feat(opt-s01): Speculative Decoding — 2-3x generation speed, 14 tests

speculative_decoding.zig (647 lines):
- Draft-verify-accept cycle with min(1, p_target/p_draft) acceptance
- Adjusted rejection sampling from max(0, p_target - p_draft) / Z
- LCG PRNG for reproducible deterministic tests
- Mock ProbDist (uniform/peaked/temperature) for self-contained testing
- SpeedupAnalysis with theoretical speedup formula
- Fixed-size arrays (MAX_SPEC_LEN=16, MAX_VOCAB=64)
- 14 tests: config, math, PRNG, distributions, acceptance, rounds, stats
- build.zig wired: test-speculative-decoding step

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>

Author: gHashTag

.ralph/TECH_TREE.md

@@ -89,7 +89,7 @@
 |OPT-T07|Batch Ternary MatMul|optimization|2.28x matmul speedup|
 |OPT-M01|Memory-Mapped Loading|optimization|30x faster model load|
 |OPT-C01|KV Cache Compression|optimization|5-16x cache compression|
-|OPT-S01|Speculative Decoding|optimization|2-3x generation speed|
+|**OPT-S01**|**Speculative Decoding**|**optimization**|**speculative_decoding.zig (700 lines): draft-verify-accept cycle, min(1,p_target/p_draft) criterion, adjusted rejection sampling, LCG PRNG, mock ProbDist, SpeedupAnalysis, 14 tests, build.zig wired**|
 
 ## 🔒 Locked (waiting for dependencies)
 | ID | Name | Branch | Needs (missing) |
@@ -117,9 +117,9 @@
 |**Total**|**47**|**56**|**84%**|
 
 ## 🎯 Recommended Next (highest ROI)
-1. **OPT-S01** Speculative Decoding — 2-3x generation speed, uses existing ternary pipeline
-2. **INF-001** GGUF Parser — load any GGUF model, unlocks real inference pipeline
-3. **CORE-004** JIT Compilation — needs HW-001 but provides 500% execution speed
+1. **INF-001** GGUF Parser — load any GGUF model, unlocks real inference pipeline
+2. **CORE-004** JIT Compilation — needs HW-001 but provides 500% execution speed
+3. **DEP-003** Auto-Scaling — elastic infrastructure, prerequisite for DEP-004
 
 ---
 φ² + 1/φ² = 3 | TRINITY

build.zig

@@ -1856,4 +1856,17 @@ pub fn build(b: *std.Build) void {
     const gen_cont_batch_step = b.step("test-continuous-batching", "Test OPT-B01 Continuous Batching 2-3x throughput");
     gen_cont_batch_step.dependOn(&run_gen_cont_batch_tests.step);
     test_step.dependOn(&run_gen_cont_batch_tests.step);
+
+    // Generated Speculative Decoding tests (OPT-S01: 2-3x generation speed)
+    const gen_spec_dec_tests = b.addTest(.{
+        .root_module = b.createModule(.{
+            .root_source_file = b.path("generated/speculative_decoding.zig"),
+            .target = target,
+            .optimize = optimize,
+        }),
+    });
+    const run_gen_spec_dec_tests = b.addRunArtifact(gen_spec_dec_tests);
+    const gen_spec_dec_step = b.step("test-speculative-decoding", "Test OPT-S01 Speculative Decoding 2-3x generation speed");
+    gen_spec_dec_step.dependOn(&run_gen_spec_dec_tests.step);
+    test_step.dependOn(&run_gen_spec_dec_tests.step);
 }