News for hackers. Stuff that matters.

A comment on Intel's New Low: Commissioning Misleading Core I9-9900K Benchmarks [video]. Parent comment

Re:Intel's New Low: Commissioning Misleading Core I9-9900K B... by iddqd on Tuesday October 09, 2018 @11:10AM (#18176542)
I like that the "Game mode" makes the AMD CPU's perform worse in game benchmarks...
[ Parent | on HN ]
Re:Intel's New Low: Commissioning Misleading Core I9-9900K B... by c2h5oh on Tuesday October 09, 2018 @12:07PM (#18177107)
Game mode was added for threadripper (workstation cpu) which has multiple dies - many games were hit heavily by numa.

It's also available for desktop parts to disable one ccx when game doesn't use more than 8 threads and you want to squize a couple more % of perf.

Game mode is disabled by default, requires AMD software to be installed to enable - they went extra mile to gimp competition..

[ Parent | on HN ]
Re:Intel's New Low: Commissioning Misleading Core I9-9900K B... by cma on Tuesday October 09, 2018 @12:22PM (#18177274)
The reviewers used Game Mode on a CPU without NUMA, so it was pointless and only meant to damage.
[ Parent | on HN ]
Re:Intel's New Low: Commissioning Misleading Core I9-9900K B... by Filligree on Tuesday October 09, 2018 @01:03PM (#18177691)
It can still help, if used in precisely the right circumstance. Not that this makes your statement any less true, I'm just being needlessly pedantic.

One amusing thing I've noticed is that, playing the same game (Minecraft), Windows runs it ~25% faster in game mode than default NUMA.

In Linux the situation is reversed; it runs 10% faster with NUMA enabled, maybe because the Java garbage collector is NUMA-aware and I'm using enough memory that it's split across NUMA nodes anyway.

But either mode is faster in Linux than Windows.

[ Parent | on HN ]