Tasty » Adam Sampson's ag0803 bookmarks

How Not To Run An A/B Test edit / delete

"Decide on a sample size in advance and wait until the experiment is over before you start believing the “chance of beating original” figures that the A/B testing software gives you."

to ag0803 honours significance statistics testing ... on 24 August 2014

The USE Method edit / delete

"For every resource, check utilization, saturation, and errors."

to ag0803 analysis benchmarking debugging networking performance ... on 24 August 2014

C++11/14 idioms I use every day | Seshbot Programs edit / delete

Generally useful (rather than obscure) features.

to ag0700 ag0803 c++ c++11 c++14 language-design ... on 24 August 2014

Acquiring high-resolution time stamps (Windows) edit / delete

For AG0803 students: here's how to get a better timing reference.

to ag0803 benchmark timer timing win32 ... on 24 August 2014

A concrete illustration of practical running time vs big-O notation - The Old New Thing - Site Home - MSDN Blogs edit / delete

One for AG0803 students: why algorithmic complexity isn't everything in a world with complex memory architectures.

to ag0803 complexity data-structures memory optimisation performance ... on 13 August 2014

mikeash.com: Friday Q&A 2014-06-06: Secrets of dispatch_once edit / delete

Nice worked example of how to implement dispatch_once.

to ag0803 atomics c++ concurrency threading ... on 26 July 2014

Floating-Point Determinism | Random ASCII edit / delete

"My hobby: injecting code into other processes and changing the floating-point rounding mode on some threads".

to ag0700 ag0803 floating-point numeric ... on 01 July 2014

Simple, Fast, and Practical Non-Blocking and Blocking Concurrent Queue Algorithms edit / delete

"Drawing ideas from previous authors, we present a new non-blocking concurrent queue algorithm and a new two-lock queue algorithm in which one enqueue and one dequeue can proceed concurrently." From 1996.

to ag0803 concurrency data-structures nonblocking queue ... on 04 June 2014

SSE – Vectorizing conditional code | Félix Abecassis edit / delete

An example of using GCC's vectorisor.

to ag0803 gcc optimisation vectorisation ... on 02 June 2014

Blosc edit / delete

"Blosc is a high performance compressor optimized for binary data. It has been designed to transmit data to the processor cache faster than the traditional, non-compressed, direct memory fetch approach via a memcpy() OS call." This is very neat -- it's a compressed data format designed so that you can decompress it into cache very quickly. AG0803 students take note!

to ag0803 cache compression memory optimisation python ... on 28 April 2014