The Performance Cost Of Spectre/Meltdown/Foreshadow Mitigations On Linux
1–10 of 33 posts
Re: The Performance Cost Of Spectre/Meltdown/Foreshadow Mitigations On Linux
#2tl;dr: 2% - 20% degradation in performance depending on specific task; AMD EPYC less affected (sometimes significantly) as compared to Xeon.
Re: The Performance Cost Of Spectre/Meltdown/Foreshadow Mitigations On Linux
#3AMD seems to fare better with the patches. Its good for AMD. I remember when AMD64 came out and AMD had a lead on intel. Its good to have have some competition.
Re: The Performance Cost Of Spectre/Meltdown/Foreshadow Mitigations On Linux
#4I guess we know why Intel tried on that benchmarking restriction. Serving static content from Nginx is a 22% drop on Intel, a 7% drop on AMD. That's huge for cloud providers!
Re: The Performance Cost Of Spectre/Meltdown/Foreshadow Mitigations On Linux
#5this is very useful to know if i've already got a server or ten thousand, but unfortunately the data doesn't help me with making a purchase decision. does amd overtake intel performance-wise after mitigations or is it only a reduced gap?
Re: The Performance Cost Of Spectre/Meltdown/Foreshadow Mitigations On Linux
#6I guess we know why Intel tried on that benchmarking restriction. Serving static content from Nginx is a 22% drop on Intel, a 7% drop on AMD. That's huge for cloud providers!
The microcode update allowed an optimization (VMX conditional cache flushes) for virtual machine hosts. It was just a lawyer having fun, I guess.
Re: The Performance Cost Of Spectre/Meltdown/Foreshadow Mitigations On Linux
#7this is very useful to know if i've already got a server or ten thousand, but unfortunately the data doesn't help me with making a purchase decision. does amd overtake intel performance-wise after mitigations or is it only a reduced gap?
See http://openbenchmarking.org/result/1808296-RA-SMELTDOWN82&ob... for the full data set.
Re: The Performance Cost Of Spectre/Meltdown/Foreshadow Mitigations On Linux
#8KPTI hurts (a lot). Syscall latency and architectural impact is dramatically increased with KPTI, in large part due to the TLB flush that is typically required.
This is probably why AMD CPUs do not have as big of a drop; KPTI is not (urgently) necessary on AMD CPUs.
Re: The Performance Cost Of Spectre/Meltdown/Foreshadow Mitigations On Linux
#9I guess we know why Intel tried on that benchmarking restriction. Serving static content from Nginx is a 22% drop on Intel, a 7% drop on AMD. That's huge for cloud providers!
The microcode update allowed an optimization (VMX conditional cache flushes) for virtual machine hosts. It was just a lawyer having fun, I guess.
The theory I heard that seems more likely than a rogue lawyer, is that it accidentally got left in from the license of a pre-release test version -- I can understand why they'd want to prevent benchmarks of unreleased software (not that I necessarily agree with that).
Re: The Performance Cost Of Spectre/Meltdown/Foreshadow Mitigations On Linux
#10AMD seems to fare better with the patches. Its good for AMD. I remember when AMD64 came out and AMD had a lead on intel. Its good to have have some competition.
Except Intel + mitigations still much faster on most workloads than AMD without mitigations.