== 3 9 . 4/ String Compression Lets make a compressed string table https://github.com/AshleyRoll/squeeze map from enum Key to Compressed String Hu�man Coding for compression Output struct: Mapping choose an arbitrary amount of work they will allow in constexpr context Complex processing like compression will hit the limits Had to make more complex implementation to cache bit streams rather than walk

units::unitratio<1>, units::base_unitratio<1>, std::ratio<0, 1>, std::ratio<1>, std::ratio<0, 1>, std::ratio<0, 1>, std::ratio<0, 1>, std::ratio<0, 1>, std::ratio<0, 1>, std::ratio<0, 1> >, std::ratio<0 std::ratio<0, 1>, std::ratio<0, 1> >; UnitTo = units::unit< std::ratio<1>, units::base_unitratio<1>, std::ratio<0, 1>, std::ratio<-1> > >; T = double]': units.h:1956:41: required from 'constexpr units::unitratio<1>, units::base_unitratio<1>, std::ratio<0, 1>, std::ratio<1>, std::ratio<0, 1>, std::ratio<0, 1>, std::ratio<0, 1>, std::ratio<0, 1>, std::ratio<0, 1>, std::ratio<0, 1> >, std::ratio<0

Time Zones <ratio>4 Agenda  Compile-Time Rational Numbers  Durations  Clocks  Time Points  Dates  Time Zones5 Compile-Time Rational Numbers  Defined in <ratio>  Work with rational in the library6 Compile-Time Rational Numbers  Define a rational number: using r1 = ratio<1, 60>; // Represents 1/60  Retrieve numerator and denominator intmax_t num { r1::num }; intmax_t r1 = ratio; // Error7 Compile-Time Rational Numbers  Arithmetic with ratio_add, ratio_subtract, ratio_multiply, and ratio_divide: using r1 = ratio<1, 60>; // 1/60 using r2 = ratio<1, 30>;

3 4 std::chrono::nanoseconds elapsed_time(uint64_t num_cpu_ticks) { using NS_PER_TICK = std::ratio<1'000'000'000, CPU_CLOCK_HZ>; return std::chrono::nanoseconds{ num_cpu_ticks * NS_PER_TICK::num 3 4 std::chrono::nanoseconds elapsed_time(uint64_t num_cpu_ticks) { using NS_PER_TICK = std::ratio<1'000'000'000, CPU_CLOCK_HZ>; return std::chrono::nanoseconds{ num_cpu_ticks * NS_PER_TICK::num 3 4 std::chrono::nanoseconds elapsed_time(uint64_t num_cpu_ticks) { using NS_PER_TICK = std::ratio<1'000'000'000, CPU_CLOCK_HZ>; return std::chrono::nanoseconds{ num_cpu_ticks * NS_PER_TICK::num

perform worse for size and speed on ARM (single core). Runtime (ms) Size (B) Throw3 Return Ratio Throw3 Return Ratio X86 2199 5 435 15536 15316 1.01 ARM 59104 128 477 14152 10356 1.4• Two main implementations Size (B) Throw1 Return Ratio Throw1 Return Ratio X86 2362 1.2 1861 14936 14056 1.1 ARM 53747 40 1343 14540 10276 1.4 Runtime (ms) Size (B) Throw3 Throw1 Ratio Throw3 Throw1 Ratio X86 2199 2262 0.97 15536 15536 14936 1.04 ARM 59104 53747 1.1 14152 14540 0.97 • Ratio when adding one exception is worse – just adding exceptions has significant overhead. • Some gain on runtime on ARM, results are not consistent

expressed with std::ratio. This is not enough for a general- purpose physical units library. CppCon 2024: Improving our safety with a quantities and units library Limitations of std::ratio 45In std::chrono::duration n, the magnitude of a unit is always expressed with std::ratio. This is not enough for a general- purpose physical units library. • std::ratio is implemented in terms of std::intmax_t • Impossible to degree CppCon 2024: Improving our safety with a quantities and units library Limitations of std::ratio 1 eV = 1.602176634 × 10−19 J 1 Da = 1.660539040(20) × 10−27 kg 45namespace si { template

0 码力 | 207 页 | 6.93 MB | 6 月前

3

parallelism Maximize data reuse & minimize communication via algorithms with low “surface-to-volume” ratio BLAS level 1 2 3 BLAS operations Dot products, vector norms, vector sum, plane rotations Matrix-vector parallelism Maximize data reuse & minimize communication via algorithms with low “surface-to-volume” ratio BLAS level 1 2 3 BLAS operations Dot products, vector norms, vector sum, plane rotations Matrix-vector parallelism Maximize data reuse & minimize communication via algorithms with low “surface-to-volume” ratio BLAS level 1 2 3 BLAS operations Dot products, vector norms, vector sum, plane rotations Matrix-vector

- std::uint16_t, - Load ID_FOR_REFRESH_RATE, - Default is 15, - Ratio is 10, - Max is 300 as tSO WHY NOT WRITING IT int main() { Parameter< std::uint16_t std::uint16_t, Load, Default <15>, Ratio <10,1>, Max <300> > t; std::cout << t.Value() << std::endl; return 0; } Parameter< std::uint16_t, Load, Default <15>, Ratio <10,1>, Max <300> > t;template class Parameter

Micro-benchmarks ▶ Operation ratio for write-dominated workload ▶ Lists: 40% Insert, 40% Delete, 20% Find ▶ Map: 40% Insert, 30% Delete, 10% Update, 20% Find ▶ Operation ratio for read-dominated workload Data Structures Live Demonstration References Demonstration Settings Micro-benchmarks ▶ Operation ratio: 33% Insert, 33% Delete, 34% Find ▶ Number of Transactions: 10K ▶ Key Range: 10K A Persistent Hash

mag<3> * QuarterNote> {} DottedHalfNote; inline constexpr struct EigthNote : named_unit<"8th", mag<ratio{1, 2}> * QuarterNote> {} EigthNote; inline constexpr struct DottedQuarterNote : named_unit<"q QuarterNoteTriplet : named_unit<"qt", mag<ratio{1, 3}> * HalfNote> {} QuarterNoteTriplet; inline constexpr struct SixteenthNote : named_unit<"16th", mag<ratio{1, 2}> * EigthNote> {} SixteenthNote; //