Best Folding Laptop CPU

QeldromaQeldroma Arid ZoneAh Member
edited July 2007 in Folding@Home
I'm going to go out on a limb and assert something:

You currently can NOT find a better folding processor than the Merom for a laptop.

With 4MB of shared cache, this CPU is averaging 1100 ppd on SMP WUs ... at ONLY 2.0GHz.

Much of my recent trips to rates of 15,000 points-per-week can mostly be owed to my laptop (whose specs are found HERE )

If you are in the market for a new laptop and want it to fold in a major way- go with the T7xxx series CPUs. It out-folds my Opteron 170 by a healthy margin- to the point where I estimate that I would have to overclock the Opteron to about 3.3GHz (about 65%) to match it! My thought: If you are going to fold like this, you’ve got to have lots of cache :) . Here are some F@H log samples (T7200 stock on left, Opteron 170 OC'd 2.4GHz right).

attachment.php?attachmentid=23581&stc=1&d=1184198411

Comments

  • DanGDanG I AM CANADIAN Icrontian
    edited July 2007
    not a surprise to me, the extra cache makes a big difference.
  • QCHQCH Ancient Guru Chicago Area - USA Icrontian
    edited July 2007
    WOW... that's impressive!!!
  • LeonardoLeonardo Wake up and smell the glaciers Eagle River, Alaska Icrontian
    edited July 2007
    Yeah, that sounds about right. My son's laptop has a T2400, which is a dual core 1.83GHz, 2MB cache. It averages around 700ppd.
  • edcentricedcentric near Milwaukee, Wisconsin Icrontian
    edited July 2007
    When my C2D and AX2 were both just running two standard installs each they ran about the same points. Now that they are running SMP the C2D is 1.5x the points of my AX2.
  • QeldromaQeldroma Arid ZoneAh Member
    edited July 2007
    edcentric wrote:
    When my C2D and AX2 were both just running two standard installs each they ran about the same points. Now that they are running SMP the C2D is 1.5x the points of my AX2.

    Yeah, but it seems to depend on the WU being done. Like with the 340x series of WUs the T7200 bogs down to a lowly 300ppd, but the AMD is still slower- just not as dramatically so. I think that how well the project programmer utilizes the capabilities and instruction sets of the host also has a large role in how well a WU does. The point system also does not seem to measure performance well.

    Since I'm travelling a lot, I've had to stop doing SMPs on my laptop. Between not being able to run continuously, having connectivity with the ports blocked, or no connectivity at all- I decided to not risk having late WUs and went back to the 5.04 console client with advanced methods set. Right now it's going through an alternating diet of snail slow 3402s and extremely fast 2620s. It's not near as lucrative as the SMPs, but it still works for the project.

    Things should settle down somewhat next month- and I should be able to run SMPs again :) .
Sign In or Register to comment.