Sandra memory bench

Archived from groups: alt.comp.hardware.overclocking (More info?)

Hey all,
Something isn't right here,I think. Sandra memory bandwidth gives me
3060/3059. Not good right?
My PC is an Athlon 64 2800+, Chaintech VNF3-250 motherboard, 1GB PC3200 DDR.
It is overclocked and overclocking an A64 system is kinda wierd.
I have my BIOS "FSB overclock" at 250 and HTT set at x3. Which gives me
HTT@750 (default is 4x200)So, according to my boot screen the CPU is running
at 2260 and actual DDR FSB is 416.PCI/AGP is locked at default. Don't ask me
how it got there.I wish I knew!!
So, Is the sandra bencmark a result of asyncronous clock settings? Thanks
15 answers Last reply
More about sandra memory bench
  1. Archived from groups: alt.comp.hardware.overclocking (More info?)

    dawg wrote:
    > Hey all,
    > Something isn't right here,I think. Sandra memory bandwidth gives me
    > 3060/3059. Not good right?
    > My PC is an Athlon 64 2800+, Chaintech VNF3-250 motherboard, 1GB
    > PC3200 DDR. It is overclocked and overclocking an A64 system is kinda
    > wierd.
    > I have my BIOS "FSB overclock" at 250 and HTT set at x3. Which gives
    > me HTT@750 (default is 4x200)So, according to my boot screen the CPU
    > is running at 2260 and actual DDR FSB is 416.PCI/AGP is locked at
    > default. Don't ask me how it got there.I wish I knew!!

    You've set the RAM to the "166MHz" or 5:6 setting. Also, the BIOS has
    calculated the RAM speed incorrectly: it should be 204.5MHz.

    > So, Is the sandra bencmark a result of asyncronous clock settings?

    No. There is no "async" in an Athlon64 due to it's architecture, no no
    penalty for running RAM at non-1:1 ratios (except the lower raw MHz, of
    course). Since you have a S754 system, your maximum theoretical speed would
    be 3273 MB/sec, so your Sandra scores are giving about 93.5% efficiency.
    This sounds about right for an A64 system.

    --
    Michael Brown
    www.emboss.co.nz : OOS/RSI software and more :)
    Add michael@ to emboss.co.nz ---+--- My inbox is always open
  2. Archived from groups: alt.comp.hardware.overclocking (More info?)

    I thought the bandwidth would be higher too.

    My OC'ed P4 on an i875 at 253MHz (3.8GHz) is giving up almost 6200mb/sec.


    "Michael Brown" <see@signature.below> wrote in message
    news:42b20d0c$1@clarion.carno.net.au...
    > dawg wrote:
    >> Hey all,
    >> Something isn't right here,I think. Sandra memory bandwidth gives me
    >> 3060/3059. Not good right?
    >> My PC is an Athlon 64 2800+, Chaintech VNF3-250 motherboard, 1GB
    >> PC3200 DDR. It is overclocked and overclocking an A64 system is kinda
    >> wierd.
    >> I have my BIOS "FSB overclock" at 250 and HTT set at x3. Which gives
    >> me HTT@750 (default is 4x200)So, according to my boot screen the CPU
    >> is running at 2260 and actual DDR FSB is 416.PCI/AGP is locked at
    >> default. Don't ask me how it got there.I wish I knew!!
    >
    > You've set the RAM to the "166MHz" or 5:6 setting. Also, the BIOS has
    > calculated the RAM speed incorrectly: it should be 204.5MHz.
    >
    >> So, Is the sandra bencmark a result of asyncronous clock settings?
    >
    > No. There is no "async" in an Athlon64 due to it's architecture, no no
    > penalty for running RAM at non-1:1 ratios (except the lower raw MHz, of
    > course). Since you have a S754 system, your maximum theoretical speed
    > would be 3273 MB/sec, so your Sandra scores are giving about 93.5%
    > efficiency. This sounds about right for an A64 system.
    >
    > --
    > Michael Brown
    > www.emboss.co.nz : OOS/RSI software and more :)
    > Add michael@ to emboss.co.nz ---+--- My inbox is always open
    >
  3. Archived from groups: alt.comp.hardware.overclocking (More info?)

    Right. According to CPU-z the memory divider is 11. If it makes a difference
    "
    Michael Brown" <see@signature.below> wrote in message
    news:42b20d0c$1@clarion.carno.net.au...
    > dawg wrote:
    > > Hey all,
    > > Something isn't right here,I think. Sandra memory bandwidth gives me
    > > 3060/3059. Not good right?
    > > My PC is an Athlon 64 2800+, Chaintech VNF3-250 motherboard, 1GB
    > > PC3200 DDR. It is overclocked and overclocking an A64 system is kinda
    > > wierd.
    > > I have my BIOS "FSB overclock" at 250 and HTT set at x3. Which gives
    > > me HTT@750 (default is 4x200)So, according to my boot screen the CPU
    > > is running at 2260 and actual DDR FSB is 416.PCI/AGP is locked at
    > > default. Don't ask me how it got there.I wish I knew!!
    >
    > You've set the RAM to the "166MHz" or 5:6 setting. Also, the BIOS has
    > calculated the RAM speed incorrectly: it should be 204.5MHz.
    >
    > > So, Is the sandra bencmark a result of asyncronous clock settings?
    >
    > No. There is no "async" in an Athlon64 due to it's architecture, no no
    > penalty for running RAM at non-1:1 ratios (except the lower raw MHz, of
    > course). Since you have a S754 system, your maximum theoretical speed
    would
    > be 3273 MB/sec, so your Sandra scores are giving about 93.5% efficiency.
    > This sounds about right for an A64 system.
    >
    > --
    > Michael Brown
    > www.emboss.co.nz : OOS/RSI software and more :)
    > Add michael@ to emboss.co.nz ---+--- My inbox is always open
    >
    >
  4. Archived from groups: alt.comp.hardware.overclocking (More info?)

    fish wrote:
    > I thought the bandwidth would be higher too.
    >
    > My OC'ed P4 on an i875 at 253MHz (3.8GHz) is giving up almost
    > 6200mb/sec.

    Yeah, but that's dual channel:

    > Michael Brown wrote:
    [...]
    >> Since you have a S754 system, your maximum
    ^^^^^^^^^^^
    >> theoretical speed would be 3273 MB/sec, so your Sandra scores are
    >> giving about 93.5% efficiency. This sounds about right for an A64
    >> system.

    --
    Michael Brown
    www.emboss.co.nz : OOS/RSI software and more :)
    Add michael@ to emboss.co.nz ---+--- My inbox is always open
  5. Archived from groups: alt.comp.hardware.overclocking (More info?)

    even with single channel, the bandwidth is still over 5700mb/sec.


    "Michael Brown" <see@signature.below> wrote in message
    news:42b2c67f$1@clarion.carno.net.au...
    > fish wrote:
    >> I thought the bandwidth would be higher too.
    >>
    >> My OC'ed P4 on an i875 at 253MHz (3.8GHz) is giving up almost
    >> 6200mb/sec.
    >
    > Yeah, but that's dual channel:
    >
    >> Michael Brown wrote:
    > [...]
    >>> Since you have a S754 system, your maximum
    > ^^^^^^^^^^^
    >>> theoretical speed would be 3273 MB/sec, so your Sandra scores are
    >>> giving about 93.5% efficiency. This sounds about right for an A64
    >>> system.
    >
    > --
    > Michael Brown
    > www.emboss.co.nz : OOS/RSI software and more :)
    > Add michael@ to emboss.co.nz ---+--- My inbox is always open
    >
  6. Archived from groups: alt.comp.hardware.overclocking (More info?)

    fish wrote:
    > even with single channel, the bandwidth is still over 5700mb/sec.

    Nope.

    In his case:
    204.5 * 2 * 64 = 26176 Mbit/s = 3272 MByte/sec
    In your case:
    253 * 2 * 64 = 32384 Mbit/s = 4048 MByte/sec

    --
    Michael Brown
    www.emboss.co.nz : OOS/RSI software and more :)
    Add michael@ to emboss.co.nz ---+--- My inbox is always open
  7. Archived from groups: alt.comp.hardware.overclocking (More info?)

    "Michael Brown" <see@signature.below> wrote in message
    news:42b2ef60$1@clarion.carno.net.au...
    > fish wrote:
    > > even with single channel, the bandwidth is still over 5700mb/sec.
    >
    > Nope.
    >
    > In his case:
    > 204.5 * 2 * 64 = 26176 Mbit/s = 3272 MByte/sec
    > In your case:
    > 253 * 2 * 64 = 32384 Mbit/s = 4048 MByte/sec
    >
    > --
    > Michael Brown
    > www.emboss.co.nz : OOS/RSI software and more :)
    > Add michael@ to emboss.co.nz ---+--- My inbox is always open
    >
    >
    Okay. Those numbers make sense. But,whats with the setting of 250 for "CPU
    overclock in mhz" in the BIOS? Sorry for being dense.
  8. Archived from groups: alt.comp.hardware.overclocking (More info?)

    On Wed, 22 Jun 2005 01:10:25 GMT, "Phil Weldon"
    <notdiscosed@example.com> wrote:

    >As for overall benchmarks, nothing beats how a user's system performs runing
    >the user's programs. Anything else is, well, beside the point.

    you are right too!
    --
    Regards , SPAJKY ®
    mail addr. @ my site @ http://www.spajky.vze.com
    3rd Ann.: - "Tualatin OC-ed / BX-Slot1 / inaudible setup!"
  9. Archived from groups: alt.comp.hardware.overclocking (More info?)

    Spajky wrote:
    > dawg wrote:
    >> Right. According to CPU-z the memory divider is 11. If it makes a
    >> difference
    >
    > yes it makes difference; your ram runs @ 91% of FSB clock if the mem
    > divider is shown properly with your version of CPU-Z ..
    >

    >>> dawg wrote:
    >>>> I have my BIOS "FSB overclock" at 250
    [...]
    >>>> CPU is running at 2260 and actual DDR FSB is 416.PCI/AGP is
    >>>> locked at default.
    >
    > your ram is running @ 227MHz (ddr454) ....

    Err, he's using a 9x multiplier (says sys clock is 250, CPU speed is 2260),
    not 10x, giving the RAM running at 82% of the sys clock speed (ie:
    204.5MHz).

    [...]

    --
    Michael Brown
    www.emboss.co.nz : OOS/RSI software and more :)
    Add michael@ to emboss.co.nz ---+--- My inbox is always open
  10. Archived from groups: alt.comp.hardware.overclocking (More info?)

    On Wed, 22 Jun 2005 18:19:18 +1000, "Michael Brown"
    <see@signature.below> wrote:

    >>> According to CPU-z the memory divider is 11.

    >> your ram runs @ 91% of FSB clock if the mem
    >> divider is shown properly with your version of CPU-Z ..

    >>>>> I have my BIOS "FSB overclock" at 250

    >> your ram is running @ 227MHz (ddr454) ....
    >
    >Err, he's using a 9x multiplier (says sys clock is 250, CPU speed is 2260),
    >not 10x, giving the RAM running at 82% of the sys clock speed (ie:
    >204.5MHz).

    IMHO wrong! ALL frequency versions of A64 have a system bus clock @
    2GHz & default FSB/mem divider @ 10 !!!! Multiplier for core clock
    frequency is another stuff ! so:

    2000 / 100 = 200 Fsb = 200MHz mem clock default
    200 fsb x 9 multi = 1,8GHz core clock default
    divider FSB/mem reported by his CPU-Z is 10/11 (if correct) in his
    case, so @ default FSB the mem clock would be 91% of fsb one (182MHz);
    overclock by 25% (x1,25) = Cpu clock 2,26GHz, Fsb 250MHz & ram clock
    227MHz ... (& Cpu-Z mostly reports true mem.clock wrong on "k8"
    family!)

    all frequency versions of s.754 Semprons have instead a system bus @
    1,6 GHz clock & default Fsb/mem divider @ 8 !!! the calculations are
    done the same! ... :-)
    /so for example same basic clock Sempron3100+ with a Fsb/mem divider
    of for example same 11 & same 25% OC, would be running at same clock,
    same Fsb, but Ram @ 182MHz .../
    --
    Regards , SPAJKY ®
    mail addr. @ my site @ http://www.spajky.vze.com
    3rd Ann.: - "Tualatin OC-ed / BX-Slot1 / inaudible setup!"
  11. Archived from groups: alt.comp.hardware.overclocking (More info?)

    "Michael Brown" <see@signature.below> wrote in message
    news:42b91f6b$1@clarion.carno.net.au...
    > Spajky wrote:
    > > dawg wrote:
    > >> Right. According to CPU-z the memory divider is 11. If it makes a
    > >> difference
    > >
    > > yes it makes difference; your ram runs @ 91% of FSB clock if the mem
    > > divider is shown properly with your version of CPU-Z ..
    > >
    >
    > >>> dawg wrote:
    > >>>> I have my BIOS "FSB overclock" at 250
    > [...]
    > >>>> CPU is running at 2260 and actual DDR FSB is 416.PCI/AGP is
    > >>>> locked at default.
    > >
    > > your ram is running @ 227MHz (ddr454) ....
    >
    > Err, he's using a 9x multiplier (says sys clock is 250, CPU speed is
    2260),
    > not 10x, giving the RAM running at 82% of the sys clock speed (ie:
    > 204.5MHz).
    >
    > [...]
    >
    > --
    > Michael Brown
    > www.emboss.co.nz : OOS/RSI software and more :)
    > Add michael@ to emboss.co.nz ---+--- My inbox is always open
    >
    >

    Thanks guys.Cleared up a few things.
  12. Archived from groups: alt.comp.hardware.overclocking (More info?)

    Spajky wrote:
    [...]
    > ALL frequency versions of A64 have a system bus clock @
    > 2GHz & default FSB/mem divider @ 10 !!!! Multiplier for core clock
    > frequency is another stuff !

    Incorrect on all three counts. Read
    http://www.emboss.co.nz/amdmults/k8mults.html
    This is documented in AMD tech docs and has been widely tested and found to
    be true (hence the discovery of the 5 exceptions).

    [...]

    --
    Michael Brown
    www.emboss.co.nz : OOS/RSI software and more :)
    Add michael@ to emboss.co.nz ---+--- My inbox is always open
  13. Archived from groups: alt.comp.hardware.overclocking (More info?)

    On Thu, 23 Jun 2005 18:17:13 +1000, "Michael Brown"
    <see@signature.below> wrote:

    >Spajky wrote:
    >[...]
    >> ALL frequency versions of A64 have a system bus clock @
    >> 2GHz & default FSB/mem divider @ 10 !!!!

    forgot to mention FX series (1,6GHz & 8 ; same as Semprons)

    > Multiplier for core clock
    >> frequency is another stuff !
    >
    >Incorrect on all three counts. Read
    >http://www.emboss.co.nz/amdmults/k8mults.html
    >This is documented in AMD tech docs and has been widely tested and found to
    >be true (hence the discovery of the 5 exceptions).

    yeah, I read that, & even the author is not 100% sure & I already know
    that Oscar Wu's table (& some others too!)

    .... There was/is a lot a mess about finding true memory clock on "k8"
    systems, I also read a lot on the Net forums, searching for a way to
    be sure about that frequency. With my way of calculation ...

    I know I could be wrong (happened many times in my life!), but I could
    be also right & all others over the Net wrong !
    (happened 4 years ago with my statement after studying
    Tualatins & BX chipset, that with some minor mods could be done
    running it on older boards; the massive "flame" tornado hit me thru
    Usenet & certain forums to my head very soon that could not be done
    :-))) . Well, few months later PowerLeap released their too-expensive
    gadget & we found another few months later a way how to do it w/o
    PowerLeap adaptor ... :-)))

    There are few things that could support my calculation (previous
    post): using some own "sane" common logic:

    - some experienced guys encountered problems with OC-ing sometimes
    memory; while proofed to handle certain high clock (downclocking
    multiplier) or high Fsb/Htt & high cpu clock too; even when getting
    better results encountered instability too soon when tried to OC
    altogether (even when lowering HTT multiplier) even if reported Ram
    clocks were lower with some programs or Bios-es.
    With my calculation could be that explained: mem.clock was
    higher than showed with some programs & expected!

    - people have serious problems calculating memory clock most "known"
    way when Cpu multiplier is a non integer number like 8,5x !!! This
    could be linked also to upper problem.
    My type of calculation does NOT have this problem!

    - this link supports the fact of basic system bus frequency:
    http://www.amdcompare.com/us-en/desktop/default.aspx
    (but first wrong declaration - memory frequency & when clicking More
    details about CPU, declaration is HT speed) & common logic tells you
    if default A64 & Sempron s.754 FSB/HTT & mem clock is 200MHz (ddr400),
    there MUST be also a default FSB (cpu)/mem. divider for same family of
    CPU, which is always an integer number !!! /Cpu multiplier can be
    instead non-integer one like 8,5x for example/

    - another my theory supporting article is my own one :-)
    /few months ago had a lot of work to find out what´s happening!/
    http://freeweb.siol.net/jerman55/HP/benchMem.htm
    (read very carefully & try!)
    With fixed totally relaxed mem.latency timings, play with some
    OC-ing & downClocking, changing other parameters; bench results will
    follow memory clock practically linearly ... :-) /first check on
    default 200MHz clocks! FSB: Mem = 1:1 (in_sync) ... :-)

    ..... I also had a lot of problems first to determine the real clock
    ram was running on "K8" core (for recalculating that 100MHz
    score-yellow column) since sometimes I was getting ridicules
    results from people believing CPU-Z or some other program was telling
    them ...

    .... but there is still a chance, that I am in missbelieve ..
    --
    Regards , SPAJKY ®
    mail addr. @ my site @ http://www.spajky.vze.com
    3rd Ann.: - "Tualatin OC-ed / BX-Slot1 / inaudible setup!"
  14. Archived from groups: alt.comp.hardware.overclocking (More info?)

    Spajky wrote:
    > On Thu, 23 Jun 2005 18:17:13 +1000, "Michael Brown"
    > <see@signature.below> wrote:
    >
    >> Spajky wrote:
    >> [...]
    >>> ALL frequency versions of A64 have a system bus clock @
    >>> 2GHz & default FSB/mem divider @ 10 !!!!
    >
    > forgot to mention FX series (1,6GHz & 8 ; same as Semprons)

    All K8 CPUs have a system clock speed of 200MHz. Socket 939 CPUs have a 5x
    hypertransport multiplier by default (resulting in a DDR'd 1GHz bus), socket
    754's have a 4x hypertransport multiplier (giving a DDR'd 800MHz bus), and
    socket 940's come in both flavours. Nothing stopping you besides BIOS
    limitations (and obviously overall chipset hypertransport limitations) from
    running it at multipliers of 0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 4.0, or 5.0
    though.

    >> Multiplier for core clock
    >>> frequency is another stuff !
    >>
    >> Incorrect on all three counts. Read
    >> http://www.emboss.co.nz/amdmults/k8mults.html
    >> This is documented in AMD tech docs and has been widely tested and
    >> found to be true (hence the discovery of the 5 exceptions).
    >
    > yeah, I read that, & even the author is not 100% sure & I already know
    > that Oscar Wu's table (& some others too!)
    >
    > ... There was/is a lot a mess about finding true memory clock on "k8"
    > systems,

    There's a lot of mess about the actual memory clock out there for three
    reasons (in no particular order):
    1) Many CPUID program writers didn't (and still don't) have a clue how to
    calculate the memory speed so did it by random guessing instead of
    researching.
    2) Sudhian media discovered that ratios and half-integer-multipliers didn't
    give the results they would expect from simple calculations, and did a
    complete train-wreck of an article. It contradicts itself multiple times and
    only serves to confuse people more. At the time, I (among many others) knew
    exactly how to calculate the real memory frequency, and pointed out to the
    author that his results were exactly in line with what would be expected.
    3) There's many people out there who THINK they know how it works, don't
    really know how it works, but still say they're right anyhow.

    > I also read a lot on the Net forums, searching for a way to
    > be sure about that frequency. With my way of calculation ...

    Hey, maybe AMD got it wrong too! After all, they only made the damn thing.
    Look on page 15 of the "AMD Functional Data Sheet, 939 Pin Package"
    document. It gives a list of real memory frequencies for various multipliers
    and ratio settings. According to you, the memory speed should not change
    when you change the multiplier. But look, the 166MHz speeds bounce all over
    the place! Care to explain how, using your method, to get a 157.14MHz memory
    clock with a 200MHz FSB, a 11x multiplier, and a 166MHz ratio setting? Note
    that the standard method has no such trouble.

    Additionally, Oskar Wu (who is probably one of the best motherboard design
    engineers in the business) did oscilloscope measurements of memory
    frequencies at a range of multipliers and ratios:
    http://www.xtremesystems.org/forums/showthread.php?t=41595
    Care to explain these numbers too?

    [...]
    > - people have serious problems calculating memory clock most "known"
    > way when Cpu multiplier is a non integer number like 8,5x !!!

    No, not at all. There's no difference between calculating for a 8.0x
    multiplier and an 8.5x multiplier. The *only* problem is with a 183MHz ratio
    setting, where AMD's engineers appeared to make a mistake in their
    calculation program resulting in the wrong divider being used.

    > This
    > could be linked also to upper problem.
    > My type of calculation does NOT have this problem!

    Your type of calculation does not give the right answer. According to your
    calculation, there should be no drop in memory speed shifting to a
    half-integer multiplier. Except there most definately is. Oops.

    [...]

    The K8 memory speed is not independent of the CPU multiplier, hence cannot
    be derived from a simple divisor of the HTT speed. QED.

    --
    Michael Brown
    www.emboss.co.nz : OOS/RSI software and more :)
    Add michael@ to emboss.co.nz ---+--- My inbox is always open
  15. Archived from groups: alt.comp.hardware.overclocking (More info?)

    On Fri, 24 Jun 2005 09:34:11 +1000, "Michael Brown"
    <see@signature.below> wrote:

    >All K8 CPUs have a system clock speed of 200MHz.

    >There's a lot of mess about the actual memory clock out there for three
    >reasons (in no particular order):

    >1) Many CPUID program writers didn't (and still don't) have a clue how to
    >calculate the memory speed so did it by random guessing instead of
    >researching.

    >Hey, maybe AMD got it wrong too! After all, they only made the damn thing.
    >Look on page 15 of the "AMD Functional Data Sheet, 939 Pin Package"
    >document. It gives a list of real memory frequencies for various multipliers
    >and ratio settings.

    > Oskar Wu (who is probably one of the best motherboard design
    >engineers in the business) did oscilloscope measurements of memory
    >frequencies at a range of multipliers and ratios:
    >http://www.xtremesystems.org/forums/showthread.php?t=41595

    > The *only* problem is with a 183MHz ratio
    >setting, where AMD's engineers appeared to make a mistake in their
    >calculation program resulting in the wrong divider being used.
    >
    >Your type of calculation does not give the right answer.

    >The K8 memory speed is not independent of the CPU multiplier, hence cannot
    >be derived from a simple divisor of the HTT speed. QED.

    Thank you for some additional ideas, since I am planning to recheck
    everything once again soon & in 3 weeks here close to me our "local OC
    website" is prepairing a big meeting & hole weekend LAN & education
    "party", since kids know me mostly IMHO won´t be problem to find among
    all setups some 10 ones with "k8" setup to allow me to make some my
    own benchmarking & re-checks and recalculations too ; I will se than
    ....
    --
    Regards , SPAJKY ®
    mail addr. @ my site @ http://www.spajky.vze.com
    3rd Ann.: - "Tualatin OC-ed / BX-Slot1 / inaudible setup!"
Ask a new question

Read More

Overclocking Memory