CPU performance issues on Dell workstations.
CPU performance issues on Dell workstations.
Great news. I wasn’t anticipating such a noticeable boost in CPU speed. The increase from 3.80 GHz to 3.60 GHz adds about 5.6%. This will likely bring a few more Cinebench scores, though the overall gain is modest. Individuals with strong benchmark results tend to excel at spotting and eliminating unnecessary background processes. The C0% metric displayed on the main screen of ThrottleStop offers a clear picture of background activity on your system. What percentage is shown when your computer is idle with ThrottleStop running? Your recent screenshot didn’t reflect the expected numbers. Compare it to my setup. The Windows High Performance or Ultimate Performance power settings can help minimize the time a CPU spends in C0 mode while handling background tasks. A swift processor is ideal. In the Task Manager, navigate to the CPU section and review task details to organize workloads by usage. Removing redundant processes can significantly improve speed. Weak benchmark outcomes often stem from unoptimized hardware. Your CPU should operate under less than 0.1% of the time in C0 mode when idle, allowing 99.9% of cycles for benchmarking. When you run the TS Bench 1 Thread test, what speed does ThrottleStop indicate? During this run, you should see a rate exceeding 3.80 GHz. Turn off any other monitoring tools and close the Task Manager during testing. If the CPU isn’t utilizing full turbo boost, restart your system, enter BIOS settings, and verify that core C states are active. Disabling these states is a common misconception. On a locked processor, disabling them also halts maximum turbo boost, which can slow performance. It restricts the CPU from leveraging the 40 multiplier under light load with only one or two cores engaged. You can check your C states in ThrottleStop to see what mode your CPU is using when idle. You don’t need to turn off SpeedStep if you want full performance from your Intel CPU. Many myths about these settings persist due to repetition. The power draw for an Intel E5-1650 v4 is around 84W, with a TDP of 140W. The TDP remains constant regardless of usage. https://ark.intel.com/content/www/us/en/...0-ghz.html If you perform a stress test using Prime95 with Small FFTs enabled, you should notice a substantial rise in power consumption and temperature. Prime95 places greater strain on the CPU compared to Cinebench tests. Are you pairing two memory modules of identical size and speed? Using dual-channel memory versus single-channel can greatly influence performance, especially when systems with identical CPUs show varying results.
I wasn't anticipating such a significant improvement in performance just from a 200mhz increase. Yes, it won't impact real-world tasks much. You won't easily spot a jump from 162 to 175 frames per second, so I'm unsure whether these adjustments made any real difference. To be honest, I opted for the 1650 V4 instead of the 1650 V3, but its speed was still comparable to the previous model, even lagging slightly behind. I'm not sure if it's due to cycle efficiency or something else, since turbo boost now works well across all cores at 3.8Ghz. I disabled several software programs and saw a noticeable boost compared to earlier screenshots. It's not as impressive as yours, but with Chrome open and typing this thread, it seems better. However, it tends to spike frequently (this one is budget). I checked the BIOS, confirmed C states were on, and tried Cinebench R15 – it reported 4Ghz on one core. I ran Prime95 with small FFTs; power use stayed at 84W, but heat increased by about 10 degrees, with the hottest core reaching 73°C. I'm using quad-channel RAM (16GB) in all four slots, and four black slots remain available for future upgrades. It doesn't seem quad-channel is the cause. One thing I'm pleased about is that turbo boost now activates at 3.8Ghz across all cores by default, without needing to enable ThrottleStop. That's a positive sign. I'm glad the processor is ramping up to 3.8Ghz automatically. Thank you for your detailed explanation – it really helped me understand. I'm also curious about the expected power consumption. If everything checks out, I think it's within normal limits. Previously, benchmarks suggested a 8-10% slowdown compared to the i7 8700 (which scored 1400), but my results show a 17% drop to 1170. The screenshot with screen capture was the worst, but without it I'd expect better. I checked Geekbench6 too and scored 5736, about 11.5% slower than the i7 8700's 6412. It seems the i7 might handle certain tasks better due to its integrated GPU. Overall, it's not a huge concern now, but I appreciate your patience and the thorough guidance. Best regards.
He works with a lightweight version of Windows 7, installing only essential benchmark tools. His GPU driver is likely missing. Users on HWBot often aim to maximize performance through clever methods. Running Cinebench with real-time priority is a common strategy. Turning off certain CPU states can slightly boost scores. He’s using dual-channel memory. His memory stats are competitive, though his CPU usage from background apps is a bigger factor. When idle with Chrome open, the test shows higher C0% than usual. Those unused CPU cycles significantly affect results. If background processes spike usage, it limits his score potential. This machine is close to one of the fastest E5-1650 v4 configurations available on HWBot.