Jump to content
HWBOT Community Forums

Lucky_n00b

Members
  • Posts

    852
  • Joined

  • Last visited

  • Days Won

    11

Posts posted by Lucky_n00b

  1. Reading a few pages back, I just can't understand why somebody would just point fingers to Massman and Vivi as if they're responsible for this set of annoying new rules.

     

    Vivi made this thread to ask for OPINIONs from different overclockers who passed the qualifier about the use of Titan and benchmark preferences. It turned out that majority ocers preferred that Titan is not used, But that DOES NOT affect the 'allow Titan' rule. So no harm done by the making of this thread here.

     

    If Massman already did try his best to explain things to MSI and MSI choose NOT to listen, then it's not massman's fault, isn't it? I totally agree that there's something that you cannot change no matter how hard you try, especially when it comes to dealing with 'high-up' people whose decisions does not come from deep understanding on how things work. I've been in many of those situation before, believe it or not.

     

    Moving on, I kinda have mixed feeling about MOA now, on one side I'm still glad there's a vendor who still cares about doing live overclocking for the OC community despite the 'drop' on Desktop PC Sales. But on the other, the new sets of rules made me loss interest on the semi-finals.

     

    Waiting for the final rules on the July 1st, let's see how it goes then..

  2. Hold off on buying EPowers for this competition. I received an email earlier today that might mean they are disallowed.

     

    More info today.

     

    IMHO, even though I don't have the proper skill to mod the titan the 'right' way with external power regulator, modding it only with variable resistors to increase volts is even more riskier. I've seen one of my friend's card going 'badum-tss' with only 1.35V @ 1250-1300Mhz-ish core, and I certainly don't wish to own a paperweight worth 1300 USD :D

     

    I think I can see why MSI is doing this, maybe they are afraid that the Epower-module gaining fame(and branding for EVGA) in their very own competition. And the availability of Epower v2.0 is very scarce at best.

     

    But I agree with Michal, how are they going to validate a vmod anyway? They might as well ban the Titan,and hold off the competition to allow the release of 780 Lightning.

  3. Now i feel kinda sad leaving those overclock-heavy-week with all the overclockers and going into 'regular working mode' back in my country. Those past two weeks was surely tiring, but it was worth it ;)

  4. Phew, back at home at last.

     

    Was nice meeting all of the overclockers, hanging around benching something or just chillin' in the oc lab, great experience! (aside for buying 4 equally-sucked retail 4770k though :D :D :D)

     

    I think the number of overclockers gathered in the Computex this year was way larger than any of those live overclocking competitions, so we should definitely do this annually ;)

     

    Until next time,

     

     

    Cheers :)

  5. Interesting topic Vivi,

     

    AFAIK, the rule about semifinals are known beforehand, but it's not like it was set in stone, right?

     

    I already bought 2 Titans(and its retail price are USD 1300-1400 each in Indonesia) and if the MSI wanted to use it for semifinal, then so be it. Many Pro overclocker say that modding Titan using bolt-on external VRM is the ultimate test of skills, so I can understand that.

     

    Anyway, these are my opinions:

     

    No Titan ( prefer non-reference 780 if any)

     

    SuperPi 32M on Haswell

    3DMark2001SE (haswell again)

    Fire Strike / Fire Strike Extreme on 780s

     

    As for the point weight, I prefer it to be 30% SuperPi 32M, 30% 3DMark2001SE, and 40% Fire Strike Extreme.

  6. There's no memory pool to share it among all cards ^

     

    Each card requires all assets (they cannot share memory).

     

    As I understand the problem, in case where there are GPU physics simulation going on, each card has to have a copy of the simulation for each card. So in 4x SLI/Crossfire. Every card has to contain the state and assets of all four simulations (or really bad things happen...) - this takes extra RAM as stuff is copied between cards.

     

    Mostly this shows up as small performance penalty for 2GB cards on Fire Strike Extreme in triple/quad sli/crossfire. Two card sli/crossfire is not a problem and even three might not show any major penalty but four and extreme preset pretty much needs 3GB cards to ensure this limitation doesn't influence performance.

     

    Performance preset is no problem unless you have only 1GB cards (should be a fairly rare case these days)

     

    ...and even then it will work. It just won't have quite as good performance as cards with more memory would have.

     

    Thx for the reply, will try it under 4-way 680 ;)

  7. @Jarnis:

     

    Nice! Waiting for the downloads :)

     

    Anyway I want to ask about this :

    (single GPU still works fine with 1GB on performance and 1.5GB on extreme but per card memory use goes up with multi-GPU)

     

    Any explanation why the GPU memory usage goes up when doing multi-GPU? (isn't it supposed to be 'lower' since the GPUs are 'sharing' their rendering task? I have very little understanding about this)

     

    Thx ;)

×
×
  • Create New...