Gigabyte GA-P35 DS3P - RAM?

bornfree

Gigabyte GA-P35 DS3P - What RAM would you recommend?

Someone told me this motherboard is "really fast", so it should have
fast ram. What do you think?

Augustus

"bornfree" <justyouandme@xemaps.com> wrote in message
news:723513fa-3bef-4365-ae98-78efdd2671c1@i12g2000prf.googlegroups.com...
> Gigabyte GA-P35 DS3P - What RAM would you recommend?
>
> Someone told me this motherboard is "really fast", so it should have
> fast ram. What do you think?

I think you should read the manual and purchase RAM according to your
planned use which may or may not include overclocking. Gigabytes website has
a detailed PDF on memory support brands and types. It's not a DDR3 board or
a 1333Mhz board. It's a DDR2 1066Mhz board. So I wouldn't recomend you go
out and buy "really fast" RAM without doing some checking.

http://www.gigabyte.com.tw/Products/...ProductID=2533

Phil Weldon

'bornfree' wrote:
> Gigabyte GA-P35 DS3P - What RAM would you recommend?
>
> Someone told me this motherboard is "really fast", so it should have
> fast ram. What do you think?
_____

You should select RAM based on the FSB speed for the CPU. And on whether or
not you intend to overclock by raising the FSB speed. And whether or not
you wish to run the Memory clock synched to the CPU clock. And whether or
not you are willing to spend the money for the diminishing performance
returns as you increase the RAM speed. And THAT is affected by the size of
the L2 cache, the applications you use, and the number of cores in the CPU.
FINALLY, after all those decisions, THEN you want to make sure the
motherboard can handle the Memory clock speed of the RAM you select.

Oh, yes, also you must select the right TYPE of RAM for the motherboard DDR2
or DDR3.

And every thing 'Augustus' posted.

Plus consider that a 32-bit Windows operating system can't use much more
than 2 GBytes of memory, especially if you one or two display adapters with
large memories (each byte of display adapter memory makes one byte of
motherboard memory between 2 GBytes and 4 GBytes unusable.

Phil Weldon


"bornfree" <justyouandme@xemaps.com> wrote in message
news:723513fa-3bef-4365-ae98-78efdd2671c1@i12g2000prf.googlegroups.com...
> Gigabyte GA-P35 DS3P - What RAM would you recommend?
>
> Someone told me this motherboard is "really fast", so it should have
> fast ram. What do you think?

Elcaro Nosille

Phil Weldon schrieb:

> You should select RAM based on the FSB speed for the CPU. ...

Not with today's chipsets. I've testet different ratios with my
Core II Extreme (non-locked multiplier) and I found that a awry
multiplier of 1:1.2 (1333/333MHz FSB and 800MHz got the highest
memory-thgoughput and the lowest memory-latency.

Phil Weldon

'Elcaro Nosille' wrote:
> Not with today's chipsets. I've testet different ratios with my
> Core II Extreme (non-locked multiplier) and I found that a awry
> multiplier of 1:1.2 (1333/333MHz FSB and 800MHz got the highest
> memory-thgoughput and the lowest memory-latency.
_____

My post stated "You should select RAM based on the FSB speed for the CPU."
My statement is correct (and a little obvious.) I am not quite sure WHAT
your statement means; did you overclock? Did you use an increased CPU
clock multiplier and a 1333 MHz stock FSB (333 MHz CPU clock)? Are you
using DDR2-800 memory? What is the rating of your selected memory? CAN
your selected memory run at 1:1 or even 2:1 Memory clock : CPU clock ratios?
Is a large increase in memory parameter timing delays necessary when your
selected memory is used at speeds above DDR2-800 equivalent?

In other words, DID you select memory based on the FSB speed for the CPU?
Perhaps your selection of CPU clock limits the memory through-put. After
all, a faster memory clock speed, even with higher memory parameter timing
delays can give greater memory through-put IF "you select RAM based on the
FSB speed for the CPU; using a 1333 MHz FSB and a 2:1 Memory clock : CPU
clock ratio will offset the disadvantage of quite a bit higher memory
parameter timing delays. If you had selected DDR2-1333 (PC10666) and used a
2:1 Memory clock : CPU clock ratio you would very likely be seeing higher
memory through-put (though perhaps not enough of an increase to justify the
added cost.) And as I posted, and you failed to fully quote:

"You should select RAM based on the FSB speed for the CPU. And on whether
or
not you intend to overclock by raising the FSB speed. And whether or not
you wish to run the Memory clock synched to the CPU clock. And whether or
not you are willing to spend the money for the diminishing performance
returns as you increase the RAM speed. And THAT is affected by the size of
the L2 cache, the applications you use, and the number of cores in the CPU.
FINALLY, after all those decisions, THEN you want to make sure the
motherboard can handle the Memory clock speed of the RAM you select."

and

"Oh, yes, also you must select the right TYPE of RAM for the motherboard
DDR2
or DDR3."

and

"Plus consider that a 32-bit Windows operating system can't use much more
than 2 GBytes of memory, especially if you one or two display adapters with
large memories (each byte of display adapter memory makes one byte of
motherboard memory between 2 GBytes and 4 GBytes unusable.)"


At any rate your comment "Not with today's chipsets" misses the point (and
is incorrect.)

Phil Weldon

"Elcaro Nosille" <Elcaro.Nosille@arcor.de> wrote in message
news:47ca09b5$0$21923$9b4e6d93@newsspool2.arcor-online.net...
> Phil Weldon schrieb:
>
>> You should select RAM based on the FSB speed for the CPU. ...
>
> Not with today's chipsets. I've testet different ratios with my
> Core II Extreme (non-locked multiplier) and I found that a awry
> multiplier of 1:1.2 (1333/333MHz FSB and 800MHz got the highest
> memory-thgoughput and the lowest memory-latency.

Elcaro Nosille

Phil Weldon schrieb:

> My post stated "You should select RAM based on the FSB speed for the
> CPU." My statement is correct (and a little obvious.) I am not quite
> sure WHAT your statement means; did you overclock? Did you use an
> increased CPU clock multiplier and a 1333 MHz stock FSB (333 MHz CPU
> clock)? Are you using DDR2-800 memory? What is the rating of your
> selected memory? CAN your selected memory run at 1:1 or even 2:1
> Memory clock : CPU clock ratios? Is a large increase in memory
> parameter timing delays necessary when your selected memory is
> used at speeds above DDR2-800 equivalent?

I've got a Gigabyte-board with an Intel P35 chipset and I get a higher
throughput and a lower latency in memory-bandwith and lower latency
when using a 1:1.2 multiplier (1333/333MHz FSB and 800/400MHz DDR2
-SDRAM clocking) than when using a 1:1 multiplier.
Your recommendation of 1:N-ratios comes from times when the chipsets
weren't as efficient as today. Today, you can forget this rule.

Phil Weldon

'Elcaro Nosille' wrote:
> I've got a Gigabyte-board with an Intel P35 chipset and I get a higher
> throughput and a lower latency in memory-bandwith and lower latency
> when using a 1:1.2 multiplier (1333/333MHz FSB and 800/400MHz DDR2
> -SDRAM clocking) than when using a 1:1 multiplier.
> Your recommendation of 1:N-ratios comes from times when the chipsets
> weren't as efficient as today.
_____

Today, you can forget this rule.Just to clarify, when you post 1:1.2 do you
mean a ratio of 1 to 2 (1/2)? [This is a somewhat murky area since various
motherboard manufacturers use two different ratios; Memory Clock : CPU
clock and CPU clock : memory clock.

Your reply still fails to quote the my entire post. Do you mean to indicate
that we agree on everything but the Memory clock : CPU clock ratio?

Motherboards based on the nVidia 680i series chipset DO show increased
memory through-put when operated at higher Memory clock : CPU clock ratios
given tight enough memory parameter delays. Though as stated, the increased
performance may not be worth the added cost. Which is why

"You should select RAM based on the FSB speed for the CPU. And on whether
or
not you intend to overclock by raising the FSB speed. And whether or not
you wish to run the Memory clock synched to the CPU clock. And whether or
not you are willing to spend the money for the diminishing performance
returns as you increase the RAM speed. And THAT is affected by the size of
the L2 cache, the applications you use, and the number of cores in the CPU.
FINALLY, after all those decisions, THEN you want to make sure the
motherboard can handle the Memory clock speed of the RAM you select."

Phil Weldon

"Elcaro Nosille" <Elcaro.Nosille@arcor.de> wrote in message
news:47ca1ff4$0$14402$9b4e6d93@newsspool3.arcor-online.net...
> Phil Weldon schrieb:
>
>> My post stated "You should select RAM based on the FSB speed for the
>> CPU." My statement is correct (and a little obvious.) I am not quite
>> sure WHAT your statement means; did you overclock? Did you use an
>> increased CPU clock multiplier and a 1333 MHz stock FSB (333 MHz CPU
>> clock)? Are you using DDR2-800 memory? What is the rating of your
>> selected memory? CAN your selected memory run at 1:1 or even 2:1
>> Memory clock : CPU clock ratios? Is a large increase in memory
>> parameter timing delays necessary when your selected memory is
>> used at speeds above DDR2-800 equivalent?
>
> I've got a Gigabyte-board with an Intel P35 chipset and I get a higher
> throughput and a lower latency in memory-bandwith and lower latency
> when using a 1:1.2 multiplier (1333/333MHz FSB and 800/400MHz DDR2
> -SDRAM clocking) than when using a 1:1 multiplier.
> Your recommendation of 1:N-ratios comes from times when the chipsets
> weren't as efficient as today. Today, you can forget this rule.

Elcaro Nosille

Phil Weldon schrieb:

> Today, you can forget this rule.Just to clarify, when
> you post 1:1.2 do you mean a ratio of 1 to 2 (1/2)?

No, I mean 1 : 1.2, i.e. the clock-rate of the DDR2-memory
(400MHz) is 1.2 times higher than the clock-rate of the
FSB (333MHz).

> Motherboards based on the nVidia 680i series chipset DO show
> increased memory through-put when operated at higher Memory
> clock : CPU clock ratios given tight enough memory parameter
> delays.

So it seems Intel did a better Job on that. I clocked my Core
II Extreme at a FSB of 400/1600MHz (core-clock 3.2GHz) so that
the memory-FSB-ratio is 1:1 and the memory-throughput was sligh-
tly lower than at 333MHz (equal memory-clocking of 400MHz aka
800MHz DDR).

And there's another thing to mention: latency is a more
important parameter than throughput in real-world-apps.

Phil Weldon

'Elcaro Nosille' wrote:
> No, I mean 1 : 1.2, i.e. the clock-rate of the DDR2-memory
> (400MHz) is 1.2 times higher than the clock-rate of the
> FSB (333MHz).
>
>> Motherboards based on the nVidia 680i series chipset DO show
>> increased memory through-put when operated at higher Memory
>> clock : CPU clock ratios given tight enough memory parameter
>> delays.
>
> So it seems Intel did a better Job on that. I clocked my Core
> II Extreme at a FSB of 400/1600MHz (core-clock 3.2GHz) so that
> the memory-FSB-ratio is 1:1 and the memory-throughput was sligh-
> tly lower than at 333MHz (equal memory-clocking of 400MHz aka
> 800MHz DDR).
>
> And there's another thing to mention: latency is a more
> important parameter than throughput in real-world-apps.
_____

Thanks for the clarification. The [,] [.] difference is sometimes
confusing. Especially when combined with differing nomenclature for the
ratio between CPU clock and memory clock (likely aggravated by meddling
sales department meddling in engineering department specifications and
performance data.

As for your
> And there's another thing to mention: latency is a more
> important parameter than throughput in real-world-apps

the larger the L2 cache and the greater the L2 cache associativity the less
important main memory latency AND through-put for 'real world' performance.
Given optimization for L2 cache size, main memory through-put AND latency
become less important for 'real-world' performance. For 'real world'
performance, memory though-put IS dependent on latency AND memory clock
speed. After all, we use computer SYSTEMS (see 'Amdahl's Rule of Thumb',
also known as 'Amdahl's Other Law': "one byte of memory and one byte of I/O
are required for each instruction per second supported by a computer"; part
of

"3. Amdahl's system balance rules
Gene Amdahl is famous for many rules of thumb.
For data engineering, there are four famous ones [6]:
10. Amdahl's parallelism law: If a computation has a
serial component S and a parallel component P, then
the maximum speedup is (S+P)/S.
11. Amdahl's balanced system law: A system needs a bit
of IO per second for each instruction per second:
about 8 MIPS per MBps.
12. Amdahl's memory law: ? ? 1: that is, in a balanced
system the MB/MIPS ratio, called alpha (? ), is 1.
13. Amdahl's IO law: Programs do one IO per 50,000
instructions."
http://research.microsoft.com/~gray/...ngineering.pdf .
Also see

Petascale Computational Systems:
Balanced CyberInfrastructure in a Data-Centric World
Gordon Bell1, Jim Gray1 and Alex Szalay2
1. Microsoft Research
2. The Johns Hopkins University

GBell@Microsoft.com, Gray@Microsoft.com, Szalay@
http://research.microsoft.com/~gray/...%20systems.pdf
..)

Phil Weldon




"Elcaro Nosille" Elcaro.Nosille@arcor.de wrote in message
news:47cae5d9$0$25508$9b4e6d93@newsspool1.arcor-online.net...
> Phil Weldon schrieb:
>
>> Today, you can forget this rule.Just to clarify, when
>> you post 1:1.2 do you mean a ratio of 1 to 2 (1/2)?
>
> No, I mean 1 : 1.2, i.e. the clock-rate of the DDR2-memory
> (400MHz) is 1.2 times higher than the clock-rate of the
> FSB (333MHz).
>
>> Motherboards based on the nVidia 680i series chipset DO show
>> increased memory through-put when operated at higher Memory
>> clock : CPU clock ratios given tight enough memory parameter
>> delays.
>
> So it seems Intel did a better Job on that. I clocked my Core
> II Extreme at a FSB of 400/1600MHz (core-clock 3.2GHz) so that
> the memory-FSB-ratio is 1:1 and the memory-throughput was sligh-
> tly lower than at 333MHz (equal memory-clocking of 400MHz aka
> 800MHz DDR).
>
> And there's another thing to mention: latency is a more
> important parameter than throughput in real-world-apps.