BenchBrothers.de

You're totally right. This one has nothing to do with the current algorithm. I just wanted to point it out because XTU is one of the main reasons why points are discussed lately.

I read your post this way: It is ok that someone can reach high ranks in his league without doing 3d. And you wouldn't be happy if someone has to start 3d to reach high ranks. Do I get this right? You discribe the current situation. An overclocker can reach a decent rank in his league without doing 3d. Now imagine an overclocker who doesn't care about 2d but loves 3d. Because of the current algorithm he is not able to reach a top spot in his league only with 3d. This individuum has to bench 2d - whether he likes it or not. Now compare this to your standpoint. You're saying that no one has to be forced to bench 3d to get a top spot in his league. But you're also saying that forcing someone to bench 2d for a top spot is ok. You see the problem? I don't want to have 3d awarded with higher points than 2d. All I want is that a global top spot in 3d is worth the same amount of points than a top spot in 2d. And that is not the case currently (at least it feels this way). And in additioin someone who wants to reach a top spot in any overclocking league has to be able to handle 2d- and 3d-benchmarks. If someone is not able (or willing) to bench 3d than he shouldn't be able to become number one in any overclocking leage - my opinion! Maybe that's the case anyway but you can reach higher ranks doing 2d without 3d than vice versa. Now I come back to the highlighted text again. You say that you need a very good 2d setup to compete in 3d. And because of this fact lots of 3d-benchmarkers also compete in 2d. In my opinion thats not the case. I think at least some do so because without points from 2d they are not be able to hold their position in ranking. As I said earlier in this thread: If you have a great 2d-setup than it is very easy to earn 3d-points. Just put in a 970 or 980 and you will get decent scores in AM3, 2k1, 2k5 and perhaps 2k6. For these benchmarks you don't need a 3-ghz-Titan X. Of course you won't reach a global first place or a wr - but you will be awarded with a decent amount of points. No zombi-modding and ln2-cooling needed. Just put in a stock relativly "cheap" air cooled card and you can achieve some good points in 3d. Now tell me in which 2d-benchmarks someone can compete with a 3-ghz-Titan X? There is none. On the other side I (and steponz) already gave examples where a 2d-overclocker can compete really easy in some specific 3d-benchmarks. You can compensate a mediocre cpu in 3d-benchmarks with a good vga and some skill up to a certain point. So a great cpu isn't mandatory in some benchmarks (unlike you stated). Now take this mediocre cpu for 2d-benchmarks - how much points will you be awarded with in 2d-benchmarks? I am sure that you won't get much points for this... To sum things up I think you're looking to the issue from a wrong perspective. No offense here, just my opinion. Lots of the "drama" here has to do with XTU. I just realized (never thought about it) that we talk about a benchmark awarded with points where only hardware of one single manufacturer can compete. All other awarded benchmarks can be used on AMD and Intel. That's not ideal (to say it frankly)... And as a side note the benchmark don't work on every Intel-cpu. I tried it on my Xeon E5-1680V2 (S2011, Ive-Bridge-E-Core) but all it said was that no supported hardware was found. To put it in simple words: Only Intel decides which hardware gets hwboints. Is that the right way?!

I'm not Andi. Don't mix up BenchBrothers.de with BenchBros... Of course the 23k results run on a higher core count. That doesn't change anything when it appears that a 23000th fastest score is worth more than a global world record. By the way: A happy new year to you all!

This will be a long post so I apologize in advance. Please see this post from the perspective of fighting for global points, not in hardware categories. The most of us overclockers do it as a hobby. So you are limited in time and you have limited money. As long as you aren't addicted to a special 3d-benchmark you have to dicide whether to go 2D or 3D. Most of us will take into account what amount of points at hwbot you can achieve with your resources. Since you can spend your Euro/Dollar/GBP/whatever only once you will make a decision which offen is based on economical facts. With that in mind with 2D you can "earn" more hwboints with the same amount of money so lots of overclockers will therefor focus on 2D. It seems that in the past such a dicision was made by many overclockers and therefor we have more submissions in 2D-benchmarks than in 3D. Because of the way the hwboint-algorithm works it becomes a vicious circle: 3D gets less and less atractiv, more and more submissions for 2D and so on. Lets say you have a budget of 5000 EUR and you will try to get as many points at hwbot as possible. The amount of money is enough to buy hardware for 2D-benchmarking, maybe you even can set up two competitive rigs (X99, Z170) and compete for top spots in SuperPI 1M/32M, wPrime 32M/1024M, Cinebench R11.5, Cinebench R15, hwbot Prime, Geekbench3, PiFast and XTU. On the 3D-part 5K will be a tough job for a rig that can compete for a top spot in Fire Strike/Vantage. For older 3D-benchmarks you need a 2nd rig which doesn't fit into your budget. Because of such a calculation again lots of overclockers will go the 2D-route. And again: More 2D-submissions, lesser points for 3D... I agree that binning vgas and then complain about the points is not the right way. But how about binning cpus? Is binning cpus any better? There are users and/or companies that can bin hundreds of cpus. And normal overclockers have to compete with this too. Lets say you have a budget for binning of 10k EUR. How much hardware can you bin? 980Ti - about 15 pieces (reference design). Titan X? About 9. 6700K? About 27. 5960X? About 11 (current prices here in Germany). Going for a top spot in 3DMark FireStrike requieres four good vgas - four out of 15 have to be good in this calculation. To go for a top spot in SuperPI 1M it needs one good cpu - one out of 27 in this calculation (if you go for a 6600K its even more). So this calculation gives you a seven times higher chance to find the "golden" piece of hardware compared for 3D. Again, here you have an advantage for 2D. And because of this mathematical advantage lots of overclockers bin cpus (even people without any kind of support). If they would save the money for binning and go for 3D, most of them would have been able to buy a decent gpu. They don't do so because they know that even the best vga is held back by a crappy cpu. So in many cases waiving on 3D is not a matter of money. Than there is the fact that even a 3 GHz-Titan X could only compete in certain 3D-benchmarks. A 7 GHz-5960X could compete in 2D-benchmarks and rule lots of the old cpu-bound 3D-benchmarks like 2k1, 2k5, 2k6 and AM3 with a mediocre vga. I know that this is a very simple calculation but again this is one more argument for going 2D. Going back in history we (me and may teammates masterchorch and hoschi) needed two QX9650 to achieve a top-10-Score in 2k6 (achieved without a single drop of ln2). We needed only three Nehalem-CPUs to find a wr-cpu. With Gulftown we needed seven or eight to find a 7-ghz-chip (6 cores, 12 threads, taking the AM3-wr at 6.8 ghz) and since Sandy Bridge even a few dozens chips aren't enough to find a decent one. That is an unhealthy development. Lets move on with another advantage for 2D. Trying to achieve a global first place in 2D doesn't necessarily need any kind of modding. Sure, delidding your Skylake-cpu is very likely. But that's the only mod needed nowadays. Go with a 5960X and you never need that. Mainboard, memory and cpu - no need to mod it. Now try to achieve a global first place in Fire Strike. You have to fully mod your vga. Soldering skills and additional parts (i.e. epower or such things which of course cost money) are needed. What's if your 5960X and your vga dies? As long as your cpu isn't physically damaged you have the chance to rma it. Try this with an epowerd Titan X. Of course this argument isn't valid for all cpus (delidding for the win) but its a true for X99 at least. Next one. Roman is offering ln2-pretested Skylakes now. You can buy a decent cpu now with warranty even with delidding and ln2-use. If it fails (and you didn't damage the cpu physically) you get back the next decent chip without going through dozens while binning. Have you ever seen this with vgas? Again you have a slight advantage when going for 2D (of course you can use such a cpu for older 3D-benchmarks and have the same advantage, but a dead epowered Titan X can't very likely be rma-ed). I know that my arguments are nearly black and white and that there are colours between. But every argument is valid for a specific situation - and every single of my arguments favours 2D. You can have a competitive rig for 2D with less money than 3D or you can bin more hardware when both "sides" spend the same amount of money. And even if someone face all disadvantages and benches 3D he gets less points for a top spot. You can see the perfect example today on the hwbot-frontpage: Yesterday 8 Pack broke the global world record in 3DMark05. And this global wr comes in in 5th position on the frontpage - beaten by a XTU-score from Bullshooter of 740 points. In the hwbot-database there are more tan 23600 faster results in XTU than the score from Bullshooter. In addition, there are six other overclockers which reached the same score earlier than him in the same hardware-category. And yet he beats a global world record on the hwbot-frontpage (and lead to believe it's a bigger achievement). Now tell me: Why someone should spend the money and time for challenging a global in 3D? I'm sorry for Bullshooter picking his very good score but it shows perfectly whats wrong with the hwbot-algorithm. The frontpage screams "Go 2D, you'll never reach anything relevant in 3D". Since this is the situation for years now lots of people see this, decide for going 2D and widen the gap between 2D and 3D even more. This has to be fixed as fast as possible! Otherwise 3D will be irrelevant in a few years - low points, less benchers, lower points, lesser benchers and so on. That has nothing to do with the argument that 2D is much more popular. Due tu the algorithm people are pushed in this direction (as it was the case when hwbot differentiated between single- and multi-vga -> every top-overclocker benched dual-gpu-cards). And since this 2D-direction has lots of other advantages (lower platform cost or higher binning budget, usually less benchmark-duration, lower cooling cost, more hwboints) nobody needs to wonder why 2D is more "popular". It's simply because how hwbot worked for the past years. hwbot reached a state where it is the only relevant database for oc-scores. No ripping.org anymore, Futuremark-HOF is irrelevant, forums like XS are nearly dead. Everyone interested in overclocking will use hwbot to share his scores. And therefor hwbot has to take responsibility for all aftereffects of the algorithm. If someone has read the whole post - thanks for this. Maybe other overclockers think the same, others may see it the other way around. Hopefully my english is good enough to all my arguments understandable. Thanks for reading.

I think it's a mix of both statements. 2D has a lower overall cost since you don't need a modern and expensive vga. Vga-cooling isn't neccessary so you can save money again (no ln2/dice-pot, no phasechange or waterblock). You can run a weaker psu and can save money again. And 2D currently gives higher points on hwbot. By the way: 3D-benchmarks usually take longer than lots of the current 2D-benchmarks (as long as you use modern cpus). So you can save time when going for 2D (and we all know how time consuming this hobby can be). A few days ago aerotracks visited me and benched his 5960X. While he did half a dozen hwbot-prime-benchmarks I couldn't have finished one single 3DMark06. If someone is interested in points and rankings he will go the 2D-route. Less money involved, more points and less time-consuming - three advantages at once. Why should someone invest more money for 3D-benchmarking as long as he isn't addicted to a particular benchmark? He has to deal with additional cost, has to invest more time and in the end he gets lesser points. Well done, mission accomplished! In my oppinion 3D-benchmarking should be the highest awarded category. You have to deal with additional hardware and additional cooling (especially when running multi-gpu), more money is involved, a single benchmark usually takes longer and every driver-release can change the game (at least for the modern 3D-benchmarks). I want to make clear that I don't want to devaluate 2D-benchmarking but for me there are no valid arguments why 2D should be awarded with higher points than 3D.

In my oppinion the popularity of a benchmark here on hwbot depends on the amount of points you can "earn" with it. Like aerotracks said, 970 isn't really worth benching. Why? Because of the algorithm hwbot is contributing points to it. And thats the case for lots of other hardware. Don't mix up popularity with the chance to get lots of points with a certain benchmark or hardware... Back in the days where 3d-benchmarks were distinguish in single-card and multiple cards, nearly every 3d-bencher tried to bench a dual-gpu-card. After hwbot started to count gpus, dual-gpu-cards lost nearly all attractivity. The same applies to multi-gpu-setups. Multi-gpu is often a tough challenge but nowadays you don't get many points so nearly no one is benching such a setup. And because nearly no one is benching such setups the algorithm of hwbot contributes less points. All decisions made at hwbot relevant to the point algorithm have effects of the "popularity" of a benchmark. People choose to bench these setups where they get the most points. So again, don't mix up popularity with the chance to get lots of points... By the way: For the current XTU-WR Dinos22 takes home a total of 429.4 points (UWP, UGP, UHP, GTPP, HTPP). 8 Pack with the current WR in 3DMark FireStrike Extreme gets 233.4 points - thats only 54.4 percent for a benchmark, that needs more attention. More hardware involved, more pots and ln2 involved and a much longer duration of the benchmark. That relationship doesn't fit anymore and even a slight variation to this algorithm won't fix this...

Thanks (as always).

And another one: http://valid.canardpc.com/show_oc.php?id=210925 ABIT KD7-E Should be the last old validation for a while...

Great, thank you.

Found a DFI LanParty UT NF590 SLI-M2R on Ebay, still looking for the rest...

Thank you very much.

I have an old validation from an mb-review from ASUS Sabertooth/GEN3 R2.0: http://valid.canardpc.com/show_oc.php?id=2793165 Please add this board. http://hwbot.org/submission/3062304_masterchorch_reference_clock_sabertooth_990fx_r2.0_342_mhz?recalculate=true

Found an OCZ PowerStream 520w and a DFI LanParty DK 790FXB-M3H5, still looking for the other stuff...

YGPM

No one? Looking for a DFI LanParty UT nF3 Ultra-D too...

still looking for this parts...

I'm currently looking for older hardware. For most of the parts I don't have a plan to overclock and benchmark them, I just want to own these parts. I'm looking for: ABIT IC7-Max3 ABIT KD7A ABIT KN9 Ultra ABIT KV8-Max3 AMD FX-62 -> found one at Ebay AMD FX-74 ASUS A7N266-C -> thanks to bartx ASUS Crosshair (NF590) -> found one at Ebay ASUS L1N64-SLI WS Chaintech CT-7NJS Ultra Zenith -> found one at Ebay Chaintech ZNF3-250 Zenith DFI LanParty DK 790FXB-M3H5 -> found one at Ebay DFI LanParty JR GF9400-T2RS -> found one at Ebay DFI LanParty KT400A DFI LanParty NFII Ultra B DFI LanParty UT ICFX3200-T2R/G DFI LanParty UT NF590 SLI-M2R -> found one at Ebay DFI LanParty UT NF680i LT SLI-T2R -> found one at Ebay DFI LanParty UT 790FX-M2R -> found one at Ebay DFI nF4x Infinity FOXCONN A79A-S FOXCONN BlackOPS -> thanks to bartx FOXCONN C51XEM2AA -> found one at Ebay FOXCONN Destroyer Intel Pentium Extreme Edition 840 -> found one at Ebay Intel Pentium Extreme Edition 965 -> found one at Ebay MSI K9N SLI Platinum -> thanks to bartx Mushkin PC-3500 Level II 2x 512 MB with black Heatspreader OCZ Powerstream 520 Watt -> found one in another forum SAPPHIRE Pure Innovation PI-A9RX480 SAPPHIRE Pure CrossFire PC-A9RD480 SAPPHIRE Pure CrossFire PC-A9RD580 SHUTTLE AN50R / AN51R SOYO P4I875P Dragon 2 SOYO SY-KT333 Dragon Ultra SOYO SY-KT400 Dragon Ultra XFX HD 4890 Black Edition All parts have to be functional and in good condition. Shipping to Germany is required, payment with PayPal is preferred. Offers welcome.

Is Skylake allowed for stages like DDR4 CL10 and fastest 3DMark99 1x 6600GT?

Like xxbassplayerxx stated the run took about 77 hours. So your estimation is roughly correct. That's sad to hear. Hope you have better luck with your next try. As you can see on the photo of my rig I used aircooling for the cpu and stock aircooling for the gpu. No problems during the run. Today I booted up the pc (I wanted to improve my score in cpu-challenge stage 1) and I realized, that the fan of my 8400 GS didn't spin. Seems to be pure luck that this didn't happen during my long AM3-run. Thank you guys. I never thought that such a low score can get this much attention. *g* To be honest I only tried the second application once. Maybe it was bad luck, maybe there is a problem too - I don't know exactly. For now I'm done with this benchmark. I will move on for some other stages to get a result. But if someone beats my score and it's enogh time left (a week? ) I will give it a try again.

Found an E760 with ECP in a german forum. Still looking for ASUS Rampage III Extreme and ASUS Rampage III Extreme Black Edition.

Still searching... Had a few offers but didn't buy anything up to now...

Sounds very promising. I'll shoot you a pm...

Ok, sorry. Didn't know that.

Would be great, thanks in advance! Sounds good. Anything to say about the board? Any problems or is it working like a charme? 170 € is with shipping included? If not, how much will be shipping to Germany (your profile says your from France?)?

Since a few days I'm looking for X58-boards. Only two boards come into account (in descending order): 1. ASUS Rampage III Extreme Black Edition 2. ASUS Rampage III Extreme 3. EVGA X58 Classified E759 4. EVGA X58 Classified E760 5. EVGA X58 Classified E762 -> found one E760 + ECP Board has to be in a good condition (no bent pins, no faulty memory slots). Any kind of isolation is ok (will be used with cascade). Shipping to Germany required. Price depends on condition (will pay up to 200 EUR, shipping included). Then I'm looking for XFX HD 4890 Black Edition 1000M. I want to buy up to four cards. Card(s) have to work without issues, isolation isn't neccessary. Price depends on condition, whereas I can buy "normal" XFX HD 4890 for about 20 to 30 EUR (shipping included). So maybe 40 to 45 EUR (shipping included) is max for Black Edition. Shipping to Germany required, too.