starred/turboquant_plus
1
0 Fork 0
mirror of https://github.com/TheTom/turboquant_plus.git synced 2026-10-22 02:53:13 -05:00
Code Issues Packages Projects Releases Wiki Activity

docs: add inline block + named-reg results to scoreboard (10 approaches total)

Browse Source
This commit is contained in:
TheTom committed 2026-03-26 22:47:53 -05:00
1 parent df1e37807f
commit 081abfff05
1 file changed
+3 -2
Download Patch File Download Diff File Expand all files Collapse all files
docs/decode-speed-hardware-analysis.md
+3 -2
View File
@@ -91,9 +91,10 @@ We tested 9 approaches total. The pattern is clear:
| 2-pair half2 + ternary | 12.0 | 2 divergent | 4 | Branches hurt |
| Select chain (zero LUT) | 11.9 | 0 | 8 | Branches kill |
| Bit-arithmetic | 11.6 | 0 | 4+ | ALU too heavy |
| Named-reg + ternary select | 10.3 | 4 uniform | 8 | Worst |
| Named-reg + ternary select | 10.3 | 4 uniform | 8 | Worst — ternary kills |
| Main (8-entry LUT) | 10.95 | 8 divergent | 0 | Baseline |
| Non-vec FA (nl=2) | 10.2 | 8 divergent | 0 | Wrong kernel |
| Inline block (FA inner loop) | 13.5 | 4 divergent | 0 | I-cache pressure |
| Non-vec FA (nl=2) | 10.2 | 8 divergent | 0 | Wrong kernel for decode |
**Any approach that replaces constant reads with branches loses on M2.** The Apple8 GPU's branch predictor/execution is worse than its constant cache. The 4-mag LUT succeeds because it reduces constant addresses (4 vs 8) WITHOUT adding branches.