Merge pull request #193 from mlemanczyk/codex/optimize-benchmarks-for-data-collection

Cache benchmark data to avoid regenerating per iteration
This commit is contained in:
Marek Lemańczyk authored and GitHub committed 2025-10-18 01:10:12 +02:00
commit b2ed0d5832
4 files changed
+409 -178

No files matched your search

@@ -9,12 +9,10 @@ namespace EvenPerfectBitScanner.Benchmarks;
public class GpuModularArithmeticBenchmarks
{
private const int MaxBatchSize = 512;
private readonly ulong[] _exponents = new ulong[MaxBatchSize];
private readonly ulong[] _moduli = new ulong[MaxBatchSize];
private readonly ulong[] _mersenneModuli = new ulong[MaxBatchSize];
private readonly int[] _mersenneBitWidths = new int[MaxBatchSize];
private readonly Random _random = new(7);
private ulong[] _exponents = Array.Empty<ulong>();
private ulong[] _moduli = Array.Empty<ulong>();
private ulong[] _mersenneModuli = Array.Empty<ulong>();
private int[] _mersenneBitWidths = Array.Empty<int>();
[Params(32, 256)]
public int BatchSize { get; set; }
@@ -22,14 +20,69 @@ public class GpuModularArithmeticBenchmarks
[GlobalSetup]
public void Setup()
{
for (int i = 0; i < MaxBatchSize; i++)
{
_exponents[i] = NextExponent();
_moduli[i] = NextOddModulus();
ModularArithmeticSampleData data = ModularArithmeticSampleCache.Instance;
_exponents = data.Exponents;
_moduli = data.Moduli;
_mersenneModuli = data.MersenneModuli;
_mersenneBitWidths = data.MersenneBitWidths;
}
int bits = _random.Next(8, 62);
_mersenneBitWidths[i] = bits;
_mersenneModuli[i] = (1UL << bits) - 1UL;
private sealed class ModularArithmeticSampleData
{
public ModularArithmeticSampleData(ulong[] exponents, ulong[] moduli, ulong[] mersenneModuli, int[] mersenneBitWidths)
{
Exponents = exponents;
Moduli = moduli;
MersenneModuli = mersenneModuli;
MersenneBitWidths = mersenneBitWidths;
}
public ulong[] Exponents { get; }
public ulong[] Moduli { get; }
public ulong[] MersenneModuli { get; }
public int[] MersenneBitWidths { get; }
}
private static class ModularArithmeticSampleCache
{
private static readonly Lazy<ModularArithmeticSampleData> Cache = new(Create);
public static ModularArithmeticSampleData Instance => Cache.Value;
private static ModularArithmeticSampleData Create()
{
Random random = new(7);
var exponents = new ulong[MaxBatchSize];
var moduli = new ulong[MaxBatchSize];
var mersenneModuli = new ulong[MaxBatchSize];
var mersenneBitWidths = new int[MaxBatchSize];
for (int i = 0; i < MaxBatchSize; i++)
{
exponents[i] = NextExponent(random);
moduli[i] = NextOddModulus(random);
int bits = random.Next(8, 62);
mersenneBitWidths[i] = bits;
mersenneModuli[i] = (1UL << bits) - 1UL;
}
return new ModularArithmeticSampleData(exponents, moduli, mersenneModuli, mersenneBitWidths);
}
private static ulong NextExponent(Random random)
{
return (ulong)random.Next(24, 63);
}
private static ulong NextOddModulus(Random random)
{
ulong value = (ulong)random.NextInt64(5L, long.MaxValue);
if ((value & 1UL) == 0UL)
{
value++;
}
return value;
}
}
@@ -69,22 +122,6 @@ public class GpuModularArithmeticBenchmarks
return checksum;
}
private ulong NextExponent()
{
return (ulong)_random.Next(24, 63);
}
private ulong NextOddModulus()
{
ulong value = (ulong)_random.NextInt64(5L, long.MaxValue);
if ((value & 1UL) == 0UL)
{
value++;
}
return value;
}
private static ulong Pow2ModBaseline(ulong exponent, ulong modulus)
{
if (modulus <= 1UL)
@@ -9,12 +9,10 @@ public class MontgomeryMultiplyBenchmarks
{
private const int MaxBatchSize = 1024;
private readonly ulong[] _operandsA = new ulong[MaxBatchSize];
private readonly ulong[] _operandsB = new ulong[MaxBatchSize];
private readonly ulong[] _moduli = new ulong[MaxBatchSize];
private readonly ulong[] _nPrimes = new ulong[MaxBatchSize];
private readonly Random _random = new(1979);
private ulong[] _operandsA = Array.Empty<ulong>();
private ulong[] _operandsB = Array.Empty<ulong>();
private ulong[] _moduli = Array.Empty<ulong>();
private ulong[] _nPrimes = Array.Empty<ulong>();
[Params(64, 256, 1024)]
public int BatchSize { get; set; }
@@ -22,13 +20,64 @@ public class MontgomeryMultiplyBenchmarks
[GlobalSetup]
public void Setup()
{
for (int i = 0; i < MaxBatchSize; i++)
MontgomerySampleData data = MontgomerySampleDataCache.Instance;
_operandsA = data.OperandsA;
_operandsB = data.OperandsB;
_moduli = data.Moduli;
_nPrimes = data.NPrimes;
}
private sealed class MontgomerySampleData
{
public MontgomerySampleData(ulong[] operandsA, ulong[] operandsB, ulong[] moduli, ulong[] nPrimes)
{
ulong modulus = NextOddModulus();
_moduli[i] = modulus;
_nPrimes[i] = ComputeMontgomeryNPrime(modulus);
_operandsA[i] = (ulong)_random.NextInt64(0, (long)modulus);
_operandsB[i] = (ulong)_random.NextInt64(0, (long)modulus);
OperandsA = operandsA;
OperandsB = operandsB;
Moduli = moduli;
NPrimes = nPrimes;
}
public ulong[] OperandsA { get; }
public ulong[] OperandsB { get; }
public ulong[] Moduli { get; }
public ulong[] NPrimes { get; }
}
private static class MontgomerySampleDataCache
{
private static readonly Lazy<MontgomerySampleData> Cache = new(Create);
public static MontgomerySampleData Instance => Cache.Value;
private static MontgomerySampleData Create()
{
Random random = new(1979);
var operandsA = new ulong[MaxBatchSize];
var operandsB = new ulong[MaxBatchSize];
var moduli = new ulong[MaxBatchSize];
var nPrimes = new ulong[MaxBatchSize];
for (int i = 0; i < MaxBatchSize; i++)
{
ulong modulus = NextOddModulus(random);
moduli[i] = modulus;
nPrimes[i] = ComputeMontgomeryNPrime(modulus);
operandsA[i] = (ulong)random.NextInt64(0, (long)modulus);
operandsB[i] = (ulong)random.NextInt64(0, (long)modulus);
}
return new MontgomerySampleData(operandsA, operandsB, moduli, nPrimes);
}
private static ulong NextOddModulus(Random random)
{
ulong modulus = (ulong)random.NextInt64(3, long.MaxValue);
if ((modulus & 1UL) == 0UL)
{
modulus++;
}
return modulus;
}
}
@@ -68,17 +117,6 @@ public class MontgomeryMultiplyBenchmarks
return checksum;
}
private ulong NextOddModulus()
{
ulong modulus = (ulong)_random.NextInt64(3, long.MaxValue);
if ((modulus & 1UL) == 0UL)
{
modulus++;
}
return modulus;
}
private static ulong OriginalMontgomeryMultiply(ulong a, ulong b, ulong modulus, ulong nPrime)
{
ulong tLow = unchecked(a * b);
@@ -26,20 +26,17 @@ public class Pow2MontgomeryModBenchmarks
private const ulong OneShiftedLeft63 = 1UL << 63;
private const long OneShiftedLeft60 = 1L << 60;
private readonly MontgomeryDivisorData[] _smallDivisors = new MontgomeryDivisorData[SampleCount];
private readonly MontgomeryDivisorData[] _largeDivisors = new MontgomeryDivisorData[SampleCount];
private readonly MontgomeryDivisorData[] _veryLargeDivisors = new MontgomeryDivisorData[SampleCount];
private readonly UInt128[] _wideModuli = new UInt128[SampleCount];
private readonly ulong[] _smallExponents = new ulong[SampleCount];
private readonly ulong[] _largeExponents = new ulong[SampleCount];
private readonly ulong[] _veryLargeExponents = new ulong[SampleCount];
private readonly ulong[] _smallCycles = new ulong[SampleCount];
private readonly ulong[] _largeCycles = new ulong[SampleCount];
private readonly ulong[] _veryLargeCycles = new ulong[SampleCount];
private readonly UInt128[] _wideCycles = new UInt128[SampleCount];
private readonly Random _random = new(13);
private readonly Random _heuristicRandom = new(113);
private MontgomeryDivisorData[] _smallDivisors = Array.Empty<MontgomeryDivisorData>();
private MontgomeryDivisorData[] _largeDivisors = Array.Empty<MontgomeryDivisorData>();
private MontgomeryDivisorData[] _veryLargeDivisors = Array.Empty<MontgomeryDivisorData>();
private UInt128[] _wideModuli = Array.Empty<UInt128>();
private ulong[] _smallExponents = Array.Empty<ulong>();
private ulong[] _largeExponents = Array.Empty<ulong>();
private ulong[] _veryLargeExponents = Array.Empty<ulong>();
private ulong[] _smallCycles = Array.Empty<ulong>();
private ulong[] _largeCycles = Array.Empty<ulong>();
private ulong[] _veryLargeCycles = Array.Empty<ulong>();
private UInt128[] _wideCycles = Array.Empty<UInt128>();
private ulong? _previousPrimeOrder;
private UInt128? _previousWidePrimeOrder;
@@ -56,34 +53,18 @@ public class Pow2MontgomeryModBenchmarks
[GlobalSetup]
public void Setup()
{
for (int i = 0; i < SampleCount; i++)
{
#if DEBUG
Console.WriteLine($"Generating sample {i}");
#endif
ulong smallModulus = NextSmallOddModulus();
_smallDivisors[i] = CreateMontgomeryDivisorData(smallModulus);
_smallExponents[i] = NextSmallExponent();
_smallCycles[i] = MersenneDivisorCycles.CalculateCycleLength(smallModulus, MontgomeryDivisorData.FromModulus(smallModulus));
(ulong largeModulus, ulong largeCycle) = NextLargeModulusAndCycle();
_largeDivisors[i] = CreateMontgomeryDivisorData(largeModulus);
_largeExponents[i] = NextLargeExponent();
_largeCycles[i] = largeCycle;
ulong veryLargeModulus = NextVeryLargeOddModulus();
_veryLargeDivisors[i] = CreateMontgomeryDivisorData(veryLargeModulus);
_veryLargeExponents[i] = NextVeryLargeExponent();
#if DEBUG
Console.WriteLine($"Calculating cycle length {veryLargeModulus}");
#endif
_veryLargeCycles[i] = CalculateCycleLengthWithHeuristics(veryLargeModulus);
Console.WriteLine($"Generating wide sample {i}");
UInt128 wideModulus = NextWideOddModulus();
_wideModuli[i] = wideModulus;
_wideCycles[i] = CalculateCycleLengthWithHeuristics(wideModulus);
}
SampleCollections cache = SampleCollectionsProvider.Instance;
_smallDivisors = cache.SmallDivisors;
_largeDivisors = cache.LargeDivisors;
_veryLargeDivisors = cache.VeryLargeDivisors;
_wideModuli = cache.WideModuli;
_smallExponents = cache.SmallExponents;
_largeExponents = cache.LargeExponents;
_veryLargeExponents = cache.VeryLargeExponents;
_smallCycles = cache.SmallCycles;
_largeCycles = cache.LargeCycles;
_veryLargeCycles = cache.VeryLargeCycles;
_wideCycles = cache.WideCycles;
}
/// <summary>
@@ -230,7 +211,6 @@ public class Pow2MontgomeryModBenchmarks
for (int i = 0; i < SampleCount; i++)
{
UInt128 cycle = CalculateCycleLengthWithHeuristics(_wideModuli[i]);
_wideCycles[i] = cycle;
checksum ^= cycle;
}
@@ -363,50 +343,199 @@ public class Pow2MontgomeryModBenchmarks
}
}
private ulong NextSmallExponent() => (ulong)_random.NextInt64(1L, 1_000_000L);
private ulong NextLargeExponent() => (ulong)_random.NextInt64(OneShiftedLeft60, long.MaxValue);
private ulong NextVeryLargeExponent()
private sealed class SampleCollections
{
Span<byte> buffer = stackalloc byte[8];
_heuristicRandom.NextBytes(buffer);
ulong value = checked(BinaryPrimitives.ReadUInt64LittleEndian(buffer) | OneShiftedLeft63);
return value;
public SampleCollections(
MontgomeryDivisorData[] smallDivisors,
MontgomeryDivisorData[] largeDivisors,
MontgomeryDivisorData[] veryLargeDivisors,
UInt128[] wideModuli,
ulong[] smallExponents,
ulong[] largeExponents,
ulong[] veryLargeExponents,
ulong[] smallCycles,
ulong[] largeCycles,
ulong[] veryLargeCycles,
UInt128[] wideCycles)
{
SmallDivisors = smallDivisors;
LargeDivisors = largeDivisors;
VeryLargeDivisors = veryLargeDivisors;
WideModuli = wideModuli;
SmallExponents = smallExponents;
LargeExponents = largeExponents;
VeryLargeExponents = veryLargeExponents;
SmallCycles = smallCycles;
LargeCycles = largeCycles;
VeryLargeCycles = veryLargeCycles;
WideCycles = wideCycles;
}
public MontgomeryDivisorData[] SmallDivisors { get; }
public MontgomeryDivisorData[] LargeDivisors { get; }
public MontgomeryDivisorData[] VeryLargeDivisors { get; }
public UInt128[] WideModuli { get; }
public ulong[] SmallExponents { get; }
public ulong[] LargeExponents { get; }
public ulong[] VeryLargeExponents { get; }
public ulong[] SmallCycles { get; }
public ulong[] LargeCycles { get; }
public ulong[] VeryLargeCycles { get; }
public UInt128[] WideCycles { get; }
}
private ulong NextSmallOddModulus()
private static class SampleCollectionsProvider
{
ulong value = (ulong)_random.NextInt64(3L, 1_000_000L);
return value | 1UL;
}
private static readonly Lazy<SampleCollections> Cache = new(Create);
private (ulong modulus, ulong cycleLength) NextLargeModulusAndCycle()
{
int bitLength = _random.Next(48, 64);
ulong modulus = (1UL << bitLength) - 1UL;
return (modulus, (ulong)bitLength);
}
public static SampleCollections Instance => Cache.Value;
private ulong NextVeryLargeOddModulus()
{
Span<byte> buffer = stackalloc byte[8];
_heuristicRandom.NextBytes(buffer);
ulong value = BinaryPrimitives.ReadUInt64LittleEndian(buffer) | OneShiftedLeft63 | 1UL;
return value;
}
private static SampleCollections Create()
{
Random random = new(13);
Random heuristicRandom = new(113);
private UInt128 NextWideOddModulus()
{
Span<byte> buffer = stackalloc byte[16];
_heuristicRandom.NextBytes(buffer);
var smallDivisors = new MontgomeryDivisorData[SampleCount];
var largeDivisors = new MontgomeryDivisorData[SampleCount];
var veryLargeDivisors = new MontgomeryDivisorData[SampleCount];
var wideModuli = new UInt128[SampleCount];
var smallExponents = new ulong[SampleCount];
var largeExponents = new ulong[SampleCount];
var veryLargeExponents = new ulong[SampleCount];
var smallCycles = new ulong[SampleCount];
var largeCycles = new ulong[SampleCount];
var veryLargeCycles = new ulong[SampleCount];
var wideCycles = new UInt128[SampleCount];
ulong low = BinaryPrimitives.ReadUInt64LittleEndian(buffer);
Span<byte> highSlice = buffer.Slice(8, 8);
ulong high = BinaryPrimitives.ReadUInt64LittleEndian(highSlice) | (1UL << 63);
ulong? previousPrimeOrder = null;
UInt128? previousWidePrimeOrder = null;
UInt128 value = ((UInt128)high << 64) | low;
return value | UInt128.One;
for (int i = 0; i < SampleCount; i++)
{
ulong smallModulus = NextSmallOddModulus(random);
smallDivisors[i] = CreateMontgomeryDivisorData(smallModulus);
smallExponents[i] = NextSmallExponent(random);
smallCycles[i] = MersenneDivisorCycles.CalculateCycleLength(smallModulus, MontgomeryDivisorData.FromModulus(smallModulus));
(ulong largeModulus, ulong largeCycle) = NextLargeModulusAndCycle(random);
largeDivisors[i] = CreateMontgomeryDivisorData(largeModulus);
largeExponents[i] = NextLargeExponent(random);
largeCycles[i] = largeCycle;
ulong veryLargeModulus = NextVeryLargeOddModulus(heuristicRandom);
veryLargeDivisors[i] = CreateMontgomeryDivisorData(veryLargeModulus);
veryLargeExponents[i] = NextVeryLargeExponent(heuristicRandom);
veryLargeCycles[i] = CalculateCycleLengthWithHeuristics(veryLargeModulus, ref previousPrimeOrder);
UInt128 wideModulus = NextWideOddModulus(heuristicRandom);
wideModuli[i] = wideModulus;
wideCycles[i] = CalculateCycleLengthWithHeuristics(wideModulus, ref previousWidePrimeOrder);
}
return new SampleCollections(
smallDivisors,
largeDivisors,
veryLargeDivisors,
wideModuli,
smallExponents,
largeExponents,
veryLargeExponents,
smallCycles,
largeCycles,
veryLargeCycles,
wideCycles);
}
private static ulong NextSmallExponent(Random random) => (ulong)random.NextInt64(1L, 1_000_000L);
private static ulong NextLargeExponent(Random random) => (ulong)random.NextInt64(OneShiftedLeft60, long.MaxValue);
private static ulong NextVeryLargeExponent(Random random)
{
Span<byte> buffer = stackalloc byte[8];
random.NextBytes(buffer);
ulong value = checked(BinaryPrimitives.ReadUInt64LittleEndian(buffer) | OneShiftedLeft63);
return value;
}
private static ulong NextSmallOddModulus(Random random)
{
ulong value = (ulong)random.NextInt64(3L, 1_000_000L);
return value | 1UL;
}
private static (ulong Modulus, ulong CycleLength) NextLargeModulusAndCycle(Random random)
{
int bitLength = random.Next(48, 64);
ulong modulus = (1UL << bitLength) - 1UL;
return (modulus, (ulong)bitLength);
}
private static ulong NextVeryLargeOddModulus(Random random)
{
Span<byte> buffer = stackalloc byte[8];
random.NextBytes(buffer);
ulong value = BinaryPrimitives.ReadUInt64LittleEndian(buffer) | OneShiftedLeft63 | 1UL;
return value;
}
private static UInt128 NextWideOddModulus(Random random)
{
Span<byte> buffer = stackalloc byte[16];
random.NextBytes(buffer);
ulong low = BinaryPrimitives.ReadUInt64LittleEndian(buffer);
ulong high = BinaryPrimitives.ReadUInt64LittleEndian(buffer[8..]) | (1UL << 63);
UInt128 value = ((UInt128)high << 64) | low;
return value | UInt128.One;
}
private static ulong CalculateCycleLengthWithHeuristics(ulong modulus, ref ulong? previousPrimeOrder)
{
MontgomeryDivisorData divisorData = MontgomeryDivisorData.FromModulus(modulus);
if (modulus <= 1UL || (modulus & 1UL) == 0UL)
{
return MersenneDivisorCycles.CalculateCycleLength(modulus, divisorData);
}
ulong order = PrimeOrderCalculator.Calculate(
modulus,
previousPrimeOrder,
divisorData,
PrimeOrderCalculator.PrimeOrderSearchConfig.HeuristicDefault,
PrimeOrderCalculator.PrimeOrderHeuristicDevice.Gpu);
if (order != 0UL)
{
previousPrimeOrder = order;
return order;
}
previousPrimeOrder = null;
return MersenneDivisorCycles.CalculateCycleLength(modulus, divisorData);
}
private static UInt128 CalculateCycleLengthWithHeuristics(UInt128 modulus, ref UInt128? previousWidePrimeOrder)
{
if (modulus <= UInt128.One || (modulus & UInt128.One) == UInt128.Zero)
{
return MersenneDivisorCycles.GetCycle(modulus);
}
UInt128 order = PrimeOrderCalculator.Calculate(
modulus,
previousWidePrimeOrder,
PrimeOrderCalculator.PrimeOrderSearchConfig.HeuristicDefault,
PrimeOrderCalculator.PrimeOrderHeuristicDevice.Gpu);
if (order != UInt128.Zero)
{
previousWidePrimeOrder = order;
return order;
}
previousWidePrimeOrder = null;
return MersenneDivisorCycles.GetCycle(modulus);
}
}
private static MontgomeryDivisorData CreateMontgomeryDivisorData(ulong modulus)
@@ -13,21 +13,86 @@ public class Pow2MontgomeryModUInt128Benchmarks
private const int SampleCount = 256;
private const ulong Pow2WindowFallbackThreshold = 32UL;
private readonly UInt128[] _moduli = new UInt128[SampleCount];
private readonly UInt128[] _exponents = new UInt128[SampleCount];
private readonly Random _random = new(19);
private UInt128[] _moduli = Array.Empty<UInt128>();
private UInt128[] _exponents = Array.Empty<UInt128>();
[GlobalSetup]
public void Setup()
{
for (int i = 0; i < SampleCount; i++)
{
UInt128 modulus = NextWideOddModulus();
_moduli[i] = modulus;
WideSampleData cache = WideSampleDataCache.Instance;
_moduli = cache.Moduli;
_exponents = cache.Exponents;
}
UInt128 exponent = NextWideExponent();
_exponents[i] = exponent;
private sealed class WideSampleData
{
public WideSampleData(UInt128[] moduli, UInt128[] exponents)
{
Moduli = moduli;
Exponents = exponents;
}
public UInt128[] Moduli { get; }
public UInt128[] Exponents { get; }
}
private static class WideSampleDataCache
{
private static readonly Lazy<WideSampleData> Cache = new(Create);
public static WideSampleData Instance => Cache.Value;
private static WideSampleData Create()
{
Random random = new(19);
var moduli = new UInt128[SampleCount];
var exponents = new UInt128[SampleCount];
for (int i = 0; i < SampleCount; i++)
{
moduli[i] = NextWideOddModulus(random);
exponents[i] = NextWideExponent(random);
}
return new WideSampleData(moduli, exponents);
}
private static UInt128 NextWideOddModulus(Random random)
{
Span<byte> buffer = stackalloc byte[16];
random.NextBytes(buffer);
ulong low = BinaryPrimitives.ReadUInt64LittleEndian(buffer);
ulong high = BinaryPrimitives.ReadUInt64LittleEndian(buffer[8..]);
if (high == 0UL)
{
high = 1UL;
}
UInt128 modulus = ((UInt128)high << 64) | low;
modulus |= UInt128.One;
if (modulus <= UInt128.One)
{
modulus += 2UL;
}
return modulus;
}
private static UInt128 NextWideExponent(Random random)
{
Span<byte> buffer = stackalloc byte[16];
random.NextBytes(buffer);
ulong low = BinaryPrimitives.ReadUInt64LittleEndian(buffer);
ulong high = BinaryPrimitives.ReadUInt64LittleEndian(buffer[8..]);
if (high == 0UL)
{
high = 1UL;
}
return ((UInt128)high << 64) | low;
}
}
@@ -63,44 +128,6 @@ public class Pow2MontgomeryModUInt128Benchmarks
return checksum;
}
private UInt128 NextWideOddModulus()
{
Span<byte> buffer = stackalloc byte[16];
_random.NextBytes(buffer);
ulong low = BinaryPrimitives.ReadUInt64LittleEndian(buffer);
ulong high = BinaryPrimitives.ReadUInt64LittleEndian(buffer[8..]);
if (high == 0UL)
{
high = 1UL;
}
UInt128 modulus = ((UInt128)high << 64) | low;
modulus |= UInt128.One;
if (modulus <= UInt128.One)
{
modulus += 2UL;
}
return modulus;
}
private UInt128 NextWideExponent()
{
Span<byte> buffer = stackalloc byte[16];
_random.NextBytes(buffer);
ulong low = BinaryPrimitives.ReadUInt64LittleEndian(buffer);
ulong high = BinaryPrimitives.ReadUInt64LittleEndian(buffer[8..]);
if (high == 0UL)
{
high = 1UL;
}
return ((UInt128)high << 64) | low;
}
private static UInt128 Pow2MontgomeryModWindowedGpuShim(UInt128 exponent, UInt128 modulus)
{
if (modulus == UInt128.One)