Rauf

There is sufficient proof along with previous experience I think. Good that you answered and explained. Mods will decide.

It doesn't happen on cold, even with unstable memory. And it's pretty easy to spot. Hope the mods agree, but they have removed these types of scores before.

It has been brought up before. When you run core or maybe primarily cache at its limit for being stable, especially on air or water cooling, you sometimes get slightly bugged runs which are way too high compared to the normal scores. It is normal to get around 5 points variance in runs, but when you get 20+ more points it is a bug as you cannot replicate it on a fully stable system.

Did a little testing tonight on the big board (XPower) on LN2: All tests 5G/5G and subs which are really good for XTU: 1933 12-12-12: 2000 12-12-12: (lack of scaling is due to the big board not handling RTL as good as smaller boards and needs higher) 1933 12-11-11: 2000 12-11-11: (again higher RTL) No way the run is legit...

I'm sorry but this looks like a bugged run. Like the ones that were removed from the skylake low clock challenge... No way you can get 1797 for real with that mem speed and cache freq...

Like others have pointed out, it depends much on return policies and if that is not good in your country it depends on second hand market. In Sweden you can try your product and then return it to get a full refund, except for return shipping. But if you do it too much you will get banned from the stores. But second hand market is bad in Sweden. Easily a 100€ loss on a single 6700K that has just been tested on air, with full warranty still left. A few tricks I have learned is that some products don't have a seal, or at least a proper one Then you have no problems with returns. But sadly CPUs have a good seal... and trays aren't available in Sweden at all. When the binning reaches Caseking levels, it is very hard to compete privately. Talking probability, it is quite easy to find the best out of 50 or 100 chips when binning yourself. That will get you a 6.4-6.5 GHz CB R15 chip. But when Caseking can find a few best of 1000 chips that do 6.55-6.6 GHz private binning is not enough. If you aim to actually beat the Caseking cpus you have to find a 1 in 5000 cpu or so, and you can't bin thousands of cpus on your own... When Kaby lake comes out I will hope to get a really good Caseking CPU, but of course I will bin as many as I can myself also and hope for luck

Because the community at hwbot is mostly made up from old overclockers who like things to stay the way they are. Whenever someone suggests something they are met with: "it has always been this way", or "can't be done" or the worst one, which is absolute silence a.k.a. no one cares. I mean, this is just the third time I bring this idea up... But if you are persistent sooner or later you get through The how to implement it needs to be thoroughly investigated. I don't think we should favour AMD to get them to make a comeback. Actually I don't think we will have to as the biggest points will never go to the latest highend cards. The lower end categories will have the majority of the submission and therefore the highest points. The key aspect in my opinion is to make 3D benching cheaper. Ideally most of the categories will evolve naturally so that each new generation of GPUs will reign in their respective category. We also need to make it so that it doesn't favour multi-gpu setups as that wouöd make it too expensive. Also we need to keep the categories to a minimum so that it doesn't kill the ranking system.

Yes, I looked at some figures. Multi-GPU setups will beat single GPUs if we categorize by ROP. Haven't looked at shaders but I doubt multi GPUs will beat any singles there. But we can still have SLI/CF like it is today, it can be linked to the number of ROPs/shaders a single card has, and then we have 2way, 3way etc for each rop/shader class. The goal is to lessen the impact of ultra high end hw, and to bring cheaper GPUs into play.

It's the prize of course. Same with 6950X, most would surely be glad if we excluded it. But excluding hw doesn't seem right. CPU categories are working well, why not do it for the GPUs as well?

It definitely requires some thought on how to do the split. ROP, cores/shaders or something else or some kind of ratio or product between rop and shaders. If we do ROP we see that titan pascal has the same amount as 980 Ti, which makes it less than ideal. We also see that 1070 and 1080 would be placed in same category. On the other hand, for AMD they have a large amount of shaders but low ROP on the Furys, which would be unfair towards them if we do just shaders. Maybe ROP x Shaders gives the best way to categorize GPUs?

Try mem at 1800, I got lower score with mem higher than that. Congrats on nr1, I might take up the fight when 1080 ti is released. Don't really care for hw-points, which is all you can bench now...

PM some of the FM guys, they will surely fix it. Datafile is actually the only thing we have to prevent photoshop WR's, we need to keep it as long as possible.

Turns out I was using another key I got from FM right before they published the keys here. Maybe they have some kind of detection against too many ip:s or something which have kicked in now for the keys published here.

I think I have used the free keys, and it has worked.

Thereby the interval... or do you mean the instruction per cycle, or whatever it is called for GPUs. We can't account for that. If some generations are inefficient, too bad. Same situation as for CPUs, too bad AMD...

Ok, the new titan is out and it upsets the rankings on hwbot like nothing else. Aircooled WR's and GFP's. And then a few months later when 1080 Ti comes everything is back to normal. Why don't we make things better to remove this "downtime" which occurs every time a new titan is released? Imagine if almost every 2D-benchmark was multithreaded and were not divided into categories by the amount of cores. 6950X would rule every benchmark, wouldn't you react to that? Couldn't we do the same kind of categories on GPU's as well? Split up the rankings based on the number of cores/shaders/streamprocessors? Now of course we can't divide by exact number of cores because there would be too many categories. But we could make intervals 0-500, 500-1000, 1000-1500, 1500-2000, 2000-2500, 2500-3000 etc. Maybe the intervals need to be tweaked and maybe they need to be changed over time but wouldn't it be a good way to lessen the impact of the titans and at the same time make 3D-benching more popular by making it possible to compete with cheaper GPUs?

XTU is definitely broken. You get bugged scores without processlasso or other software. And really, don't bash processlasso. It just sets cpu priority, if a benchmark can't handle that, it is bad software... As for other bugged benchmarks, I haven't really noticed anything. Maybe that just means you and other mods are doing a good job As for legacy 3D I have actually suggested that it should be moved to 2D category for globals. Which would be lower points in R7... But I love the legacies so I would bench them most anyway.

I don't know that to decrease 2D hw points have been requested. I think if anything to balance things out 2D hw-points should be given a boost as it is much more difficult to achieve. 3D hw-points in many cases is just about decent GPU-clocks and skyhigh cpu-clocks. Also, you can't fix a broken benchmark by altering the points it gives in the rankings. XTU should be put on hold until intel fixes it so that it doesn't bug out AND (wishful thinking) so that it actually measures some kind of performance.

If you are looking to do overclocking as a job you might find it easier to become president. Anyway, I'm 32 and I love the competition. Pushing to do great scores. I like the current gen hardware, so there is always something new coming out to keep you interested.

Amazing cpu!

Yes, that was my proposal also.

So we'll just agree on that only cowards will use Titan XP

Ok, I will do the best I can with 980 Ti

"GPU of user's choice" This was decided before the new titans were released? Are they allowed...?

It will be better in R7, but nonK xtu will still be problematic. The first person to get the "max" score gets the good points, which is fair in a way. The problem is that it cannot be beaten because it is limited in a very artificial way. The score cannot be beaten and lives on forever. A key aspect of any competitive ranking is that you can affect not only your own ranking but also take down someone else.