RSAPIAIN_DMC
New member
Hello.
Just noticed some couple days ago that there are a lot of AMD (and Intel) speculative execution vulnerabilities
For Zen1 and newer there is a speculative execution vulnerability, that would be fixed usually by server manufacturer; In my case I have Zen3 machines.
The issue is that these speculative execution vulnerabilities, the fix tends to impact performance of some CPU/Memory intensive applications,and sometimes by a noticeable number: 30%, or even 50% or more time to finish a task
I would like to know if someone has some numbers of WRF impact performance: how much extra time it takes to run a 120 hour simulation with hourly output, after a set or even specific fixes are applied, using a given domain and the same data.
Or this has not been measured by someone?
Thank you in advance.
Some benchmarks I found regarding the impact of the fix "Inception", in different workloads
Just noticed some couple days ago that there are a lot of AMD (and Intel) speculative execution vulnerabilities
For Zen1 and newer there is a speculative execution vulnerability, that would be fixed usually by server manufacturer; In my case I have Zen3 machines.
The issue is that these speculative execution vulnerabilities, the fix tends to impact performance of some CPU/Memory intensive applications,and sometimes by a noticeable number: 30%, or even 50% or more time to finish a task
I would like to know if someone has some numbers of WRF impact performance: how much extra time it takes to run a 120 hour simulation with hourly output, after a set or even specific fixes are applied, using a given domain and the same data.
Or this has not been measured by someone?
Thank you in advance.
Some benchmarks I found regarding the impact of the fix "Inception", in different workloads
