>>> Corsair XMS 7 stick shootout! <<<

Page 3 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

canadianpsycho

Diamond Member
May 23, 2001
3,417
0
0
Originally posted by: THUGSROOK
seems to be more of a difference then just color.
the black cant do cas222 at all....
the platinum cant do cas232 437ddr....
etc.

theres a big difference in these sticks.

:)

AFAIK XMS memory is just like intel CPUs... Tested to meet a certain spec... I don't think the PT makes a difference, just probably luck of the draw I'd say.
 

THUGSROOK

Elite Member
Feb 3, 2001
11,847
0
0
here's some benchmarks @ 2.7ghz 451ddr ....

3DMark2001se = 12828
X-Isle = 135.0 fps avg.
Sandra Memtest = 3385/3257
UT 2003 Demo = 169.2 / 71.1

:D
 

RaynorWolfcastle

Diamond Member
Feb 8, 2001
8,968
16
81
got my stick of XMS3200C2 (silver heatspreader) today, will post OC results tomorrow sometime.

Also, I received my new 1600+ and it's an AGOIA Y (wk 13)... things are falling into place, schweet :D
 

Lizardman

Golden Member
Jul 23, 2001
1,990
0
0
Originally posted by: oldfart
memory ratios? overclockability? stability? all unknown.
Yeah, the 845G is a good example. Got a new mem ratio, but overclockability and stability are not as good as 845D/E.



I wounldnt say that is always true. I can run 166fsb on a BG7 with 4:5 ratio. That seems pretty high to me and its 100% stable. It depends on many more things than just mobo. Maybe those people who couldnt OC there 845G very well had subpar memory or a shitty power supply. I have not tried a 845D/E mobo, but with the success I have had on this one I dont want to.
:)
 

THUGSROOK

Elite Member
Feb 3, 2001
11,847
0
0
going from 7 - 5 isnt worth much at all.
it gives the least performance out of all the possible settings.
obviously 5 will not get as far as 7 does too.
mix that with a cas 2s and youre better off running @ 6 or 7.

ill be honest ~ a cas2 setting is only equal to 2fsb (in performance) tops. you are better off gettting your highest fsb @ cas 2.5-3-3-7 and then tweak for better cas settings. its a bit more complicated then im making it out to be :confused:

:)
 

MichaelD

Lifer
Jan 16, 2001
31,528
3
76
I know this is a P4 thread, but I'm a little confused about something. You say that the "black heatspreader v2.1 can't do 166MHz 2-2-2 at all" That is what it's rated at! It's rated CAS 2.5 @ 200MHz, but CAS 2, 2-2-2 @ 166MHz.

I just received my 512mb stick of this stuff (yeah the black one, v2.1) not even an hour ago and now I see this thread! I'm either not understanding something here, or I got ripped off. This is supposed to be some of the BEST stuff you can buy and I paid $161 for mine. Please help me understand here.

I am intending on running this stick in an Epox 8K5A2+ mobo at 166MHz FSB 2-2-2-2 timings. It had better do that, at least. down the road, I was hoping for maybe even 180 w/max timings.
 

THUGSROOK

Elite Member
Feb 3, 2001
11,847
0
0
ok some benchmarks! .......looking.......
ah here! (xms3000 c2 black stick)

Cas 2.0-3-3-6 = passed upto 164fsb:437ddr = 12712 3dmarks
Cas 2.0-3-2-7 = passed upto 164fsb:437ddr = 12740 3dmarks
Cas 2.5-3-2-7 = passed upto 165fsb:440ddr = 12777 3dmarks
Cas 2.5-3-3-7 = passed upto 169fsb:451ddr = 12833 3dmarks

as you can see ~ youll need to run your own tests and decide for yourself what is best for you :)

MichaelD ~ youre missing the fact im using a 3:4 memory ratio to boost memory speeds. youre thinking im running in 1:1 mode ~ im not ;)

HTH :)
 

MichaelD

Lifer
Jan 16, 2001
31,528
3
76
Originally posted by: THUGSROOK
ok some benchmarks! .......looking.......
ah here! (xms3000 c2 black stick)

Cas 2.0-3-3-6 = passed upto 164fsb:437ddr = 12712 3dmarks
Cas 2.0-3-2-7 = passed upto 164fsb:437ddr = 12740 3dmarks
Cas 2.5-3-2-7 = passed upto 165fsb:440ddr = 12777 3dmarks
Cas 2.5-3-3-7 = passed upto 169fsb:451ddr = 12833 3dmarks

as you can see ~ youll need to run your own tests and decide for yourself what is best for you :)

MichaelD ~ youre missing the fact im using a 3:4 memory ratio to boost memory speeds. youre thinking im running in 1:1 mode ~ im not ;)

HTH :)


Thanks for scaring the living crap out of me! :Q For a second there, I thought I just blew away $161! Sorry about my ignorance....thank you and I will depart this thread that I had no business being in in the first place. *goes to change shorts as they are very soiled*
 

jose

Platinum Member
Oct 11, 1999
2,079
2
81
Thugsrook,

Whats the increase in benchmark performance between
1:1 & 3:4 @ 169fsb in 3dmark ??


Regards,
Jose
 

THUGSROOK

Elite Member
Feb 3, 2001
11,847
0
0
this will answer SCK's earlier question too...

3dmark2001se

169fsb 1:1 338ddr cas 2-2-2-5 = 12523
169fsb 1:1 338ddr cas 2-2-2-6 = 12558
169fsb 1:1 338ddr cas 2-2-2-7 = 12534

169fsb 3:4 451ddr cas 2.5-3-3-7 = 12833

:D
 

RaynorWolfcastle

Diamond Member
Feb 8, 2001
8,968
16
81
Originally posted by: SupermanCK
and what is turbo?

Most aggressive settings for everything. Trp=2, Tras=5,Trcd=2 1T command rate and all options to fastest settings. I'm sorry but I never understand what people say when they say CAS2-3-2 :eek:.

Sandra brings it up as:
"Logical/Chipset Memory Banks
Bank 0 Setting : 256MB DDR-SDRAM 7-1-1-1R 4-1-1-1W 2.0-2-2CL 1CMD
Bank 1 Setting : 256MB DDR-SDRAM 7-1-1-1R 4-1-1-1W 2.0-2-2CL 1CMD"

 

oldfart

Lifer
Dec 2, 1999
10,207
0
0
So, the Readers Digest version is....?
From what I see in your testing, the 3200C2 is no better than the 3000C2. As you know (I was PM'ing Thugs about a system I was working on), I had similar results. The XMS3200C2 I had was pretty decent, but nothing really special.

I could not get the system completely stable @ 160 FSB DDR426. This was 2.6 or 2.7v DDR, most conservative timings. What was interesting is Memtest would pass, but I'd get Prime95 errors after an Hr or two. 160 FSB 1:1 was completely stable, so it must have been the ram. 160 FSB cas2 3:4 would not even POST.

I didn't have as much time as I would have liked to continue tweaking it. The guy was anxious to pick it up, and I was pretty limited on time. If I had more time, maybe I could have got it a little higher.

In the end, I wound up:
1.6A @ 158 FSB 2.53 GHz
2.5,3,2,7 timing
Completely stable, ran Prime95 7 Hr no errors.

He is VERY happy with it.

 

WA261

Diamond Member
Aug 28, 2001
4,631
0
0
i can onlly run 210fsb on this 8k3a+ for some reason..wont let me select higher.....sandra mem benches at 2900/2700
 

RaynorWolfcastle

Diamond Member
Feb 8, 2001
8,968
16
81
Originally posted by: WA261
i can onlly run 210fsb on this 8k3a+ for some reason..wont let me select higher.....sandra mem benches at 2900/2700

wow, 210 MHz is pretty sweet.... My video card freaks out when I get in the high 180's on my 8K3A (I pass memtest and prime95 but I get artifacts in any 3d apps)
 

WA261

Diamond Member
Aug 28, 2001
4,631
0
0
i do not dare keep it there 24/7...i was only seeing how it would do in a bench.....i had no problems, but i do not want to risk a hdd crash...so i usually keep it around 185