Jump to content
HWBOT Community Forums

SF3D

Members
 View Profile  See their activity

  • Posts

    1650

  • Joined

  • Last visited

 Content Type 

Posts posted by SF3D

    MaxxMem, The Truth

    in General overclocking

    Posted

    I think you read my posts wrong. I have not written anything about bugs with memory clock or reference clock :)

     

    If I understand you right, you are trying to say, that splmann had to remove his score from database even it was not bugged.

    I can not talk about other scores in the database, but that one particular score was bugged. It was so much off for the clocks, that there is no need to continue discussion about it. I think your friend should get over that one wrong submit he did.

     

    I will not continue to discuss about this, cause it seems to lead nowhere. I have nothing personal towards any overclockers related to this maxxxmem discussion. If someone feel insulted, I apologize from him. I have done overclocking long time and it is totally normal to have bugged scores now and then. When you see them, do not post them, that is my rule.

    MaxxMem, The Truth

    in General overclocking

    Posted

    Then the database is wrong, if there is so many bugged results

     

    Maxxxmem is on the database to add something extra to HWbot. What part was missed last time? It does not give any points, so why would anyone post bugged scores in to the rankings. If someone post score to get larger E-penis, it is just funny.

     

    Memory clock + reference clock + maxxxmem are there just for fun. I really don't get it, why they give so serious issues for somebody.

    MaxxMem, The Truth

    in General overclocking

    Posted

    If some benchmark will easily give false results, it is purely based on individual honour after that point. If you really know that you have got bugged score and submit it, then you are a loser in my eyes. It is simple as that. If you do it by mistake, it is no ones fault. Then others can point out the facts, why the score is not real and it can be removed or deleted.

     

    Maxxxmem does not get any points in the HWbot rankings, so it is there to add something extra for the database.

    If someone really like to get top spot in that ranking and then shout it to the whole world, it is just fine for me. As long as the score is totally ok.

    With sandy bridge the maxxxmem benchmark scales directly with core speed + memory timings and speed. The real scores scales with small steps like 2 to 3 points and it is very easy to see when something goes wrong.

     

    If you liked to proof something with this thread, I guess you did it already. You got some bugged scores + normal ones and we have now seen them. It is your decision which one of them you submit to HWbot database.

    That is the thing what matters after all :)

    MaxxMem, The Truth

    in General overclocking

    Posted

    I thin chew have some points here.

     

    You need to have same platform and huge amount of testing. Then you learn what is bugged run and what is not with that certain platform. It is almost impossible to start shouting which one is bugged and which one is not, if there is no 50h+ testing behind.

     

    I am sorry to say, but this thread is pointless. You can not prove anything with this sort of poll :(

    Sandy Bridge + Quad CF

    in Sandy Bridge (P67/Z68) OC

    Posted

    Hi all!

     

    4 GPU's on sandy bridge platform is easy with 6990's, but I have only 1 sample.

    After perfect day with family, I got some time to bench.

    I put 2 x 6970 card in to setup and did few benchmark runs, just for fun.

     

    My CPU did not like to overclock at all with these cards installed, so I had to lower clocks down to 5460 :(

     

    Setup:

    Intel Core i7 2600K

    Asus MIVE

    Kingston HyperX 2133 CL8

    Antec HCP 1200W x 2

    AMD 6990 x 1

    AMD 6970 x 2

     

    Here are the pictures and scores:

     

    setup%20p.jpg

     

    3 GPU 3Dmark11:

    3dmark11%203CF%2014481.png

     

    3 GPU 3Dmark03 (200K+):

    3Dmark03.png

     

    4 GPU 3Dmark11:

    3dmark11%2016611.png

     

    4 GPU Hwbot Unigine:

    Hwbot%20heaven.png

     

    I was surprised about the performance this setup could deliver. 6990 and 6970 are not that bad after all. CF scaling is making these cards looking good in these tests,

    but still they are far away from competitors.

    For gamers this sort of setup would be already too much even with 3 x 30" displays :)

     

    That's the sunday benching from this week.

    LGA1155 facts, stats and list of mainboards

    in Sandy Bridge (P67/Z68) OC

    Posted

    If some P67 board will be released with OC name on it and all features have been ripped off, it should cost like 70€.

     

    Overclocking is mostly limited by CPU on this platform, so what kind of gains we would get from OC board?

     

    Current boards with price around 100€ can overclock CPU's very well and they have all the features basic user can need.

     

    This board cost around 100€ in Finland: http://hwbot.org/community/submission/2123102_drnip_pifast_core_i7_2600k_13.3_sec

    Nooooo....Huge amount of unnecessary features!!! http://www.asus.com/product.aspx?P_ID=oztxS9uaDrwQwO2Y

     

    I hope at least someone will get my point :)

×
×
  • Create New...