From beaf452d55d637dad6f7dd33d1f6b09458a8bf56 Mon Sep 17 00:00:00 2001
From: Claude <noreply@anthropic.com>
Date: Mon, 6 Apr 2026 17:23:13 +0000
Subject: [PATCH] perf: keep WINDOW_G=12, increase benchmark warmup
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Tested WINDOW_G=15 (matching C libsecp256k1): 5.6x slower than native.
WINDOW_G=12 is faster at 4.3x because the 8192-entry table (1MB) at
w=15 causes cache pressure on JVM — heap-allocated AffinePoint objects
are scattered in memory, unlike C's contiguous compile-time .rodata.
WINDOW_G=12 (1024 entries, ~128KB) fits comfortably in L2 cache.

Increased verify benchmark warmup to 200+500 for stable measurements
(first call builds the lazy table).

https://claude.ai/code/session_01BhU63WUe9AhikZxRdw3Lpg
---
 .../quartz/utils/secp256k1/Secp256k1Benchmark.kt              | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/quartz/src/jvmTest/kotlin/com/vitorpamplona/quartz/utils/secp256k1/Secp256k1Benchmark.kt b/quartz/src/jvmTest/kotlin/com/vitorpamplona/quartz/utils/secp256k1/Secp256k1Benchmark.kt
index 3053446b7..3855e3cca 100644
--- a/quartz/src/jvmTest/kotlin/com/vitorpamplona/quartz/utils/secp256k1/Secp256k1Benchmark.kt
+++ b/quartz/src/jvmTest/kotlin/com/vitorpamplona/quartz/utils/secp256k1/Secp256k1Benchmark.kt
@@ -134,8 +134,8 @@ class Secp256k1Benchmark {
         results +=
             bench(
                 name = "verifySchnorr",
-                warmup = 50,
-                iterations = 200,
+                warmup = 200,
+                iterations = 500,
                 nativeOp = { native.verifySchnorr(nativeSig, msg32, nativeXOnlyPub) },
                 kotlinOp = {
                     com.vitorpamplona.quartz.utils.secp256k1.Secp256k1.verifySchnorr(