Gskill Perfect storm, Kingston Hyper-X C8 2000, OCZ Blade, etc.
Printable View
thx for the info
You know with primary timings of 9-11-9-X you could be right there, I had previously thought they were Hypers but binned a little lower than the GTX2's but I can't find what I was reading that gave me that impression.
Well even without the GTX1's, he still has a decent list of ram to choose from.
Hoping to put the REX tomorrow for the first time under LN2, hope that she's better than my first board :)
Good luck mate, have fun with 01.
Just playing with this board again since I picked up a 2 x 1GB GTS kit and 2 x 1GB JNL kit. I thought my 2 x 2GB hypers were picky but these micron sticks are insane, they just require so much more voltage compared to what I'm used to. I pulled off 6-6-4-X @ DDR3-1800 with my hypers, but that is not happening with the microns. The JNL outclock the GTS at 8-8-8-X but they don't respond the voltage well, and don't like tight timings at all. The GTS does respond to voltage, but they have very abruptly ending stability windows where it will go from stable to incredibly unstable within 20 -30 MHz or so.
I'll keep playing but this is not really what I was expecting from 2 x 1GB kits.
Some testing results from my GTS and JNL kits.
For GTS settings were:
7-6-5-15 | 1T | tRFC = 60 | subs mostly auto, some one step tighter
These don't clock on X48 as well as they seem to on 790i. At 1.95v on 790i, they out clocked this kit at 2.20v on X48.
2.05v
SPI32M
http://i720.photobucket.com/albums/w...trfc60205v.jpg
Everest
http://i720.photobucket.com/albums/w...151Ttrfc60.jpg
2.10v
SPI32M
http://i720.photobucket.com/albums/w...trfc60210v.jpg
Everest
http://i720.photobucket.com/albums/w...151Ttrfc60.jpg
2.15v
SPI32M
http://i720.photobucket.com/albums/w...trfc60215v.jpg
Everest
http://i720.photobucket.com/albums/w...151Ttrfc60.jpg
2.20v
SPI32M
http://i720.photobucket.com/albums/w...trfc60220v.jpg
Everest
http://i720.photobucket.com/albums/w...151Ttrfc60.jpg
Fastest SPI32M runs so far (LSC enabled, OCX SPI tweaker 680MB file size, single partition HDD, no max mem)
Boot from BIOS
http://i720.photobucket.com/albums/w...dDDR3-1828.jpg
Memset Tweaked
http://i720.photobucket.com/albums/w...8287-6-5-1.jpg
Max Bandwidth so far (boot from BIOS)
http://i720.photobucket.com/albums/w...1968MHz8-7.jpg
SPI32M - max FSB & Ram freq at 8-7-6-16 1T so far (boot from BIOS / ram limited)
http://i720.photobucket.com/albums/w...8-7-6-161T.jpg
The JNL was far less exciting timings wise, so this is all I'll show. It clocks quite high though, managed an everest at 2136 MHz 8-8-8-16 1T so I'll work on a SPI32M with >2000MHz with this kit when I get the chance.
SPI32M - max ram freq at 8-8-8-16 @ 1.85v
http://i720.photobucket.com/albums/w...DDR3-1998M.jpg
Very nice and interesting tests CryptiK, thanks a lot for sharing! :)
Not good Board is limited to 1.9250Vcore, if I set any higher it doesn't post... did some 2D 6.3Ghz runs and 6.1Ghz 3D... Q822A435 inbound, hope it does a bit better than my current CPU. Another board i will not biuy as my previous REX wasn't Vcore limited but a far worse FSB monster (615 max stable...)
Will bench this weekend prolly with Massman some 01
Sorry to hear that....was hoping to see some inspirational numbers...:D
Quick test in the wee hours on new cascade "Deep Purple":D before I had to fly....next time NB on cold as well. I can boot @ 6 ghz no problem.....as soon as I changed memory things got better, limited with Corsair 2x1 Dom Gt's....have some Diablos to test as well....:yepp: Hitting 6 was a BIG rush for me...waited a Looooooong time!! The 1.944 not necessary but quickly tried to scale higher, no go. More testing required, hoping memory and cold NB will do the trick!!
Great cpu bro! :D
There must be more in that cpu, can't wait to see where it will land but i'm pretty sure it will be over 6.5GHz. Together with the Diablo's it should be capable of doing wicked things, the only thing needed is a REX that will do some good FSB, like the one i had for example.
Let's hope yours will be as good. ;)
Zeus..I sure hope so....being able to boot @ 6+ is a real treat. This is my original RE....I'm quite surprised the poor thing still works after all the torture I gave it on the SS and my first E8600. Was going to snag another board (but the price went up) because I never was able to do really high FSB with this one...that's the best effort yet. Too bad your board died. So, here is the plan.......:cool:
Get home....do the domestic scene...keep momma bear happy, play with the little one till she drops from exhaustion...which is hard...:rofl: Format some 2D & 3D disks, XP & Win 7. Have you come down (open invitation) stuff in the Diablos and have some fun. (could really use your 2D skills) Have a new Otti pot for the NB, had it under water but hoping that cooling it right down with some LN2 will allow us some more FSB. Whadda yah think????? :D
Sounds like a plan. :)
I'd say install Vista Ultimate 32bit instead of Win 7 for 3D and things are ready to become big. ;)
I might pop by too watch the masters at work :up: and swap ya board Expat :D
I can't speak for Zeus but you've got the wrong guy if you think I'm in that class....:rofl::ROTF: I never thought the board had it in it.....High FSB was always a problem for me. The only reason I think its still alive is due to my OCD insulation & a dab of Owens Corning grease in the socket. Once I've got the "shed" built (thank you Belgian slow poke law makers:down:) you can come by and "contribute" :D you Belgian Team GURU!!
I know I'm a bit late to the party but I have a REX now as well.
First kit I'm testing is the blue Cellshock DDR3-1866 CL8 with D9JNL chips, it's the same kit CryptiK is using.
This is the highest 32M-stable speed I managed on CL7-7-7-14 1T, VDIMM is 2.04V:
http://img594.imageshack.us/img594/3...dr31920cl7.jpg
Now I'm working on CL8-8-8-24 1T, but I can't even get it 32M-stable at 1000MHz with 1.90V, with the same subtimings CryptiK is using. Even on 2T with even more relaxed subtimings it's impossible.
The specific DRAM skews that worked well at CL7 don't work at CL8.
What's your secret? :D
Got a new one /w new E8600, can't wait to try it out.[had it for some time.] Got some Kingston 1x2 1866 cas 8, micron gts. Hope that works out.
I have the swify block for the stock board cooling. Good enough, or should I go further cooling the board?[full coverage blocks= $220]
I want to try dice on it this summer. Keeping it on a bench for a while.
Thanks Jor3lBR! :)
Leeghoofd, I'm working on it! :p:
Today I changed the thermal compound under the heatsink, the NB was only covered 1/4 by the original paste. Temps dropped about 13° C.
Some screws were so tight I was afraid of damaging some traces in case of slipping off a screw, but everything survived.
This is still the D9JNL kit, testing CL8-8-8-24 2T, 1.95V:
http://img546.imageshack.us/img546/5...dr32100cl8.jpg
32M at 1000MHz CL8-8-8-24 1T is still not possible, how high are you guys running vNB for that kind of stability? I'm at 1.44V right now. Too low?
Then I switched RAM kits for the black Cellshock DDR3-1800 CL8-7-6-21 kit, which is using D9GTR.
So far I managed a 32M run at 920MHz CL7-6-5-15 1T at 2.14V, which is not that great...
Subtimings are very loose ( auto ), more tuning and possibly higher clocks tomorrow!
http://img690.imageshack.us/img690/3...dr31840cl7.jpg
where did you find them selling these new? thought these boards were discontinued a while ago.
i also had the swiftech block that goes right on top of the HS assembly. although it does work somewhat, it's still not as good as putting on a block that directly cools the NB. with that sock heatsink assembly, it looks like an awesome cooling solution, but heat must still travel though several layers before it gets transferred to the water. full cover WC board is much better, although it also costs much more.
edit: here, post #38 from this thread accurately describes this lol
http://www.xtremesystems.org/forums/...7&postcount=38
Try the new 0002 bios mate ( it has Tref values ) it could help you squeeze them few extra mhz... I never realy tried 500FSB stability on this board... but went straight for 600FSB and around 2000mhz. I topped at +/- 1980Mhz no matter what NB voltage. Memtest would always error out after a few passes...
For high clocks seems the 266 strap is favourable on my board. This REX is in 3D better than my previous one, but FSB is fully blocked at 635... bummer