Feature request - CPU Temp

[KilroySmith]

Just built a Ryzen 5900x system (it's amazing how cheap last years technology can be), and while waiting for Memtest to finish running, I realized that there's one killer feature missing. CPU Temp.
No, I don't want to KNOW CPU temp - Memtest already does a good job of displaying that. What I want is a configurable item called "Desired CPU Temp", which uses spare CPU cycles (and/or spare CPU cores) to attempt to drive the processor temp up to the setting. I know that silicon gets slower as it gets hotter; so what I want to be able to do is run MemTest with my CPU at 90C (with Memtest throttling the CPU-intensive tasks to stay around the temp setting) to assure that memory errors don't start cropping up when I'm running an ASIC simulation using all cores at 100% (who am I kidding; let's go with "running a AAA game that drives the CPU and GPU temps through the roof"). Laying on the floor for hours pointing a hair dryer at the CPU while Memtest runs is one solution, I guess; plugging/unplugging the CPU fan based on the displayed temperature is another. But those seem somewhat...redneck.

Anyway, I really appreciate the availability of Memtest86. I've been using it for a lot of years - but only about once every 5 years or so when I build a new machine. This new machine is replacing my workhorse Phenom II X4 machine (OK, so it's been more than 5 years for this one). Heck, it was great (well, actually, disappointing) when I enabled XMP on my DDR4 and had Memtest tell me it was failing - it would have been remarkably frustrating to find that out without Memtest. Tweaked a few values, and now things are looking great. Thanks.

[David (PassMark)]

Main aim of MemTest86 is to test the RAM (not so much the CPU).
But yes, being able to lift the temperature of the entire system may expose some additional errors on some marginal systems. Pretty much any additional code that runs on the CPU however would slow down the memory tests to some degree (all code requires some use of cache & RAM). It would also be unlikely that arbitrary temperature targets could be hit, while at the same time as doing a RAM test. i.e. 90C might be impossible to hit, depending on the cooling solution.
It is on our to do list to look at at some point in the future however.

[KilroySmith]

I guess I'm a bit of an outlier; with 12/24 cores on this CPU there's a few doing nothing but twiddling their thumbs during a test. That's undoubtedly not true for the vast majority of users, and I understand that adding code to generate heat is probably gonna slow down the vast majority of users. Just thought I'd throw it out there, and I appreciate the response.