stays at 2240.- In
resolve_kv_cache_block_sizesthe (align-mode) mamba block is now 4480 ≠cache_config.block_size(2240), so its back-off branch fires and forceshash_block_size = LCM = 4480. The drafter group is 2240, and2240 % 4480 ≠ 0→ assert. - But
GCD = 2240divides every group (4480 and 2240) and is the correct finer hash granularity (vLLM'shash_block_size < block_sizemerge-up, #29143). The back-off only exists to disable fine hashing for non-align mamba — itsblock_size != cache_blocktest is a buggy proxy that also trips when an align-mode block was merely scaled up.
Fix (patch_prefix_align.py): make the back-off
align-aware — only back off when mamba_cache_mode != "align". In align mode it falls
through to the GCD path. One-condition change; no drafter-geometry surgery, no extra memory.
(vLLM #45181's pad-don't-scale path does not apply here: our pages are an exact 2×, so
unify scales rather than pads — #45181 only adds a branch for the non-divisible case.)
The KV reshape already reads padded/strided pages correctly in this image
(get_kv_cache_block_dim → physical_block_dim), so there is no acceptance risk — the
old 1.47-accept "pad" regression was a different, since-fixed reshape bug, not this path.
Validated (prefix-ON + DFlash n=12, flash_attn): READY; DFlash accept ~7.7 tok/step
on code (240 accepted / 36 drafts; per-position 34→11, unchanged from prefix-off);
warm-prefix TTFT 2.30 s → 0.18 s (~13×) on a 4k shared prefix with a real
prefix_cache_hits_total bump; coherent output; KV pool 421 888 tokens (≈ the prefix-off
~427k, no regression). On by default (win for multi-turn / long-context, neutral for
single-turn c=1); set PREFIX_CACHE=0 to disable.
2. DFlash vs MTP — acceptance is task-dependent
MTP-2 (the native head) drafts 2 tokens sequentially, so acceptance caps at ~3. DFlash block-drafts 12 in one parallel forward, so its acceptance fills the block on predictable traffic. Same harness, unpatched, flash_attn:
| Workload | accept (MTP-2 / DFlash) | tok/s (MTP-2 / DFlash) |
|---|---|---|
| Prose | 2.24 / 2.3 | 33.7 / 33.2 (tie; use DFlash n=4) |
| Code | 2.77 / 5.4 | 40.5 / 54.5 |
| Counting | 3.00 (maxed) / 11 | 43.7 / 124.5 |
| Hermes (real) | 2.88 / 8.66 | 39.9 / ~81 |
The "DFlash caps at 2.3 / 33 tok/s" story was a prose-benchmark artifact.
On agent/code traffic DFlash pulls ~2× ahead because MTP is acceptance-saturated.
n (num_speculative_tokens) is task-dependent: prose → 4, agent/code → 12+.
3. Methodology — two non-comparable harnesses
bench_decode.py= decode-only tok/s (excludes TTFT). Good for c=1 kernel comparisons. This is the ~81 Hermes number.bench_albond.py= end-to-end (completion_tokens / total wallclock incl. prefill, non-streaming, 5 prompts, run-1 discarded). This reproduces albond's own method and is directly comparable to his published 51.58.
Apples-to-apples (albond's method): DFlash n=12 unpatched = 53.7 tok/s cross-prompt mean — already above his fully-patched MTP stack (51.58).
4. The dense-bandwidth levers and the amortization law
albond's non-MTP wins are always-on bandwidth cuts. We ported the two that transfer to vLLM 0.23 + DFlash:
- hybrid INT4+FP8 (
patch_inc_hybrid.py): the Intel base already stores attention as INT4 (0.5 B/param, better than albond's FP8 attention), so the only thing to gain is the BF16 shared experts → calibrated FP8 (144 layers, 0.48 GB saved). The dispatch patch adds anINCConfig.maybe_update_configoverride (AEON 0.23's hook signature takeshf_config=, unlike albond's 0.19) that detects FP8 dense layers and dispatchesFp8LinearMethodfor them. - int8 lm-head (
patch_int8_lmhead_v3.py): the 248 320-row vocab projection is the single largest dense read (1.5 GB BF16, every token). A batched int8 w8a16 Triton GEMV reads it at ~227 GB/s (vs bf16 ~6.5–8.8 ms) — ~2× faster, argmax-exact. Prior ports failed not on the kernel but on integration: zeroing the lm-head weight corrupted the drafter-shared head (garbage), and a per-row loop for B>4 was slower under spec. v3 uses one batched kernel; on the first warmup forward it builds the int8 copy, then frees the now-dead bf16 weight (~1.4 GiB) andempty_cache()s so the block returns before vLLM sizes the KV pool. This is safe here because the DFlash drafter shares this same int8 lm_head module (verified: exactly one int8 build; the drafter checkpoint carries no lm_head/embed tensors) andtie_word_embeddings=False(so.weightis not aliased toembed_tokens) — the bf16 copy has no remaining reader (the bias fallback does int8-GEMV + add-bias instead). SetSPARK_KEEP_BF16_LMHEAD=1to restore the keep-bf16 behavior. WithVLLM_MEMORY_PROFILER_ESTIMATE_CUDAGRAPHS=0(reclaims the CUDA-graph over-estimate), this lifts thedenseKV pool 376,518 → 426,610 tokens (+13 %) at the same0.82headroom, validated coherent with acceptance 4–12.
Why the denominator matters: the 0.48 GB shared-expert saving is 0.7 % of the 71 GB on disk but ~8 % of the ~6 GB active per-token footprint (the disk is mostly sparse routed experts). Shared experts and the lm-head are dense — read every token — so at base decode the savings land in full:
| Config | base (acc 1) | albond-bench (acc 6.4) | Hermes (acc 8.3) |
|---|---|---|---|
| INT4 baseline | 28.2 | 53.7 | ~81 |
| + hybrid-FP8 | 30.4 (+7.8%) | 57.0 (+6.1%) | ~80 |
| + int8 lm-head | 32.7 (+16%) | — | — |
| + both | 36.0 (+28%) | 59.0 (+10%) | ~80–87 (noise) |
The levers compose additively (step savings 2.6 + 4.9 ≈ 7.7 ms). But the uplift decays monotonically with acceptance:
Under speculative decode the verify forward processes ~
acceptpositions and reads each dense weight once, amortized across them. So a dense-weight cut that is +X % at base is ~+X/accept % under spec.
+28% base (accept 1) → +10% albond-bench (accept 6.4) → ~0% Hermes (accept 8.3)
Consequences
- For the agent path (
dflash), DFlash's own high acceptance already amortizes the dense levers to ~null. Its remaining bottleneck is routed-expert verify-batch reads (each of ~13 verify positions routes to different experts) — untouched by any dense-weight quant. To push Hermes further you must attack that: a smaller/faster drafter, lowernat equal acceptance, or sub-INT4 routed experts. - For base / low-acceptance serving (
dense), the stack is a real +28 % (36 tok/s) and is the recommended config there.
5. Things that did NOT help c=1 decode
- FLASHINFER target backend — null both short-context and Hermes (attention isn't the bottleneck on this GDN/mamba-heavy MoE; most layers are linear attention). albond's "+16 %" was on his dense-attention MTP path.
- b12x / native FP4 MoE — null at c=1 (a throughput/concurrency lever, not a latency one; at batch 1 the active-expert GEMM is tiny).
- FLA sm121 big-tile shmem fix — real bug, but prefill/TTFT only; c=1 decode uses the GDN recurrent path, a different kernel. Kept (free TTFT win).
- PR#38325 swapAB FP8 GEMM — marginal (+0.76 %), only with the FP8 checkpoint.
Production recommendation
Ship dflash for the agent (DFlash unpatched, ~81 tok/s, ~2× MTP, > albond's
patched 51.58). Reserve dense for base / low-acceptance serving (+28 %).
The dense patches are upside there, not a requirement for the agent to win.