Most users expect reliable and smooth performance from their PC. The boot process takes just seconds, programs open with little delay, and videos run smoothly. But sometimes your PC crashes without warning, or copying files takes minutes. In most cases, the cause of poor performance is hardware, not software. But it's not always clear where the bottleneck is right away. Is your memory module defective or is your CPU overheating? Is your hard drive or SSD working at its capacity limit? Or are your applications simply slow because they retrieve data from the Internet? If your PC is running slower than usual, you can find out the cause with a special tool that shows you the current performance of each component. Read on to find out more. Using the Task Manager tool is a good place to start Task Manager is a useful tool for monitoring performance. Microsoft has significantly revamped and improved the program in recent years. To open Task Manager, press Ctrl>Alt>Delete and select Task Manager (or you can type "Tasks" in the Windows search window). In the app, click Settings at the bottom left and enable the Always on top option. Then switch to the Performance tab and open the CPU, RAM, GPU, Ethernet, and/or WLAN tabs in turn. Keep the tabs open and work with your PC as usual, observing the information about your components. If they are constantly above 80%, something is wrong. Monitor your CPU with Core Temperature and Performance Monitoring If your computer keeps crashing or running noticeably slower, the cause is often an overheated CPU. The trigger could be a poorly installed or defective fan. The processor's protection circuitry will immediately reduce the clock frequency if the temperature rises too quickly. If this is not enough, it will deactivate the CPU to prevent damage and turn off the PC without warning. Therefore, you should use Core Temp to check your processor temperature. This freeware offers a constantly updated database with all old and current CPU models from AMD and Intel and displays their data after launch. It also retrieves the individual core temperatures and displays them in the tool interface and at the bottom of the taskbar. The program window displays the maximum allowable CPU operating temperature in Tj. The maximum and current temperatures are as follows: If they are constantly above 80 degrees, you may have a cooling error. Now you need to look inside your computer and check if the fan is spinning and if it is firmly attached to the processor. If your PC is simply running a task too slowly, your CPU may be to blame. In this case, observe the usage curve in Task Manager. Another useful tool is the Windows Performance Monitor. You can access it by typing "perfmon" in the search box. Once launched, open Performance > Monitoring Tools > Performance Monitor in the left column. You will see a constantly updating graph with Processor Time at the bottom. This value indicates how long it takes the CPU to execute a thread. A thread in this case is a specific program action such as entering text, saving, printing in a word processor, etc. The less processor time the CPU spends executing a thread, the better. If the performance monitor value is consistently above 80-90%, then your CPU is too slow and should be replaced. Test your RAM for defects and errors Repeated PC crashes may be caused by memory modules in the main memory, and not just the CPU. Defects due to manufacturing errors or overheating are not uncommon. Windows crashes when individual memory addresses become inaccessible. Memtest allows you to check the installed RAM for errors. However, using the software requires some prior consideration. Memtest can only check free RAM, because Windows does not allow access to memory that is already occupied elsewhere. Therefore, it is best to restart your computer and start Memtest without any other programs open. Windows dynamically moves subsystems around in memory, so the tool will end up checking all your RAM sooner or later. Memtest can check up to 2-3.5 GB of contiguous memory. Therefore, check the amount of currently free memory in Task Manager under Performance > Memory and divide the MByte value by 2,000. The result will be the number of Memtest instances that should be opened in parallel. Enter 2,000 MB as the storage volume for each of them. Run each instance until it reaches 100% coverage. According to the manufacturer, this will detect 95% of all memory errors. If Memtest encounters an error, the program will abort and print a message. In this case, it is best to replace the memory latch. Check the temperature and capabilities of your graphics card Graphics cards can cause crashes, especially when used heavily in PC games. In this case, the GPU can overheat and cause memory errors. You can read the temperature in Task Manager under Performance > GPU. However, a better tool than the onboard tools is GPU-Z, which collects and displays additional information about your hardware. As a rule of thumb, your GPU shouldn't exceed about 80 degrees, otherwise you risk faulty hardware. If GPU-Z detects higher values, make sure the card fans are working and tightly fastened. It depends on the graphics card itself whether it can be replaced. OCCT can be used to test memory on graphics cards. In the left interface, click Test > VRAM. Select the correct graphics card and test duration. The default setting is 30 minutes, which is sufficient even for well-equipped cards. Then click the start icon at the bottom. If the program diagnoses a memory error, you will have no choice but to install a new graphics card. If the game runs slowly and stutters at a low frame rate, it may simply be that your GPU is under strain. This can be tested with the 3DMark benchmark, although the free demo version is very limited. The truth about long wait times when copying If the Windows boot process suddenly takes a disproportionately long time, despite the presence of an SSD, you should check your data carrier. Crystaldiskmark has proven to work as a speed test. The tool runs several tests and measures the time it takes to write and read data. If the values are noticeably low, we recommend taking a closer look. Crystaldiskmark not only recognizes hard drives and SSDs, but can also measure USB drives and sticks. SSD-Z is better for a more detailed analysis. This tool shows the technical data and current temperature of the storage media, and also lists the S.M.A.R.T. values of the drive. SSD-Z also comes with its own benchmark, but it is less accurate than Crystaldiskmark.