Page 1 of 2

Who has the worlds fastest memory?

Posted: Fri Dec 08, 2006 10:26 am
by Apoptosis
Image

It has been a number of years since an opportunity like this to take place, so I am going to take full advantage of it – Corsair v. Kingston v. OCZ… Just like the old days back in 2002 and 2003!

The Contenders:

Corsair’s 2GB PC2-9136 Memory Kit
· CL5 Timings 5-5-5-15
· 1142MHz
· 2.1V
· True Heat Sinks w/ AIRFLOW Cooling (will use DOMINATOR AIRFLOW)
· $459.99 @ Newegg

Kingston’s 2GB PC2-9600 Memory Kit
· CL5 Timings 5-5-5-15
· 1200MHz
· 2.35V
· $525 MSRP

OCZ’s 2GB FLEX XLC PC2-9200 Memory Kit
· CL5 Timings 5-5-5-18
· 1150Mhz
· 2.35V
· Water Cooled (will use nautilus 500)
· $539 @ ATAcom


The Test System Configuration:

Will be Intel Core 2 Duo Based with an Intel E6300 Processor that is stable well beyond 450MHz FSB and will be run on the 680i motherboard.

Which kit will overclock the most? Vote for the winner in the poll!

Posted: Fri Dec 08, 2006 10:33 am
by kenc51
OCZ FTW
Kingston will be second

:edit:
What will be your max voltage?

Posted: Fri Dec 08, 2006 10:36 am
by Bwall
Another for OCZ but the Corsair will be close behind.

I'd better get to help with this!!

Posted: Fri Dec 08, 2006 11:38 am
by Apoptosis
damn all voted for OCZ so far... interesting and one sided!

Posted: Fri Dec 08, 2006 11:59 am
by dicecca112
hey hey I voted for corsair

Posted: Fri Dec 08, 2006 12:31 pm
by Digital Puppy
OCZ...cause it just looks sweet! 8)

Posted: Fri Dec 08, 2006 12:43 pm
by cadaveca
I think myself that the kingston modules may go the furthest, given active cooling. watercooling is not gonna make that large of a difference when clocking, merely has an affect on lifetime of the modules in a cramped case.

Corsairs 9136 are D9GKX, i think, which may prove to scale over 1400mhz, given proper treatment. However, i also beleive that all three modules share the same IC's, but it's the cooling, voltage, and burn-in method, as well as the pcb, that will make the difference.

Corsair's PCB is very good...

Dunno about the OCZ flex modules, but i may pick up a set this weekend, possibly, finances allowing.

no active cooling for 2.35v, 1200mhz on the kingston...same votlage as teh OCZ modules...but with far less cooling...may prove to be the best.

Posted: Fri Dec 08, 2006 12:58 pm
by hainer36
im going with kingston

i love mine, and i will back that up anyday

Posted: Fri Dec 08, 2006 1:16 pm
by Apoptosis
cool thanks for the votes guys... let's get some more people involved! Speak up!

Posted: Fri Dec 08, 2006 5:32 pm
by Dragon_Cooler
i am rooting for kingston, however it might start to loose the battle when it gets OCed.

but the question was faster memory, not fastest memory oced!!

Posted: Fri Dec 08, 2006 6:48 pm
by FZ1
I've always been a Corsair fan and they seem to be the underdog at this point so I'm going with them. I'll take the sticks when they win :finga:

Posted: Fri Dec 08, 2006 7:08 pm
by Apoptosis
Well -- I started benchmarking the modules and after a conference call with NVIDIA I was advised to wait to complete testing. Today's launch of the 97.44 video card driver has a bunch of game fixes and next week i was told that NVIDIA would release a new BIOS for the 680i series of motherboards. It will be named P23 and will update P21. In it will be a number of updated memory profiles and memory optimizations, so if I do my testing this weekend before the BIOS is completed it wouldn't be fair to all the modules. For example Corsair's PC2-9136 modules have been tested by NVIDIA for use on the 680i, but testing on the OCZ FLEX XLC modules have not...

I got the kits to 1.2GHz default voltages with not much work, so I have a feeling the top kits will be hitting 1.3-1.4GHz.

Posted: Sat Dec 09, 2006 3:08 am
by HONkUS
Im puttin my money on OCZ.

But in all seriousness my memory is way faster than anything out there because I modded it with a flux-capacitor and its cooled by the cold blackness of space.





not really.

Posted: Sat Dec 09, 2006 10:24 am
by stev
I went with Corsair since we had just purchased a Corsair microSD for the Sonic Slider (not remix) virgin mobile phone.

After much reviews on the cell phone web sites dealing with the 1Gb transflash/microSD's being very slooooooooooow, one reader mentioned the Corsair 1Gb card didn't have the issue at all. Most people went with the 512Mb to prevent the slooooow transfers of data (music, games, photos, etc.)

So, we purchased the Corsair 1Gb microSD and were amazed at the speed indeed! :mrgreen:

Since the Corsair memory microSD proved itself in the smallest possible memory card size in the retail market today (even smaller than xD), I voted for Corsair. :)

I can't wait for the results. It's a bummer that I've been sooo busy this December to stop in more often.

Posted: Sat Dec 09, 2006 10:58 am
by eva2000
OCZ 1st
Corsair 2nd
Kingston 3rd

:)

Not just about raw mhz but clock for clock performance too. I'd like to see you squeeze the timings to tightest possible and clock them as high as you can and see how far they go :)

I haven't played with 680i boards yet, Asus should be sending me a Striker extreme next week so if you get you're results up please share the settings used too :mrgreen:

also memset for 680i is out http://www.xtremesystems.org/forums/sho ... tcount=334 and http://www.xtremesystems.org/forums/sho ... tcount=339 be sure to include screenies in tests

Posted: Wed Dec 13, 2006 8:57 am
by Apoptosis
Well i've had no luck getting any of these modules to run anything higher than 1100MHz as I've found the 680i motherboard has some issues with SATA and some memory settings in the BIOS. I can set the Corsair PC2-9136C5 modules to "SLI Memory" with no OC and the system hangs at the Windows XP splash screen... Actually over 10 kits did that. It doesn't matter if the timings are 5-6-6 or 3-3-3 nothing works with the memory unlinked or even linked for that matter. I tried 1:1 at 1066FSB, so the memory was 1066MHz and none of the kits I have would even load windows! I've tried CL3, Cl4, CL5 timings and have tried 2.0-2.4V on the modules with no luck... 5-6-6-18 2T at 1066MHz at 2.4V should be possible on nearly all decent IC's.

NVIDIA is aware of many of the problems and actually has been working on BIOS P22 and P23 to resolve some of them. Due to the number of open issues they are working on, I don't think P22 might will ever become public as P23 is in QA and might be out this week if it passes. In the mean time I got my hands on P22 (I'm sorry, but I can't send it to you so don't even ask) for internal testing. The automatic BIOS update shows P22 as not having a matching BIOS ID and it came without the .exe, so I had to write my own .bat and .exe to get the P22 BIOS loaded... That's another reason why I'm not sending it out.

I just loaded it up a few minutes ago...


Image

Image

Maybe now I can test some of these modules... I'll let you know.

Posted: Wed Dec 13, 2006 9:25 am
by Apoptosis
No luck moving from P21 to P22... I'm still having the same SATA issues and when running the memory over 800MHz I can't load windows 99% of the time.

I just saw this over at evga:
Both EVGA and NVIDIA have reproduced this problem and are close to finding the root issue. We hope everybody can be a little more patient and stay tuned for the new update about the resolution. We are dedicated to resolving this issue, just as we resolved the Creative X-Fi + 8800GTX SLI issue. Resolving the SATA issue is the number one priority for both EVGA and NVIDIA. EVGA does not want to mislead its customers by trying this and that or offering any unprofessional blind patches to fool customers around.

If you are experiencing this problem, please try the following in the meantime as a temporary fix until a solution is released.
• Set your memory to 800MHz 2T with 2.1v
• Disable NCQ
• Disable LinkBoost

Please note that this issue only effects certain system configurations and does not pertain to all systems using the 680i motherboard.
I kinda find it funny that they are calling this system configuration errors...

I've tried four models of hard drives, 20+ memory kits, half dozen power supplies, five processors (quad-core and dual-core), and several video card combinations from 8800GTX, 8800GTX SLI, 7800GTX, 7900GTX, 7900GTX SLI and nothing works... I guess it's my configuration though... LOL :finga:


UPDATE: Having a hard drive plugged into SATA Master Channel 2 will cause the system to hang on the post screen with a post code of 75. You can't enter the BIOS or anything... Interesting...

Posted: Wed Dec 13, 2006 9:28 am
by dicecca112
to them system configuration = user error. God forbid their product isn't perfect

Posted: Wed Dec 13, 2006 9:55 am
by Apoptosis
This P22 BIOS is a complete bust for me...

SATA Hard Drive Plugged Into Channel 1 Master:
Image

The system posts and then gets stuck here:

Image

If I move the HDD to Channel 2 Master I get a funky symbol and a hard lock. Can't enter the BIOS or anything at this point.

Image

eVGA and NVIDIA said to disable Linkboost -- Did nothing for me and I even slapped in an ATI X1950XTX and still nada... Told me to check SATA cables at one point... tried three different ones... nada... Hooked up a CD-ROM and when i went into the BIOS set the first boot device to HDD and disabled the second and third devices it still tried to boot off the CD-ROM and got stuck there...

I guess my PC Power & Cooling 1000W power supply isn't enough or something... I'm soo frustrated with this board... from not being able to use PS2 mice to kentsfield overclocking issues and now this...

NVIDIA is working on getting me another board... I'm about ready to shoot this one. Looks like this memory review is going to have to wait. I even talked Corsair about this as I'm testing the memory they designed to use with the 680i board and they too have had bads boards come into their lab. Once board will run 1143MHz memory and then the next won't. at least I'm not the only one here...

The million dollar question is why are some boards bad and some good? Obviously if Corsair swaps out just the board and everything is fine then it's a board issue and not a configuration issue. If it is a board issue then what is causing the 680i to have a hit or miss track record?

oh and P22 fixed one thing on the Corsair PC2-9136 modules... it now reads the EPP correctly and sets the timings to 5-5-5-15 as before it was 5-6-6-15.

Posted: Wed Dec 13, 2006 10:16 am
by hainer36
just send all you got to me, ill test em and send back the weaklings :P