Page 9 of 16 FirstFirst ... 4567891011121314 ... LastLast
Results 161 to 180 of 304
  1. #161  
    Okay, I ran pBench on my 700p. I have installed the following apps: xWeather, Chatter, VeriChat, Speedy, and pBench. No VersaMail or Multimail was installed. I have a Verizon unit purchased on 6/13/06.

    Arithmetic: 466%
    Heap Management: 356%
    Data Access: 760%
    Graphics: 987%

    I ran it multiple times and the scores varied +/- 10 points each time. I performed a soft reset prior to running the benchmark, to make sure nothing weird was running in the background, and phone functionality was enabled. The Treo was plugged into the powered USB cable and charging at the time the benchmark was run.

    Since I haven't looked at the stock 650 numbers (not 2000 Man's overclocked custom ROM 650 numbers, which are absolutely meaningless to the adults on this forum), how do these numbers compare?

    Also, I might point out that pBench has not been updated since 2003. Benchmark applications, in order to be absolutely accurate, must be updated with every new product revision to make sure it is using the most current system calls and routines to properly test the newer hardware. Case in point: MacBench. With every new rev. of the Mac, even running on the same CPU, MacBench would consistently score lower on newer, faster hardware until it was updated.
  2. #162  
    Folks,

    I have Palm development experience and I have a theory why the 700p may skip worse with pocket tunes -- though I don't use this app much.

    The Palm OS does not multitask apps, but it does run threads to handle TCP/IP and other things (like the old grafitti).

    The EvDO speeds are faster and causing the TCP/IP stack to need more CPU time. So when a background app uses the data connection it is faster and requires more CPU cycles to handle the connection.

    So if your e-mail program checks every 5 minutes (like I do with Versamail which works pretty well), it likeley causes PTunes to skip because it is getting data a lot more quickly with EvDO speeds.

    I am really happy with the 700p. It has only crashed once in 3 weeks of use and is a much better device.

    Glenn
  3. #163  
    My 700p is MUCH slower than my 650. Without a doubt. Having said that, here is a speedy screenshot. Those taking screenshots should really include the 650 as one of the devices to compare against. Those are official results in the speedy db of a stock 650.

    650: 1376
    700p: 1500

    The 700p is faster by whatever standards speedy measures. In real use, it's slower and the difference is noticeable. Hopefully Palm releases an update that addresses this. Otherwise, I'd consider going back to the 650.
    Attached Images Attached Images
    A new Avatar to commemorate Silly Season.
  4. ink883's Avatar
    Posts
    872 Posts
    Global Posts
    883 Global Posts
    #164  
    Quote Originally Posted by skfny
    The 700p is faster by whatever standards speedy measures.
    Speedy measures CPU speed. Pbench has a measurement for data access.
    We are after the data access measurement, as that is what is causing the slow downs.
    Visor --> Visor Platinum --> Treo 300 --> Treo 600 --> Treo 650 --> Treo 700p --> Treo 755p --> Treo 800w --> Palm Pre
  5. ink883's Avatar
    Posts
    872 Posts
    Global Posts
    883 Global Posts
    #165  
    Quote Originally Posted by dcpmark
    I've had mine since 5/27, and it is not any slower today than it was on 5/27, and I don't have long lags switching between apps or white screens of death. I do get pTunes skipping when I try to multitask, but it works perfectly when I leave it alone, with or without the screen on, leds blinking, etc.
    Have you installed any 3rd party apps or are you running pretty much stock?
    Visor --> Visor Platinum --> Treo 300 --> Treo 600 --> Treo 650 --> Treo 700p --> Treo 755p --> Treo 800w --> Palm Pre
  6. ink883's Avatar
    Posts
    872 Posts
    Global Posts
    883 Global Posts
    #166  
    Quote Originally Posted by SonnyS

    Arithmetic: 466%
    Heap Management: 356%
    Data Access: 760%
    Graphics: 987%

    how do these numbers compare?
    Your numbers look pretty good for a 700.
    much faster then my 700 (543%), but not as fast as my 650 (950%).

    Do you have many 3rd party apps installed?
    Visor --> Visor Platinum --> Treo 300 --> Treo 600 --> Treo 650 --> Treo 700p --> Treo 755p --> Treo 800w --> Palm Pre
  7. #167  
    I have very few 3rd party apps. Only xWeather, Chatter, VeriChat, and the two benchmarking apps pBench and Speedy. Missing Sync installed an app as well, but I think that's for browsing the contents of my SD card from the desktop.

    I did a soft reset to make sure no apps were running in the background. So perhaps the data access speed is inverseley proportional to the amount of memory available in the machine? It might also be impacted by 3rd party applications running in the background. Anyway, no real complaints here... maybe I'll experience this phenomenon when I have a lot of apps installed, but I have limited need for apps.

    So it would seem that my data access score is about 20% below that of a stock 650.

    Mod Edit - If you want to continue posting here I suggest you leave other users out of your replies and stick to the subject of the Treo and only the Treo.

    You can consider this your one and only warning.
    Last edited by lgreenberg; 07/05/2006 at 07:08 PM.
  8.    #168  
    Quote Originally Posted by ink883
    Hate to be a fly in the ointment, but Speedy doesn't test memory speed (data access).
    Speedy seems to only measure CPU speed. Thus, there will not be much of a difference between the 650 and 700. In fact, since the 700 cpu is slightly more advanced it would be normal to see a small performance gain on the 700. Think about the difference between CPU and Hard drive benchmarks.

    However, what Jeff Gibson and others are saying is that the 700 is slow to access data in memory. Pbench is the only benchmark I am aware of that measures data access speed. Thatís the measurement we are after, thatís what is causing the slow down, lags, and ptunes skipping.

    If you look at the numbers I posted earlier you will notice that I am comparing my stock 650 and 700 (no custom roms, no over-clocking). Data access 950 v. 543, thatís a 47% decease in access times. That is significant.

    For those of you that say your 700 is just as fast as your 650 play ptunes in the background and open other apps, try to surf the net, compose an email. Or you could use a benchmark program that measures data access speed and not CPU speed.

    If there is another benchmark program that does data access let me know and I would be more then happy to post comparison results.
    Bingo. I've been saying it for a while now. The main difference seems to be in the database access speeds which can only be measured using pBench. Jeff Gibson and I have found the same thing (as have the 700p owners with IQs higher than 80pts.)

    Mod Edit - don't lower yourself to their level. Leave other members out of your reply.

    Please remember the old saying, "two wrongs do not make a right"
  9. #169  
    [QUOTE=2000 Man]Exactly. The 700p is much slower than the 650 in real life and it's also much slower in the benchmarks. Coincidence? I think not.

    The 700p is slower and suffers from extremely inconsistent performance. It resets often and is particularly problematic when updating or adding a new contact. Very distressing however I am certain that Palm and Verizon will address this in a patch release. Hopefully soon.
  10.    #170  
    Quote Originally Posted by ink883
    Speedy measures CPU speed. Pbench has a measurement for data access.
    We are after the data access measurement, as that is what is causing the slow downs.
    Exactly. That seems to be where the main problem is. It's an interesting theory about Palm possibly needing to increase the block size of the NVFS to speed up performance at the expense of free memory. I hope that isn't the solution and they're able to find a true bug in the system somewhere and erradicate it.
  11. dcpmark's Avatar
    Posts
    560 Posts
    Global Posts
    562 Global Posts
    #171  
    Quote Originally Posted by ink883
    Have you installed any 3rd party apps or are you running pretty much stock?
    I've installed the following third party apps:

    ababall
    snappermail
    datebk5
    ZLauncher
    LudusP 1.6
    Causerie
    Trivia 3.9.0 (on SD card)
    SplashPhoto
    SplashID
    TCPMP
    Traffic
    Speedy
    Keytest
    FlightStatus
    Joust
    SpyHunter
    My Little Tank
    Hellfire
    Cherrysoft Strippoker 2 (on SD card)
    Opera Mini
    IBM Java
    iSilo (on SD card)
    MaryoBros (on SD card)
    MegaBowling (on SD card)
    Missing Sync
    pTunes
    Vegas Video Poker
    Zap!2016
    Palm Reader (on SD card)
    PowerOne Calculator (on SD card)

    Except where noted, everything is in the main memory. I have 40M still left.
  12. #172  
    Both 650 and 700p units not fully charged, but at 75% battery. Only soft reset at begining of tests. No custom rom.

    here are my Pbench...First one 650 unclocked and second 650 overclocked. The overclocked one is very close to the OPs number at the begining of the thread. Coincedence? Or more evidence of overclocked 650?

    The third one is Pbench of 700p. With only snap running in background Still not 50% slower than overclocked 650.

    Forth one is Palm internals to show snap usage on 700p while running in background.

    fifth one is VFS Mark on 700p. this was actually faster on 700p than the 650.

    The OPs name was left out as per the Mods lurking about and to protect the guilty...<===can I say that?

    i would have posted these earlier, but I actually had to do some work today.

    Chuck

    My edit...definately not 50% slower than unclocked 650....hmmm...
    Attached Images Attached Images
  13.    #173  
    Thanks for posting those. Your 700p's numbers are pretty low compared to the 650. If you want to speed your 650 up a lot more, try running a custom ROM. I'm surprised that you can't notice the speed difference between the two, even with such a large gap in the performance numbers. Most others have been able to feel the difference.

    Looks like I was right all along about the data access being the main problem on the 700p. pBench is an incredible benchmarking tool and should be the only one used from now on. Jeff Gibson confirmed these results using the Palm File Manager and he found the following:

    Quote Originally Posted by jeffgibson
    What I am finding is that database access times are significantly greater on the 700P than they are on the 650. This is causing slow downs in every application I have on the 700P. You can really see the impact if you use Palm File Browser. It takes forever to have the file actually written to the device.

    Jeff
  14. #174  
    Quote Originally Posted by 2000 Man
    Thanks for posting those. Your 700p's numbers are pretty low compared to the 650. If you want to speed your 650 up a lot more, try running a custom ROM.

    Looks like I was right all along about the data access being the main problem on the 700p. pBench is an incredible benchmarking tool and should be the only one used from now on. Jeff Gibson confirmed these results using the Palm File Manager and he found the following:

    Question for you....

    How do you explain your Pbench mark of 1400 on a unclocked 650. My overclocked 650 is almost dead on to your 1400 unclocked number....

    700p is still not 50% slower than unclocked 650, is it?

    chuck
  15. #175  
    Quote Originally Posted by Mods

    Mod Edit - If you want to continue posting here I suggest you leave other users out of your replies and stick to the subject of the Treo and only the Treo.

    You can consider this your one and only warning.
    OMG...Is this the voice of GOD....

    sheeesh

    Can I go use the bathroom now???

    Chuck
  16.    #176  
    My 650 gets 1384 and my 700p got 844 on the pBench data access. Mind you, this was on a 700p with NOTHING installed except pBench and had never even been hotsynced. It's basically as fast as it was ever going to get without doing a custom ROM.

    Ink's numbers from earlier in the thread were 950 on the 650 and 543 on the 700p!

    Looks like I was right about the data access being the culprit.
  17. ink883's Avatar
    Posts
    872 Posts
    Global Posts
    883 Global Posts
    #177  
    Quote Originally Posted by chckhbrt
    My edit...definately not 50% slower than unclocked 650....hmmm...
    Those a good numbers, only a little over 20% slow down. There definately seems to be a slow down as people are ranging from 20%-47%

    Its a little fustrating that some people are only experiencing a 20% slowdown and still have many apps installed.

    I really wasn't expecting to see such a wide range in data access scores, I thought no more then 10% in either direction.
    Visor --> Visor Platinum --> Treo 300 --> Treo 600 --> Treo 650 --> Treo 700p --> Treo 755p --> Treo 800w --> Palm Pre
  18.    #178  
    Quote Originally Posted by ink883
    Those a good numbers, only a little over 20% slow down. There definately seems to be a slow down as people are ranging from 20%-47%

    Its a little fustrating that some people are only experiencing a 20% slowdown and still have many apps installed.

    I really wasn't expecting to see such a wide range in data access scores, I thought no more then 10% in either direction.
    I guess that 20% slower than the 650 = "lightning fast" for the 700p.
  19. ink883's Avatar
    Posts
    872 Posts
    Global Posts
    883 Global Posts
    #179  
    Quote Originally Posted by 2000 Man
    I guess that 20% slower than the 650 = "lightning fast" for the 700p.
    As a general rule, anything less then 20% isn't very noticable to the user. That would explain why some don't notice any slowdown.

    For example most users can't tell the difference between an 800 mhz processor and 1000 mhz processor unless they are running benchmarks.
    Visor --> Visor Platinum --> Treo 300 --> Treo 600 --> Treo 650 --> Treo 700p --> Treo 755p --> Treo 800w --> Palm Pre
  20.    #180  
    Or doing things like playing pTunes in the background, loading large apps like the phone app or Chatter, etc.
Page 9 of 16 FirstFirst ... 4567891011121314 ... LastLast

Posting Permissions