CAS 2 vs. CAS 3 speed differences in RAM with Athlon 64?

[Yousuf Khan]

Great post, but shouldn't that be 20ns?

My bad, you're right 20ns.

Yousuf Khan

 

[Yousuf Khan]

Okay, I don't want to throw a monkey wrench in things here, but
there's also CAS2.5, which is what I got for my new Athlon 64. It
was the same price as the CAS3 when I bought it, so I thought it
wouldn't hurt.

Yes, actually CAS2.5 was the original "high-performance" DDR-RAM in the
early days of DDR's introduction, with CAS3 being the economical version.
Eventually they got that down to CAS2, and CAS2.5 is now the intermediate
performance level.

I don't know exactly what it means, but I'd imagine it's somewhere
between CL2 and CL3 in performance, meaning you'd be lucky if the
performance difference in any real-world situation wasn't lost in the
noise.

It's all gotta do with the actual clock rate vs. the DDR clock rate. For
example, DDR333 is being fed by a 166Mhz clock, which gets doubled to 333Mhz
internally. So something that has a latency of 5 cycles in the DDR clock
rate, would be equivalent to having a latency of 2.5 cycles in the actual
clock rate.

Yousuf Khan

 

[ANTant]

Thanks to everyone who posted comments even if they were technical that
went over my head. It looks like I will be getting CAS 3 on Thursdya to
save money.

Hello.

I am going to upgrade my old gaming box to an Athlon 64 system. I
noticed the memory CAS speeds. Does CAS 2 (faster) and CAS 3 (slower)
make any big speed differences for 1 GB of PC3200 RAM on a Socket 754
ATX motherboard with an AMD Athlon 64 3000+/2.0 Ghz to 3400+ 2.4 Ghz
(both with 512 KB socket and 754 CPU)? CAS 2 is expensive so... I am
wondering if getting 3 is really worth the price. I am mainly gaming,
watching movies, using the Internet, etc. Gaming is the big one to
note.

You can see my current Athlon XP 2200+ system at
http://alpha.zimage.com/~ant/antfarm/about/computers.txt ... I will be
replacing the motherboard (ASUS K8V SE Deluxe or a MSI K8T Neo F...),
CPU, RAM, and sound card (getting an Audigy 2).

Thank you in advance.

--
"Oh, look what Kyle got me, it's a red Mega... Ants in the pants? Ants
in the pants?! Ants in the Pants?!! ..." --Eric Cartman in South Park's
Damien Episode (Season 1; Episode 8)
/\___/\
/ /\ /\ \ Phillip (Ant) @ http://antfarm.ma.cx & http://aqfl.net
| |o o| | E-mail: (e-mail address removed) or (e-mail address removed)
\ _ / Nuke ANT from e-mail address if your e-mail was returned.
( )

 

[Thundercracker]

This was assuming PC3200 like Kingston brand. I guess I will go with CAS
3 then to save money and since the speed difference isn't that big. I
have until Wednesday night to decide.

I got a stick of Corsair CAS 2.5 for the same price as other manufacturers' CAS3.

 

[Yousuf Khan]

Of all the explanations I have read concerning memory timings, that
is far and away the most concise and practical one I have ever had
the pleasure to read. Thanks, Yousuf!

Thanks, but it's also a bit of an oversimplification. I ignored two other
latencies that are also used in determining timings. One is RCD, i.e. the
"RAS to CAS Delay", which is the amount of time delay needed between
activating a row strobe and then activating a column strobe. The other is
RP, which is "RAS Precharge", which is the time needed to switch from one
row to another row.

The following are the methods for fully describing the latencies inside a
DRAM. There's a 3 item style (used mostly with SDR SDRAM), and there is also
a 4 item style (used mostly with DDR SDRAM):

3-item: t_CL, t_RCD, t_RP (e.g. 3-2-2)
4-item: t_CL, t_RCD, t_RP, t_RAS (e.g. 2.5-2-2-7)

where,
t_CL: CAS latency, the time from when the column strobe is activated to when
you get data
t_RCD: RAS-CAS delay, the time from when you first activate a row strobe to
the time you activate a column strobe
t_RP: RAS Precharge time, the time delay between closing one row and opening
the next row
t_RAS: RAS activation length, amount of time that a row can remain active
before you have to recharge it (this one is better if its longer)

I got most of this information from the following website:

http://www.ccs.neu.edu/course/com3200/parent/NOTES/DDR.html

It's a somewhat technical website, not really for laymen, but pretty
detailed though. It also describes the various DDR ram commands available.

Yousuf Khan

 

[George Macdonald]

Due to the mainly linear nature of memory accesses, you will only need to
activate the CAS only once in awhile and keep it open, while you cycle
through the memory using the RAS. This is known as a burst transfer. Really,
RAS is the more important measure of the speed of RAM, because it's done so
much more often.

Uhh, you got that backwards Yousuf: RAS happens on the Activate command and
CAS on the Read/Write command. A burst is a max 8 chip-words, i.e. 64bytes
in current systems, though CAS cycles can be pipelined... if they are to an
open page.

But RAS timings don't really change from module to module,
it's always stuck at about 3 cycles in DDR-RAM. So the only timing that
might be slightly better or worse is the CAS timing, so that's what they
advertise. Really, CAS timing is sort of a measurement of what kind of a
delay you can expect for a burst of data to begin coming to you after you've
told it to start, since CAS is one of the first activities done.

Now a DDR-200 (aka PC1600) runs at 100Mhz, DDR-266 (PC2100) runs at 133Mhz,
DDR-333 (PC2700) runs at 166Mhz, and DDR-400 (PC3200) runs at 200Mhz. CL3
would mean 3 cycles for CAS, and CL2 would mean 2 cycles. So a DDR-200 CL2
module would have a CAS latency time of 2/(100) = 50ns; while a DDR-400 CL3
would have a CAS latency of 3/(200) = 15ns. So a DDR-400 even with a CL3 has
considerably lower latency than a CL2 DDR-200.

So yes, CAS latency does make a small difference, but we're only talking
about nanoseconds at the beginning of a burst.

Then again, chips/modules with lower CAS timing capability tend to have
better performance on other parameters. Whether they're worth the extra,
depends on the size of "extra".

Rgds, George Macdonald

"Just because they're paranoid doesn't mean you're not psychotic" - Who, me??

 

[George Macdonald]

Thanks to everyone who posted comments even if they were technical that
went over my head. It looks like I will be getting CAS 3 on Thursdya to
save money.

Whatever mbrd & memory you get check for compatibility. If the MSI mbrd,
the Taiwan site, http://www.msi.com.tw/index.htm, has a Recommend List for
several add-ins including memory modules for their mbrds.

FWIW I've found their recommended Kingmax memory to work well in the K8T
Neo boards for a reasonable price. I got the Kingmax PC3500 CL=2.5 for a
coupla $$ extra. Of course after you get it, beat the hell out of it with
memtest86 before you even think about installing an OS.

Rgds, George Macdonald

"Just because they're paranoid doesn't mean you're not psychotic" - Who, me??

 

[Mirek Fidler]

FWIW I've found their recommended Kingmax memory to work well in the
K8T

Neo boards for a reasonable price. I got the Kingmax PC3500 CL=2.5 for a
coupla $$ extra. Of course after you get it, beat the hell out of it with
memtest86 before you even think about installing an OS.

BTW, are you using one or two sticks?

I am planning to buy two sticks of Kingmax PC3500 cl 2.5 512MB (for 1GB
total), but I am a little bit afraid whether it will work. I plan to use
it with K8T800 (or K8M800) chipset, either SK83G, MSI K8TM / MM or DFI
K8M800.

Mirek

 

[assaarpa]

it would only run at 3-3-4. This made a noticible difference in
performance.

Windows didn't snap open as fast, and games ran slower.

Bullshit.

 

[assaarpa]

I have all mem settings in bios at most agressive possible and it runs

better then with slower ram.
This is ofcourse my opinion,can't back it up with figures.

Psychological effect can be very dramatic and YMMV obviously, LOL.

 

[George Macdonald]

BTW, are you using one or two sticks?

I am planning to buy two sticks of Kingmax PC3500 cl 2.5 512MB (for 1GB
total), but I am a little bit afraid whether it will work. I plan to use
it with K8T800 (or K8M800) chipset, either SK83G, MSI K8TM / MM or DFI
K8M800.

Interesting question. I only have the one 512MB DIMM in each system. I
think maybe I'll try an experiement with two... just to see.

Rgds, George Macdonald

"Just because they're paranoid doesn't mean you're not psychotic" - Who, me??

 

[turtle]

Thanks to everyone who posted comments even if they were technical that
went over my head. It looks like I will be getting CAS 3 on Thursdya to
save money.

Great, now what about overclocking the FSB ?

AS I understand it, faster FSB at higher CAS blows away expensive CAS2
at low FSB.

How fast you plan to run that Nforce3 250Gb motherboard ?

 

[Mirek Fidler]

Interesting question. I only have the one 512MB DIMM in each system. I
think maybe I'll try an experiement with two... just to see.

Please! Thanks!

Mirek

 

[George Macdonald]

Please! Thanks!

Yup I transferred one of the Kingmax PC3500 CL=2.5 512MB DIMMs over and ran
with 1GB. This was in a MSI K8T Neo-FSR with an Athlon64 3200+
(Clawhammer), i.e. 1MB L2 cache, and all default BIOS settings for memory
timings, voltages and clocking etc. Note I did *not* use the MSI BIOS
Setup option for "Performance Settings", which AFAIK does some mild
overclock but I don't know of what... so strictly standard setup.

I ran memtest86 V3.1a for 5 full passes with default tests, which took
about 1hr 40mins and got no errors... about as long as I'd normally let it
run before assuming a new system was good enough for an OS install. I also
booted into Windows XP just to be sure it would work OK but did no real
testing there.

I'll add that I would normally buy Crucial memory given a free choice but
MSI's recommendations there seemed to be for Micron part numbers, so hard
to correlate... plus I think the "Crucial/Micron" part mentioned was for
ECC. I looked at all their other recommendations, did a bit of searching
and came up with the fact that the Kingmax PC3200 CL=2.5 was said to be
fine with two 512MB DIMMs at DDR400 and some others were only good with one
512MB DIMM at DDR400. and the Kingmax was well priced. For an extra $4. or
so the PC3500 seemed a good bet with *maybe* a little extra headroom...
plus that nice red PCB matches the red of the MSI mbrd.

That is one sweet system - good luck with it.

Rgds, George Macdonald

"Just because they're paranoid doesn't mean you're not psychotic" - Who, me??

 

[Mirek Fidler]

Yup I transferred one of the Kingmax PC3500 CL=2.5 512MB DIMMs over and ran
with 1GB. This was in a MSI K8T Neo-FSR with an Athlon64 3200+
(Clawhammer), i.e. 1MB L2 cache, and all default BIOS settings for memory
timings, voltages and clocking etc. Note I did *not* use the MSI BIOS
Setup option for "Performance Settings", which AFAIK does some mild
overclock but I don't know of what... so strictly standard setup.

I ran memtest86 V3.1a for 5 full passes with default tests, which took
about 1hr 40mins and got no errors... about as long as I'd normally let it
run before assuming a new system was good enough for an OS install. I also
booted into Windows XP just to be sure it would work OK but did no real
testing there.

I'll add that I would normally buy Crucial memory given a free choice but
MSI's recommendations there seemed to be for Micron part numbers, so hard
to correlate... plus I think the "Crucial/Micron" part mentioned was for
ECC. I looked at all their other recommendations, did a bit of searching
and came up with the fact that the Kingmax PC3200 CL=2.5 was said to be
fine with two 512MB DIMMs at DDR400 and some others were only good with one
512MB DIMM at DDR400. and the Kingmax was well priced. For an extra $4. or
so the PC3500 seemed a good bet with *maybe* a little extra headroom...
plus that nice red PCB matches the red of the MSI mbrd.

That is one sweet system - good luck with it.

Thank you! Looks like I will buy kingmax

Mirek

 

[Evgenij Barsukov]

Interesting question. I only have the one 512MB DIMM in each system. I
think maybe I'll try an experiement with two... just to see.

I recently had experience with OEM 3200 512MB memory (On Biostar nForce3 mobo, AMD64 2800)
hat turned out to use Micron CAS2 ICs - it would run with 3-3-3-8 settings only,
but stable with one DIMM. However, any attempt to run system
with second same manufacturer DIMM failed regardless of settings.
Replacing one DIMM did not help, so I am replacing the whole memory now hopefuly
to something better.

It looks like two DIMMs are definetely more finiky than one, even despite this
Biostar mobo has got into best 3 from 10 in reviews of memory compatibility.

Regards,
Evgenij

--

__________________________________________________
*science&fiction*free programs*fine art*phylosophy:
http://sudy_zhenja.tripod.com
----------remove hate_spam to answer--------------

 

[The little lost angel]

I ran into this problem with my last system. I bought one stick of ram, that
I was able to run at 2-2-3. When I added ram, I bought a similar type, but
it would only run at 3-3-4. This made a noticible difference in performance.
Windows didn't snap open as fast, and games ran slower.

I really think it's a placebo kind of effect, you know you're using
RAM with slower spec and therefore feel slower even though in reality
there really isn't much difference. Not too long again, I tried
fooling around with my processor and memory settings to get a bit more
performance out of my aging system for Photoshop work and such. But
apart from the numbers reported by synthetic benckmarks, I seriously
don't feel any noticeable difference between running memory at the
fastest 2-a-b-c vs a slower 3-x-y-z.

--
L.Angel: I'm looking for web design work.
If you need basic to med complexity webpages at affordable rates, email me
Standard HTML, SHTML, MySQL + PHP or ASP, Javascript.
If you really want, FrontPage & DreamWeaver too.
But keep in mind you pay extra bandwidth for their bloated code

 

[DDStech]

No, there is deffinately a difference. Like another poster stated, The
numbers can't quantify it, but, there is a noticable difference. You can't
benchmark every aspect of every usage of windows performace. Perhaps,
because opening windows is only a small burp of data, that the CAS2 pops it
open much faster than what a prolonged 512mb data stream would register
(sissoft sandra).

Additionally, my videocards have not always been on the cutting edge, so
memory performance is notable here. Perhaps the frame rate might rise to
some extent, but, there is a noticable smoothness about CAS2 over CAS3. You
know how you used to get those "pauses" as memory was swapped from system to
video, well, that was almost eliviated if not seriously reduced. I would
assume these days, that is not so much of an issue. However, these days
video cards come with 128 to 512mb's of onboard memory. I assume this is to
eliminate these lag moments.

Perhaps some people are just more sensitive to these timming differences
than others?

 

[Alan Walpool]

< - snip - >

DDStech> Perhaps some people are just more sensitive to these timming
DDStech> differences than others?

The main advantage of lower CAS numbers is your system maybe able to
overclock and/or increase your memory timing numbers.

We need human benchmarks because opinions are more accurate than
objective data. Must be the time of the year the political season.
It's amazing but that it what is so nice about usenet. ;-)). Now we
need to discuss how slow ECC memory really is and how ECC memory is a
real waste of money and we have finished a nice discussion about the
usual myths of memory, again.

There is a silver lining here what is I don't have a clue. Oh no!

Feel the memory Luke!

Later ;-)

Alan

 

[Matt Anderson]

Mhz means Megahertz, which refers to the number of clock ticks per second.
The prefix "mega" is 1 million, so 266 mhz means 266,000,000 clock ticks.
So CAS 2 refers to 2 clock ticks, whereas 2mhz refers to 2 million clock
ticks. CAS 2 takes 2 clock ticks to refresh, whereas CAS 3 takes 3 ticks.
Your FSB runs at a rate of 266 million ticks per second.

That's the difference between ticks and mhz. The mega just refers to
millions of ticks. It's friday and my brain's shutting off, so the
explanation wasn't the clearest.