Heavy frame failure in gaming laptops
Heavy frame failure in gaming laptops
When attempting to play games, my performance spikes briefly with high FPS for a few minutes, followed by a sharp drop in frame rate (120+ to 25-35) across most titles. I updated drivers but saw no improvement. A laptop cooler helped slightly, but the issue persists. I’m eager to regain normal frame rates. System specs: CPU – Intel Core i7-9750H, GPU – Nvidia RTX 2060 (Mobile), SSD – Samsung MZVLB256HBHQ-000L2 256GB (19.5 GB free), HDD – WD WD10SPZX-00HKTT0 1TB (24.1 GB free), RAM – Samsung M471A1K43DB1-CTD 89C2 8ATF1G64AZ-2G1B1 16GB, MBD – Lenovo 81Q6, OS – Windows 11 Pro 22H2. Benchmark link: https://www.userbenchmark.com/UserRun/51506192
That sounds like a classic case of thermal throttling. I'd open up some monitoring software like AIDA to monitor temperatures when you notice this happening. https://www.howtogeek.com/739732/what-is...hrottling/ https://forums.aida64.com/topic/1478-ben...hrottling/
Lenovo employs unusual throttling techniques on certain laptops. Try Run ThrottleStop, enable the Log File option, and run a 15-minute test. The log will track CPU performance and reasons for throttling. Paste this information onto www.pastebin.com and share the link. Include screenshots of ThrottleStop to understand your system setup and get advice. Most throttling issues stem from low power limits. Display the TPL and FIVR windows.
You recently tried CoD: Vanguard and experienced significant performance problems. Your frame rate dropped from over 120 fps to around 40 fps within the first five minutes, then fell below 15 fps. You monitored GPU temperature and CPU usage closely and found your GPU consistently stayed under 70°C, possibly aided by a cooler. Your CPU utilization rose to about 20% once the frame rate dropped to 15. The logs you received are saved as logs03272022.txt. For checking TPL and FIVR on Windows, you may need to consult official documentation or community resources since these terms aren't standard.
When I first checked the log, the top entries in the far right column were TEMP, TEMP, TEMP. Your processor is frequently reducing its speed due to heat. Maintaining steady or optimal performance becomes challenging when the CPU stays too hot. It’s time to open your laptop and clean it up. Think about swapping out the thermal paste if needed. Choose a paste that has been effective for laptops. Some common thermal pastes designed for desktop CPUs might only last a week or two if applied directly to a laptop CPU. Here are key points to keep in mind. Once thermal throttling persists, it often shifts to power limit throttling. The 9750H model has a 45W TDP and adjusts based on that cap. Despite this, the CPU doesn’t always operate at its worst levels. Look into the Nvidia GPU settings within ThrottleStop options to see temperature and performance details in your log. If frame rates are extremely low, GPU throttling is likely the main issue. Some makers release updates that reduce GPU throttling thresholds, which can lead to significant slowdowns. Many users then opt for a new gaming laptop. Companies aren’t naive. Consider using GPU-Z to check the current GPU throttling temperature. The GPU may run at half or a quarter of its full speed when this temperature is reached. A fresh ThrottleStop log that includes GPU data should clarify what’s happening. It’s common to suggest BIOS updates, but not everyone understands the risks involved. The latest BIOS might have been installed automatically after switching to Windows 11.
Thanks! I just purchased some thermal paste and was looking at GPU-Z. I’ll let you know once I have it, and I’m here to assist if anything comes up.
Just finished applying the thermal paste again. I’m worried something went wrong. Could have misplaced it during installation? Here’s a brief CPU status log – it didn’t look well. The GPU stayed under 63°C, which is odd given it reached over 90°C recently. I don’t see temperatures everywhere in the logs. During gameplay, my CPU spiked to 60°C and fluctuated between 45-60°C quickly. I have screenshots of my settings; they might help. I haven’t changed anything yet, just hoping for improvement. Also, GPU-Z showed throttling temps, and NZXT CAM reported different readings than the other tools. Should I rely on NZXT CAM or trust GPU-Z/ThrottleStop more? Thanks for your help – I’m really confused. logs20220329.txt
You are employing a specific thermal compound. Certain well-known options perform poorly when applied directly to laptop internals. Reviewing the ThrottleStop log reveals peak CPU core temperatures exceeding 90°C in the TEMP section. The NZXT chart you shared doesn’t reflect such high readings. Did NZXT operate concurrently with your ThrottleStop data collection? If both applications ran simultaneously, avoid using NZXT.
TIME STAMP MULTIPLE COLD CORE MOD BAT_MW TEMP NVIDIA GPU VID POWER 2022-03-29 09:55:58 36.03 84.8 100.0 0 87 960 53 1.2463 74.1 2022-03-29 09:55:59 36.36 67.4 100.0 0 90 960 53 1.2570 74.7 2022-03-29 09:56:00 36.72 60.1 100.0 0 79 1155 54 1.2952 63.9 2022-03-29 09:56:01 38.63 51.0 100.0 0 87 1155 54 1.1947 65.8 2022-03-29 09:56:02 37.19 40.8 100.0 0 87 1845 56 1.2588 70.0 EDP 2022-03-29 09:56:03 39.40 39.6 100.0 0 88 1845 56 1.2839 57.3 2022-03-29 09:56:04 38.66 28.8 100.0 0 84 1830 56 1.2930 65.6 2022-03-29 09:56:05 39.05 44.5 100.0 0 86 1830 56 1.2695 62.2 2022-03-29 09:56:06 39.40 20.4 100.0 0 86 1215 55 1.3041 57.6 2022-03-29 09:56:07 39.81 23.0 100.0 0 87 1215 55 1.2766 49.7 2022-03-29 09:56:08 39.11 40.0 100.0 0 88 1125 55 1.2839 57.9 2022-03-29 09:56:09 37.70 29.0 100.0 0 88 1125 55 1.2439 68.3 PL2 2022-03-29 09:56:10 39.09 34.9 100.0 0 92 960 55 1.2737 68.1 2022-03-29 09:56:11 39.15 44.8 100.0 0 82 960 55 1.2917 65.7 2022-03-29 09:56:12 39.80 33.6 100.0 0 83 960 55 1.2477 53.5 2022-03-29 09:56:13 39.61 33.9 100.0 0 83 960 55 1.2988 57.9 2022-03-29 09:56:14 39.60 35.7 100.0 0 81 960 55 1.2610 54.1 2022-03-29 09:56:15 39.92 18.9 100.0 0 81 960 55 1.3202 47.9
The detailed log confirms stable CPU performance throughout. The issue lies with the Nvidia graphics card. At the beginning of this segment it operates at 960MHz, climbs to 1845MHz, then drops back down. Such fluctuations aren’t ideal for consistent gaming. The GPU temperature remains acceptable, indicating throttling isn’t caused by heat. It’s likely a power limit change from a recent update. GPU-Z displays the thermal threshold at the default 87°C, which is within normal range. The logged temperatures hover around 55°C, ruling out any genuine thermal problems.
Most users with a 9750H model opt for Windows 10 rather than Windows 11, where the GPU often throttles aggressively. If you recall installing updates around that time, it’s possible they altered power constraints. You can access GPU-Z to monitor real-time performance in-game and verify if power limits are being adjusted. Lowering voltage via BIOS or FIVR settings may help, but persistent throttling usually signals a firmware or driver issue. Restoring older BIOS versions might be necessary.
Consider checking for updates and adjusting voltage offsets to see if temperatures stabilize. If performance remains affected, further diagnostics with manufacturer tools are recommended.
My machine runs Windows Insider, focusing on the Dev channel. I’m worried these updates might have damaged it. Over the past six months, I haven’t applied any updates except from November to February. I wonder if the newest patches would have caused more harm than good. Would a clean install of the stable Windows 11 help fix my performance problems? If it does, I’m planning to revert to a stable build. I haven’t received the latest release yet, so I’ll test if it stabilizes my GPU. If not, I’ll try to exit Insider as soon as possible. After resolving a major bug that blocked updates in late October, I feel it’s time to stick with Windows 11 and revert to older builds. The NZXT temps shown while the PC was off seemed lower than those reported in ThrottleStop—my confusion remains about which readings are accurate.
I understand ThrottleStop accurately captures the highest core temperatures from Intel processors. Rely on what it displays. If NZXT shows readings that fall 5°C to 10°C below ThrottleStop’s numbers, NZXT isn’t reliable. Numerous tracking tools overlook peak core temps correctly. ThrottleStop stands out in this area. The Windows Defender issue appears to affect Windows 11. I still prefer Windows 10 and haven’t encountered any problems, so I’m not planning a switch. Consider a clean install of a stable Windows 11 version and see if the issues disappear. If they persist, try reverting to Windows 10. Make sure the GPU bios are up to date. Switching between different OS versions probably won’t help. The situation is already resolved. Edit – I added NZXT for a quick check. During a CPU-Z benchmark, ThrottleStop logged a peak of 91°C while the CPU ran at full capacity. NZXT recorded 29°C. Once the test ended, NZXT finally updated to 36°C. It completely overlooked the actual peak. Avoid using NZXT for CPU temperature tracking—it’s the least accurate I’ve tested. With my processor reaching over 260W under heavy load, it’s definitely warmer than 29°C.