Welcome to the Mechanical Sympathy group.  This group is for discussing how to code sympathetically to and measure the underlying stack/platform so good performance can be extracted.

Showing 1-20 of 706 topics
Write release post Francesco Nigro 6/17/17
references for an efficient clock + counter implementation? Rick Richardson 6/6/17
Why does the G1 GC shrink number of regions? Alexandr Nikitin 6/5/17
Aeron zero'ing buffer? Peter Veentjer 5/30/17
processor.max_cstate, intel_idle.max_cstate and /dev/cpu_dma_latency Himanshu Sharma 5/26/17
Isolating low latency application on CPU-0? Michael Mattoss 5/26/17
Mystery: Why does the JVM show more latency for the same block of code after a busy spin pause? J Crawford 5/15/17
Memory model of mprotect Yichao Yu 5/5/17
Falcon JIT compiler - enhanced LLVM backend for JVM Daniel Eloff 5/5/17
Why would SocketChannel be slower when sending a single msg instead of 1k msgs after proper warmup? J Crawford 4/20/17
How would you approach the problem with large data over MQ Vero K. 4/17/17
Socket buffer and rates streaming Vero K. 4/16/17
How to recompile native JDK code (C/C++) and deploy on my JVM? ( J Crawford 4/12/17
strange behaviour in a virtual environment Erik Svensson 4/11/17
Why a Java method invocation is slower when you call it somewhere else in your code? J Crawford 4/9/17
Linux futex_wait() bug... [Yes. You read that right. UPDATE to LATEST PATCHES NOW]. Gil Tene 3/18/17
New Java Web Server based on mechanical sympathy Nathan Tippy 3/16/17
Using a SEDA-like architecture to reduce icache misses Avi Kivity 3/16/17
Using JDK 9 Memory Order Modes - Doug Lea (VarHandles) Benoît Paris 3/15/17
Is it me or is the Java memory model spec poorly written? Sharath Gururaj 3/10/17
More topics »