» Tag
performance
240 postsRunloom Brings Go-Style Fibers to Python Concurrency
Runloom is an experimental extension for free-threaded CPython 3.13t that brings Go-style, work-stealing fibers to Python, matching Go's networking throughput in benchmarks.
Netflix Cuts Cassandra Partition Latency to Milliseconds via Per-ID Splitting
Netflix's AI team splits wide Cassandra partitions per ID asynchronously during reads, cutting tail latency from seconds to milliseconds with zero app changes.
Stream Compaction on NEON: Hand-Vectorizing copy_if
NEON lacks a compress instruction, so compilers can't vectorize copy_if. This piece shows how tbl-based lookup tables restore SIMD throughput with a 30x speedup.
CommitBrief — AI code reviews, right in your terminal
A provider-agnostic, local-first CLI that reviews your staged changes, a historic range, or a whole GitHub pull request. Zero telemetry, no server. Free and open source.
commitbrief.comWhy Rust services hold onto memory: glibc, jemalloc and munmap
A Rust service's RSS stayed flat after load tests. The cause wasn't a memory leak but glibc's arena-based allocator behavior; switching to jemalloc fixed it.
Allocating on the Stack
Go 1.25 and 1.26 enhance stack allocations, reducing garbage collector load and improving memory efficiency.
Taffy: A Flexible, High-Performance UI Layout Library
Taffy is a flexible, high-performance UI layout library written in Rust.
Hiding Memory Latency in eBPF: Solutions to Avoid Stalling
Exploring methods to hide memory latency with eBPF and its importance for engineers.
Self-hosting Kimi K3: Cost and Performance Insights
Analysis of cost and performance when self-hosting Kimi K3.
Ninfer: High-performance single-GPU inference
NInfer offers a high-performance C++/CUDA inference engine for RTX 5090.
Frontier-class LLM Inference on a Laptop CPU
cpubrrr surpasses llama.cpp on Apple M4 Max CPU for frontier-class LLMs.