Massman

1.25Vgpu, don't remember how high Vmem

Afterburner. This run is at approx -110°C. Using the slim pot, the temp can be controlled to 1°C ... but lower than -130°C load (and -120°C idle) make the memory coldbug.

Just getting to know the board. I've no idea how the HT Link will affect 3D performance, or whether CPU power will do anything. Going to take the setup apart again and do some hardmods first ...

It's not looking all too bad ... this is on air cooling. Vdimm is set to 2.1V or, as Asrock bios states ... 'ULTRA HIGH' Note: use orange DIMM slots for dual channel overclocking.

First it didn't boot with my 2x1GB Crucial, so I tried 2x512MB sticks (which worked perfectly). Now the crucials are working again ... but booting after crash remains an issue. Also will need to figure out loads of hardmods, I think :-/

I've bought this board months ago, but it shipped with the totally wrong bios making it impossible to boot the X4 Deneb straight away. So, bought an old Sempron CPU, flashed the bios, mounted the X4 Deneb again and everything should be working. Sadly enough, that wasn't the case: gave up on the board for a few weeks, then last weekend decided to give it another spin. After much trying, I finally got it to boot just the one time ... and after much more trying I finally got it working somewhat properly ... Just wondering if there are other guys or gals out there who have touched this board and if they are willing to share some info. I had a great time with the LGA775 Asrock variant last time ... hoping this board can do the same magic!

Not much, really. Quickly tested my own retail and crosscheckt coldbug with Xpower: both boards act the same. I'm a bit 'pissed' that the board laying on my desk here is rev 0.2, which is not retail. This means that any issues I report can be swept from the table stating the board isn't retail. In fact, it's possible that my overclocking results don't resemble the overclocking results on retail ...

Yes ... that's what I'm saying. Removing NOW is no solution ... never said anything about the future.

Removing 980X and Clarkdale ES results IS retro-active ... ?

First of all I want to stress this: I am replying here as 'Massman' and not 'HWBOT'. I'm also not informing you on the latest developments in the project, but just thinking out loud and presenting my view on the suggestions Just so that you guys know ... That's a good question and certainly a topic that we need to think about. My opinion is different from the one of other staff members as I believe that we shouldn't be forcing anyone to join the league or not to join the league. Ideally, all the people who are seeded heavily would move by themselves to the league and we'd not have this issue. I think that this 'pro league' should be seen more of a 'marketing league', by which I mean that this league overclockers play who receive hardware to be used for marketing a product. This, in other words, excludes those people who get a mainboard once in a while for review, as a gift or just to get feedback. There's a distinct difference between being a PRO overclocker and a power user, I think. In the staff forums, SF3D made a good (and more importantly: doable) suggestion to rate PRO scores the same way you'd rate the 'amateur' scores, without both affecting each other. Team rankings wouldn't be affected then, which is good. Still thinking about the hardware points ... this may be a little bit more tricky. As for joining the PRO league ... I'm not sure about this either. I think there should be some minimum requirements to join as we don't want newcomers to join just because they want to. Nothing has been done in terms of coding, so nothing to show yet.

I asked this, but no. Reason is quite simple: the XPower is the only board that has been designed for Gulftown. Not that the other boards would explode and burn down your house, but overclocking results will very likely to almost certainly be much inferior to the XPower.

Technically and theoretically, it certainly is no problem. I think not only within the staff this topic has been discussed many times, but also at many overclocking meetings. It's also, I think, a possible route for overclocking to take in a non-specified timeframe. That's not the issue. Please sit down and have a look at all the extra things you need to take into account: - How many people do you allow in the leagues - Based on what criteria can they enter the league - Based on what criteria do they drop from the league - How can we control submissions when serious business is on the line - How to integrate vendor interests in overclocking into this (I'm especially worried in how to prevent people from cheating if all happens from home and how to make this league interesting for non-overclockers to watch) We don't have to start running before we can stand-up. This kind of competition requires a LOT of thinking and it's simply something that HWBOT is not capable off at this point in the competition. Splitting up PRO and Amateur ranks is already a very big step for the overclocking community ... let's make sure that this step is done properly before we add new ideas and features to it.

This is ... erh ... madness!

Upgrade will be send to you. No #2 / #3 prize ... you need to be first to win!

Uncle, No, not entirely of course. The logic behind "removing ES from the rankings now is not a good solution" is based on more than just one argument. 1) Our notion of 'ES' is based on the detection by CPU-Z, hence the applicability of the rule is relative to CPU-Z's capability to detect it 2) There is no way to detect VGA ES samples through GPU-Z 3) Introducing the 'ES forbidden' rule means removing them across the board, including Yorkfield, Wolfdale, Conroe and older as we want consistent rules 4) Removing the ES mid-game in unfair to those who have spend a month's paycheck or more on trying to acquire one 5) It doesn't solve the underlying issue of marketing-related overclocking scores dominating amateur rankings. Furthermore, for the 980X, the issue is the difference in stepping rather than the difference between ES and retail. I think no one cares about the B1 ES samples as they are a lot worse than the current average retails. Also, it's not about trusting or having faith in the overclocking community (I reckon whoever lost faith earlier will have gained loads of it seeing Hipro5 share the coldslow), but rather experience. If there are two things I've learned from X years of HWBOT it's that a) people will go very far to have an edge over the competition and b) people will consider people cheaters for the slightest oddity in submissions/scores/explanation. This means that IF it's possible to get ES's to look like retail and it gives an edge, people will do it and even if no one would do it, people would accuse people of doing it. A good example of this is the assumption by someone that any 6.7GHz retail shown by AndreYang will not actually be retail.

Let's just focus on step 1 first. Having a system where people can lose their place in an "upper league" and drop back to the "lower league" is a much more complicated issue than splitting up rankigs. Yes, even more complicated

IMHO We have to keep in mind, however, that the word 'pro' suggests that there's some kind of financial reward. As of today, only very few overclockers are getting paid for their overclocking escapades (in fact, I don't think any overclocker has a salary based on overclocking results), which means that the 'pro league' is more about having acces and connections that really getting paid. Making it mandatory to participate in competitions is quite diffuclt in this scenario, as even for those in the 'pro league' this remains a hobby. Another solution would be to have a ranking based on both 'free overclocking' (what we do now) and the 'pro comps' (for extra points). If it would be balanced 50/50, pro comps wouldn't be mandatory to compete in, but necessary if you want the top spot in the pro league. I don't think we are at this stage yet that we can talk about relegating someone from the PRO league based on the overclocking results. Step 1 is to relieve the normal overclockers out there from the burden of competing with heavily supported guys. Step 2 might be a dedicated OC league with relegation etc.

Fixed.

I kind of disagree with this. Getting hardware for review is different from getting hardware for overclocking. People who do reviews often have to make sure everything can be send back in a decent state and the products are not send for overclocking, but for mere testing. The amateur league is, in most people's minds, a place where ne freebies are given away. This is, I think, an utopia because 1) it's too strict: attending a live final (something you might have worked very hard for to get to) leads to being moved to PRO. It's not because you attend one final or receive one board you're a pro 2) We can simply not control or even check up on the flow of hardware coming from manufacturers. No freebies = uncontrolable. I don't think the amateur league should be a league for people who have no skill. If you have no skill, you simply can't overclock ...

There's always been the intention to decrease/remove manufacturer influence in amateur rankings. This also never has been said otherwise. Removing (points from) ES tomorrow is just not the solution we're looking for ...

Yes ... but this implies we also need to make it interesting for the overclockers who go there. If it's not interesting, they just won't submit anymore and that would render the league's existence also pointless . I guess it shouldn't be a problem convincing the 'top' guys to move over (at least I hope), but depending on how we define PRO, some smaller guys could be forced to move without them wanting to.

Currently some discussion going on in the staff forums, playing with some ideas there. I'll give a heads up later this week when we got the initial problems solved.

The issue is "should want". It's one thing to split up things at the side of HWBOT, but if this pro league has totally no support from any vendor, it's existence is pointless. After all, if vendors don't like/need the pro league, they can just continue to manipulate amateur rankings.

That sidenote was not about winning or losing, it's only to elaborate on how the voting systems can give a wrong impression. As Tin said: as there are more non-smokers than smokers, a vote on disallowing smoking will very likely end up against smoking. Possibly because it's the right way, but also possibly because there's a majority of non-smokers. In this way, a group of people can always block another group of people, even if the outcome doesn't make sense. It's a similar bias of those who have a crazy ES voting yes and those who have 100 weak ES voting totally no. A voting poll can always be used as guideline, but should not be used as tool to determine the path taken. If the latter was the case, you'd simply say "the mob knows best", which has been proven incorrect scientificly .

Tin refers to a reason why this POLL is biased: there are more people who don't have an ES than people who do have an ES (just like there are more non-smokers than smokers). Because the group is a lot larger, it's likely that the outcome of the poll is affected by group size. Compare it to this: Group A = I like apples = 20 people Group B = I don't like apples = 1000 people Poll: should be remove apples from our shopping list? Since Group B is larger, more people will vote to remove apples. Actually, in Belgium we have a law to prevent large groups to win voting over small groups. This is a known phenomenon