Get the latest tech news
Playing with BOLT and Postgres
playing with BOLT and Postgres
But it’s actually backed by a hierarchy of memory types with vastly different performance (throughput, latency, energy costs, …). Note: I wonder if it would be possible to do “poor man’s Stabilizer” by randomly adding local variables to functions, to change the size of the stacks. To measure the impact of BOLT optimization, I ran a couple traditional benchmarks - pgbench for OLTP, and the TPC-H queries for OLAP.
Or read this on Hacker News