Question Intel i7-8565U clock speed problems

May 14, 2019
Hi all,

I have recently purchased a lenovo E590 laptop, with an Intel i7-8565u, advertised as "up to 4.6GHz". I think I am understanding correctly that the U models for laptops are always balancing power consumption, temperature and performance. So, I am not expecting to see this running at 4.6GHz all the time.

However, I do expect it to get there, for periods of operation when I am running intense multi-threading software, which I happened to be trying to do at the moment. The software is in java, written by me, it is a puzzle solver, and it runs the cpu-intense solving on 8 threads simultaneously, taking between 7 and 12 seconds to complete. After noticing that windows reported 1.8GHz as the CPU speed, I started digging into the CPU speeds to see if this was the best it could do. So, I tried a number of tools to report the CPU speeds, both benchmarks and informational. This is what I get.

  • Purchased: Lenovo E590 Laptop, with Intel i7-8565u upto 4.6GHz
  • Windows System reports: Intel i7-8565u @1.8GHz 2.00GHz (I assume this to be the normal and turbo speeds?)
  • CPU-Z reports: Intel i7-8565U @1.8GHz. The Core Speed then reports a range of 1.8GHz to 3.8Ghz, averaging around 3.0 I'd say, while the system is idle.
  • Intel Processor Identification Utility: Intel i7-8565U @1.8GHz. Reported: 3.78GHz, Expected: 1.8GHz
  • UserBenchMark: Intel i7-8565U Base clock 2 GHz, turbo 2.25 GHz (avg)
  • Bios: 1.8GHz
Please note, before running the above, I made sure: I am plugged in. I have modified my "balanced mode" power plan so that Minimum and Maximum processer state when plugged in = 100%. I turned off the PCI Express power mgmt. I made sure in the bios - when plugged in, it is set to maximum performance. I disabled a number of services that use a bit of background CPU.

In order to see what's happening when I'm running my solver, I run CPU-Z, and show the four core speeds. It starts out by jumping quickly to between 3.5 to 4.0 then drops back to 1.8GHz when the main business of the solver starts. This seems odd to me. These threads are independent, and the only thing that will make them quicker is faster clock speed so I don't understand why it won't use the full speed of the chip for the short 10s burst.

Any thoughts experts out there? Does anyone know of a user benchmark I can run that will geneuinely max out the CPU and push it into turbo mode?

Thanks for your help, Jason