Sign in with
Sign up | Sign in
Your question

Can system benchmarking highlight at-risk components?

Tags:
  • Benchmark
  • CPUs
  • Systems
  • Components
  • SSD
  • GPUs
Last response: in Systems
Share
October 1, 2014 6:06:47 PM

Do components test out of a specified range when they are at-risk of failure? Or can they test well day one and fail day two?

Of course, there are exceptions to everything; however, the information I seek is general, not a one-off situation.

More about : system benchmarking highlight risk components

a b à CPUs
October 1, 2014 6:30:36 PM

Some of both. Benchmark testing can find things like poor cooling, or questionable memory that, during routine use, might not be discovered for a long time. But a power supply will work find today and when powered on tomorrow decide it is "a good day to die".

I recommend that memtest be run for several hours on a new build prior to installing the OS. This allows stability testing of the hardware without the hassle of an OS instal.
m
0
l
October 2, 2014 6:40:39 PM

Thanks for the response. I am most concerned with CPU, GPU, and SSD. I'm also more interested in component integrity. For instance, can I say something about the component with benchmarking scores, not necessarily the integrity of the system as a whole? Of course, assuming that we can isolate the component objectively (if testing a GPU, assume that there are no CPU issues and the benchmark applies strictly to the GPU).
m
0
l
a b à CPUs
October 2, 2014 6:56:06 PM

jmn319 said:
Thanks for the response. I am most concerned with CPU, GPU, and SSD. I'm also more interested in component integrity. For instance, can I say something about the component with benchmarking scores, not necessarily the integrity of the system as a whole? Of course, assuming that we can isolate the component objectively (if testing a GPU, assume that there are no CPU issues and the benchmark applies strictly to the GPU).


I don't know that scores will tell you as much as failure. Again, you might see some thermal problems on a GPU that benchmarking might show as low scores. But that isn't necessarily a component integrity issue as much as poor airflow in the case. I prefer to design it right, then verify that the system works rather than obsess about every last decimal place in benchmark scores.
m
0
l
!