The Story of Phenom's Erratic Performance

A few months ago I called AMD with a problem. In testing for AMD's Phenom re-launch, I encountered a major issue: Phenom performance actually degraded since I first tested it. There were two benchmarks in particular that saw performance go down: SYSMark 2007 and Adobe Photoshop CS3. SYSMark gave me scores that were around 10% lower than what they were when Phenom launched, despite these being B3-stepping parts. Photoshop was far worse, with performance being around half of what it was at the Phenom launch.

I originally attributed the changes to something strange that happened with the move to Vista SP1. I theorized that somehow the TLB fix was being applied to B3 stepping parts and negatively impacting performance, but WinRAR and memory tests didn't support the thought. AMD couldn't figure out what was happening so I chalked it up to a problem with my testbed or testing methodology, something I'd have to revisit at a later time.

The SYSMark issue actually went away on its own; I swapped from my 780G motherboard to a 790FX without reinstalling Windows to see if it was a 780G/integrated graphics issue, the performance problems remained. But upon swapping back, once again without a Vista reinstall, my SYSMark scores magically jumped around 10%. The "fix" lasted long enough for me to finish the benchmarks for the Phenom re-launch review, but sure enough the problem reappeared when I tried to re-run one test after I'd gotten everything I needed for the review. I never did figure out what was causing my Photoshop performance issues however.

The Culprit: Cool'n'Quiet

When I started testing for today's review, I ran into the same issue again. I always start by benchmarking SYSMark first, since the suite takes forever to complete on a single CPU. As soon as I got my first results, I realized my problem was back. Determined to find the cause I tried everything...again. The one thing I didn't change however was the Cool'n'Quiet setting in the BIOS, but I did try it this time.

Cool 'n Quiet is the marketing name for AMD's on-the-fly clock speed/voltage adjustment. Depending on the software load on the CPU, AMD's Cool'n'Quiet will adjust the p-state of the CPU cores - which includes adjusting core voltage and clock speed. If you're running a processor intensive game or application, CnQ will make sure your CPU runs at full speed, but if you're just typing a text document it will underclock/undervolt the CPU.

Phenom's CnQ is a more advanced version of what was in the Athlon 64 X2, it not only allows for individual core clock speed adjustment but is able to transition between states faster than previous versions of CnQ (at least that's what AMD implies).

SYSMark 2007 Preview Overall Score CnQ/EIST Enabled CnQ EIST/Disabled Performance Increase from Disabling CnQ/EIST
AMD Phenom X4 9350e 101 113 11.8%
AMD Athlon X2 6400+ 121 123 1.7%
Intel Core 2 Quad Q9450 153 156 2.0%

 

 

Disabling CnQ increased my SYSMark scores by around 12% and cut my Photoshop CS3 render times in half (58.7s with CnQ enabled, 35.2s with CnQ disabled); enabling CnQ had the opposite effect. Gary ran similar numbers using PCMark Vantage and found a 5% difference. AMD originally insisted that the problem was because SYSMark introduces unrealistic pauses into its benchmark (so called "think" times or periods of time while the system is waiting for user input), but since we found the same issue in other benchmarks (PCM Vantage and our Photoshop test), we believe this is more than just a SYSMark issue.

The SYSMark problem was mostly repeatable, it would consistently produce lower scores with CnQ enabled on the Phenom CPUs. The Photoshop scores were a bit more erratic - the problem went away for a little while but has since returned and won't go away again, even with CnQ disabled. It is worth mentioning that the majority of our benchmarks wasn't impacted by the problem, but that doesn't mean that it won't impact your daily usage.

Note that the same problem doesn't plague the Athlon X2, this appears to be a Phenom/K10 issue only. As a reference we ran some numbers with Intel's SpeedStep enabled vs. disabled and didn't see any similar behavior.

I had found the source of my problems, but I didn't understand why it caused them.

Intel Makes 45nm Affordable The Mystery of the Missing Performance
Comments Locked

36 Comments

View All Comments

  • Gikaseixas - Wednesday, July 2, 2008 - link

    Other sites tested it already and could hit 3.2 - 3.6 speeds. Hopefully Anandtech will be able to overclock a Phenom to it's limits this time around.
  • Googer - Wednesday, July 2, 2008 - link

    Missing from the benchmarks is the Intel Core 2 Quad Q9550 Yorkfield 2.83GHz 12MB. How would this chip stack up against all others tested?

    http://www.newegg.com/Product/Product.aspx?Item=N8...">http://www.newegg.com/Product/Product.aspx?Item=N8...
  • RamarC - Wednesday, July 2, 2008 - link

    the q9550 isn't in the same price range as the other processors so that's why it wasn't included. as for performance, either subtract or add 10% to the q9450's figures.
  • DanD85 - Wednesday, July 2, 2008 - link

    Are you absolutely sure about that? As hothardware thinks differently:
    "By altering its multiplier and increasing the CPU voltage to 1.45v, we were able to take our Phenom X4 9950 to an respectable 3.1GHz using nothing but a stock AMD PIB cooler. Higher frequencies were possible, but we couldn't keep the system 100% stable, so we backed things down to 3.1GHz. While running at that speed, we re-ran some tests and also monitored core temperatures and found that the chip never broke the 60ºC mark, and hovered around 58ºC under load - at least according to AMD's Overdrive software. That is one heck of an overclock and relatively cool temperatures for a Phenom in our opinion. If the majority of chips have the same amount of headroom as ours, we suspect the 9950 Black Edition will be appealing to AMD CPU enthusiasts looking for the best the company has to offer."
    http://www.hothardware.com/Articles/AMD_Phenom_X4_...">http://www.hothardware.com/Articles/AMD...nom_X4_9...
  • KaarlisK - Wednesday, July 2, 2008 - link

    Maybe by setting affinity for Photoshop's threads to a certain core, it would be possible to verify whether Vista's thread management is part of the cause?
  • Zoomer - Wednesday, July 2, 2008 - link

    I know its a novel concept, but what about running some benches in XP to see if it's another Vista issue?
  • Rhoxed - Wednesday, July 2, 2008 - link

    Increasing the NB core (IMC) clock (in Phenom it runs async from the Core Speed unlike Athlon which is Sync) drops latencies (especially L3) and increases memory performance/throughput, which in turn improves system performance. The Phenom starts to come to life when you hit a 2.6GHz core speed with a NB core clock at 2200MHz+. Depending on the application and CPU, increasing NB core speeds (getting up to 2200MHz+) can result in performance differences from 3%~12% in most cases.

    Upping my NB/HT to 2400MHz over the stock 2000 at the same clockspeed (2800) i net a 15%~ increase (on a 9850BE)
  • RamarC - Tuesday, July 1, 2008 - link

    i'm a developer and want to upgrade my win2k3/ss2k5 server to a quad core. since it currently has a 3.4ghz p4d, a phenom 9x50 would be a big step-up (even though i don't have any performance issues). but the p4d has been very reliable and i don't want to have to deal with flaky hardware issues when i'm pushing code out the door. should i just bite the bullet and shell out the extra cash for a p45+q9450?
  • Calin - Wednesday, July 2, 2008 - link

    I have an AMD based PC at home, and I look forward to another AMD-based pc (780G and Phenom X3 or X4).
    These being said, I think for a server you really really should go for an Intel configuration. Also, at 3.4 GHz a P4D probably is one hell of a power draw.
    Compared to your current server, and based on what I think you need, I don't think a quad core would help you - a dual core would probably be enough, and Intel has those aplenty.
  • Zoomer - Wednesday, July 2, 2008 - link

    From the article, it seems like sticking to the cheaper, sub 100W TDP cpus and not overclocking is the way to go.

Log in

Don't have an account? Sign up now