1 ram is 2700;1 ram is 2100; so do they both read at 2100?

G

Guest

Guest
Archived from groups: microsoft.public.windowsxp.hardware (More info?)

I have one ram module that is PC2100 and the other is PC2700. They both show
in system properties and in the free program called CPUz. I am wondering
though, is the 2700 module reading at the slower 2100 level?
 
G

Guest

Guest
Archived from groups: microsoft.public.windowsxp.hardware (More info?)

There is a possibility. Either the PC2700 runs at PC2100
speed or the PC2100 is overclocked to run at PC2700. You
have to check your front side bus(FSB) speed for the
answer. Download Sisoft Sandra from www.downloads.com to
find out, or go to your BIOS. The speed of PC2100 is
266MHz, and PC2700 is 333MHz. The FSB determines the
speed. So check to see if the FSB is 266MHz or 333MHz.
Good Luck!
>-----Original Message-----
>I have one ram module that is PC2100 and the other is
PC2700. They both show
>in system properties and in the free program called CPUz.
I am wondering
>though, is the 2700 module reading at the slower 2100
level?
>
>
>.
>
 

Jim

Distinguished
Mar 31, 2004
2,444
0
19,780
Archived from groups: microsoft.public.windowsxp.hardware (More info?)

Memory can only run at ONE speed and set of timings, therefore, the lesser
of the two is always used to configure it. In this case, PC2100 (266MHz
DDR) runs slower than PC2700 (333MHz DDR). CPU-Z (I'm not familiar with it)
is most likely measuring *actual* runtime configuration, NOT the
specification (i.e., potential) of each module.

HTH

Jim


"Waheeba" <Waheeba@aol.com> wrote in message
news:uSHTko7JEHA.3980@TK2MSFTNGP10.phx.gbl...
> I have one ram module that is PC2100 and the other is PC2700. They both
show
> in system properties and in the free program called CPUz. I am wondering
> though, is the 2700 module reading at the slower 2100 level?
>
>
 
G

Guest

Guest
Archived from groups: microsoft.public.windowsxp.hardware (More info?)

To explain to the OP a little more. SDRAM is designed to
sync dynamically with the system bus, what ever the speed of
the FSB is the speed the RAM will run at. The RAM speed
ratings are the fastest speeds that the RAM will operate at
error free. RAM is tested and rated statistically. CPU
speeds are also rated in the same manner, which is why
over-clocking works, some of the CPU that are sold as 2 GHz
can run at 3 GHZ and some will fail at 2.2 GHz.


--
The people think the Constitution protects their rights;
But legislators see it as an obstacle to be overcome.


"Quang" <anonymous@discussions.microsoft.com> wrote in
message news:231201c427be$05122040$a601280a@phx.gbl...
| There is a possibility. Either the PC2700 runs at PC2100
| speed or the PC2100 is overclocked to run at PC2700. You
| have to check your front side bus(FSB) speed for the
| answer. Download Sisoft Sandra from www.downloads.com to
| find out, or go to your BIOS. The speed of PC2100 is
| 266MHz, and PC2700 is 333MHz. The FSB determines the
| speed. So check to see if the FSB is 266MHz or 333MHz.
| Good Luck!
| >-----Original Message-----
| >I have one ram module that is PC2100 and the other is
| PC2700. They both show
| >in system properties and in the free program called CPUz.
| I am wondering
| >though, is the 2700 module reading at the slower 2100
| level?
| >
| >
| >.
| >
 
G

Guest

Guest
Archived from groups: microsoft.public.windowsxp.hardware (More info?)

Thank you Jim and Quang.

Would I really notice the difference between 2700 and 2100 anyway? Would I
notice the difference in gaming or music editing?

"Jim" <null@null.com> wrote in message
news:xCxhc.25914$Yf6.9141@fed1read07...
> Memory can only run at ONE speed and set of timings, therefore, the lesser
> of the two is always used to configure it. In this case, PC2100 (266MHz
> DDR) runs slower than PC2700 (333MHz DDR). CPU-Z (I'm not familiar with
it)
> is most likely measuring *actual* runtime configuration, NOT the
> specification (i.e., potential) of each module.
>
> HTH
>
> Jim
>
>
> "Waheeba" <Waheeba@aol.com> wrote in message
> news:uSHTko7JEHA.3980@TK2MSFTNGP10.phx.gbl...
> > I have one ram module that is PC2100 and the other is PC2700. They both
> show
> > in system properties and in the free program called CPUz. I am wondering
> > though, is the 2700 module reading at the slower 2100 level?
> >
> >
>
>
 

Jim

Distinguished
Mar 31, 2004
2,444
0
19,780
Archived from groups: microsoft.public.windowsxp.hardware (More info?)

Ideally, it's more important and advantageous that the CPU FSB and memory
run at the same speed. IOW, the FSB and memory "clocks" are synchronous.
Since you didn't mention anything about your processor, it's hard to be
precise here, I can only speak in general terms.

Suppose you have an Athlon XP 2600 "Barton" 333MHz (DDR), that's 166MHz
"clock". If you installed only PC2100, which is 133MHz "clock", then
(assuming no overclocking) you'd be running the CPU FSB @ 166MHz, but the
memory at 133MHz, i.e., "asynchronously". As a general rule, computers run
more efficiently when the FSB and memory clocks are "in sync". The Barton
would be better matched to the PC2700 (166MHz "clock"). On the other hand,
suppose you had an Intel P4 533MHz (133MHz "clock") CPU. In this case, the
PC2100 (133MHz "clock") would be a better match, since it runs synchronously
w/ the CPU FSB.

That's why you really can't discuss memory selection outside the context of
the CPU (or more precisely, the FSB). There's a relationship between the
CPU FSB and memory that's relevant too, not just the pure speed potential of
the memory. Trying to discuss the best configuration of your PC w/o regard
to both, it like trying to get a quality picture w/ your camera and
discussing only the lens aperture or shutter speed. It takes BOTH to get
the right exposure, not just one or the other.

Some people do claim better performance when the FSB and memory are
asynchronous, but lots of empirical testings has shown that in most cases,
unless the differences are substantial, the benefits are so slight as to be
hardly worth the effort. Here's an example of a recent examination of
running two identical configurations, one 250/250 (sync), vs. 250/200
(async):

http://www.cluboverclocker.com/reviews/memory/corsair/xms3200LLPRO/index.htm

The benefits of 250/250 (PC4000) vs. 250/200 (PC3200) are so slight, who
could be bothered. Of course, the fact that latencies are a real problem at
these speeds contributes significantly to the lack of performance gains,
most PC2700 and PC2100 could probably run w/ closer timings, and so there
might be a slightly better case for the PC2700 in an asynchronous
configuration.

Of course, some people over-buy memory (PC2700 when PC2100 will do) if they
intend to overclock the CPU FSB. This way, as they increase the CPU FSB,
the memory remains underclocked and doesn't become a bottleneck to
successful OC'ing of the FSB. Whether this is a concern or interest to you
is beyond this discussion, I just want to complete the picture, since there
are people who *do* buy faster memory for such reasons.

As an aside, anytime you *mix* memory speeds (PC2700 and PC2100), it's
doubtful your motherboard will support the faster of the two by default (By
SPD). If it will support the faster speed of the two, and that's iffy, it
would only happen if you *manually* set the Memory Freq (and timings) in the
BIOS. Most any motherboard is going to default to the most conservative
configuration.

But for most mere mortals, you match the CPU FSB and memory speed for
typical, normal, everyday operations.

HTH

Jim


"Waheeba" <Waheeba@aol.com> wrote in message
news:OSe%23HF8JEHA.952@TK2MSFTNGP12.phx.gbl...
> Thank you Jim and Quang.
>
> Would I really notice the difference between 2700 and 2100 anyway? Would I
> notice the difference in gaming or music editing?
>
> "Jim" <null@null.com> wrote in message
> news:xCxhc.25914$Yf6.9141@fed1read07...
> > Memory can only run at ONE speed and set of timings, therefore, the
lesser
> > of the two is always used to configure it. In this case, PC2100 (266MHz
> > DDR) runs slower than PC2700 (333MHz DDR). CPU-Z (I'm not familiar with
> it)
> > is most likely measuring *actual* runtime configuration, NOT the
> > specification (i.e., potential) of each module.
> >
> > HTH
> >
> > Jim
> >
> >
> > "Waheeba" <Waheeba@aol.com> wrote in message
> > news:uSHTko7JEHA.3980@TK2MSFTNGP10.phx.gbl...
> > > I have one ram module that is PC2100 and the other is PC2700. They
both
> > show
> > > in system properties and in the free program called CPUz. I am
wondering
> > > though, is the 2700 module reading at the slower 2100 level?
> > >
> > >
> >
> >
>
>
 
G

Guest

Guest
Archived from groups: microsoft.public.windowsxp.hardware (More info?)

Probably not. Again, understand the RAM does not control
the speed in any way, it has no function that sets the speed
of the computer. If your FSB is running at a speed higher
than the rated speed of your RAM, the RAM will operate at
that speed or it will have errors, run hot and burn up or it
may fail on boot.

There is no harm in having PC2700 and PC 2100 in the same
system and it might all run at 333 FSB, but if the slow
module can't keep up the system will crash.


--
The people think the Constitution protects their rights;
But legislators see it as an obstacle to be overcome.


"Waheeba" <Waheeba@aol.com> wrote in message
news:OSe%23HF8JEHA.952@TK2MSFTNGP12.phx.gbl...
| Thank you Jim and Quang.
|
| Would I really notice the difference between 2700 and 2100
anyway? Would I
| notice the difference in gaming or music editing?
|
| "Jim" <null@null.com> wrote in message
| news:xCxhc.25914$Yf6.9141@fed1read07...
| > Memory can only run at ONE speed and set of timings,
therefore, the lesser
| > of the two is always used to configure it. In this
case, PC2100 (266MHz
| > DDR) runs slower than PC2700 (333MHz DDR). CPU-Z (I'm
not familiar with
| it)
| > is most likely measuring *actual* runtime configuration,
NOT the
| > specification (i.e., potential) of each module.
| >
| > HTH
| >
| > Jim
| >
| >
| > "Waheeba" <Waheeba@aol.com> wrote in message
| > news:uSHTko7JEHA.3980@TK2MSFTNGP10.phx.gbl...
| > > I have one ram module that is PC2100 and the other is
PC2700. They both
| > show
| > > in system properties and in the free program called
CPUz. I am wondering
| > > though, is the 2700 module reading at the slower 2100
level?
| > >
| > >
| >
| >
|
|
 
G

Guest

Guest
Archived from groups: microsoft.public.windowsxp.hardware (More info?)

>-----Original Message-----
>Ideally, it's more important and advantageous that the
CPU FSB and memory
>run at the same speed. IOW, the FSB and memory "clocks"
are synchronous.
>Since you didn't mention anything about your processor,
it's hard to be
>precise here, I can only speak in general terms.
>
>Suppose you have an Athlon XP 2600 "Barton" 333MHz (DDR),
that's 166MHz
>"clock". If you installed only PC2100, which is
133MHz "clock", then
>(assuming no overclocking) you'd be running the CPU FSB @
166MHz, but the
>memory at 133MHz, i.e., "asynchronously". As a general
rule, computers run
>more efficiently when the FSB and memory clocks are "in
sync". The Barton
>would be better matched to the PC2700 (166MHz "clock").
On the other hand,
>suppose you had an Intel P4 533MHz (133MHz "clock") CPU.
In this case, the
>PC2100 (133MHz "clock") would be a better match, since it
runs synchronously
>w/ the CPU FSB.
>
>That's why you really can't discuss memory selection
outside the context of
>the CPU (or more precisely, the FSB). There's a
relationship between the
>CPU FSB and memory that's relevant too, not just the pure
speed potential of
>the memory. Trying to discuss the best configuration of
your PC w/o regard
>to both, it like trying to get a quality picture w/ your
camera and
>discussing only the lens aperture or shutter speed. It
takes BOTH to get
>the right exposure, not just one or the other.
>
>Some people do claim better performance when the FSB and
memory are
>asynchronous, but lots of empirical testings has shown
that in most cases,
>unless the differences are substantial, the benefits are
so slight as to be
>hardly worth the effort. Here's an example of a recent
examination of
>running two identical configurations, one 250/250 (sync),
vs. 250/200
>(async):
>
>http://www.cluboverclocker.com/reviews/memory/corsair/xms3
200LLPRO/index.htm
>
>The benefits of 250/250 (PC4000) vs. 250/200 (PC3200) are
so slight, who
>could be bothered. Of course, the fact that latencies
are a real problem at
>these speeds contributes significantly to the lack of
performance gains,
>most PC2700 and PC2100 could probably run w/ closer
timings, and so there
>might be a slightly better case for the PC2700 in an
asynchronous
>configuration.
>
>Of course, some people over-buy memory (PC2700 when
PC2100 will do) if they
>intend to overclock the CPU FSB. This way, as they
increase the CPU FSB,
>the memory remains underclocked and doesn't become a
bottleneck to
>successful OC'ing of the FSB. Whether this is a concern
or interest to you
>is beyond this discussion, I just want to complete the
picture, since there
>are people who *do* buy faster memory for such reasons.
>
>As an aside, anytime you *mix* memory speeds (PC2700 and
PC2100), it's
>doubtful your motherboard will support the faster of the
two by default (By
>SPD). If it will support the faster speed of the two,
and that's iffy, it
>would only happen if you *manually* set the Memory Freq
(and timings) in the
>BIOS. Most any motherboard is going to default to the
most conservative
>configuration.
>
>But for most mere mortals, you match the CPU FSB and
memory speed for
>typical, normal, everyday operations.
>
>HTH
>
>Jim
>
>
>"Waheeba" <Waheeba@aol.com> wrote in message
>news:OSe%23HF8JEHA.952@TK2MSFTNGP12.phx.gbl...
>> Thank you Jim and Quang.
>>
>> Would I really notice the difference between 2700 and
2100 anyway? Would I
>> notice the difference in gaming or music editing?
>>
>> "Jim" <null@null.com> wrote in message
>> news:xCxhc.25914$Yf6.9141@fed1read07...
>> > Memory can only run at ONE speed and set of timings,
therefore, the
>lesser
>> > of the two is always used to configure it. In this
case, PC2100 (266MHz
>> > DDR) runs slower than PC2700 (333MHz DDR). CPU-Z
(I'm not familiar with
>> it)
>> > is most likely measuring *actual* runtime
configuration, NOT the
>> > specification (i.e., potential) of each module.
>> >
>> > HTH
>> >
>> > Jim
>> >


That's all very well and good, but you seem to have over-
looked what the physicists and RAM designers call the
"Henderson Factor" (vis-a-vis the inverse squares rule,
of course).
 

Jim

Distinguished
Mar 31, 2004
2,444
0
19,780
Archived from groups: microsoft.public.windowsxp.hardware (More info?)

If you're referring to the camera analogy (aperture size vs. shutter speed),
I wasn't characterizing the *nature* of the relationship, but the fact there
*is* a relationship (something that many people don't even realize exists).
Indeed, the inverse squares rule does not apply to the topic of FBS and
memory (not everything in nature and physics does). If I somehow left the
impression this was the case, I apologize. I was merely indicating "it
takes two to Tango". Please, no lessons in dance theory.

Frankly, only one in a thousand who stroll through this NG will "catch" it,
or even understand what you and I are discussing, which reminds me of
another rule, "know your audience", but don't feel too bad, a common error
of the elite.

Jim



"Wislu Plethora" <anonymous@discussions.microsoft.com> wrote in message
news:266401c427d9$97be8ad0$a001280a@phx.gbl...
>
> >-----Original Message-----
> >Ideally, it's more important and advantageous that the
> CPU FSB and memory
> >run at the same speed. IOW, the FSB and memory "clocks"
> are synchronous.
> >Since you didn't mention anything about your processor,
> it's hard to be
> >precise here, I can only speak in general terms.
> >
> >Suppose you have an Athlon XP 2600 "Barton" 333MHz (DDR),
> that's 166MHz
> >"clock". If you installed only PC2100, which is
> 133MHz "clock", then
> >(assuming no overclocking) you'd be running the CPU FSB @
> 166MHz, but the
> >memory at 133MHz, i.e., "asynchronously". As a general
> rule, computers run
> >more efficiently when the FSB and memory clocks are "in
> sync". The Barton
> >would be better matched to the PC2700 (166MHz "clock").
> On the other hand,
> >suppose you had an Intel P4 533MHz (133MHz "clock") CPU.
> In this case, the
> >PC2100 (133MHz "clock") would be a better match, since it
> runs synchronously
> >w/ the CPU FSB.
> >
> >That's why you really can't discuss memory selection
> outside the context of
> >the CPU (or more precisely, the FSB). There's a
> relationship between the
> >CPU FSB and memory that's relevant too, not just the pure
> speed potential of
> >the memory. Trying to discuss the best configuration of
> your PC w/o regard
> >to both, it like trying to get a quality picture w/ your
> camera and
> >discussing only the lens aperture or shutter speed. It
> takes BOTH to get
> >the right exposure, not just one or the other.
> >
> >Some people do claim better performance when the FSB and
> memory are
> >asynchronous, but lots of empirical testings has shown
> that in most cases,
> >unless the differences are substantial, the benefits are
> so slight as to be
> >hardly worth the effort. Here's an example of a recent
> examination of
> >running two identical configurations, one 250/250 (sync),
> vs. 250/200
> >(async):
> >
> >http://www.cluboverclocker.com/reviews/memory/corsair/xms3
> 200LLPRO/index.htm
> >
> >The benefits of 250/250 (PC4000) vs. 250/200 (PC3200) are
> so slight, who
> >could be bothered. Of course, the fact that latencies
> are a real problem at
> >these speeds contributes significantly to the lack of
> performance gains,
> >most PC2700 and PC2100 could probably run w/ closer
> timings, and so there
> >might be a slightly better case for the PC2700 in an
> asynchronous
> >configuration.
> >
> >Of course, some people over-buy memory (PC2700 when
> PC2100 will do) if they
> >intend to overclock the CPU FSB. This way, as they
> increase the CPU FSB,
> >the memory remains underclocked and doesn't become a
> bottleneck to
> >successful OC'ing of the FSB. Whether this is a concern
> or interest to you
> >is beyond this discussion, I just want to complete the
> picture, since there
> >are people who *do* buy faster memory for such reasons.
> >
> >As an aside, anytime you *mix* memory speeds (PC2700 and
> PC2100), it's
> >doubtful your motherboard will support the faster of the
> two by default (By
> >SPD). If it will support the faster speed of the two,
> and that's iffy, it
> >would only happen if you *manually* set the Memory Freq
> (and timings) in the
> >BIOS. Most any motherboard is going to default to the
> most conservative
> >configuration.
> >
> >But for most mere mortals, you match the CPU FSB and
> memory speed for
> >typical, normal, everyday operations.
> >
> >HTH
> >
> >Jim
> >
> >
> >"Waheeba" <Waheeba@aol.com> wrote in message
> >news:OSe%23HF8JEHA.952@TK2MSFTNGP12.phx.gbl...
> >> Thank you Jim and Quang.
> >>
> >> Would I really notice the difference between 2700 and
> 2100 anyway? Would I
> >> notice the difference in gaming or music editing?
> >>
> >> "Jim" <null@null.com> wrote in message
> >> news:xCxhc.25914$Yf6.9141@fed1read07...
> >> > Memory can only run at ONE speed and set of timings,
> therefore, the
> >lesser
> >> > of the two is always used to configure it. In this
> case, PC2100 (266MHz
> >> > DDR) runs slower than PC2700 (333MHz DDR). CPU-Z
> (I'm not familiar with
> >> it)
> >> > is most likely measuring *actual* runtime
> configuration, NOT the
> >> > specification (i.e., potential) of each module.
> >> >
> >> > HTH
> >> >
> >> > Jim
> >> >
>
>
> That's all very well and good, but you seem to have over-
> looked what the physicists and RAM designers call the
> "Henderson Factor" (vis-a-vis the inverse squares rule,
> of course).
 
G

Guest

Guest
Archived from groups: microsoft.public.windowsxp.hardware (More info?)

Ok. I have a Pentium 4, 1.8ghz and this is the info the CPU-z program gave
me.

FSB = 100.9mhz
Bus Speed = 403.6mhz

Then under a section called Timing, it says:
FSB: DRAM is 3:4
Frequency is 134.5

What do you think about this? Any suggestions?



"Jim" <null@null.com> wrote in message
news:QOyhc.26366$Yf6.14892@fed1read07...
> Ideally, it's more important and advantageous that the CPU FSB and memory
> run at the same speed. IOW, the FSB and memory "clocks" are synchronous.
> Since you didn't mention anything about your processor, it's hard to be
> precise here, I can only speak in general terms.
>
> Suppose you have an Athlon XP 2600 "Barton" 333MHz (DDR), that's 166MHz
> "clock". If you installed only PC2100, which is 133MHz "clock", then
> (assuming no overclocking) you'd be running the CPU FSB @ 166MHz, but the
> memory at 133MHz, i.e., "asynchronously". As a general rule, computers
run
> more efficiently when the FSB and memory clocks are "in sync". The
Barton
> would be better matched to the PC2700 (166MHz "clock"). On the other
hand,
> suppose you had an Intel P4 533MHz (133MHz "clock") CPU. In this case,
the
> PC2100 (133MHz "clock") would be a better match, since it runs
synchronously
> w/ the CPU FSB.
>
> That's why you really can't discuss memory selection outside the context
of
> the CPU (or more precisely, the FSB). There's a relationship between the
> CPU FSB and memory that's relevant too, not just the pure speed potential
of
> the memory. Trying to discuss the best configuration of your PC w/o
regard
> to both, it like trying to get a quality picture w/ your camera and
> discussing only the lens aperture or shutter speed. It takes BOTH to get
> the right exposure, not just one or the other.
>
> Some people do claim better performance when the FSB and memory are
> asynchronous, but lots of empirical testings has shown that in most cases,
> unless the differences are substantial, the benefits are so slight as to
be
> hardly worth the effort. Here's an example of a recent examination of
> running two identical configurations, one 250/250 (sync), vs. 250/200
> (async):
>
>
http://www.cluboverclocker.com/reviews/memory/corsair/xms3200LLPRO/index.htm
>
> The benefits of 250/250 (PC4000) vs. 250/200 (PC3200) are so slight, who
> could be bothered. Of course, the fact that latencies are a real problem
at
> these speeds contributes significantly to the lack of performance gains,
> most PC2700 and PC2100 could probably run w/ closer timings, and so there
> might be a slightly better case for the PC2700 in an asynchronous
> configuration.
>
> Of course, some people over-buy memory (PC2700 when PC2100 will do) if
they
> intend to overclock the CPU FSB. This way, as they increase the CPU FSB,
> the memory remains underclocked and doesn't become a bottleneck to
> successful OC'ing of the FSB. Whether this is a concern or interest to
you
> is beyond this discussion, I just want to complete the picture, since
there
> are people who *do* buy faster memory for such reasons.
>
> As an aside, anytime you *mix* memory speeds (PC2700 and PC2100), it's
> doubtful your motherboard will support the faster of the two by default
(By
> SPD). If it will support the faster speed of the two, and that's iffy, it
> would only happen if you *manually* set the Memory Freq (and timings) in
the
> BIOS. Most any motherboard is going to default to the most conservative
> configuration.
>
> But for most mere mortals, you match the CPU FSB and memory speed for
> typical, normal, everyday operations.
>
> HTH
>
> Jim
>
>
> "Waheeba" <Waheeba@aol.com> wrote in message
> news:OSe%23HF8JEHA.952@TK2MSFTNGP12.phx.gbl...
> > Thank you Jim and Quang.
> >
> > Would I really notice the difference between 2700 and 2100 anyway? Would
I
> > notice the difference in gaming or music editing?
> >
> > "Jim" <null@null.com> wrote in message
> > news:xCxhc.25914$Yf6.9141@fed1read07...
> > > Memory can only run at ONE speed and set of timings, therefore, the
> lesser
> > > of the two is always used to configure it. In this case, PC2100
(266MHz
> > > DDR) runs slower than PC2700 (333MHz DDR). CPU-Z (I'm not familiar
with
> > it)
> > > is most likely measuring *actual* runtime configuration, NOT the
> > > specification (i.e., potential) of each module.
> > >
> > > HTH
> > >
> > > Jim
> > >
> > >
> > > "Waheeba" <Waheeba@aol.com> wrote in message
> > > news:uSHTko7JEHA.3980@TK2MSFTNGP10.phx.gbl...
> > > > I have one ram module that is PC2100 and the other is PC2700. They
> both
> > > show
> > > > in system properties and in the free program called CPUz. I am
> wondering
> > > > though, is the 2700 module reading at the slower 2100 level?
> > > >
> > > >
> > >
> > >
> >
> >
>
>
 

Jim

Distinguished
Mar 31, 2004
2,444
0
19,780
Archived from groups: microsoft.public.windowsxp.hardware (More info?)

Your memory is currently running asynchronously to your FSB (hence why your
FSB/DRAM ratio is 3:4, 100 / 3 * 4 = 133). FSB is 100MHz (clock), while the
memory is 133MHz (clock), or PC2100 (133MHz clock x 2 (DDR) x 8 (bytes wide)
= 2100). So the fact you have another PC2700 module installed is causing
*that* module to be underclocked to match the PC2100, which is to be
expected.

You frankly have quite a mixed environment there. Your CPU is an older P4
w/ 100MHz (clock), while the most current generation of P4's are 200MHz
(clock). In between these two, we had a generation of P4's 133MHz (clock).
So you're back two generations. At the time of your P4's heyday, PC1600
would have been an ideal match, i.e., synchronous (100MHz clock x 2 (DDR) x
8 (bytes wide) = 1600). But the PC2100 does work, if asynchronously.

So everything is as expected, your system is configured properly, its
normative given the components you have and their limitations. If you
suspect that your PC2100 memory *might* be able to run a little faster,
something closer to PC2700, you could try overclocking it by specifying a
FSB/DRAM ratio of 3:5 (100 / 3 * 5 = 166). But that assumes several
factors. a) the PC2100 module can run @ PC2700 speed, b) that you have
some room for slight voltage changes should it proof necessary to help
stabilize the memory c) that's there even a 3:5 FSB/DRAM ratio available
(maybe not)!, d) your motherboard support 166MHz memory speed (possibly
could, but perhaps only w/ a faster CPU, e.g., 133MHz clock, w/ a 4:5
FSB/DRAM ratio, and e) you're willing to OC the memory w/ the risk of
damaging it if you don't know what you're doing!

IOW, everything is fine on your setup, but if you're looking to explore the
possibility of getting up to PC2700 memory speeds, you might have to OC, or
else swap another PC2700 for the current PC2100. Btw, even then, w/ that
100MHz P4 (this is really holding you back), you still need to verify that
your mobo supports the necessary FSB/DRAM ratio (3:5). The only other
option would be to OC the FSB too, in hopes of running the CPU @ 133MHz
(clock), two sticks of PC2700 @ 166MHz, using a FSB/DRAM ratio of 4:5 (133 /
4 * 5 = 166).

Totally confused at this point? ;) If you take nothing else away from this,
know all is OK on your config.

HTH

Jim



"Waheeba" <Waheeba@aol.com> wrote in message
news:OyM6mD$JEHA.2580@TK2MSFTNGP12.phx.gbl...
> Ok. I have a Pentium 4, 1.8ghz and this is the info the CPU-z program gave
> me.
>
> FSB = 100.9mhz
> Bus Speed = 403.6mhz
>
> Then under a section called Timing, it says:
> FSB: DRAM is 3:4
> Frequency is 134.5
>
> What do you think about this? Any suggestions?
>
>
>
> "Jim" <null@null.com> wrote in message
> news:QOyhc.26366$Yf6.14892@fed1read07...
> > Ideally, it's more important and advantageous that the CPU FSB and
memory
> > run at the same speed. IOW, the FSB and memory "clocks" are
synchronous.
> > Since you didn't mention anything about your processor, it's hard to be
> > precise here, I can only speak in general terms.
> >
> > Suppose you have an Athlon XP 2600 "Barton" 333MHz (DDR), that's 166MHz
> > "clock". If you installed only PC2100, which is 133MHz "clock", then
> > (assuming no overclocking) you'd be running the CPU FSB @ 166MHz, but
the
> > memory at 133MHz, i.e., "asynchronously". As a general rule, computers
> run
> > more efficiently when the FSB and memory clocks are "in sync". The
> Barton
> > would be better matched to the PC2700 (166MHz "clock"). On the other
> hand,
> > suppose you had an Intel P4 533MHz (133MHz "clock") CPU. In this case,
> the
> > PC2100 (133MHz "clock") would be a better match, since it runs
> synchronously
> > w/ the CPU FSB.
> >
> > That's why you really can't discuss memory selection outside the context
> of
> > the CPU (or more precisely, the FSB). There's a relationship between
the
> > CPU FSB and memory that's relevant too, not just the pure speed
potential
> of
> > the memory. Trying to discuss the best configuration of your PC w/o
> regard
> > to both, it like trying to get a quality picture w/ your camera and
> > discussing only the lens aperture or shutter speed. It takes BOTH to
get
> > the right exposure, not just one or the other.
> >
> > Some people do claim better performance when the FSB and memory are
> > asynchronous, but lots of empirical testings has shown that in most
cases,
> > unless the differences are substantial, the benefits are so slight as to
> be
> > hardly worth the effort. Here's an example of a recent examination of
> > running two identical configurations, one 250/250 (sync), vs. 250/200
> > (async):
> >
> >
>
http://www.cluboverclocker.com/reviews/memory/corsair/xms3200LLPRO/index.htm
> >
> > The benefits of 250/250 (PC4000) vs. 250/200 (PC3200) are so slight, who
> > could be bothered. Of course, the fact that latencies are a real
problem
> at
> > these speeds contributes significantly to the lack of performance gains,
> > most PC2700 and PC2100 could probably run w/ closer timings, and so
there
> > might be a slightly better case for the PC2700 in an asynchronous
> > configuration.
> >
> > Of course, some people over-buy memory (PC2700 when PC2100 will do) if
> they
> > intend to overclock the CPU FSB. This way, as they increase the CPU
FSB,
> > the memory remains underclocked and doesn't become a bottleneck to
> > successful OC'ing of the FSB. Whether this is a concern or interest to
> you
> > is beyond this discussion, I just want to complete the picture, since
> there
> > are people who *do* buy faster memory for such reasons.
> >
> > As an aside, anytime you *mix* memory speeds (PC2700 and PC2100), it's
> > doubtful your motherboard will support the faster of the two by default
> (By
> > SPD). If it will support the faster speed of the two, and that's iffy,
it
> > would only happen if you *manually* set the Memory Freq (and timings) in
> the
> > BIOS. Most any motherboard is going to default to the most conservative
> > configuration.
> >
> > But for most mere mortals, you match the CPU FSB and memory speed for
> > typical, normal, everyday operations.
> >
> > HTH
> >
> > Jim
> >
> >
> > "Waheeba" <Waheeba@aol.com> wrote in message
> > news:OSe%23HF8JEHA.952@TK2MSFTNGP12.phx.gbl...
> > > Thank you Jim and Quang.
> > >
> > > Would I really notice the difference between 2700 and 2100 anyway?
Would
> I
> > > notice the difference in gaming or music editing?
> > >
> > > "Jim" <null@null.com> wrote in message
> > > news:xCxhc.25914$Yf6.9141@fed1read07...
> > > > Memory can only run at ONE speed and set of timings, therefore, the
> > lesser
> > > > of the two is always used to configure it. In this case, PC2100
> (266MHz
> > > > DDR) runs slower than PC2700 (333MHz DDR). CPU-Z (I'm not familiar
> with
> > > it)
> > > > is most likely measuring *actual* runtime configuration, NOT the
> > > > specification (i.e., potential) of each module.
> > > >
> > > > HTH
> > > >
> > > > Jim
> > > >
> > > >
> > > > "Waheeba" <Waheeba@aol.com> wrote in message
> > > > news:uSHTko7JEHA.3980@TK2MSFTNGP10.phx.gbl...
> > > > > I have one ram module that is PC2100 and the other is PC2700. They
> > both
> > > > show
> > > > > in system properties and in the free program called CPUz. I am
> > wondering
> > > > > though, is the 2700 module reading at the slower 2100 level?
> > > > >
> > > > >
> > > >
> > > >
> > >
> > >
> >
> >
>
>
 
G

Guest

Guest
Archived from groups: microsoft.public.windowsxp.hardware (More info?)

>-----Original Message-----
>If you're referring to the camera analogy (aperture size
vs. shutter speed),
>I wasn't characterizing the *nature* of the relationship,
but the fact there
>*is* a relationship (something that many people don't
even realize exists).
>Indeed, the inverse squares rule does not apply to the
topic of FBS and
>memory (not everything in nature and physics does). If I
somehow left the
>impression this was the case, I apologize. I was merely
indicating "it
>takes two to Tango". Please, no lessons in dance theory.
>
>Frankly, only one in a thousand who stroll through this
NG will "catch" it,
>or even understand what you and I are discussing, which
reminds me of
>another rule, "know your audience", but don't feel too
bad, a common error
>of the elite.
>
>Jim
>
>

I don't feel bad at all. In fact, I feel great. I fell
asleep midway through your overly-verbose and self-absorbed
original response. "Henderson effect" and reference to
the inverse squares rule were a trap that you obligingly
stepped right into--a common error of the pseudo-elite.