Page 1 of 5 1234 ... LastLast
Results 1 to 25 of 118

Thread: A short story behind Corsair Dominator @ DDR3-2462 CL9

  1. #1
    Xtreme X.I.P.
    Join Date
    Dec 2003
    Posts
    5,065

    Cool A short story behind Corsair Dominator @ DDR3-2462 CL9

    Hi there, guys!

    Some of you probably already read on other hardware sites on the Web that we hit DDR3-2462 CL9 with Corsair Dominator. I decided that I might have to "explain myself", to those of you I grew up with in Xtreme Bandwidth section on XS, place I call home This is also the chance for me to come by again, and say hi, even though I constantly pay attention to you guys pushing the limits here everyday - awesome 15k @ 2200 Jody, sweet 11m @ 4G Spi32M dinos, nice 1200+ on 9136C5 George, and more!

    So here it goes...

    All this was a it "on the side" activity during some test we've been doing at very high DDR3 speeds on Intel X48 chipset. Then I decided, what a hell, let's go, and see how far I can crank up those babies. Natually, I had to find a spot for SuperPi 32M. Nothing tweaked here, no copy-waza, maxmem and all the good stuff, not this time at least. Fruits of the work below



    Setup:

    Motherboard: Asus P5E3 Premium
    Memory: Corsair Dominator 2133C9 "in the works"
    Processor: Intel Core 2 Duo E8400
    CPU cooling: Corsair Nautilus 500
    NB cooling: stock, 120mm fan blowing across

    Code:
    Voltages for: 
    DDR3-2462 CL9     DDR3-2268 CL9    DDR3-2228 CL9 SuperPi 32M
    
    vDD = 2.30v       vDD = 2.10v      vDD = 2.10v
    vNB = 1.71v       vNB = 1.71v      vNB = 1.71v
    vPLL = 2.30v      vPLL = 2.10v     vPLL = 2.10v
    vFSB = 1.30v      vFSB = 1.30v     vFSB = 1.30v

    As you can see, there's not much extreme about this setup's cooling.


    Absolute MAX achieved DDR3 speed

    --- Click image for DDR3-2462 CL9 CPU-Z Validation ---



    MAX achieved DDR3 speed in dual-channel configuration (2 DIMMs)

    --- Click image for DDR3-2268 CL9 CPU-Z Validation ---


    Dual SuperPi 32M at DDR3-2228 CL9 (2 DIMMs)

    --- Click image for larger version ---


    SuperPi 32M at DDR3-2228 CL9 (2 DIMMs)

    --- Click image for larger version ---



    So, that's the story! Hit me up here, if you need an urge to ask a question


    Greetings to all XB pushers,


  2. #2
    Xtreme X.I.P.
    Join Date
    Aug 2004
    Location
    Chile
    Posts
    4,230
    Great achivement, i have some question about volts specially about vPLL you needed that much for high FSB? also with GTL_Ref did you use for the FSB? which BIOS? how about CAS8 tests?.

    Really impresive Bachus and nice to see you here

  3. #3
    Xtreme Legend
    Join Date
    Mar 2005
    Location
    Australia
    Posts
    17,403
    awesome results particularly because it's Intel chipset i must say

    great RAM and great mobo
    Team.AU
    Got tube?
    GIGABYTE Australia
    Need a GIGABYTE bios or support?



  4. #4
    Tyler Durden
    Join Date
    Oct 2003
    Location
    Massachusetts, USA
    Posts
    5,668
    Holy $h!t!!!! 2462MHz with 2.3v on X48! That is awesome Bachus!

    And very nice to see you again, you should really show your face more. All we have these days from corsair is some guy with a yellow beard? j/k
    Formerly XIP, now just P.

  5. #5
    Xtreme Addict
    Join Date
    Apr 2005
    Location
    Calgary, AB
    Posts
    2,249
    Excellent BA, good to see you in XB again.
    MB Reviewer for HWC
    Team OCX Bench Team

  6. #6
    Aussie God
    Join Date
    Feb 2005
    Location
    Copenhagen, Denmark
    Posts
    4,678
    Nice
    I want a board that can go that high !!!

    Bach, what chips?
    Competition ranking;
    2005; Netbyte, Karise/Denmark #1 @ PiFast
    2008; AOCM II, Minfeld/Germany #2 @ 01SE/AM3/8M (w. Oliver)
    2009; AMD-OC, Viborg/Denmark #2 @ max freq Gigabyte TweaKING, Paris/France #4 @ 32M/01SE (w. Vanovich)
    2010: Gigabyte P55, Hamburg/Germany #6 @ wprime 1024/SPI 1M (w. THC) AOCM III, Minfeld/Germany #6 @ 01SE/AM3/1M/8M (w. NeoForce)

    Spectating;
    2010; GOOC 2010 Many thanks to Gigabyte!


  7. #7
    Xtreme X.I.P.
    Join Date
    Dec 2003
    Posts
    5,065
    Quote Originally Posted by metro.cl View Post
    Great achivement, i have some question about volts specially about vPLL you needed that much for high FSB? also with GTL_Ref did you use for the FSB? which BIOS? how about CAS8 tests?.

    Really impresive Bachus and nice to see you here
    Hi metro! After 550FSB/2200DDR, increasing vPLL from Auto (around 1.95-2.00v) to 2.20v was a must, and then 600FSB/2400DDR was a no go, unless I bumped it to 2.30v. On another hand, 533FSB/2133DDR is a walk in the park . CPU GTL Ref was at 66%, and BIOS I used in essence 0301. These modules would run whatever I throw at them @ 2000C8 at 1.90v on P5E3 Premium.
    Quote Originally Posted by dinos22 View Post
    awesome results particularly because it's Intel chipset i must say

    great RAM and great mobo
    Yeah, it will do Asus P5E3 Premium has a lot of potential if set up right. All it needs is some kick a$$ RAM

    Quote Originally Posted by EnJoY View Post
    Holy $h!t!!!! 2462MHz with 2.3v on X48! That is awesome Bachus!

    And very nice to see you again, you should really show your face more. All we have these days from corsair is some guy with a yellow beard? j/k
    Thanks, man! They don't let me out anywhere anymore, hehe! But I secretly watch what's up on XS everyday

    Quote Originally Posted by Eldonko View Post
    Excellent BA, good to see you in XB again.
    Heellooo...

    Quote Originally Posted by M.Beier View Post
    Nice I want a board that can go that high !!!

    Bach, what chips?
    Hey! Hmm, chips... Something else for a change

  8. #8
    Aussie God
    Join Date
    Feb 2005
    Location
    Copenhagen, Denmark
    Posts
    4,678
    IC's
    Im not talking about "Sour Onion princles"

    Btw... Im suprised by the big difference between single and dualchannel.... This must be due chipset...?
    Bachus, do you have a RAM-tester? I imagine Corsair have a few....? Can you please by chance test some IC's on it?

    JNL / GTS / JNM / GTR, see if ANY hit 1500mhz+ (DDR3-3000) at CL9-9-9-xx

    Thanks,
    Marc
    Competition ranking;
    2005; Netbyte, Karise/Denmark #1 @ PiFast
    2008; AOCM II, Minfeld/Germany #2 @ 01SE/AM3/8M (w. Oliver)
    2009; AMD-OC, Viborg/Denmark #2 @ max freq Gigabyte TweaKING, Paris/France #4 @ 32M/01SE (w. Vanovich)
    2010: Gigabyte P55, Hamburg/Germany #6 @ wprime 1024/SPI 1M (w. THC) AOCM III, Minfeld/Germany #6 @ 01SE/AM3/1M/8M (w. NeoForce)

    Spectating;
    2010; GOOC 2010 Many thanks to Gigabyte!


  9. #9
    Xtreme Legend
    Join Date
    Mar 2005
    Location
    Australia
    Posts
    17,403
    Quote Originally Posted by bachus_anonym View Post
    Thanks, man! They don't let me out anywhere anymore, hehe! But I secretly watch what's up on XS everyday
    quite the contrary
    from what i could see you are the one not happy when they pull you out of the lab

    see you soon mate...catching a plane in a couple of hours
    Team.AU
    Got tube?
    GIGABYTE Australia
    Need a GIGABYTE bios or support?



  10. #10
    Diablo 3! Who's Excited?
    Join Date
    May 2005
    Location
    Boulder, Colorado
    Posts
    9,414
    Fantastic ram there Wonder if you could gain anything by cooling the northbridge slightly below ambient.. sounds like it's time to setup a waterloop with some ice cubes

  11. #11
    Registered User
    Join Date
    Aug 2007
    Location
    Romania
    Posts
    92
    Quote Originally Posted by bachus_anonym View Post
    Hey! Hmm, chips... Something else for a change
    could it be a non-micron chip?

    congrats on the WR!
    [SIGPIC][/SIGPIC]

  12. #12
    Xtreme Cruncher
    Join Date
    Jul 2003
    Location
    Finland, Eura
    Posts
    1,744
    Want to see chips these sticks use


    http://mato78.com - Finnish PC Hardware news & reviews
    BulldogPO @ Twitter


  13. #13
    Xtreme Member
    Join Date
    Apr 2007
    Location
    I.D
    Posts
    308
    wow....amazing roxxxies corsair
    what chip used? jnm/jnl or gtt

  14. #14
    Wanna look under my kilt?
    Join Date
    Jun 2005
    Location
    Glasgow-ish U.K.
    Posts
    4,455
    isnt 2.00vPLL for auto a bit high, especially for 45nm?

    congrats on the MHz I appreciate the question might be straying towards off-topic, but seeing as 2.20vPLL was damanging Team Japans 65nm chips back in the day, I think its an important area to have info on, before people are inspired by your results and try it for themselves...



    Cheers!

    K
    Quote Originally Posted by T_M View Post
    Not sure i totally follow anything you said, but regardless of that you helped me come up with a very good idea....
    Quote Originally Posted by soundood View Post
    you sigged that?

    why?
    ______

    Sometimes, it's not your time. Sometimes, you have to make it your time. Sometimes, it can ONLY be your time.

  15. #15
    Champion
    Join Date
    Jan 2007
    Location
    Romania, lab501.ro
    Posts
    1,705
    The clocks are amazing. What do you think will be the highest rated kit that would go on the retail market? 2133C9 or do you guys have even more up your sleeve?
    Weissbier - breakfast of champions



  16. #16
    Xtreme X.I.P.
    Join Date
    Nov 2002
    Location
    Shipai
    Posts
    34,549
    Quote Originally Posted by bachus_anonym View Post
    Xtreme Bandwidth section on XS, place I call home
    your an outdoor guy then i guess? havent seen you for ages! good to see your back... for a few days at least

    Quote Originally Posted by M.Beier View Post
    Bachus, do you have a RAM-tester? I imagine Corsair have a few....? Can you please by chance test some IC's on it?
    JNL / GTS / JNM / GTR, see if ANY hit 1500mhz+ (DDR3-3000) at CL9-9-9-xx


    Quote Originally Posted by Monstru View Post
    The clocks are amazing. What do you think will be the highest rated kit that would go on the retail market? 2133C9 or do you guys have even more up your sleeve?
    besides michal metro and jody, i dont know anybody who managed to get 2133 stable... i mean rock stable, 24/7 stable... not bench stable
    even with a handpicked board, handpicked memory and a looooot of tweaking.. so before you ask for more, wait for the 2133 id say
    I think chipsets are limiting DDR3 clockspeeds more than the memory chips... and p45 doesnt clock ddr3 any better than current chipsets it seems

    nice results michal!
    Did you try 790? jody reached almost the same 32m clocks as you with 888 on 790
    What was the max 32m cas7?
    and a question you probably wont reply to, how many boards did you try hehehe
    Last edited by saaya; 05-21-2008 at 01:55 AM.

  17. #17
    Xtreme Addict
    Join Date
    Sep 2003
    Location
    Edinburgh - UK
    Posts
    1,363
    Damn Bachus that's some scary VPLL voltages you're running there

    Congrats on the speeds though

  18. #18
    Champion
    Join Date
    Jan 2007
    Location
    Romania, lab501.ro
    Posts
    1,705
    Quote Originally Posted by saaya View Post
    your an outdoor guy then i guess? havent seen you for ages! good to see your back... for a few days at least



    besides michal metro and jody, i dont know anybody who managed to get 2133 stable... i mean rock stable, 24/7 stable... not bench stable
    even with a handpicked board, handpicked memory and a looooot of tweaking.. so before you ask for more, wait for the 2133 id say
    I think chipsets are limiting DDR3 clockspeeds more than the memory chips... and p45 doesnt clock ddr3 any better than current chipsets it seems

    nice results michal!
    Did you try 790? jody reached almost the same 32m clocks as you with 888 on 790
    What was the max 32m cas7?
    and a question you probably wont reply to, how many boards did you try hehehe
    You are right, I know that. But I was thinking that once technology gets more mature we're really going to see high clocks with good latency on DDR3 (just like DDR2 was at the beginning and then the clocks went sky high). Maybe that will also come with a new chipset.
    Weissbier - breakfast of champions



  19. #19
    Xtreme X.I.P.
    Join Date
    Nov 2002
    Location
    Shipai
    Posts
    34,549
    Quote Originally Posted by Monstru View Post
    You are right, I know that. But I was thinking that once technology gets more mature we're really going to see high clocks with good latency on DDR3 (just like DDR2 was at the beginning and then the clocks went sky high). Maybe that will also come with a new chipset.
    well 790 is a step in the right direction, but it solves one problem, the fixed mem divider and limitation by fsbs, and creates 2 new, gtl and vtt problems hehehe
    p45 doesnt seem to bring any improvements...
    so im hoping for the next new chipset... nehalem!

    any hints on nehalem clocking michal?

  20. #20
    Xtreme X.I.P.
    Join Date
    Dec 2003
    Posts
    5,065
    Thanks guys!

    Quote Originally Posted by M.Beier View Post
    Bachus, do you have a RAM-tester? I imagine Corsair have a few....? Can you please by chance test some IC's on it?

    JNL / GTS / JNM / GTR, see if ANY hit 1500mhz+ (DDR3-3000) at CL9-9-9-xx
    Man, never enough? We're constantly looking into the future, in an effort to bring fastest, and most exciting memory for extreme overclockers, but my humble prognosis is that you will not see DDR3-3000 kits making its way anytime soon, hehe. Even if DDR3-3000 becomes reality at a reasonable voltage on a component level, you will still need a platform that could support DIMMs built with it. I'd wouldn't hold your breath until DDR4 for that...
    Quote Originally Posted by Monstru View Post
    The clocks are amazing. What do you think will be the highest rated kit that would go on the retail market? 2133C9 or do you guys have even more up your sleeve?
    Always, always As you can see, we know how to make DDR3-2133 happen, and we'll go from there
    Last edited by bachus_anonym; 05-21-2008 at 10:12 AM.

  21. #21
    Xtreme X.I.P.
    Join Date
    Dec 2003
    Posts
    5,065
    Quote Originally Posted by K404 View Post
    isnt 2.00vPLL for auto a bit high, especially for 45nm? congrats on the MHz I appreciate the question might be straying towards off-topic, but seeing as 2.20vPLL was damanging Team Japans 65nm chips back in the day, I think its an important area to have info on, before people are inspired by your results and try it for themselves...

    Cheers!

    K
    Quote Originally Posted by Johnny Bravo View Post
    Damn Bachus that's some scary VPLL voltages you're running there Congrats on the speeds though
    I know Good thing is, that's only for 600FSB. Up to 550FSB, it's not so bad, actually.
    Quote Originally Posted by saaya View Post
    Did you try 790? jody reached almost the same 32m clocks as you with 888 on 790
    What was the max 32m cas7?
    and a question you probably wont reply to, how many boards did you try hehehe
    Yep, I've seen his 15k @ 2200C8+ on C73. Very impressive! I know we did above 2000C8, but I have not pushed it too far on C73 yet.
    Regarding amount of boards I tried. Honestly, only two and they are very close, within a few MHz. Those are regular retail boards with retail E8400s. As I mentioned, things seem to get a bit bumpy above 550FSB/2200DDR3 on Intel X48 boards, and a lot of voltage seems to be needed, from what I can tell, once you approach 600FSB. Plus few tweaks here and there. We're still talking on air, that is.
    Last edited by bachus_anonym; 05-21-2008 at 10:17 AM.

  22. #22
    Champion
    Join Date
    Jan 2007
    Location
    Romania, lab501.ro
    Posts
    1,705
    Quote Originally Posted by saaya
    well 790 is a step in the right direction, but it solves one problem, the fixed mem divider and limitation by fsbs, and creates 2 new, gtl and vtt problems hehehe
    p45 doesnt seem to bring any improvements...
    so im hoping for the next new chipset... nehalem!
    Yes, indeed. But you know that Nvidia chipsets always clocked RAM different than Intel ones, with much more accent on CR 1. Nevertheless, it is indeed a step forward.

    Quote Originally Posted by bachus_anonym
    Always, always As you can see, we know how to make DDR3-2133 happen, and we'll go from there
    This is really good news, I think that this is a major bump forward for all the DDR3 market, as other producers will have to come up with something similar to stay in the race. And that is always good for us, users
    Weissbier - breakfast of champions



  23. #23
    Banned
    Join Date
    Feb 2008
    Location
    Buffalo, NY
    Posts
    1,607
    Spectacular speed!! Thats very impressive, and makes me need to save even more cash for nehalem setup, for some nice ddr3

  24. #24
    Aussie God
    Join Date
    Feb 2005
    Location
    Copenhagen, Denmark
    Posts
    4,678
    Quote Originally Posted by bachus_anonym View Post
    Thanks guys!

    Man, never enough? We're constantly looking into the future, in an effort to bring fastest, and most exciting memory for extreme overclockers, but my humble prognosis is that you will not see DDR3-3000 kits making its way anytime soon, hehe. Even if DDR3-3000 becomes reality at a reasonable voltage on a component level, you will still need a platform that could support DIMMs built with it. I'd wouldn't hold your breath until DDR4 for that...
    Always, always As you can see, we know how to make DDR3-2133 happen, and we'll go from there
    Exactly
    Never is enough! For daily I use a laptop, only benchmarks of importance for desktop, so never is enough for that one

    No, the reason is, I have belives in nehalem will reveal the true power of DDR3, and would be very nice to know which ram to gather...
    GTR / GTS puts more load to NB, due tighter subtimings, compared to JNL / JNM - this might cause them to overclock lower, but it would be rather interesting to see what is the better when memorycontroller is out of the picture....
    Yet, it will still be impossible for me/others to know which is better, as we dont have any information on the nehalem.
    Competition ranking;
    2005; Netbyte, Karise/Denmark #1 @ PiFast
    2008; AOCM II, Minfeld/Germany #2 @ 01SE/AM3/8M (w. Oliver)
    2009; AMD-OC, Viborg/Denmark #2 @ max freq Gigabyte TweaKING, Paris/France #4 @ 32M/01SE (w. Vanovich)
    2010: Gigabyte P55, Hamburg/Germany #6 @ wprime 1024/SPI 1M (w. THC) AOCM III, Minfeld/Germany #6 @ 01SE/AM3/1M/8M (w. NeoForce)

    Spectating;
    2010; GOOC 2010 Many thanks to Gigabyte!


  25. #25
    uncore challenged...
    Join Date
    Feb 2006
    Location
    ontari-ho (canada)
    Posts
    1,655
    first off, as everyone else has said, it is GREAT to see you back in action here at the house of Bachus. we (a lot of great clockers) have been trying our best to keep this place popping but it just doesn't have that same feeling without your excellent work. i don't have to say you inspire us with your work Michal so don't be a stranger

    Quote Originally Posted by bachus_anonym View Post
    Yep, I've seen his 15k @ 2200C8+ on C73. Very impressive! I know we did above 2000C8, but I have not pushed it too far on C73 yet.
    thank-you. i have also managed some very interesting numbers with dual 32M / single 32M / 1M with this Ballistix kit and a G.Skill D9GTR kit. check my sig for the OC Report of the Ballistix PC3-16000 kit, but definitely not on X48 like you have.

    oh yeah, and head over to the 32M low clock if you have time...some results from you over there would be nice too
    i don't quote in my signature, but best WR ever...
    Quote Originally Posted by Jor3lBR View Post
    It holds the current WR for the least vcore required to run 4500Mhz stable (1.32vcore)
    i can't even make that shyt up ^^^

Page 1 of 5 1234 ... LastLast

Bookmarks

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •