Disabled Hyperthreading for CPU temps?

[dangee]

Before I begin, I am not that great with computers so I am not at all very knowledgeable.

 

I recently bought a Lenovo Legion Y540, with a 9th Gen Intel Core i7 9750h, and a GTX 1650. I was testing my CPU temps with HWiNFO, since I already could see my GPU temps under load on MSI Kombuster, and I noticed that some of my CPU cores were reaching 90+ degrees Celsius literally within a few seconds of starting a CPU stressors, it also said that they were throttling. After some research, I disabled hyperthreading in my BIOS menu and immediately saw the temps not even passing 70 degrees, as well as running the game better than before. Is it possible that with the hyperthreading the CPU was constantly throttling and making the performance worse than without and so with it disabled and the lower temps it wasn't throttling and allowing better performance? I read that disabling the hyperthreading feature should have a large impact on gaming, but I haven't noticed it, in fact the opposite has happened.

 

Secondly, before trying the method I tried above, I simply tried to change to Processor Power Management in the advanced power plan settings. I had the minimum set to 100% and the maximum set 100% as well. Changing the minimum seemingly did nothing from what I could tell, and changing the maximum simply from 100% down to 99% (as well as the minimum far below 99% at the same time) caused a massive change. Under stress tests the CPU would barely reach 70 degrees if even that. I didn't pay much attention to performance here, since the temps I was seeing confused me enough and I undid the changes. I don't understand how this is at all possible, it doesn't make sense to me. Can anyone explain this?

 

I'd also like to add that I did these 2 tests separately. Also, I have literally not seen my GPU temperature exceed 65 degrees once ever, and it is extremely rare for it to go above 60. For the first part, is my conclusion correct? And as for the second, can anybody explain the results of my tests?

[jaslion]

Depending on the game you may take a performance hit as 6 cores for some games is already a close call. A lot of laptops have under performing cpu cooling solutions so when the game you are playing is actually using a lot of the cpu including some of the 6 extra threads it will generate more heat and it will downclock the cpu. Pretty much in your case you might as well have gone for the i5 model then instead of the i7 as you are gaining nothing from it now.

 

The sudden push for 4+core cpu's in laptops has led to a lot of designs for the 4 core models being reused to save money thus causing the higher core count cpu's to massively throttle. As most designs are made to be used for 3-4 years and we are now finally at the stage where a laptop case design refresh is starting to appear in the market.

[TrigrH]

19 minutes ago, dangee said:

I'd also like to add that I did these 2 tests separately. Also, I have literally not seen my GPU temperature exceed 65 degrees once ever, and it is extremely rare for it to go above 60. For the first part, is my conclusion correct? And as for the second, can anybody explain the results of my tests?

Hyperthreading increases processor performance by essentially increasing processor utilization as each physical core can more efficiently be loaded. Using "more" of your cpu will increase power and thermal output. If thermals get too hot the processor will lower is core clock speed to avoid overheating. This is what was happening when you had hyperthreading enabled.

 

The best solution would be to replace the thermal paste with liquid metal so you can keep the processors temps down while keeping hyperthreading on.

[dangee]

14 hours ago, TrigrH said:

The best solution would be to replace the thermal paste with liquid metal so you can keep the processors temps down while keeping hyperthreading on.

I have read online about applying LM, but I have seen lots of conflicting information. It doesn't seem that complicated but I see some information warning that using LM is conductive and corrosive and can damage your components, among other things. I can't seem to find a definitive answer since lots of forums will have people saying it's great and others saying it's a terrible idea. I definitely won't make a decision based on one person's word, so do you have any sources or thoughts I can use look into this? 

[Guest]

15 minutes ago, dangee said:

I have read online about applying LM, but I have seen lots of conflicting information. It doesn't seem that complicated but I see some information warning that using LM is conductive and corrosive and can damage your components, among other things. I can't seem to find a definitive answer since lots of forums will have people saying it's great and others saying it's a terrible idea. I definitely won't make a decision based on one person's word, so do you have any sources or thoughts I can use look into this? 

You don't have to use liquid metal, it's just that it conducts heat nearly as well as solder.

 

There's a risk. Follow directions, and you'll be fine. Find a good guide.

[TrigrH]

5 minutes ago, dangee said:

I have read online about applying LM, but I have seen lots of conflicting information. It doesn't seem that complicated but I see some information warning that using LM is conductive and corrosive and can damage your components, among other things. I can't seem to find a definitive answer since lots of forums will have people saying it's great and others saying it's a terrible idea. I definitely won't make a decision based on one person's word, so do you have any sources or thoughts I can use look into this? 

Liquid metal contains gallium, which means it cannot be used with aluminum.

Liquid metal is conductive so it is advised on laptops and devices with SMDs around the chip that a coating (like nail polish) is applied around the chip (not on it).

Liquid metal is safe to apply on nickel and copper, however when used on copper overtime it can "soak" into the surface and may need to be reapplied ONE more time in the future.