Jump to content
HWBOT Community Forums

l0ud_sil3nc3

Members
 View Profile  See their activity

  • Posts

    3536

  • Joined

  • Last visited

  • Days Won

    36

 Content Type 

Posts posted by l0ud_sil3nc3

    Fastest MFR based SuperPI 32M runs - Full out and 4G/5G

    in Haswell (Z87/Z97) OC

    Posted

    Hahaha :D

     

    //edit: George, since I know I can trust your judgement, how much stretch is there on the MFR performance wise? Higher frequency is not going to help as you keep dragging the RTLs up too, but what about lower tCL and tWL? tRDRD should be possible at 4 with cold, I guess?

     

     

    I think tighter timings would help especially at lower frequencies, but as Dumo has stated tighter is not always faster.

     

    I could not get 3100 with those timings to pass, but the good thing about my kit is it seems to like voltage, 1.9 on air is fine so I am hoping they do well on cold.

     

    Hiwa is correct being single sided hurts the most, but I think if you could get 4 dimms (full bank) to run it would cancel out the single sided performance hit, but I am sure the IMC would have a tough time coping with full bank, decently high mhz and tight timings (can't believe I just said tight timings and MFR in the same sentence:p).

     

    Thinking about strapping the pot on these suckers but so much more PSC to bin atm.

    All Haswell degradation info in here please.

    in Haswell (Z87/Z97) OC

    Posted

    Well I know for a fact I have degraded a few cpu's some with high voltages and others with what I would consider safe voltages, the most current one was damaged with 2.5 input and 1.74 vcore while simply ramping up the multiplier via software in Windows. All this was at -130c so don't think temp should have been an issue.

     

    Vdimm was 2.1v and after rebooting once it hit x59 it could never boot 5Ghz again at any uncore or voltage.

    MOA 2013 Discussion

    in HWBOT Competitions

    Posted

    Lol - this gets better and better. Haswell sux ass, benching titans on air - lets have a low score challenge!!!

     

    The only world records I see coming are the number of dead cpus and gpus for bunnyextraction scores....

     

     

    I can see it now, MSI is going to get so much crap for dead retail Haswell, but for once I don't think it will be the boards that have anything wrong with them, rather the cpu's just fail hard, so fragile and easy to damage.

     

    I don't think we even have to have the Titan death talk lol

     

    No more LN2 runs on Haswell from me except at 4/5Ghz as they don't seem to die there.

     

    Semi Finals will be interesting to say the least:)

    MOA 2013 Discussion

    in HWBOT Competitions

    Posted

    Regardless I consider 2600K higher than 3570K to MOA benchmarks, and rules may have dual interpretation, they are clear.

    2600K should be allowed.

     

    Rbauss why do you even care what happens in B Class, you have all ready secured your self a ticket it makes me wonder why you have such a vested interest in not allowing a 2600/2700K.

     

    Everyone else has/is securing a nice Sandy as it will be the winning setup, not sure why you expect it would be any different.

    MOA 2013 Discussion

    in HWBOT Competitions

    Posted

    to follow your logic, if costa rica haswell hit the shelves mid July, only malay chips can be use during the semis?

     

     

    All ready out apparently, I will troll the local store tomorrow and see what is going on as mine should have them since I am very close to the San Jose Fry's where the first pics of a retail Costa cpu were re posted at XS.

    Asrock Z87 OC Formula initial PSC testing

    in Memory Heaven (air/extreme)

    Posted

    How's the performance scaling with frequency in comparison to Ivy Bridge? I remember even Samsung based sticks starting to cap out on performance around DDR3-2800C9 as more frequency wouldn't really provide better performance.

     

    Well at 5Ghz on Ivy I would say 6:15:9xx would be absolutely killing it and on Haswell I think 5:55:9xx is the fastest time to date, so maybe 21-23 seconds with similar cpu and memory clocks between both architectures.

     

    For Samsung I am sure you would need 3000+ c9 for it to be competitive with ultra tight secondaries and tertiaries, but where do such sticks exist, cause I know mine won't do anywhere near that at c9.

×
×
  • Create New...