45x vs. 25x speed CF card in a netBook - Scion

This is a discussion on 45x vs. 25x speed CF card in a netBook - Scion ; Just upgraded from a 256M 25x to 512M 45x compactflash (both Transcend). An RMRFile non-case-sensitive search for "phase" through 1068 Data files (135M), searching only Data type files, took 1m50s on the 25x card, 1m37s on the 45x card. (didn't ...

+ Reply to Thread
Results 1 to 9 of 9

Thread: 45x vs. 25x speed CF card in a netBook

  1. 45x vs. 25x speed CF card in a netBook

    Just upgraded from a 256M 25x to 512M 45x compactflash (both Transcend).

    An RMRFile non-case-sensitive search for "phase" through 1068 Data files
    (135M), searching only Data type files, took 1m50s on the 25x card,
    1m37s on the 45x card. (didn't format the CF in the psion. Windows said
    it was "FAT" so I just wrote the files to it.)

    Much as I expected, only a slight increase in speed.

    Wonder if the netBook can be overclocked :-)

    Paul



  2. Re: 45x vs. 25x speed CF card in a netBook

    Paul Taylor wrote:
    > Just upgraded from a 256M 25x to 512M 45x compactflash (both
    > Transcend).
    >
    > An RMRFile non-case-sensitive search for "phase" through 1068 Data
    > files (135M), searching only Data type files, took 1m50s on the 25x
    > card, 1m37s on the 45x card. (didn't format the CF in the psion.
    > Windows said it was "FAT" so I just wrote the files to it.)
    >
    > Much as I expected, only a slight increase in speed.
    >
    > Wonder if the netBook can be overclocked :-)
    >
    > Paul


    So are you going to sell the 256MB card - I could use one I think

    Dr. Dweeb.



  3. Re: 45x vs. 25x speed CF card in a netBook

    Sorry, gonna keep the old card for when I get a digital camera.
    They're not too expensive though:
    http://orcalogic.co.uk/asp/prodtype....=765&ft=d&st=1

    P.



  4. Re: 45x vs. 25x speed CF card in a netBook

    Paul Taylor wrote:
    > Just upgraded from a 256M 25x to 512M 45x compactflash (both
    > Transcend).
    >
    > An RMRFile non-case-sensitive search for "phase" through 1068 Data
    > files (135M), searching only Data type files, took 1m50s on the 25x
    > card, 1m37s on the 45x card. (didn't format the CF in the psion.
    > Windows said it was "FAT" so I just wrote the files to it.)
    >
    > Much as I expected, only a slight increase in speed.
    >
    > Wonder if the netBook can be overclocked :-)
    >
    > Paul


    135/110 ~ 1.23MB/s
    135/97 ~ 1.39 MB/s

    This converts to a 13.4% improvement produced by an 80% theoretical IO speed
    improvement. I suspect that your test might be CPU bound. Perhaps we need
    a better test. :-)

    Dr. Dweeb.

    Since these values are



  5. Re: 45x vs. 25x speed CF card in a netBook

    > 135/110 ~ 1.23MB/s
    > 135/97 ~ 1.39 MB/s
    >
    > This converts to a 13.4% improvement produced by an 80% theoretical IO

    speed
    > improvement. I suspect that your test might be CPU bound. Perhaps we

    need
    > a better test. :-)
    >
    > Dr. Dweeb.


    Yep. Needs a PXA255 (looks like a bug-fixed PXA250):
    http://www.intel.com/design/pca/prodbref/252780.htm (mentions MPEG4)
    http://www.intel.com/design/support/faq/PXA_FAQ.htm

    Then again... It's interesting to look at the LCD controller specs!
    p266 of the 'Processor Developer's Manual' says
    "Display sizes up to 1024x1024 pixels, recommended maximum of 640x480"
    Hmm, OK for a phone I guess.
    I thought I read that the nB-PRO struggles with video? Maybe this is why?
    Looks like you need to use a separate 'graphics accelerator'?
    (shows this in a block diagram of the SA-1111 companion chip)

    http://www.intel.com/design/pca/appl...sors/index.htm

    Paul.



  6. Re: 45x vs. 25x speed CF card in a netBook

    Paul Taylor wrote:
    >> 135/110 ~ 1.23MB/s
    >> 135/97 ~ 1.39 MB/s
    >>
    >> This converts to a 13.4% improvement produced by an 80% theoretical
    >> IO speed improvement. I suspect that your test might be CPU bound.
    >> Perhaps we need a better test. :-)
    >>
    >> Dr. Dweeb.

    >
    > Yep. Needs a PXA255 (looks like a bug-fixed PXA250):
    > http://www.intel.com/design/pca/prodbref/252780.htm (mentions MPEG4)
    > http://www.intel.com/design/support/faq/PXA_FAQ.htm
    >
    > Then again... It's interesting to look at the LCD controller specs!
    > p266 of the 'Processor Developer's Manual' says
    > "Display sizes up to 1024x1024 pixels, recommended maximum of 640x480"
    > Hmm, OK for a phone I guess.
    > I thought I read that the nB-PRO struggles with video? Maybe this is
    > why? Looks like you need to use a separate 'graphics accelerator'?
    > (shows this in a block diagram of the SA-1111 companion chip)
    >
    > http://www.intel.com/design/pca/appl...sors/index.htm
    >


    Well, I meant more along the lines of a non-CPU bound test for the netBook,
    rather than changing devices !! :-)

    There is a little tester program that while not too great, is useful for
    comparisons. You could also try measuring the speed of dumping a 32MB file
    from the internal memory to the CF card. I will test mine with some big JPG
    files and see what I report.

    Dr. Dweeb
    > Paul.




  7. Re: 45x vs. 25x speed CF card in a netBook

    "Dr. Dweeb" wrote in message
    news:c2cd55$216m$1@news.cybercity.dk...
    > Paul Taylor wrote:
    > >> 135/110 ~ 1.23MB/s
    > >> 135/97 ~ 1.39 MB/s
    > >>
    > >> This converts to a 13.4% improvement produced by an 80% theoretical
    > >> IO speed improvement. I suspect that your test might be CPU bound.
    > >> Perhaps we need a better test. :-)
    > >>
    > >> Dr. Dweeb.

    > >
    > > Yep. Needs a PXA255 (looks like a bug-fixed PXA250):
    > > http://www.intel.com/design/pca/prodbref/252780.htm (mentions MPEG4)
    > > http://www.intel.com/design/support/faq/PXA_FAQ.htm
    > >
    > > Then again... It's interesting to look at the LCD controller specs!
    > > p266 of the 'Processor Developer's Manual' says
    > > "Display sizes up to 1024x1024 pixels, recommended maximum of 640x480"
    > > Hmm, OK for a phone I guess.
    > > I thought I read that the nB-PRO struggles with video? Maybe this is
    > > why? Looks like you need to use a separate 'graphics accelerator'?
    > > (shows this in a block diagram of the SA-1111 companion chip)
    > >
    > > http://www.intel.com/design/pca/appl...sors/index.htm
    > >

    >
    > Well, I meant more along the lines of a non-CPU bound test for the

    netBook,
    > rather than changing devices !! :-)
    >
    > There is a little tester program that while not too great, is useful for
    > comparisons. You could also try measuring the speed of dumping a 32MB

    file
    > from the internal memory to the CF card. I will test mine with some big

    JPG
    > files and see what I report.
    >
    > Dr. Dweeb
    > > Paul.

    >
    >


    FWIW, I just tried copying a 7787406 byte pdf from C to each of the CF
    cards.
    (don't have a lot of free RAM)

    Write times:
    45x = 19s:
    7787406/19 = 409863 B/s /1024 = 400 kB/s /150 = actual 2.67x
    25x = 19.4 s:
    7787406/19.4 = 401413 B/s /1024 = 392 kB/s /150 = actual 2.61x

    Obviously the RMRFile search will show a read-time-based difference.
    These (up to) 25x, 45x speeds are nominally x150 KB/s write speeds,
    the read speeds are typically something like up to 20% faster.

    P.



  8. Re: 45x vs. 25x speed CF card in a netBook

    Paul Taylor wrote:
    > "Dr. Dweeb" wrote in message
    > news:c2cd55$216m$1@news.cybercity.dk...
    >> Paul Taylor wrote:
    >>>> 135/110 ~ 1.23MB/s
    >>>> 135/97 ~ 1.39 MB/s
    >>>>
    >>>> This converts to a 13.4% improvement produced by an 80% theoretical
    >>>> IO speed improvement. I suspect that your test might be CPU bound.
    >>>> Perhaps we need a better test. :-)
    >>>>
    >>>> Dr. Dweeb.
    >>>
    >>> Yep. Needs a PXA255 (looks like a bug-fixed PXA250):
    >>> http://www.intel.com/design/pca/prodbref/252780.htm (mentions MPEG4)
    >>> http://www.intel.com/design/support/faq/PXA_FAQ.htm
    >>>
    >>> Then again... It's interesting to look at the LCD controller specs!
    >>> p266 of the 'Processor Developer's Manual' says
    >>> "Display sizes up to 1024x1024 pixels, recommended maximum of
    >>> 640x480" Hmm, OK for a phone I guess.
    >>> I thought I read that the nB-PRO struggles with video? Maybe this is
    >>> why? Looks like you need to use a separate 'graphics accelerator'?
    >>> (shows this in a block diagram of the SA-1111 companion chip)
    >>>
    >>> http://www.intel.com/design/pca/appl...sors/index.htm
    >>>

    >>
    >> Well, I meant more along the lines of a non-CPU bound test for the
    >> netBook, rather than changing devices !! :-)
    >>
    >> There is a little tester program that while not too great, is useful
    >> for comparisons. You could also try measuring the speed of dumping
    >> a 32MB file from the internal memory to the CF card. I will test
    >> mine with some big JPG files and see what I report.
    >>
    >> Dr. Dweeb
    >>> Paul.

    >>
    >>

    >
    > FWIW, I just tried copying a 7787406 byte pdf from C to each of the CF
    > cards.
    > (don't have a lot of free RAM)
    >
    > Write times:
    > 45x = 19s:
    > 7787406/19 = 409863 B/s /1024 = 400 kB/s /150 = actual 2.67x
    > 25x = 19.4 s:
    > 7787406/19.4 = 401413 B/s /1024 = 392 kB/s /150 = actual 2.61x
    >
    > Obviously the RMRFile search will show a read-time-based difference.
    > These (up to) 25x, 45x speeds are nominally x150 KB/s write speeds,
    > the read speeds are typically something like up to 20% faster.
    >
    > P.


    Well, I would now try copying the file FROM the CF card to the main memory.
    I predict interesting results, not consistent with your RMR experiment

    Dr. Dweeb.



  9. Re: 45x vs. 25x speed CF card in a netBook

    > > FWIW, I just tried copying a 7787406 byte pdf from C to each of the CF
    > > cards.
    > > (don't have a lot of free RAM)
    > >
    > > Write times:
    > > 45x = 19s:
    > > 7787406/19 = 409863 B/s /1024 = 400 kB/s /150 = actual 2.67x
    > > 25x = 19.4 s:
    > > 7787406/19.4 = 401413 B/s /1024 = 392 kB/s /150 = actual 2.61x
    > >
    > > Obviously the RMRFile search will show a read-time-based difference.
    > > These (up to) 25x, 45x speeds are nominally x150 KB/s write speeds,
    > > the read speeds are typically something like up to 20% faster.
    > >
    > > P.

    >
    > Well, I would now try copying the file FROM the CF card to the main

    memory.
    > I predict interesting results, not consistent with your RMR experiment
    >
    > Dr. Dweeb.


    I did, but they were so short (5.6 s?) that I couldn't tell much difference
    (by hand with a stopwatch.

    P.



+ Reply to Thread