Explore micro-optimization techniques for Go code in this conference talk from GopherCon 2018. Learn how to improve performance by writing more efficient Go, starting with a hash function implementation. Follow along as the speaker optimizes code from a direct RFC implementation to surpassing the standard library's performance. Discover the impact of API design on performance, avoid microbenchmarking pitfalls, and gain an introduction to profiling. Acquire practical tips for optimization, including algorithmic improvements, program tuning, and diving deep into performance analysis. Examine additional examples from standard libraries, particularly in cryptography, and other high-performance Go projects to further enhance your optimization skills.
Overview
Syllabus
Introduction
What is BlakeToHash
The problem
Blake
Hash Interface
Go Benchmark Ecosystem
Go Testing
Benchmarks
WebView
Math Bits
G
Inlining
Inliner
Hard stops
Round function
inline function
G is too expensive
Get rid of function calls
Triple addition
Budget code
Dont chase this
Whats next
P prof
Assembly
Bounce Checking
SSA
Go Compiler
Linear code
Bounce checks
Hash state
Preventative hints
Eliminate bounds checks
Performance improvement
Eliminate allocations
Slices
Finalize
Pattern Matching
Sum
Some
Slice
Pvalues
Diminishing Returns
More Optimization
Worth it
Taught by
Gopher Academy
