github/tfhe-rs
1
1
Fork 0
mirror of https://github.com/zama-ai/tfhe-rs.git synced 2026-01-07 22:04:10 -05:00
Code Issues Packages Projects Releases Wiki Activity

chore: fix scalar benchmarks generating fixed values

Browse Source 
- this would not give an average runtime for scalar benchmarks and for
small precisions could give super good timings (for lucky values)
- the timings for other precisions could still be favorable or unfavorable
depending on the value that was drawn
This commit is contained in:
Arthur Meyre
2025-11-21 19:07:43 +01:00
committed by IceTDrinker
parent fc78450245
commit caf5e9d879
1 changed files with 16 additions and 13 deletions
Download Patch File Download Diff File Expand all files Collapse all files

29
tfhe-benchmark/benches/integer/bench.rs
View File

@@ -465,23 +465,26 @@ fn bench_server_key_binary_scalar_function_clean_inputs<F, G>(
match get_bench_type() {
BenchmarkType::Latency => {
let bench_data = LazyCell::new(|| {
let (cks, sks) = KEY_CACHE.get_from_params(param, IntegerKeyKind::Radix);
let clear_0 = gen_random_u256(&mut rng);
let clear_1 = rng_func(&mut rng, bit_size) & max_value_for_bit_size;
let ct_0 = cks.encrypt_radix(clear_0, num_block);
(sks, ct_0, clear_1)
});
let bench_data =
LazyCell::new(|| KEY_CACHE.get_from_params(param, IntegerKeyKind::Radix));
bench_id = format!("{bench_name}::{param_name}::{bit_size}_bits_scalar_{bit_size}");
bench_group.bench_function(&bench_id, |b| {
let (sks, ct_0, clear_1) = (&bench_data.0, &bench_data.1, bench_data.2);
let (cks, sks) = (&bench_data.0, &bench_data.1);
b.iter(|| {
binary_op(sks, ct_0, clear_1);
})
b.iter_batched(
|| {
let clear_0 = gen_random_u256(&mut rng);
let clear_1 = rng_func(&mut rng, bit_size) & max_value_for_bit_size;
let ct_0 = cks.encrypt_radix(clear_0, num_block);
(ct_0, clear_1)
},
|(ct_0, clear_1)| {
binary_op(sks, &ct_0, clear_1);
},
criterion::BatchSize::SmallInput,
)
});
}
BenchmarkType::Throughput => {