paper(1H): reflect D8 single-op cost model in §2/§3.4 + figure/diagram regen

Two strands bundled as the 1H-codesign-paper refresh unit:

(A) This session — single-op cost-model reflection (depends on 2d8271c):
  - §2 Table 2 (tab:hw): split "FIXED per command" into "FIXED per
    single-op command" (8 cycles) and "FIXED per composite command"
    (40 cycles); §2 dispatch-overhead prose updated to the two-class split.
  - §3.4 (sec:gemm-vs-async): rename paragraph headers + prose to
    async-full / async-tiled; "atomic" -> "single-op" throughout; reframe
    mechanism #3 from the old DMA-only fast-path to the single-op
    fast-path. Headline narrative now: even with EVERY single-op cmd
    (96 DMA + 48 dot + 47 add) charged the light 8-cycle FIXED, composite
    still wins ~2.8x at K=3072 purely on command-count structure (1 vs
    192 commands) -- down from the pre-D8 ~6.3x, and explicitly NOT a
    modelling artifact. Numbers refreshed from the regenerated sweep:
    async-full 3.83->3.91, async-tiled 1.14->~2.53, under-tile corner
    1.06->1.21, depth-2 vs depth-inf spread <1%. New figure wired in.
  - build/main.pdf rebuilt (tectonic); pdftotext-verified (no broken
    refs; Table 2 split, single-op terms, 2.8x/2.53/192-host-commands
    all present).

(B) Prior-session paper work riding along uncommitted: §4 all-reduce
    deep-edit, §5 GQA, §6 discussion trims; milestone_1h_ccl.py plot
    label "FSIM" -> "H2 2025 SW queue baseline"; regenerated diagrams
    under docs/diagrams/** and gemm output PNGs under
    1H_milestone_output/gemm/. (Composite-window gemm plots are
    unaffected by D8 — D8 only changes single-op dispatch FIXED, which
    the composite window excludes.)

All TODO items for the D8 single-op extension are now complete and
pushed across 3 commits (2d8271c cost-model+ADR+tests, 821bbf2 bench
harness, this paper refresh). Full regression green (826 passed, 1
skipped). No remaining work.

NOTE for review (carried from 2d8271c): ADR-0065's "2x CPU-offload win"
headline for GQA decode opt2 may want a refresh to the post-D8 ~1.87x.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

src/kernbench/benches/milestone_1h_ccl.py

@@ -882,7 +882,7 @@ def emit_topology_diagram(out_dir: Path | None = None) -> str:
 # single-device reference marker in the top panel (hardcoded 366 µs; no
 # external data file). Reads summary.csv written by _render_latency_sweep.
 
-_FSIM_EXT_LABEL = "FSIM (single device): 366 µs"
+_FSIM_EXT_LABEL = "H2 2025 SW queue baseline (single device): 366 µs"
 _FSIM_EXT_LATENCY_NS = 366_000.0
 _CMP_COLORS = {
     "ring_1d": "tab:blue",
@@ -1007,7 +1007,10 @@ def emit_comparison_fsim_plot(out_dir: Path | None = None) -> str | None:
     ax_bot.legend(handles_bot + handles_top, labels_bot + labels_top,
                   loc="upper left")
 
-    fig.suptitle("Multidevice allreduce (ring, Mesh, 2DTorus) vs FSIM latency")
+    fig.suptitle(
+        "PE_IPCQ multidevice allreduce (Ring, Mesh, 2DTorus) "
+        "vs H2 2025 single-device SW-queue baseline"
+    )
     fig.tight_layout()
     out = out_dir / "comparison_mesh_vs_ring_vs_2DTorus_vs_theoretical_vs_fsim.png"
     fig.savefig(out, dpi=120)