Massman

I'm just talking about the "World Records", not all results (that would be impossible)

Fixed

Looking into this (part 2)

My point is not to make OC more exciting, but more from a general point of view why the community (or even press) accepts records done behind closed doors without any external peer validation that the score was in fact achieved at that moment. Competitive overclocking and exciting livestreams is another can of worms.

Just for kicks I made a small table with the best results achieved in HWBOT contests and then live events. Contest (online and offline) Live Contest/Event

Verifiability is an important aspect of legitimizing a fact. That applies universally, in sports but also in science and research. Someone mentioned that he found it weird that the overclocking community accepts "records" behind closed doors with no 3rd party presence to verify the score was actually set. I thought about it for a while, and he made a good point. It would be interesting to make a list of the "World Records" made during contests (online) and then another list of "World Records" in a live environment.

We didn't recalculate the ranking until yesterday when the staff confirmed the leaderboard. The recalculation happened this morning (Taiwan time).

A result with verification is pretty useless, imho ...

The scores are still in the database, both Karl's and Sam's. They've not been deleted. As for Karl's, he made the request to make the scores invisible before his passing. Sam also asked to make them invisible (I believe). We can disable the points for the submission, and I don't think they would take points for other users. I need to check that, though ... can't remember exactly.

In any other sports records always require validation by spectators or a specialized firm. Usain Bolt can run 9.1s 100M in his backyard, but it would not count (in fact no one even tries to claim this). Even speed records of the fastest cars or bikes are always video documented. Why is this not so for overclocking? I can't remember there's been a major community push for this ever? Why? :celebration: :celebration:

After we get some feedback here, I'll move the thread to the public forums ...

The exact points implementation is up for discussion. The main benefit of using the same algorithm is that it's 1) efficient and 2) easy to understand. It would give the #1 from a specific class a wild amount of points, but then again ... if you beat 3000 others the points are probably deserved

Ever since LN2 became a real thing in the world of overclocking, we've had people complain that the reward for running LN2 is too high compared to decent air overclocking. To make a long story short: I was thinking on how to improve this and this is what the brainstorming led to. Below you can find some slides with the explanation. I know some of you guys don't like slides, but I need it to explain to dev and others exactly what I mean. So please bear with it. TL;DR - Global Class Points use the same algorithm like the Global Points, but only takes into account participation within the League Class (eg: Enthusiasts). Each score is graded overall (Combined) and per class; the maximum of these values count as global points for your score. In practice, the effects are: being #1 in your class will give you a lot more points (sometimes up from 5pts to 66pts) - especially effective in classes lower than Extreme top air cooled submissions will appear on the front page an LN2 submission will always receive more points than an equally ranked air submission (because combined always has higher participation than any class specific)

Love those Asrock boards ... I wrote an article on AGP overclocking with one of the variants like 7 years ago (OMG, can't believe I'm getting old) http://www.madshrimps.be/articles/article/670/Tweaking-the-fastest-AGP-Based-System-Part-1#axzz3f9tSCfo7 Did some theoretical stuff on memory bandwidth in that article too. Reading through it again, it appears memory bandwidth was the key. I can't remember messing around too much with dividers other than 1:1 because of instability and compatibility issues.

Let me look at this when I get in office later today

Awesome Christian!

One of the issues with SB is the low limit on memory clocks due to the DDR3-2133 multiplier being the max. So you'll be looking at either Hyper or PSC/BBSE for top performance. Hyper didn't do particularly well on Z77, but I think PSC/BBSE was fine. I think many Z68 boards were more mature for SB than P67 ones.

I think he mentioned somewhere that his position in the HWBOT rankings didn't really help him move up in the "oc world". Ie landing work, paid consultancy, or sampling. I'm sure the recognition of his work is there, though.

There should be a delete button at the top of the sidebar menu of any XTU submission. You have to be logged in to see the options, though

You can submit as many scores as you want, Br3tt96. Especially if you're improving your score . Dry ice isn't allowed in this competition as that would push you straight into the Extreme League and this competition is only for Rookies.

WOW, Gubben going for his second Round of full points!

Thanks Stasio! - GIGABYTE X99 UD4 --- F13b: download - GIGABYTE X99 UD5 WiFi --- F10c: download - GIGABYTE X99 Gaming 5P --- F2e: download

Hi Trodas, I fixed some links between your HWBOT account and your forum account. Can you try clearing the cache from HWBOT and the forum, then log in via the front page and check if you can comment ?

The info is in other threads. No pin shortings, 'just' using exploiting an older microcode.