hpc-patterns

High-performance computing patterns for C++20 including cache-friendly data structures, SIMD vectorization, memory management, thread parallelism, lock-free data structures, and NUMA-aware allocation.

Install with Tessl CLI

npx tessl i github:ysyecust/everything-claude-code --skill hpc-patterns

What are skills?

Quality

73%

Does it follow best practices?

Impact

Pending

No eval scenarios have been run

Optimize this skill with Tessl

npx tessl skill review --optimize ./skills/hpc-patterns/SKILL.md

SKILL.md

Review

Evals

Evaluation results

100%

17%

Particle Simulation Performance Optimization

SoA layout and cache alignment

Criteria

Without context

With context

SoA field arrays

100%

Separate position arrays

100%

No primary AoS

100%

alignas(64) usage

100%

Cache-line padding

100%

Aligned allocator

100%

high_resolution_clock

50%

100%

Chrono duration

100%

Benchmark output

100%

Contiguous field storage

100%

Without context: $0.3792 · 20m 24s · 13 turns · 105 in / 6,370 out tokens

With context: $1.1369 · 49m 37s · 28 turns · 503 in / 14,536 out tokens

84%

56%

High-Performance Vector Math Library

SIMD vectorization and parallel processing

Criteria

Without context

With context

__restrict__ pointers

100%

#pragma omp simd

100%

Alignment assertion

#pragma omp simd aligned

AVX2 intrinsics include

100%

AVX2 fmadd usage

100%

Aligned allocation

100%

std::jthread workers

100%

hardware_concurrency

100%

Chunk-size division

100%

high_resolution_clock

100%

Without context: $0.3632 · 16m 41s · 17 turns · 137 in / 4,764 out tokens

With context: $1.2504 · 50m 20s · 37 turns · 510 in / 13,759 out tokens

90%

18%

Concurrent Task Scheduler and Lock-Free Event Queue

Thread pool and lock-free data structures

Criteria

Without context

With context

jthread workers

hardware_concurrency default

100%

packaged_task submission

100%

future return

100%

Lock-free push CAS release

58%

100%

Lock-free pop CAS acquire

58%

100%

CAS relaxed failure order

100%

Memory pool allocation

100%

high_resolution_clock

100%

GFLOPS calculation

100%

Without context: $0.7276 · 42m 2s · 16 turns · 121 in / 12,969 out tokens

With context: $1.2865 · 43m 47s · 25 turns · 234 in / 19,977 out tokens

Evaluated: 5 days ago
Agent: Claude Code
Model: Claude Sonnet 4.6

Table of Contents

Particle Simulation Performance Optimization High-Performance Vector Math Library Concurrent Task Scheduler and Lock-Free Event Queue

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.