Tuning Cool'n'Quiet: Maximize Power And Performance, Part 2

Earlier this week we discussed the inner-workings of AMD's Cool'n'Quiet technology and what you could do as an enthusiast with the right tools to further-optimize an Athlon or Phenom processor.

Now it's time to talk about performance. Like most of the processor-oriented reviews you read, we're going to use several benchmarks to measure the speed of our various samples. But we're going to go even further this time around. In addition to performance numbers and test results, we're also going to include power consumption measurements for each benchmark.

The purpose is actually quite simple. Different applications utilize processors differently. Some will make full use of all available cores, while many others will not. This also means power consumption with each scenario will likely be different. By measuring the power consumption of single- and multi-threaded applications, we can get a reasonable idea of the performance and power consumption in a handful of scenarios, addressing whether or not our optimizations had a positive effect.

There’s also a different type of scenario we wanted to test. Many benchmarks apply a 100% load to each core, so the processor will be in its highest p-state during the test. We wanted to observe performance with idle p-states, or at least intermediary p-states. After all, the processor is idle through much of time we spend on the desktop, whether you're Web browsing, writing e-mails, playing music, or watching standard definition video--all of which hardly load the processor. For our purposes, we'll be measuring system power consumption when playing high definition video with and without hardware acceleration.

We're also going to look at the core/processor utilization levels. Why? To see whether or not the benchmarks really make use of all available processor resources. We can see just how often the processor is idle and to what degree. We can also see whether or not all cores are fully-utilized. To do this, we record processor utilization by using Vista's built-in Performance Monitor.

These measurements were taken with different power profiles: No Management (we turned off Cool'n'Quiet in the BIOS), Balanced (Cool'n'Quiet is enabled in the BIOS and the Windows power policy is set to Balanced), and Optimized (Cool'n'Quiet is enabled in the BIOS, the Windows power policy is set to Balanced, and the processor is running with lower core and northbridge voltages, shorter p-state transition times, and synchronous clock changes or “Ganged” in K10Stat). These settings are selected because we want to focus on performance and consumption with power management features enabled.

This thread is closed for comments
    Your comment
  • nzprogamer
    GO AMD go
    i am telling you my next build AMD/ATI
    """I WILL BE BACK"""
  • jedimasterben
    I'd be interested to see the tests performed on Windows 7 to see what the effect of reducing thread "jumping" would be.
  • cnox
    Dammit...how can this part 2 article be posted before the Building the Balanced PC Part 2?

  • melangex3
    Great Stuff. Keep up the good work. This is the type of review that will keep me coming back. How about throwing in the ever popular 720 BE and the new 620 or 630 just for giggles?
  • Ryun
    jedimasterbenI'd be interested to see the tests performed on Windows 7 to see what the effect of reducing thread "jumping" would be.

    I was thinking the same thing as well.

    Also, were the BIOSs all updated? The asynchronous clocks problem you're experiencing with Athlon II X2 was supposed to be fixed with updated CPU microcode.
  • Summer Leigh Castle
    620 and 720? :D
  • redgarl
    I must admit that lately AMD is impressive. I got a PII X3 720 BE unleashed at PII X4 20 fully stable with an Asus M4A78T-E latest BIOS. Let simply add that my 2 radeon 4850 OC in Crossfire are running as fast as 2 stock 4870...

    If you take into account that the 2 cards only cost 82$ each for a total of 165$ for the two... I can hardly believe that so little money can give so much results.
  • JimmiG
    With my Phenom X4 9650, I found Cool n Quiet to be pretty much worthless without tweaks. There were huge performance drops across the board, especially with tasks that didn't use all four cores, or only loaded cores partially. Videos and games would stutter and skip every couple of frames, compressing files would take longer etc. I basically had a 1.1 GHz CPU that would sometimes run at 2.3 GHz, if it felt like it. Too bad there was no tweak guide available then. I just disabled CnQ which solved all problems but made the system use more power and run hotter.

    With my 955BE, I haven't really had a need to tweak CnQ. It might cause a slight performance hit in some rare cases, but generally when I need a 3.2 GHz CPU, that's what it delivers.
  • tacoslave
    Nice, amd owns in the graphics department now with that $1.2 billion im sure amd is heading to pwn BOTH markets.
  • saint19
    Good!!!, I have my 955 to 3.8GHz at 1.5V....
  • marraco
    For this test, we've selected a 616MB folder full of files (the installer for Adobe Photoshop CS4)

    Bad choice. You are "compressing" already compressed files.
    A better choice would be to copy 600 MB from windows "program files" folder, and play with it.
  • eyemaster
    marracoBad choice. You are "compressing" already compressed files.A better choice would be to copy 600 MB from windows "program files" folder, and play with it.

    Bad choice if you're trying to compress file, but not a bad choice if you're just trying to stress a CPU. It will still have to do a whole lot of calculations regardless of the files.
  • b23h
    Thank you very much for this article. It came at the perfect time for me. I’ve just upgraded from a 65 watt AMD Athlon 64 X2 4600 to the AMD Phemon II X3 720. Since I’m running a fanless CPU heatsink (ZEROtherm BTF95) I was concerned that the 95 watts of power of the 720 would be too much for the BTF95. I was planning on underclocking the CPU in order to approximate what I thought the heatsink could handle. However with the help of the article I don’t need to underclock it at all. Using the chart of the 710 I estimated some beginning settings for the 720. While I may be able to further lower the voltages I’ve stressed test my current settings by running a program called the Intel Burn Test plus an immoderate amount of Borderlands.

    The 720 seems to have four p-states. The defaults were 1.25/1.15/1.05/.95 I am currently running the CPU at 1.15/1.125/1.0250/.9 I really haven’t thoroughly stress tested all the possibilities so I expect I may still be able to optimize further P-states one through three.

    The timing of this article was excellent for me and I appreciate all the information I’ve gotten at Tom’s Hardware Guide all these years.
  • volks1470
    Typo! Phenom II X3-X4 720-965 only have 6MB of L3 cache, not 8MB. Not a big deal but for a second there I though the 955 had more L3 cache than my 965.
  • 4ILY45
    melangex3Great Stuff. Keep up the good work. This is the type of review that will keep me coming back. How about throwing in the ever popular 720 BE and the new 620 or 630 just for giggles?

    YES PLEASE!!!! :)
  • shreeharsha
    I still need to sell my intel Pentium 660 system to build a AMD system. I am a AMD fan (converted) stuck with intel processor system.
  • JohnnyLucky
    Thanks for a very informative article that helps put things in perspective.
  • chaitanya_mkin
    Im Telling u AMD wll be the king in HISTORY theres no other name than AMD.
    AMD the BEST
  • chaitanya_mkin
    And this comparision is great for amd users, ofcourse for me tooooo cause im using AMD since 6 years.(my brain is AMD ATHLON X2 6000+ WINDSOR 3.02GHz)
  • arnawa_widagda

    We've updated this article (and the first part) with results from an AM3 motherboard and an Athlon II X4 620. The p-state settings we tested with the Athlon II X4 620 can be found in the first part (along with some power consumption numbers).
  • ohim
    http://it.anandtech.com/IT/showdoc.aspx?i=3784&p=7 , this benchmark cought my attention, while intel loses few performance on Linux, AMD gains almost double the performace ? wtf ?
  • Cespenar
    Many users opted for the Phenom X4 940 to steer away from a motherboard and RAM expense.
    I wonder how one of those rigs would stand up to identical scrutiny.
  • maddy143ded
    this article is nice help to me.
    I was searching for something like this.
    you have tested the Athlons, Athlon II X2/X4, Phenom II X3/X4.
    I would have liked to see some tests for Phenom II X2. as essentially this one is a Deneb with 2 cores disabled and a lower 80W TDP istead of 125W tdp.
    it would have given some idea on how much disabling the cores save us power.
    hope thats included in your next update.
  • TrackSmart
    Question for Tom's: Which Phenom II X4 945 processor did you use? AMD sells a 95W TDP version for almost the same price as the 125W TDP version. If you are interested in powersavings, that's clearly the one to buy. It should offer lower power usage with stock settings. And who knows how low you could take it, since it's a binned part. That would have been worth testing given the negligible price premium (less than $5 typically) compared to the 125W TDP version.