• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

1:1 vs 5:4

B

burningrave101

[H]F Junkie
Joined
Sep 9, 2003
Messages
11,825
There is a sticky for system memory in the General Hardware forum. If you want more information on memory you should look there.

If you buy Corsair TwinX or Kingston HyperX and it doesn't have BH-5 chips which almost none of them do except for the example the current line of corsair pc3500C2 does then you wont be running 2-2-2-5 stable.

Those selections your talking about use CH-5 chips which are rated 2-3-2-5 and wont hold very tight timings in overclocking usually.
 
So pretty much you have to pay the same amount to get Low Latency DDR400, as you do to get High Latency DDR500?

WELL THAT'S GAY! lol.
 
I have Geil Golden Dragon PC3500, and it runs cas 2, 7-3-3 at 267mhz fsb 5:4, and currently I'm testing cas 2.5, 8-4-4 just to see how high it will go. I'm at 275 with 5:4 (440mhz) at those timings, I'm pretty sure I can lower them, but I'm still testing stability of mobo and cpu.
 
Anyone have the capability to test something for me? How does cas 2, 7-3-3 timings at 267mhz fsb 5:4 ratio compare to cas 3, 8-4-4 timings at 267mhz fsb 1:1 ratio?
 
best i can do is my kingston 3500 bh's i can get 262 fsb 1:1 2-2-2-6,

With my geil 4300 i can get 295fsb 1:1 with the usual slack timings.

The kingston kills my geil. as far as benching. but it depends on the chip im using. Kingston for my 3.2, and the geil for the 2.4 or 2.6 i have.
 
Originally posted by Mojo
So pretty much you have to pay the same amount to get Low Latency DDR400, as you do to get High Latency DDR500?

WELL THAT'S GAY! lol.
Click to expand...

Um well Mushkin PC3500 lvl 2 which is the best overclocking RAM there is that runs 2-2-2-5 costs around $283 for 1GB.

Corsair PC4000 with Platinum-Silver Heat Spreaders thats rated for 3-4-4-8 costs around $337 for 1GB.
 
if you read the article i mentioned above you'll see they stress that running memory in 1:1 is better than 5:4 or 3:2. they also mention that the new ddr2 memory coming out will have cas5 chips. they will still be faster and have greater bandwidth than today's fastest chips. get the ddr500. you won't be sorry.
 
Originally posted by lodingi
if you read the article i mentioned above you'll see they stress that running memory in 1:1 is better than 5:4 or 3:2. they also mention that the new ddr2 memory coming out will have cas5 chips. they will still be faster and have greater bandwidth than today's fastest chips. get the ddr500. you won't be sorry.
Click to expand...

running 1:1 with crap timings is only better when you are using programs that demand a crap load of memory bandwidth, however games are not programs that do, so running 5:4 with tight timings will give you better performance in games.

but you're right running memory in 1:1 IS better, but only when memory timings are the same, when you factor in crappy timings like 3,4,4,8 compared to tight timings like 2,2,2,5 then running 5:4 is better.

the thread starter should save his money and get the ddr400
 
sandra mem & pcmark mem tests are faster with crap timings, 1:1.
all other.. favour good timings..

about ddr2.. they will be "faster" because there is no official word
about ddr1-533 or even ddr1-500.
and of course ddr2-533 has greater bandwidth than ddr1-400 :)
plus, thanks to its 2x lower working frequency, ddr2 is able to scale more than ddr1.. imagine.. ddr2 running @ 200mhz would be called ddr2-800..



Originally posted by lodingi
if you read the article i mentioned above you'll see they stress that running memory in 1:1 is better than 5:4 or 3:2. they also mention that the new ddr2 memory coming out will have cas5 chips. they will still be faster and have greater bandwidth than today's fastest chips. get the ddr500. you won't be sorry.
Click to expand...
 
not true. its working frequencies are exactly the same. you are thinking of qdr which effectively works at 4x the rate of the fsb. ddr-2 is just a reworked, tweaked version of ddr with shitty latency timings. yes it can work at higher frequencies but it also runs a lot hotter and is inefficient (because of worse timings).


edit: ddr2 @ 200mhz = ddr-400/pc3200 with SHITTY ASS TIMINGS.
 
Originally posted by Mojo
Will these do 2-2-2-5?
http://www.ncix.com/products/index.php?sku=10798&vpn=KHX3200AK2/1G
Click to expand...
FEATURES:
Power supply : Vdd: 2.6V ± 0.2V, Vddq: 2.6V ± 0.2V
Double-data-rate architecture; two data transfers per clock cycle
Bidirectional data strobe(DQS)
Differential clock inputs(CK and CK)
DLL aligns DQ and DQS transition with CK transition
Programmable Read latency 2 (clock)
Programmable Burst length (2, 4, 8)
Programmable Burst type (sequential & interleave)
Timing Reference: 2-3-2-6-1 at +2.6V
Edge aligned data output, center aligned data input
Auto & Self refresh, 7.8us refresh interval(8K/64ms refresh)
Serial presence detect with EEPROM
High Performance Heat Spreader
PCB : Height 1.250? (34.75mm), double sided component


Maybe. You might try google for some reviews and see if anyone has gotten them to run tighter.
 
Originally posted by snowwie
no.

hyperx 3200A uses CH-5 chips as far as anyone knows, and ch-5 can only do 2-3-2-5
Click to expand...
edit: Yes. BH-5 can do 2-2-2-5, but CH-5 can only do 2-3-2-5.
 
So only the BH-5 chips can do 2-2-2-5 (IE: Mushkin, and Corsair)?

You say that the Kingston's advertised speeds are 2-3-3-6 or whatever, but the Corsair's advertised latencies are 2-3-2-6 on that site.
 
Originally posted by snowwie
no, I mean ch-5 ;)
Click to expand...
Man I don't know what the hell is wrong with me - I totally misread your post.

I was thinking that only bh-5 can do 2-2-2-5
 
Originally posted by EnderW
Man I don't know what the hell is wrong with me - I totally misread your post.

I was thinking that only bh-5 can do 2-2-2-5
Click to expand...

that is correct

i'm confused
 
that anand thread is really confusing


which one is it saying is better 5:4 or 1:1
 
Originally posted by PureBooYah
running 1:1 with crap timings is only better when you are using programs that demand a crap load of memory bandwidth, however games are not programs that do, so running 5:4 with tight timings will give you better performance in games.

but you're right running memory in 1:1 IS better, but only when memory timings are the same, when you factor in crappy timings like 3,4,4,8 compared to tight timings like 2,2,2,5 then running 5:4 is better.

the thread starter should save his money and get the ddr400
Click to expand...

I don't quite understand why you say 1:1 is better when memory timings are the same, yet your system specs are:

Asus P4C800-E w/P4 2.4c @ 3.51ghz (292fsb) w/MCX4000
1Gb (2x512mb) Mushkin Black Level II PC3500 (DDR468, 5:4, 2,3,2,6, vdimm mod)

You have PC3500 memory capping at 234mhz, with an impressive 2-3-2-6 at that speed. A 5:4 (CPU FSB:RAM) ratio yielding an awesome 3.51ghz effective CPU speed. Since if you were to change to a 1:1 ratio, your system will not change the RAM timings (2-3-2-6 @ 234mhz). Your statement is:

1:1 with X timing > 5:4 with same X timing in games.

So why not change to 1:1 ratio? You'd be doing 234mhz cpu fsb : 234mhz RAM, 2-3-2-6. All you lose is 702mhz effective cpu speed =P
 
what about if i have a 3.0 and i wish to get to 3.6. with a voltmodded mobo with bh5 chips i can run 240 fsb 1:1 at 2-2-2-5. is it worth 283 dollars?
 
Originally posted by Unit-0
I don't quite understand why you say 1:1 is better when memory timings are the same, yet your system specs are:

Asus P4C800-E w/P4 2.4c @ 3.51ghz (292fsb) w/MCX4000
1Gb (2x512mb) Mushkin Black Level II PC3500 (DDR468, 5:4, 2,3,2,6, vdimm mod)

You have PC3500 memory capping at 234mhz, with an impressive 2-3-2-6 at that speed. A 5:4 (CPU FSB:RAM) ratio yielding an awesome 3.51ghz effective CPU speed. Since if you were to change to a 1:1 ratio, your system will not change the RAM timings (2-3-2-6 @ 234mhz). Your statement is:

1:1 with X timing > 5:4 with same X timing in games.

So why not change to 1:1 ratio? You'd be doing 234mhz cpu fsb : 234mhz RAM, 2-3-2-6. All you lose is 702mhz effective cpu speed =P
Click to expand...

what i meant with 1:1 being better than 5:4 when memory timings are the same is

for example:

275fsb 1:1 2,2,2,5 timings will beat out
275fsb 5:4 2,2,2,5 timings (which is obvious)

i didn't mean lowering the cpu fsb down to where you can run 1:1, i meant keeping the fsb and timings the same and just changing the ratio. i was just pointing out the obvious to carry on my point with the rest of the post. :p
 
Originally posted by mtbaird
how does running at 3:2 work?
Click to expand...

It works the same way as the 1:1 or 5:4 but it runs your memory the slowest and thus will hurt the performance the most. 5:4 is a happy medium if your racking up the FSB real high since you can keep your 2-2-2-5 timings.

You would have to have a 300FSB just to get your memory running at 200MHz with 3:2.
 
You must log in or register to reply here.
Back
Top