Increase every core's frequency to its highest setting
Increase every core's frequency to its highest setting
I aimed to enhance your laptop's performance, similar to the second model shown. The processor in question is an i5 1035G1, which can reach up to 3.6GHz as advertised. However, it seems locked and you can't get it to run at its full boost speed even when under heavy load. In my second laptop, simply navigating to the settings and enabling all cores worked, but this didn’t happen here. (The first two images) Do you have any suggestions on how to proceed? (It’s causing noticeable lag even for basic tasks.)
It doesn’t appear to function properly even when fully loaded. You might want to test another approach or check for updates. Also, consider whether the issue is specific to Windows or if it persists across systems.
I've turned it off all the time and never encountered any problems. Both laptops I use have it disabled—it's not the processor overheating, more likely the GPU or something similar. Occasionally it might also be due to a faulty sensor or a Dell laptop that's malfunctioning and triggers BD ProChat without reason. At least what I've read online supports that. Even at low temperatures, you'd probably need to push them past 130°C to notice any real issues, and they’d likely crash before reaching 120°C. I really want to try high-temperature testing, but the internal sensors just fail under heat, which is frustrating.
If this is the situation, PROCHOT 97°C would appear red but that’s just theory—it needs real testing to confirm and solid data to support it. Statistics alone don’t explain everything; 20 years or 50 years are needed for clarity. I’ve already worked with 105°C on 1156 Westmere temps, and I’d be much more concerned about running at 1.5v versus 1.2v at 105c (especially with 1156 Westmere CPUs), since degradation comes from both voltage and heat. Death is certain at 200°C regardless of voltage, and a CPU will fail at 2.2v+, no matter the temperature. At 105°C, standard voltage is okay (~1.2v), but 1.5v is definitely risky. It’s not just about understanding—it’s basic knowledge.
For another case with high TJmax over 1v, look at the RX 6000 cards that have 110°C max temps. Disabling thermal limits would let me see actual temperatures instead of static readings. Same applies to not being able to detect below 0°C—those issues matter too.
Laptop model available at the link provided. It appears the BIOS might be limited by power settings, which is common for business devices. Updating BIOS isn’t straightforward on this model, and switching to Linux would be a good alternative for your programming needs. The Windows update changed processor counts unexpectedly, so checking BIOS options or consulting the manufacturer’s support could help.
It makes sense. I’m used to running at a steady speed, so if it works without throttling, 60 versus 105°C won’t affect performance much. They’re similar, but this needs real testing on the actual chip. I still need funds for more unusual projects, like modifying a card to bypass temperature limits and see its breaking point. I’m curious about how voltage impacts things—1V at 115°C probably won’t cause much damage, but 1.5V might wear it out quickly. Degradation is the right term; you can’t just force a CPU to stop working permanently. It’s more about hitting it with enough stress until it fails, like running it at full speed for a short time. Voltage and heat both matter, especially in components like DDR4 DRAM or GPUs. I’ve pushed my Samsung module up to 2.6V on an X58A and it still worked. For higher speeds, I’m thinking about phase systems that use the Joule-Thomson effect at low temps—maybe something like a cascade phase or a liquid nitrogen cooler. That would be risky, but if it keeps running smoothly, it could work. I’m also considering how heat affects VRMs and RAM, which might cause other problems even if temps look okay. It’s just a temp sensor right now. Still, experimenting is key—don’t let anyone follow outdated numbers without proof. I’d rather test with a fake chip first before using my main board. If you’re looking for safe limits, aim for around 1.5V or lower at high temps, and don’t forget to check the cooling setup. For now, it’s all about balancing speed, voltage, and heat management.
That is not true. Someone that does not understand the difference between PROCHOT and BD PROCHOT should not be giving advice on this subject. A processor can still thermal throttle to protect itself against any damage whether BD PROCHOT is enabled or not. More misinformation. Do some testing. Intel CPUs continue to run at full speed while using maximum turbo boost. They do not throttle 1 MHz before reaching the thermal throttling temperature. Any throttling before an Intel CPU reaches the thermal throttling temperature is not thermal throttling. @Shailesh Vats The Utilization data that the Task Manager graphs does not accurately track the CPU speed or what your CPU is doing. Do not bother posting Task Manager screenshots. Utilization is not the same as usage. It is also physically impossible for a 1035G1 to run at 0.19 GHz. Your first ThrottleStop screenshot shows that the reason for throttling is BD PROCHOT. This is a common problem with many laptops. Your CPU is idle and only running at 48°C. This is obviously not a temperature related problem. The only way to fix a BD PROCHOT throttling problem is to clear the BD PROCHOT box on the main screen of ThrottleStop. BD PROCHOT throttling is almost always caused by a sensor that has gone bad. When this type of throttling first starts to happen it might be an intermittent problem. A broken sensor is never going to fix itself. Always run ThrottleStop and always leave the BD PROCHOT box clear. Check the MMIO Lock box which is near the top right of the TPL window. This will disable the secondary MMIO turbo power limits. This one simple tweak was designed to significantly improve the performance of many laptops with 10th Gen G series CPUs. The FIVR window will show the default maximum turbo ratios. Most Intel non K series CPUs slow down as more cores become active. The 1035G1 can use the 36 multiplier when 1 or 2 cores are active but this automatically decreases to the 33 multiplier when all cores are active. These CPUs cannot use the 36 multiplier when all cores are active. There is no way to get beyond this Intel limitation. Disable Windows core isolation memory integrity. I will make some more suggestions about maximizing performance after you post screenshots of the FIVR and TPL windows.
For testing purposes it’s around 40$ for a used POLARIS today, so ¯\_ (ツ) _/¯ I’m talking about instant death—just hit the temperature or adjust the voltage and the CPU shuts down right away. That’s what happened with the i3 540 (rip 2$) I showed earlier, but the 1156 model seems a bit more sensitive than the 1366 version. My approach is to crank the frequency up to its maximum, then gradually reduce the timing signals while increasing the primary voltage. This lets me see how it reacts before reaching any limits.
If things go beyond what I can handle, I’ll start by loosening the primary connections and boosting the voltage a bit higher. That’s the first step since the other parts rely on that. After that, I’ll push the frequency further, which usually means relaxing the primaries and raising the voltage again.
From my experience, I don’t just stop there. I tighten the subs next, and I did this for a 2000C9 DDR3 at 1156 at 1.8V because I wanted to sell it as a bundle. It ran smoothly at 1.5V at 11:00 but I kept it stable at 1.8V since that was the highest I could get reliable.
So, I’ll keep increasing frequency and voltage until I hit either the IC limit or the MOSI/MBO limits. In that case, I reach the IC limit because even 2.4V wouldn’t stabilize at 3000, so I settle for 2.1V at 2800. Then I focus on fine-tuning the subs and finally the primaries.
I haven’t tested RAM yet, let alone optimized it properly. Just did a few trials without a solid setup, but I’m planning to expand that later. It’s more about having someone else run extended tests, and honestly, with all these tweaks it’s not just practical—it’s more of a challenge. The thrill is in the experimentation, like adjusting settings for a custom loop or a cascade/jet phase cooler.
And yes, I’m aiming for lower temps—everything will freeze, so I write it down at midnight and now it’s late 2:38am. My words are filtered, and I don’t sleep until I’m done.