+ Reply to Thread
Results 1 to 18 of 18
  1. #1
    Comfortably Converged
    Join Date
    08-10-2005
    Location
    North Central Ohio
    Posts
    567

    Benchmarking the i730 and i760

    I did this over on my blog a day or so ago and I will post here as well.


    Got a copy of SPB Benchmark. Many thanks to SPB Software. One of you asked for some speed comparisons to the 730. They were pretty predictable.

    I ran two tests with my 730 and 760 both hard reset. Yes, I hard reset my 760.

    Keep in mind I have never done this before. I ran two tests. Some kind of CPU speed test and a test to see how fast they could copy a 1mb file.

    If you want more tests or if I did something wrong just let me know I will have time tonight to run more tests.

    The pic with four test results is with the i760 only seeing the difference in using superdave's Autostep prog.
    Attached Thumbnails Attached Thumbnails Benchmarking the i730 and i760-760730-test-1.jpg   Benchmarking the i730 and i760-760730-test-2.jpg   Benchmarking the i730 and i760-760-withwithout-autostep.jpg  

  2. #2
    Fisher of Men mwfielder's Avatar
    Join Date
    10-30-2003
    Location
    Arlington, TX
    Posts
    4,752
    for those of us checking the forum from the pda (who can't view pics), can you just post some of the summary findings in text format?

  3. #3
    Registered User
    Join Date
    03-19-2005
    Location
    Allentown PA
    Posts
    1,376
    Could you explain the 3rd picture? It's not clearly labeled.
    Thanks
    - Andrew - PDAPhoneHome.com

  4. #4
    Comfortably Converged
    Join Date
    08-10-2005
    Location
    North Central Ohio
    Posts
    567
    The third pic is two tests run with just the i760 with and without Autostep
    CPU test with Autostep 4918 ms
    CPU test without Autostep 4917 ms

    Copy 1mb file with Autostep 1285 ms
    Copy 1mb file without Autostep 1619 ms

    Later I will run some more tests.

  5. #5
    Comfortably Converged
    Join Date
    08-10-2005
    Location
    North Central Ohio
    Posts
    567
    Heres full results on two tests. The test that compares the 760 with and without Autostep had Autostep running at 520.
    Attached Thumbnails Attached Thumbnails Benchmarking the i730 and i760-760withwithoutautostep.jpg   Benchmarking the i730 and i760-full-results730760.jpg  
    Last edited by sixftunda; 09-29-2007 at 11:13 PM.

  6. #6
    Finally received the i760... morgan2005's Avatar
    Join Date
    01-22-2006
    Posts
    27
    Wow those bench marks are surprising, some of those results are up to 4 times better on the i760 than the i730. I think the only category the i730 was better was in graphics.
    Last edited by morgan2005; 09-30-2007 at 12:20 PM.

  7. #7
    Registered User
    Join Date
    09-13-2007
    Posts
    54
    what's autostep for us morons?

  8. #8
    Comfortably Converged
    Join Date
    08-10-2005
    Location
    North Central Ohio
    Posts
    567
    Its a program created by superdave to make the processor go a certain speed. WM5 had problems with "speed stepping". As you can see from the tests the 760 does pretty good without it.

    Oh and its right here if you want it. Search is your friend.

  9. #9
    Moderator iProb8's Avatar
    Join Date
    11-01-2005
    Location
    Phoenix, Arizona
    Posts
    3,449
    Thanks for posting these interesting results. If I'm reading them correctly, they indicate the i760 is faster than the i730 in benchmarking. But does the i760 feel faster than the i730 in your day-to-day use?
    -Jay
    The Fine Print:Nothing in this post (or any of my other posts) is intended to constitute legal advice or the establishment of an attorney-client relationship. For purposes of this forum, I'm just another nerd like you. :-)

  10. #10
    Comfortably Converged
    Join Date
    08-10-2005
    Location
    North Central Ohio
    Posts
    567
    tytnII review

    Check out the benchmarks of the TyTnII compared to the 760! Spb's Benchmark Index rated the 760 higher if only by a few points. I am going to send my results to Jon Westfall to make sure we ran the same tests before I jump for joy. I will let you know what he says. Hopefully he will send me his XML file so I can put them side by side.

    Here... I put the pics of the results side by side to make it easier
    Last edited by sixftunda; 10-02-2007 at 07:06 PM.

  11. #11
    Registered User
    Join Date
    01-13-2006
    Location
    New Jersey
    Posts
    1,525
    You think you could try the i730 with XCPUScalar?
    Make some nice money from home:
    http://www.tvtravelbiz.com/ds9916

  12. #12
    Comfortably Converged
    Join Date
    08-10-2005
    Location
    North Central Ohio
    Posts
    567
    yep give me an hour or so it takes 35-40 mins to tun a full test

  13. #13
    Comfortably Converged
    Join Date
    08-10-2005
    Location
    North Central Ohio
    Posts
    567
    Here ya go Dominick
    Autostep was run at 520
    XCPUScalar was run at 500. On the trial copy I got that was the highest it would go. I thought it used to go to 624? Also I did not hard reset on the XCPU test (didn't feel like reloading progs tonight). I did soft reset before I started so there were no running progs. If you think that would make a huge difference I could do it again
    Attached Thumbnails Attached Thumbnails Benchmarking the i730 and i760-5waytest.png  

  14. #14
    Registered User
    Join Date
    01-13-2006
    Location
    New Jersey
    Posts
    1,525
    Quote Originally Posted by sixftunda View Post
    Here ya go Dominick
    Autostep was run at 520
    XCPUScalar was run at 500. On the trial copy I got that was the highest it would go. I thought it used to go to 624? Also I did not hard reset on the XCPU test (didn't feel like reloading progs tonight). I did soft reset before I started so there were no running progs. If you think that would make a huge difference I could do it again

    It can go to 624 yes but it sucks the battery of course. If you have a lot of programs it will make a difference, but its just interesting to see the results either way with or without you doing it. In looking at the results, it seems the i760 is better for program use than for graphics. I think it would be better if you could do 624. Thanks for the work on the benchmarks man.
    Last edited by Dominick_7; 10-02-2007 at 10:43 PM.
    Make some nice money from home:
    http://www.tvtravelbiz.com/ds9916

  15. #15
    Comfortably Converged
    Join Date
    08-10-2005
    Location
    North Central Ohio
    Posts
    567
    I just checked again v3.03 only lets me go to 500. It's a trial but the pop-up at the beginning only say user settings and auto scale are disabled.
    BTW, if the other progs aren't running, how would they interfere with the test? Just curious.

  16. #16
    Registered User
    Join Date
    01-13-2006
    Location
    New Jersey
    Posts
    1,525
    Quote Originally Posted by sixftunda View Post
    I just checked again v3.03 only lets me go to 500. It's a trial but the pop-up at the beginning only say user settings and auto scale are disabled.
    BTW, if the other progs aren't running, how would they interfere with the test? Just curious.
    Now that I think of it, it may not effect the tests, I just know from experience once you load up a BUNCH of stuff like I do things can tend to run a bit slower.
    Make some nice money from home:
    http://www.tvtravelbiz.com/ds9916

  17. #17
    Registered User mrailing's Avatar
    Join Date
    07-19-2004
    Location
    Indiana
    Posts
    8,410
    Quote Originally Posted by Dominick_7 View Post
    It can go to 624 yes but it sucks the battery of course. If you have a lot of programs it will make a difference, but its just interesting to see the results either way with or without you doing it. In looking at the results, it seems the i760 is better for program use than for graphics. I think it would be better if you could do 624. Thanks for the work on the benchmarks man.
    Some i730's can run at 624, not all. Since this is overclocking the chip, beyond what it is suppose to run even at max, I would say this isn't something that should be done in a real test comparison. It's kinda like saying, let me take this Intel chip, overclock it with some liquid nitrogen cooling to like 8 GHz and then compare it to a normally clocked AMD chip, see it runs faster...

    When doing real world comparisons, companies generally use the default processor speed, without ANY additional software, such as something like AutoStep or XCPU. But this would be a drawback to the i730, since we know for a fact that the processor isn't scaled correctly. But if you do use a scaling program, it should be set to the max that the processor would actually hit, which is 520 MHz, since realistically running it at 624 MHz on a daily basis isn't something that can happen, since it can overheat, and it drains the battery extremely fast.

    Also note that these are different CPUs, and that the Samsung is meant to run faster at a slower clock speed, with less battery drain, and less heat. It's also dependent upon the OS, since WM6 is suppose to be faster than WM5 in general. So to get a true test, we need to get a WM6 i730, so how about is Samsung and Verizon, give us WM6 for the i730 so we can compare....

  18. #18
    Registered User
    Join Date
    01-13-2006
    Location
    New Jersey
    Posts
    1,525
    Quote Originally Posted by mrailing View Post
    Some i730's can run at 624, not all. Since this is overclocking the chip, beyond what it is suppose to run even at max, I would say this isn't something that should be done in a real test comparison. It's kinda like saying, let me take this Intel chip, overclock it with some liquid nitrogen cooling to like 8 GHz and then compare it to a normally clocked AMD chip, see it runs faster...

    When doing real world comparisons, companies generally use the default processor speed, without ANY additional software, such as something like AutoStep or XCPU. But this would be a drawback to the i730, since we know for a fact that the processor isn't scaled correctly. But if you do use a scaling program, it should be set to the max that the processor would actually hit, which is 520 MHz, since realistically running it at 624 MHz on a daily basis isn't something that can happen, since it can overheat, and it drains the battery extremely fast.

    Also note that these are different CPUs, and that the Samsung is meant to run faster at a slower clock speed, with less battery drain, and less heat. It's also dependent upon the OS, since WM6 is suppose to be faster than WM5 in general. So to get a true test, we need to get a WM6 i730, so how about is Samsung and Verizon, give us WM6 for the i730 so we can compare....
    Good points about the testing and speed of the i730 processor. I agree with what you're saying about having a fair test, but thats the reason why I suggested a higher speed on the i730 for the tests because the i730 wasn't able to handle scaling properly. Because of that I would think it would be at an automatic disadvantage...besides the fact that it also doesn't have WM6, that's supposed to run faster making it have yet another disadvantage. True about sucking the battery on 624 too. Now I could be wrong, but if I remember correctly didn't the guy who made XCPUScalar in this forum say that 624 wasn't true over clocking since it was designed to go higher? One would wonder how that could be since the battery goes so fast like you mentioned. Concerning WM6 on the i730, now that's what I'm talking about I would think Samsung and Verizon would surely be able to empathize with what you said about that and get WM6 on the i730 for us just so we can benchmark it
    Make some nice money from home:
    http://www.tvtravelbiz.com/ds9916

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts