Walter_Kai

After dropping it down to 3300MHz, seems like the computer is now able to perform sustained load. Temperature now sits below 85C and can run at clock of around 2800MHz. According to the program, CPU has never throttled. Not sure why it recorded minimum at 1995MHz or it's not pinned at 3300MHz, but it ain't throttling according to the program. This is with the service panel back on, hopefully I can use it without it freezing now.

The post I found about the 5587 is either you, or someone sharing a very close mindset with you haha. To be, that is a disgusting attitude, no only is your product faulty, you are blaming your customers for the issues you created as a reputable entity. Any computer regardless of "gaming" or not, if it can't sustain a load that it was designed for by hardware configuration, is an engineering failure. Throttling is an acceptable concept for me, as long as it's done properly. If it drops to 0.8MHz, whether it is just bad configuration, or cooling design failure, then it becomes unacceptable. Laptops are crammed units and it makes sense there will be limitation, just has to be designed properly. Say an iPhone, you don't see it freezing cause of heat management, unless you are using it in full brightness under a 40C sun. Thanks for the recommendation on Throttlestop, but to me, if it only tells me the issue but doesn't fix it, it serves no purpose and there is no point in putting effort into it. Perhaps it's time to move on from Dell in my next purchase.

Thanks for the info, it makes more sense now why the seemingly "powerful" i7 performance just drops performance so drastically. It just doesn't make sense they are pinning it to 4.1GHz when it can't sustain it. Surely their engineering department is aware of the 6core, hyperthreading to 12 will result in thermal requirement. Have they not test it when they designed it? I have seen extreme answers on the Dell official technical support, ranging from blaming the laptop is only designed for excel and word since it's not an Alienware, to actually working on it and released an updated Dell Power Management software with much more aggressive fan curves. I know it seems lazy and I hope not to upset the hardcore gaming enthusiast, but I can't be bothered playing with all those undervolting / throttle stop management / testing etc. I rather limit it's power output and get it "fixed". Best I can do is disassemble the laptop to clean the fans, I paid a computer shop to do thermal re-paste end of last year. From what they sell, I think it's artic silver, or thermal grizzly, one of the better known brands. I can only assume they did it with those. This 5fps drop has been happening way before I re-pasted it and ever since I repasted it, so unlikely the thermal paste's problem. Will try tomorrow with Halo Infinite, hopefully it doesn't do this 5fps stuff again with the service panel back on.

Not sure if it's just how Dell did it, when it drops to 5fps, the CPU frequency drops to 0.89GHz, then go back up to 2.8 or 3.8 when it "Cools" down. As far as I can tell, that's the main thing that's killing my FPS. GPU is always like 78C and as far as I can tell, the laptop's 180W powerbrick is doing it's job? It's a laptop, cooling is not great but in my opinion Dell f*cked the 5587 cooling design, under engineered it and now there's no way to fix it, but not to push it so hard. Ran HWinFO64 with BeamNG Drive, after dropping the CPU turbo frequency, max temp is 93C on the i7 cores, as oppose to 99C before, and this is with the service panel back on. 3D mark is a good idea, but not that hard core, just wondering how much generally I would notice on a day to day use, not so much scientifically putting a % in performance drop. I did in previously have undervolting going on, but didn't really notice the difference and once Dell updates come in, all the settings gets erased, so it's not pragmatic for me. I use this laptop docced with Type-C to external monitor, keyboard and mouse, and take it with me on the go when I need to. 70% personnel use, 10% gaming, 20% excel/word work related stuff. Having to remove the service panel or do a bunch of "out of factory" tuning is not ideal for me.

Dell G5 5587 struggles with thermal. Unless you drop the service panel and put a fan into it, it will thermal throttle and drops to 5fps before it recovers or crashes. New thermal paste did nothing. Ended up dropping the maximum processor state to 98% and maximum frequency to 3.3GHz, as opposed to the designed 4.1GHz. There is no point pinning it to 4.1 when it can't sustain it. It's got an i7-8750H, which is decent. Just wondering how much performance am I sacrificing by doing this?

Just stressed the computer again and I see the avg temp is about 78 to 80C across the 6 cores and maximum is 94 to 98 across 6 cores. So while they are not great obviously, but I guess for a laptop, it is what it is

Sorry for the late updates, been trying out a few things. The GPU according to Throttle stop never thermal throttles, it's only the CPU that's thermal throttling. After disabling full screen optimisation, overriding the DPI stuff and shutting off Nvidia In-game overlay, Modern Warfare (2019) never experiences frame drops anymore. It's locked at 60FPS despite the CPU thermal throttling according to throttle stop. Yeah it's maximum at 99, not sure if it's actually forever at 99, I assume not, cause otherwise the computer should have exploded and I see the avg at like 77 or 80+. So I guess maybe Dell messed with the fans or something and say do not kick in 100% unless it reaches 99C or something? The computer had its heatsink, fans, thermal paste changed out in late 2018 due to a faulty fan that came with the laptop. I saw the Dell technician replace everything and the thermal paste for the CPU and GPU are re-applied onto the heat sink, so I don't think it's an application issue. Normally the laptop idals at 40 and it's when i really push it, then the CPU will heat up. I know I know, a gaming laptop is expensive and the heat management is a joke, but I need to stay very mobile and I still want to game when I am free, so.......Not much alternatives really I chose Dell cause of the game experience I had previously with the DELL 14R (7420), which lasted 7 years before I changed it cause of the GPU not keeping up. The 14R still runs like a dream and it can even play GTA V. The attached pic is the OLD heatsink module, as you can see the thermal paste is pre-applied in a perfect square, so I don't think you can mess up the paste application, cause well I assume a machine must have done it? I saw the technician wiping off stuff off the dye of the CPU and the GPU, so I assume it's paste and not pads? IMG_0973.HEIC

Um...I guess it depends on what Dell has decided to use originally then

I own a Dell G5 (5587) and the thermals are as you expect poor. (77C for GPU and 99 for CPU MAX recorded). I know it's a fundamental issue with cramming powerful hardware in a "small" chassis and there isn't too much to do about it. I have: 1. Undervolted the CPU 2. Set thermal management to ultra performance 3. Use Balanced power mode and set CPU max performance at 99%, in hope the Turbo boost won't pin it down at 4GHz) 4. Raised the laptop so it sucks more air in 5. Cleaned the laptop fan and heatsink fins (Fair bit of dust but I guess that's what happens to laptops) My questions are as follows: 1. Is there such thing as "service life-time" for laptop thermal paste? Do you have to replace them after a certain period of use? say 5 years? 2. Does applying aftermarket (e.g. thermal-grizzly or IC-Diamond) thermal paste actually give you as much as 10C reduction in thermals? One of my fans died during warranty and the dell technician brought me a new heatsink module with fans on and the process of swapping it (which I guess would be identical when re-pasting) is horrifying. Seeing the amount of force required to crack the screws and the number of those tiny easily fall off thermal pads really puts me off in re-pasting. I guess what I am trying to say is, I don't want to do repasting if the benefit is a 3-5C degrees reduction. My laptop in total cost 2K with some upgrades and I don't want to send it to the dumpster for negligible improvements.

I have already undervolted the CPU and prefer not to mess with the GPU cause of the Max-Q stuff. In terms of temp, max temp for GPU was 77 ish and the CPU is 99. If anything I think I should fix the CPU first RIP

Yeah...thermals aren't exactly excellent in the G5

lol, of course it's one of the two, but which one though. Mine is a Dell G5 (5587), so it does get toasty under load. I think I see it averaging at about 72 and peaking at 75 for the GPU when playing BeamNG Drive and playing YouTube simultaneously.

It's doing 52atm, but I am not after the current temp as I use another program to view it. I just want to know the thermal throttle temp for my GPU, not sure which tab to read

What's the current temp for? It is definitely not currently running 78C, it's not doing anything atm

I want to find out what the Maximum Temp of which my GPU will run without thermal throttling so I got GPU-Z. Under Temperature Limit, there are 1. Current 2. Minimum 3. Default 4. Maximum 5. Adjustment Range Can someone tell me which one should I be looking at? 78 and 96 are quite a large difference.