Is it essential for Prime 95 Small FFT using AVX to assess the stability of a 9900k OC?
Is it essential for Prime 95 Small FFT using AVX to assess the stability of a 9900k OC?
I've seen discussions about whether this test imposes too much stress on the CPU and if it's appropriate for a 9900k. There are still some differing opinions on the matter.
Your measurements are quite accurate, but there are always adjustments and variations between CPUs and their performance.
Prime95 small FFT using AVX helps maintain CPU cooler thermal stability.
Using Asus Realbench 8hr or Cinebench R20 with infinite loops ensures voltage stability.
If both tests pass without problems, it means everything is working well.
Be mindful, since P95 with AVX on can be challenging for many coolers, it's best to turn off AVX during the initial few tests.
AVX is an enhanced vector instruction addition for x86 architecture. AVX performs 256-bit floating point operations. AVX2 extends this capability to 256-bit integers. AVX-512 further increases it to 512-bit on both floating point and integer data types.
Only a limited number of CPUs support AVX-512, with the 9900k being one example.
Running Prime95's small FFT is a realistic workload. AVX, AVX2, and AVX-512 are extremely demanding for CPU usage and cause significant temperature rises, often exceeding a 130% synthetic load.
Games utilize AVX in minor quantities to aid physics calculations, such as simulating rock chunks during explosions. I’m not aware of any games using AVX2 or AVX-512. These features are typically reserved for professional software development.
For a home PC, it’s largely unnecessary to enable these features during testing, as real-world gaming loads won’t achieve 100% utilization and small AVX usage will be quickly overshadowed.
Here’s a revised version of your message:
Thanks for the thorough reply. Yes, I’m struggling to pass this with my 4.9 all-core overclocked setup. I frequently encounter BSODs or freezing issues. I believe the problem isn’t related to thermal throttling since my temperatures are reaching their maximum around 80°C. I increased the voltage to 1.3V, which seems excessive for a 4.9 core, but the voltage droop during P95 drops it back to 1.145V. This is the lowest I’ve observed in these tests. I’d prefer not to increase the voltage further or risk destabilizing the system again.
Besides the testing I’ve already done, do you have any recommendations on additional checks I should perform to confirm stability? I want to be confident before proceeding.