Super User is a question and answer site for computer enthusiasts and power users. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I noticed that if I lower my CPU clock speed via P-states/SpeedStep, my RAM slows down. However, through CPU-Z, my HT link, which is what connects my CPU to the memory, is still running at the same clock speed. What causes the RAM speed to slow down?

Here's what I did:

  • I used the AMD catalyst utility to underclock the cpu to 800mhz.
  • I ran Geekbench.
  • The Geekbench score for memory went significantly down.
share|improve this question
Please add some measures to your post, describing exactly what you did. – harrymc Jan 27 '13 at 20:33
Do you mean the AMD overdrive utility? I'm trying to figure out if your underclocking technique would have affected your hypertransport speed – Marcus Chan Jan 28 '13 at 2:44
Its the utility that comes with my graphics card. However if I use the built in windows one (under advanced power settings and max processor state) it has the same effect. Also, it does not affect my hypertransport speed according to cpuz. All that changes is my core speed and multiplier. It doesnt affect my HT or my NB speed – agz Jan 28 '13 at 6:40
First, as Geekbench runs in the CPU, one would expect that slowing that down will result in lower memory ratings. Second, your BIOS may bind RAM speed to that of the CPU. To check, go into the BIOS and see if there is any option for 'Memory Clock' or 'Memory Frequency'. If there isn't, then probably for your computer CPU speed also determines RAM speed. – harrymc Jan 28 '13 at 9:46
My ht/nb frequency isn't bound to the cpu frequency according to cpu-z. – agz Jan 28 '13 at 20:24

From the Geekbench 2 benchmark description page:

Memory benchmarks measure not only the performance of the underlying memory hardware, but also the performance of the functions provided by the operating system used to manipulate memory.

  • Read Sequential loads values from memory into registers.
  • Write Sequential stores values from registers into memory.
  • Stdlib Allocate allocates and deallocates blocks of memory of varying sizes using functions from the C Standard Library.
  • Stdlib Write writes a constant value to a block of memory using functions from the C Standard Library.
  • Stdlib Copy copies values from one block of memory to another using functions from the C Standard Library.

I'm guessing that because Geekbench is (in some parts) testing how quickly it can load data from CPU registers into RAM, that test would (of course) be slower with a reduced CPU clock. And, of course, as harrymc mentioned, reduced CPU speed would make the whole test run slower.

In general, score-based benchmarks like Geekbench tell you very little about how your system is running. There's very little way to isolate "memory performance" as a separate entity, since there are so many variables and most of them are directly related to the performance of the rest of your system.

share|improve this answer

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .