General

Profile

Erik Lindahl

  • Email: erik@kth.se
  • Registered on:
  • Last connection: 11/17/2017

Issues

Projects

Activity

11/24/2017

02:58 PM GROMACS Revision c6ce8efc: Detect AVX-512 FMA units to choose best SIMD
Add a test program that times AVX-512 code to
detect single vs. dual AVX-512 FMA units.
Added CMake code to always co...

11/21/2017

07:42 AM GROMACS Revision c4cd9968: Separate SIMD flag detection from management
This provides separate cmake functions to detect the flags
required to compile each SIMD instruction set. These
funct...

11/18/2017

01:53 AM GROMACS Task #2294: Require identical hardware on nodes on parallel runs
As long as somebody volunteers to maintain it, I'm fine having it. However, arguing that it's an important feature an...

11/17/2017

06:28 PM GROMACS Task #2294: Require identical hardware on nodes on parallel runs
I was not intending this for release-2018, but I wanted to create an entry for the future while I remember.
Multis...

11/14/2017

02:41 PM GROMACS Task #2294: Require identical hardware on nodes on parallel runs
We might actually have more similar opinions than we think. If we compare it to running a binary compiled for a diffe...
02:17 AM GROMACS Feature #2295 (New): Don't write a ton of log info to stderr
We have gradually started to use stderr a bit as a backup copy of the logfile, which should be avoided.
However, wh...
01:53 AM GROMACS Task #2294: Require identical hardware on nodes on parallel runs
The main reason is of course that it adds complexity to the code, which costs time & effort.
Right now one problem...

11/13/2017

04:59 PM GROMACS Task #2294 (New): Require identical hardware on nodes on parallel runs
Some Gromacs routines currently try to cater to nodes with different hardware.
However, as hardware has gotten mor...

11/12/2017

03:50 PM GROMACS Revision 2af9c550: Only issue FFT warning messages on changes
Similar to other CMake modules, we should only issue
warnings at the first invocation, or if the FFT library
was chan...

09/14/2017

09:51 AM GROMACS Bug #2243 (Resolved): Reference SIMD exp doesn't flush to zero for very large negative inputs
Applied in changeset commit:36311d95c2d8b0ebd889d96b681795c4c34634dd.

Also available in: Atom