Sign in with
Sign up | Sign in
Your question

Why is my 2600k's multiplier fluctuating at a seemingly random rate?

Last response: in CPUs
Share
April 1, 2012 7:32:22 PM

I am aware that the multiplier of Sandy Bridge processors is lowered depending on its inherent utilization; utilizing simple tasks such as browsing Chrome and playing video drops the voltage core to a relatively calm 0.9 volts and the multiplier to 16, however, it seems to determine when to increase such in a relatively odd manner; sometimes, it takes quite a bit of time to detect when such an increase is necessary; and in some instances; does not seem to increase to the full 3.8 gigahertz that it is rated for; despite such tasks being single-threaded and all bins available to such.

Is this a standard feature of the chipset; or am I somehow utilizing a processor incorrectly?
a c 126 à CPUs
April 1, 2012 8:29:24 PM

This is very nromal. Its Speedstep + EIST + C states. All features that Intel has used for quite a while. They enable the CPU to change the multiplier anf therefore the clock speed relitive to the load on the CPU. As well the voltage will also change with the clock speed to lover temperatures and of course lower power usage.

As for the 3.8GHz, you may never see it as it may come real fast and go real fast. I for one am overclocking so I am always at 4.5GHz. I go between 1.6GHz and 4.5GHz all the time.

But overall you are doing nothing wrong. Just keep going knowing your CPU is working perfectly fine.
m
0
l
a c 185 à CPUs
April 1, 2012 10:49:35 PM

^^Exactly.
m
0
l
Related resources
a b à CPUs
April 1, 2012 11:33:11 PM

Not all tasks, like internet browsing or playing Sudoku, needs the 4.5GHz : ))
m
0
l
a c 185 à CPUs
April 2, 2012 12:11:56 AM

nikorr said:
Not all tasks, like internet browsing or playing Sudoku, needs the 4.5GHz : ))

:ange: 
m
0
l
April 9, 2012 12:03:31 AM

nikorr said:
Not all tasks, like internet browsing or playing Sudoku, needs the 4.5GHz : ))


Does simulating metal deformation and rendering scenes in Krakatoa with 3.5 million particles require such?

This is a rhetorical question; of course; I'm simply questioning the fact that using 100 percent of my processor does not invoke the entire theoretical overclock on a single-threaded application.
m
0
l
a c 185 à CPUs
April 9, 2012 12:05:24 AM

You will see a huge increase in rendering etc with a much higher cpu frequency. Especiailly 3.5mil particles, that's a lot! :ouch:  Even me simulating 1mil with realflow starts lagging, that's why i overclocked...
m
0
l
!