|ron Posted February 14, 2011 Posted February 14, 2011 (edited) Yesterday the AOC (me, 7ornado and TheKing) had the first session with Sandy@LN2... it has been a great day, for the results and for the fun as well We had only 2d benches, because after 5 hours of ln2, we spotted water around the socket zone, so the mobo refused to post... a day or two and it will be ok Here we go! Let's start with the 32M: Server2003 installed, no Waza, no maxmem ecc, we obtained this: 6m 20,937sec RAM@1184@9-12-9-24 Trfc80 At this time, we couldn't scale more with the Kingston kit so 7ornado told me "come on, let's pass to another bench, those kingston sticks aren't going to go anywhere better than this"... "Uhmm... no, we must have those sticks go even if they aren't so good! Let's mount our Ram Pot!". And here it is... (audio is of course in Italian, but don't worry, we were saying really stupid things ) CAS from 9 to 8, 0,5 seconds better... Going on... 7-11-9-24, Trfc80 Given that 32M is sensible to ram frequencies, why don't we try some strange combination of bclk/ram frequency? And here we go again Ram@2270mhz at 9-13-9-21 but this time BCLK@106.4MHz and multi at 50x! Nervousism incoming :asd: Our last step with the 32M, multi at 49x, BCLK@108.6 32M stable!!!! Final time, without Copy Waza, 6min 19.219sec, ram@2300MHz at 10-13-10-21! (definitely we need better chips, Hyper maybe ) We were so happy being stable at 32M at 108.6mhz BCLK, so we tried the max bclk frequency... I can say it's 109.2MHz, the validate is coming later today This is a good BCLK freq, 8 place in 2600K rankings Hwbot Edited February 14, 2011 by |ron Quote
|ron Posted February 14, 2011 Author Posted February 14, 2011 (edited) A sub 7s PI 1M: Pifast@13,81sec: And finally Wprime 1024M e 32M: I hope to be back soon to update this topic with some 3d submission, we still have half dewar to play with Edited February 14, 2011 by |ron Quote
|ron Posted February 14, 2011 Author Posted February 14, 2011 Guys, here it is the 109.2MHz validate Quote
Crew Leeghoofd Posted February 14, 2011 Crew Posted February 14, 2011 very nice work lads on the budget motherboard who needs UD7 or M4E ? Quote
|ron Posted February 14, 2011 Author Posted February 14, 2011 Ehy leeg, thanks mate! We are very satisfied with that mobo! The only weird thing is that today it blocks during post sequence with code "58" when I try to use one of the first two slots near to the socket. Everything well if I use one or both the "external" slots... Any idea? Quote
Hondacity Posted February 14, 2011 Posted February 14, 2011 please don't let water short anything...it could ruin some mosfets and caps... anyways nice work from you guys Quote
|ron Posted February 14, 2011 Author Posted February 14, 2011 Thanks Hondacity! I just tried to revive those 2 slots by spraying them with Isopropyl Alcohol, let's wait a day or two and see if it comes back to life Quote
|ron Posted February 28, 2011 Author Posted February 28, 2011 Hey thanks bob! For the mobo issue... it wasn't the mobo. It was a partially broken MC, I never saw a half broken part of a cpu, but that's the case. Anyway, cpu chaged, today we benched at around 5,57ghz, waaaay better Quote
|ron Posted March 15, 2011 Author Posted March 15, 2011 I didn't open a new thread just because it's an update. We changed the cpu and the rams, so this is our last run at 32M: Tonight we'll try to go under 6mins by playing with subtimings! We had no luck trying to use the 2400mhz ram setting... Quote
Massman Posted March 15, 2011 Posted March 15, 2011 DDR3-2400 divider only works on early ES samples (C1) Quote
|ron Posted March 15, 2011 Author Posted March 15, 2011 argh, I didn't know it, thanks for the clarification Allright, then we must work on the rams a little bit more, to break the 6mins barrier Hope to have good news later tonight! Quote
|ron Posted March 16, 2011 Author Posted March 16, 2011 No way to squeeze out more than this: Too bad we didn't break the 6mins barrier... we need some more mhz here and there Quote
|ron Posted March 16, 2011 Author Posted March 16, 2011 No way to squeeze out more than this: Too bad we didn't break the 6mins barrier... we need some more mhz here and there Quote
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.