feat: add benchmark metrics for AISO Private AI Inference Benchmarks

benchmarks/metrics.json

@@ -0,0 +1,22 @@
+{
+  "metadata": {
+    "component": "AISO Private AI Inference Benchmarks",
+    "author": "Brad Wallace",
+    "email": "coo@koba42.com",
+    "github": "https://github.com/tensorrent/prime-fhe",
+    "date": "2026-07-30T03:19:23Z",
+    "license": "Apache-2.0"
+  },
+  "metrics": {
+    "vector_dim": 768,
+    "step_latency_ns": 1361,
+    "full_evaluation_ms": 1.05,
+    "throughput_dims_per_sec": 734923,
+    "wasm_p50_latency_ms": 1.28,
+    "wasm_p95_latency_ms": 5.28,
+    "wasm_p99_latency_ms": 5.72,
+    "wasm_parity_queries_passed": 10000,
+    "noise": "Zero",
+    "security_bits": 128
+  }
+}