Links tagged “performance”
44 links, newest first.
malisper.me
Postgres was designed when the disk was the bottleneck.
github.com
GigaToken: ~1000x faster language model tokenization - a thousandfold speedup on the least glamorous step in the pipeline, with benchmarks in the repo.
purplesyringa.moe
Quadrupling code performance with a "useless" if - branch prediction, aliasing and autovectorisation, taken apart with benchmarks.
fergusfinn.com
What happens when you run a CUDA kernel? - the whole journey from launch to silicon, one layer at a time.
jayfreestone.com
You might not need a service worker - the browser APIs that often make it redundant.
brooker.co.za
Meet Alice. Alice is impatient.: Marc Brooker on why your average user waits longer than your average response time. Queueing theory you can feel.
danielchasehooper.com
A fast, growable array with stable pointers in C.
tomtunguz.com
The Hungry, Hungry AI Model
zhihaojia.medium.com
Read more →
thenewstack.io
Read more →
spring.io
Weighs partitioned topic routing against many direct exchanges, and shows how exchange-to-exchange bindings keep a large RabbitMQ topology flexible.
itnext.io
Throughput, resource cost and feature comparison of Kubernetes CNI plugins on a 10Gbit/s bare-metal cluster, re-run on Kubernetes 1.14 in April 2019.
cybertec-postgresql.com
Why aggregating first and joining the lookup table afterwards beats joining five million rows then grouping, with the execution plans that explain the gap.
jaegertracing.io
Documentation portal for Jaeger, the open-source distributed tracing platform, covering architecture, deployment, storage backends and operations.
peterbe.com
Benchmarks several ways to remove duplicates from a Python list, and finds list(dict.fromkeys(seq)) fastest when insertion order must be preserved.
arkwright.github.io
Account of scaling a React server-side rendering service on fixed hardware — round-robin load balancing, a client-side rendering fallback under load, and the risks of component caching.
explain.depesz.com
Paste a PostgreSQL EXPLAIN plan and get it back as a readable annotated tree that highlights the costly nodes; plans can be shared or anonymised.
numba.pydata.org
Open-source JIT compiler that translates a subset of Python and NumPy code to machine code through decorators, with no separate compilation step.
pypi.python.org
Python module that monitors a process's memory usage and reports memory consumption line by line.
