Froyo Kernels - 7-1-2010 - ULV/LV/MV

Discussion in 'ChevyNo1' started by ChevyNo1, Jul 1, 2010.

  1. ChevyNo1

    ChevyNo1 Premium Member
    Premium Member Developer

    Joined:
    Dec 28, 2009
    Messages:
    5,354
    Likes Received:
    5
    Trophy Points:
    153
    Ratings:
    +5
    Sounds like a SetCPU issue then, as it is controlling the CPU speed ...
     
  2. infinity8x3

    infinity8x3 Member

    Joined:
    Feb 22, 2010
    Messages:
    193
    Likes Received:
    0
    Trophy Points:
    16
    Ratings:
    +0
    so in theory, the ULV kernel is tons faster than stock and at the same time uses less power = longer battery life?

    If so then this is amazing!
     
  3. otter34

    otter34 Member

    Joined:
    Jun 9, 2010
    Messages:
    660
    Likes Received:
    0
    Trophy Points:
    16
    Ratings:
    +0
    I'm not so sure that it's using less power. If I have it set at 550 and the tests show it's running over 1100 consistently, then is it possible it's using MORE power?

    I emailed the developer.
     
  4. yusky03

    yusky03 Member

    Joined:
    May 11, 2010
    Messages:
    94
    Likes Received:
    4
    Trophy Points:
    8
    Ratings:
    +4
    I can't seem to get 'conservative' to ramp up the cpu clock at all. It just gets stuck on 250Mhz or 1.25Ghz.

    Up Threshold I had at 80% then set it to 40% and still stays at 250Mhz or 1.25Ghz for some reason.

    Down Threshold is at 20%

    I have OS Monitor in my task bar and I can see the CPU load.

    Short Answer: ULV uses the least amount of power but is just as fast as the MV and LV Kernels.

    Long Answer: http://www.droidforums.net/forum/bugless/55673-ulv-lv-mv-wth.html#post594012
     
    #28 yusky03, Jul 2, 2010
    Last edited: Jul 2, 2010
  5. mkraffert

    mkraffert Member

    Joined:
    Feb 10, 2010
    Messages:
    264
    Likes Received:
    0
    Trophy Points:
    16
    Location:
    Denver, CO
    Ratings:
    +0
    Chevy, I have a question. Does it make sense that I get much higher Quadrant scores (by 200 points) with a 1.25 LV kernel than with a 1.25 ULV kernel? Every thing else is identical except the kernel voltage.

    Sent from my Droid using Tapatalk
     
  6. yusky03

    yusky03 Member

    Joined:
    May 11, 2010
    Messages:
    94
    Likes Received:
    4
    Trophy Points:
    8
    Ratings:
    +4
    I basically have the same setup as you. I can get 19.6Mflops but I only get 1466 Quadrant.
     
  7. ChevyNo1

    ChevyNo1 Premium Member
    Premium Member Developer

    Joined:
    Dec 28, 2009
    Messages:
    5,354
    Likes Received:
    5
    Trophy Points:
    153
    Ratings:
    +5
    That's interesting ... will have to do some testing now!
     
  8. audaibnjad

    audaibnjad New Member

    Joined:
    Jun 30, 2010
    Messages:
    1
    Likes Received:
    0
    Trophy Points:
    1
    Ratings:
    +0
    I'm not sure why you think Quadrant should report various values. Perhaps you could elaborate on that? For the rest of this post, I'm going to assume that you think if you double the CPU speed, the Quadrant score should (approximately) double as well.

    Note that Quadrant tests things like filesystem & graphics performance. It is extremely unlikely that overclocking the CPU will increase filesystem performance (unless it also manages to overclock the bus to the flash, but even that may not increase speeds as much as you think).

    There's a separate graphics processor for doing accelerated 3d as well. Overclocking the CPU will help for CPU-bound issues, but I don't believe that the graphics core is being overclocked. So the graphics scores won't go up by much.

    You should not expect Quadrant scores to scale as fast as CPU clock speed.

    Also, Froyo (2.2) is faster than Eclair (2.1). So even if you stick with 550MHz, you'll see better performance out of 2.2.
     
  9. mkraffert

    mkraffert Member

    Joined:
    Feb 10, 2010
    Messages:
    264
    Likes Received:
    0
    Trophy Points:
    16
    Location:
    Denver, CO
    Ratings:
    +0
    That's about what I was getting with the ULV kernel. The screenshot is from the LV kernel. That's the difference I'm seeing.
     
  10. Tallica

    Tallica Premium Member
    Rescue Squad Premium Member

    Joined:
    Mar 17, 2010
    Messages:
    3,259
    Likes Received:
    1
    Trophy Points:
    103
    Location:
    Middleboro, MA
    Ratings:
    +1
    While that might be true when I set the kernal to a higher max speed slot the quadrant and linpack scores go up significantly, and if I set the min speed slot higher then 250 the scores go up again. At 1.25max-1.25min the Linpack and Quadrant scores are nuts...aproaching 2000 for quadrant
     
  11. otter34

    otter34 Member

    Joined:
    Jun 9, 2010
    Messages:
    660
    Likes Received:
    0
    Trophy Points:
    16
    Ratings:
    +0
    I just flashed a P3, 250-1200 kernal and have problems w/ that as well when set to 500 (I don't have time to test different settings right now). I had a quadrant score of 1146, 1132, 1050 and 1100 when I had it set to 250 min, 500Mhzmax!!! I emailed the developer again, but he hasn't yet responded to my first email that I sent 16 hours ago...maybe he's in a different time zone and needed 16 hours sleep last night. I've been there lol
     
  12. virtueoftheabsurd

    Joined:
    Jun 21, 2010
    Messages:
    66
    Likes Received:
    0
    Trophy Points:
    6
    Ratings:
    +0
    Love these kernels. I'm running 1Ghz ULV, everything is smooth as butter. I know my phone can handle 1.25, but I'm more concerned with battery than performance. I know that I should just install 1.25 and only clock it up to 1.0, but I'm not sure my self-control is that good.

    Thanks Chevy!
     
Search tags for this page

download kernel chevy no 1 mv