Experiencing major performance issues with my new Dell G7 7700 - i7-10750H / RTX 2070 Max-Q. Almost all games, even ones almost a decade old, trigger the GPU to throttle down to 300MHz and bring frame rates crashing to 10-15fps. The issue is erratic but frequent - sometimes it takes close to an hour to occur and other times it happens within a few minutes.
I reached out to Dell support via WhatsApp. After spending hours troubleshooting with them, collecting GPU logs, etc. and waiting sometimes 24+ hours for each response, they tried to close my ticket, saying "this is normal behavior in response to heat issues". Anyone who would have run into this issue during beta testing of this laptop before Dell started selling it would NEVER have considered this "normal" or in any way acceptable. I contacted Dell "premium" support via phone and spent over an hour troubleshooting with one of their reps. He stated the issue was definitely not heat related and was definitely not normal. After we were unsuccessful at getting the issue resolved on the phone, I collected some additional data for him and sent it his way via email. That was 7 days ago and since then, I haven't received any response at all, from him or from anyone else at tech support.
This is an issue that is apparently affecting a lot of people. See my post https://www.reddit.com/r/Dell/comments/ilm377/g7_7700_frame_rate_dropsgpu_throttling/ regarding this, and the Dell forum post at https://www.dell.com/community/Inspiron/G7-7500-RTX-2070-Drops-Framerate/m-p/7670471 - a Reddit search for "G7 7700" will also yield a huge number of reports.
It's not a thermal issue as the throttling has happened for me in the mid-60C temp range and has occurred all the way up to 78C. At the time the issue occurs, CPU temps are around 75C (disable Intel Turbo Boost in the BIOS to get these more sane temperatures). Some users have speculated that it's a voltage issue, but mine seems to trigger generally in the 0.96-0.98V range and there have been long periods of time where the GPU has sat at higher voltages than that without throttling. If this is some sort of hardware limitation where the GPU is pulling back to protect itself, I'd expect it would happen consistently upon hitting the same temperature or the same voltage every time it occurs.
Some people have had success limiting the GPU clock speed with MSI Afterburner and turning off Intel Turbo Boost. Some people have had success with uninstalling Alienware Command Center and Overclock Tools (though then you don't get macro keys, performance mode, or control over keyboard/ambient lighting). At least one user was able to get his replaced and that resolved his issue (he had a defective motherboard, GPU, and wifi adapter). Regardless of the fix, it's clearly affecting a lot of people with this generation of laptop as well as the previous generation.
<Non-public info removed>
Since Dell apparently doesn't read or reply directly to these posts except on a rare basis, I wanted to share that Dell released a BIOS update yesterday (9/14/20) that is specifically targeted at this issue. I tested the update yesterday for a couple of hours; throttling was still occurring, but clock speeds went down to a more sane level and FPS in games dropped to a more playable ~30fps.
It requires more testing due to the erratic nature of the issue, but the update looks promising. Hopefully this will be a permanent fix for this problem.
Hi, my experience - part of previous thread: Turning of the intel turbo boost was a first thing that I have personally tried and reinstalled almost everything with exception of windows. But turning off the turbo boost makes the game in this case ms flight simulator 2020 unusable. Which defeats the purpose of having a GAMING laptop. As I have stated above loosing few percent on frame rate form 50+ in country side or 30in new York is not real issue for flight simulator, but dropping to 10-12 in newYork by disabling the turbo boost or dropping to 10 with the GPU at 300Mhz is an issue for a expensive gaming hardware. The lack of functional throttling on the GPU is the issue now. after the 1.3. bios patch the processor works well and throttles nicely. the laptop is great, but unfortunately the throttling on GPU just does not work and that is sad.
There is a way to under-volt the CPU (there is solution on the net ), unfortunately the same in not certain for the GPU. I could investigate, but I have a workaround so I do not have the incentive to do so now. The work around has to do with stupidity in clock - voltage setting for the GPU. there is a curve that sets the voltage based on gpu frequency and the curve maxes out at about 1.2v which the laptop is completely incapable of cooling. so in MS FS 2020 it will go to safe mode in less then a minute if the laptop is already warm.
Now in my case and this has to be tried for each board and your ambient temperature. (so expect few tries) The GPU can take about .8 volts without overheating and going to the dreaded 300mhx (sometimes to 900mhx) in my case. And it happens to be at 1600MHz, so effective speed 1575MHz as opposed to the 1890MHz at 1.2 V and it has no effect on game speed as the utilization of the GPU is not 100% at the 1890 by far.
So the utility that works is MSI afterburner, you have to manually set the curve to flatline from your desired value. Start form left and set every other point to the sane value. In my case 1600 and it has to be set exactly in curve editor. If you accidentally set 1601 at the 1.2v range you get 1.2v and overheat. The frequency offsetting also does not wort as it also puts max volts and lover frequency to the GPU and overheats.
Bottom line is that this is still not a working laptop, but you can make it usable (why is there no temperature limit other than the safety is beyond me. The solution above is greatly dependent on ambient temperature while limiting the max temperature on the GPU would not be)
<Profanity removed> this is disheartening. I bought my G7 7700 in August and was having random freezing in Photoshop all the time and weird issues trying to start the system from a cold boot. I would usually have to press the power button 2 to 3 times for it to respond. No issues in gaming with Quake Live, Overwatch and the latest DooM.
Last week Thursday it froze on photoshop once again (usually once a day) but this time wasn't able to boot at all after forcing it to shut down. The lights would power up, the fans would get loud, you could see the screen backlight power up but no bios/logo screen at all. Hooked it up to a monitor and nothing. Called support and after an hour and half they said I need a motherboard replacement. Since then I've had 2 cancelled appointments from Dell, I guess the motherboard replacement part is scarce.. probably cause this laptop has a faulty design is my guess. Now they are saying I can expect a new appointment by the 29th. I'm seriously pissed. Of all the laptops I chose I picked a lemon that Dell is choosing to pretend is not s problem. So to recap, Bought August 6th 2020, froze just about every day in photoshop, failed to start Oct 15th 2020 and won't get looked at til Oct 29th.. and that's just a maybe if the part becomes available.
Is there anything we can do? After reading your post I'm worried a motherboard replacement isn't going to do much and that the system is going to fail again. I've been doing IT now for 20+ years. I've been buying Dell desktops, servers and Sonicwall devices for the company I work for those last 20 years and I feel like an idiot for having faith in a company like Dell. I had a small window to return the laptop but I was sure that bios updates would fix the issues and that the bad reviews were all just user error. Boy was I wrong.
What can be done? Is it too late and we are now stuck in warranty limbo for 1 year until it expires? I feel like Dell needs to make this right.