Jump to content

DrMikeNZ

Member
 View Profile  See their activity

  • Posts

    1,559

  • Joined

  • Last visited

 Content Type 

Everything posted by DrMikeNZ

  1. DrMikeNZ
    What are the system specifications, what game(s) and random background application(s) have you observed this issue happen with? Does task manager still work when the windows are frozen?
  2. DrMikeNZ
    There will always be some sloshing, I have never seen an AIO that had completely displaced all of the air in the loop with the coolant. Does it feel/sound like it is more air or more water? If the air from the radiator isn't getting warm then the cooler may need to be replaced. Double check that the pump RPM readback in the BIOS is actually for the pump. It might be worth installing Corsair Link if you haven't already to double check in there. If the radiator is laying horizontally mounted at the top of the case exhausting upwards then a typical air pocket in the loop would spread out in the top of the radiator and have minimal overall effect. If air got into the pump, the pump will often be unable to displace it, giving the loop a shake with the radiator at the top can help displace an air bubble and prime the pump. This would likely only be a solution if the cooler had stopped working properly due to the air being moved into the pump when the cooler was removed for cleaning, and not likely to help if it was faulty beforehand. If the cooler is within its warranty period it would help to contact Corsair to request an RMA if you haven't already.
  3. DrMikeNZ
    Is it summer or winter where you are, what is the ambient temperature? Is the radiator an intake or exhaust on the case, and do the graphics card(s) exhaust their heat out the back of the case or into the case? I have had a large number of AIO coolers fail over the years. Even if the pump is running it is not a guarantee that the water is flowing properly, or that the cooler is able to efficiently conduct the heat. How quickly does the air exhausting from the radiator take to heat up after the stress test starts? If the air exhausting from the radiator is cold, then the cooler will either need to be remounted, or replaced. If the air exhausting from the radiator is taking a long time to get warm then the cooler might need to be replaced. The most common issue I have seen with AIO coolers is oxidation of the heat blocks creating an insulating layer between the CPU and the water. This can result in increased idle and peak thermals. Oxidation on the heat block and radiator can be dislodged by the moving water and accumulate at restrictions in the loop and eventually block the water flow. This can further reduce the efficiency of the cooler. A less common problem that I have seen in a couple of AIO coolers is water evaporation. While the AIO coolers are sealed loops, the water can slowly pass through the rubber tubing, this evaporation process is accelerated if CPU temperatures are kept above 60°C for extended periods of time. Generally this should not be an issue until about 3-5 years though.
  4. DrMikeNZ
    Yes. There isn't the best standardisation on how measurements are taken, or whether or not the measurements are even performed with calibrated test tools, and there is no accreditation authority for this. A range of dBA and CFM is typically used on PWM fans, where they are designed to have the fan speed controlled. In my experience the Corsair fans that I have used have all been around 10-20dB louder than the manufacturer claim (possibly they had tested at 10m rather than 1m).
  5. DrMikeNZ
    Under sources, click the little plus symbol and add what you want to capture (display capture, game capture etc).
  6. DrMikeNZ
    It really depends on your workload, in all of the applications I run on my R7 1700, RAM has made no difference. I get the same performance with the RAM at stock JEDEC 2133CL15, or overclocked to 3333CL14. However, if there is a lot of cross core communication with the workload, for example a physic engine computation in an un-optimised game, then the infinity fabric can cause bottlenecks which the RAM clock increase can alleviate. I would just stick with 3.8GHz. If your cooler can handle it and you are comfortable then you could increase you cpu voltage to push it a little further, although with Ryzen once you hit the wall you typically need a significant power increase for an insignificant performance boost.
  7. DrMikeNZ
    Unlikely. On the Wraith Spire my R7 1700 I was only able to maintain reasonable temperatures up to about 1.38V.
  8. DrMikeNZ
    You can try 1.425V if you have sufficient cooling.
  9. DrMikeNZ
    Have you tried increasing your core voltage?
  10. DrMikeNZ
    Set cinebench priority to run as realtime. Run it 3 times with each configuration and take the average.
  11. DrMikeNZ
    I use a lot of programs that heavily utilise AVX2 instructions, which Ryzen requires 2 cycles to complete and thus have relatively halved performance. Although, I am not aware of any specific programs that run on Intel and not on AMD, software would have to be pretty poorly coded to not work on AMD hardware. Is there an alternative software package for the task?
  12. DrMikeNZ
    The chance of there being an issue is very small, although it is best to always have a plan in case something goes wrong. Personally, if my water cooling loop leaked and killed everything in the system it would be pretty sad, although I have several spare computers to use in the downtime and have contingency budget set aside to replace it all. Consider replacing the thermal paste, and/or try undervolting it.
  13. DrMikeNZ
    That sounds pretty average for the R7 1700X.
  14. DrMikeNZ
    I personally use a 3200CL14 kit running at 3333CL14. In my workloads it made zero difference compared to 2133CL15, but in several games (not all) you will notice a small improvement, I personally wouldn't consider the benefit worth the cost. Depending on your luck with silicon lottery though, you may not even get the rated speed out of the RAM due to the IMC on the CPU or poor RAM overclocking still present on some motherboards.
  15. DrMikeNZ
    Are you putting the system under load when you are checking the clock speed? Make sure the windows power profile is set to high performance. Try increasing vcore to 1.3-1.35V. How many weeks does a system need to be running at under full load to call it stable? The only crashes I get on my Ryzen system are have been traced to driver failures related to the killer network card on the motherboard, and it is typically 3-4 weeks between crashes.
  16. DrMikeNZ
    When the system crashed, did it create a dump file? If it did, try checking it with a diagnostic tool such as bluescreenview to see if it can help identify the root cause. It is possible that a background process or driver has a bug which is causing the issue, finding what program it is and uninstalling it, or reinstalling the drivers would be a more long term solution than changing the page file size. Run memtest86 as @RadiatingLight said.
  17. DrMikeNZ
    8GB is plenty for >400 tabs in my workload. About 2/3rds of the tabs are scientific journal websites, and 1/3rd of the tabs are pdf documents of journal articles opened from previously mentioned websites.
  18. DrMikeNZ
    It doesn't really matter how many tabs are open, it depends on what is in the tabs. I typically would have 400+ tabs open in chrome at any one moment at work, don't have any issues on an i7 3770 with 8GB RAM.
  19. DrMikeNZ
    The Lenovo website has the $977.99 base model Y720 as having 6GB VRAM. It only has a HDD, 8GB RAM, and a 1080p panel. Tiny battery considering the components, pretty average overall.
  20. DrMikeNZ
    Use HWinfo to read what the clock speeds on each of the cores and temperature are when under load. Ram speed has next to no impact on cinebench performance. While is is a good idea for gaming to have fast RAM, if cinebench scores are low the issue is elsewhere.
  21. DrMikeNZ
    That sounds like the CPU wasn't aligned in the socket properly. Asus used a similar mounting tool to guide CPUs into sockets on some of the LGA2011v3 X99 boards, small defects in the mounting bracket often resulted in incorrect mounting pressure and made it more difficult to install the CPU. I would not have expected TR4 to have these issues, but it could be a similar problem.
  22. DrMikeNZ
    That sounds like it could be a PSU issue. Do you have a spare power supply to try?
  23. DrMikeNZ
    Fullscreen windowed is a mode that most games (not all) have and is different to windowed or fullscreen. It creates a borderless window on the screen. It looks the same as fullscreen, however the OS doesn't unload applications in the background. This can be particularly useful on multi-monitor configurations helping with smoothing out alt tabbing, and sometimes tearing.
  24. DrMikeNZ
    Are there any error LEDs visible on the motherboard when it crashes? Does it crash running CPU stress tests such as prime95 or cinebench? Was the OS freshly installed, or was the boot drive transferred from the old to new computer? Try reinstalling the GPU drivers fresh, uninstall the old drivers in safe mode with DDU. If that doesn't work, run memtest86 overnight.
  25. DrMikeNZ
    Do you really expect people to go out of their way to track down your previous topic to read it? A faulty power connection could result in unstable power delivery, crashes, and damaged components. High temps and clock speed at idle is usually due to the GPU not actually idle, either driver bug or background processes using the GPU.
×