Move the 3 decode CPU-pipelining variants (opt1 current / opt3 sw-pipe /
opt2 ex_composite) up into the TL;DR as full standalone kernels alongside
the full prefill ring kernel, with the comparison table. S5.6 is reduced to
a brief anchor (still referenced by S8/SB) pointing to the TL;DR code +
keeping the recommend/cost-model linkage. Docs only.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The prior commit updated S0/S10/S5.5 terminology and added S5.6 (3 decode
variants) but left the TL;DR with the old 'all G heads replicated' wording
and no pointer to S5.6. Sync the TL;DR: 'Q replicated (G heads stacked into
the GEMM M-dim, M-fold)', '1 Q head per CUBE', and a one-line pointer to the
3 CPU-pipelining variants in S5.6. Docs only.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- Terminology: 'Q replicated' (all G query heads stacked into the GEMM M-dim;
M-fold explained) for decode; 'one Q head per CUBE' precise for prefill.
- New S5.6: three decode CPU-pipelining variants — opt1 current CompositeCmd
(has GEMM-engine bubble), opt3 software pipelining (issue next Q.Kt before
this tile's softmax; Sj in persistent double buffer; ships now, no new cmd),
opt2 ex_composite split into two (#1 = existing GEMM+scale reads K first;
#2 = softmax+P.V+accumulator merge, the only new flash-epilogue machinery,
gives DMA K-before-V priority). MATH engine already has max/sum/exp — the
new part is the stateful flash accumulator, not the ops.
- S2.1/SB: shared prefill/decode KV layout = contiguous CxP blocks (prefill
causal-skip needs contiguous; avoids prefill->decode reshard; short-context
under-use caveat). S8 item 4 sizing note for the two-composite split.
Prefill note: opt2/opt3 give little for prefill (causal if can't enter a
composite; recv_async already overlaps). Docs only; KO mirror deferred.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Decode-SP and prefill-SP are structurally different and cannot share one
kernel (principle: move the smaller thing):
- Decode: O=[G,d] tiny, KV cache big -> keep KV statically sharded/resident,
G heads replicated (M-fold), move only (m,l,O) via the 2-level reduce (S4).
- Prefill: O=[S,d] big -> shard heads (1 query head per CUBE, C=G),
rotate KV (Ring KV, S5.5), no (m,l,O) reduce; each CUBE writes its own head.
Rewrites TL;DR (two kernels), S0 (head map differs by case), S0.5.4 (output
head distribution differs -> downstream out-proj impact), S4 (scoped to
decode; S4.1 = intra-CUBE KV-split + PE reduce, the only way decode uses P
PEs), S5.1 (decode skeleton), S5.5 (head-parallel Ring KV), S9/S10/S11, and
adds SB items (two head mappings, output asymmetry, prefill within-CUBE PE,
C=G coupling, reconcile with _attention_mesh_mlo_2d). KO mirror deferred
until the design stabilizes (adr-proposed is mirror-exempt). Docs only.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Placement pivot (user-approved): a CUBE Group = C CUBEs within one SIP that
jointly own one KV head + its G=8 query heads. KV sequence sharded 2 levels
(Level-1 inter-CUBE over C, Level-2 intra-CUBE PE over P=8 = C*P ranks); G
folded into matmul M. C is a knob (8/4; C=1 = single-CUBE). Reverses the old
'a query head never spans CUBEs' non-goal (held only because the baseline
was H_kv=1). device=SIP; the for-kv loop is gone (head picked by CUBE coord).
Reduction is a 2-level reduce-to-root (not all-reduce): Level-2 PE tree ->
Level-1 center-root CUBE-mesh reduce, adapting lrab_hierarchical_allreduce's
inter-CUBE pattern as reduce-only + log-sum-exp. Data-driven (send on local
P.V completion, no global barrier) + level-pipelined; per-level topology
configurable (tree for decode, ring for long prefill).
Rewrites SS0/2/4/5/0.5/8-11, pseudocode, and adds SSB items (4-SIP config,
mesh partition, C knob, invariant-reversal check, index-math test). KO mirror
updated. Topology grounded in topology.yaml (4x4 CUBE mesh/SIP, 8 PE/cube).
Docs only; no production code changed.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- ADR-0060: GEMMs (Q.Kt, P.V) via existing tl.composite (scheduler-managed
tiling + K/V DMA streaming); softmax merge + IPCQ tree reduction stay in
kernel. Front TL;DR pseudocode of the final composite kernel; new section
B lists open design items (DDD sync, K pre-transpose, dma_read lever,
kernel-vs-scheduler tiling, ring path).
- ADR-0062: redefined from a new load_async op to global lazy tl.load
(non-blocking + auto-wait on first use; API unchanged; goldens regenerate).
- ADR-0064 (new): per-op-type CPU issue cost model (composite ~40ns >>
primitive) so the hybrid's CPU-saturation win becomes measurable
(currently dispatch_cycles=0 hides it). Cost-model impl deferred.
- KO mirrors for ADR-0060/0062/0064 (-ko suffix, adr-proposed).
Rationale: non-blocking CompositeCmd offloads tiling to PE_SCHEDULER,
decoupling CPU issue-rate from execution so the CPU can saturate the
engines; the prior 'composite = no latency benefit' claim was an artifact
of dispatch_cycles=0. Docs only; no production code changed.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Verified against sim_engine data path (memory_store, data_executor):
- GQA reuse does NOT need a broadcast op. The baseline's h_q==h_kv limit
is its head-packing reshape hack, not a missing primitive. Correct fix:
per-KV-head loop with G folded into matmul M dim (byte-conserving
reshape) — runs today, timing correct (m=G*T_q), data mode runs.
- ADR-0061 broadcast demoted from 'the blocker' to optional convenience.
- Surfaced tl.trans = reshape-not-transpose (memory_store reshapes;
data_executor np.matmul on reshaped operands) -> numeric parity is
bounded; verification is structural/timing/determinism-first (matches
SPEC perf-model purpose). Optional tl.transpose deferred.
- Reordered DDD phase plan (P1 GQA needs no new feature; P3 scratch_scope
is the key scale feature); added open decisions 10.10 (transpose) and
10.11 (GQA-via-M-fold finding).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Reframes the proposed GQA FlashAttention design onto kernbench's actual
execution model (greenlet tl API + IPCQ), replacing the composite-centric
mechanism that does not match the simulator:
- Records relationship to existing baseline kernels (_attention_mesh_kv/mlo,
milestone-gqa-llama70b) and their 3 deliberate limitations.
- Mechanism is greenlet tl (per-op latency; no fusion benefit), not
composites; running (m,l,O) is Python handles; reduction is tl.send/recv.
- Tree reduction (log N) replaces baseline all-to-all fan-out (N-1).
- Pseudocode rewritten in real tl.* signatures; depends on ADR-0061/62/63.
- Rejects composite-IPCQ-push + composite-carried-state + flash-composite
with documented efficient alternatives.
- Adds verification plan.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Proposed prerequisites surfaced while evaluating the GQA fused-attention
ADR against the actual kernbench tl/sim_engine implementation:
- ADR-0061 tl.broadcast: data-faithful GQA head reuse (fixes the
MemoryStore nbytes check that forces h_q==h_kv==1 today).
- ADR-0062 tl.load_async: non-blocking HBM tile load for KV prefetch
(KV-load-bound decode/long-context overlap).
- ADR-0063 tl.scratch_scope: per-tile scratch recycling (removes the
1 MiB bump-allocator ceiling that caps context at S=16).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
ywkang
Yangwook Kang