websmile

Solid

I checked and obviously the rumbles not only had no stages when I found them but also zero limitations, which I did not see. I fixed this, should be OK now and removed results of non rookies and Intel stuff. Might be my last task here at admin panel^^

Only amd, sorry, use rookie rumble #51

to everyone, health and luck at 2018 to all members of the hwbot family

Nope, a mod who apllies to the rules would block it. You did not only submit it wrong due to your big expertise (8x cpu instead of one cpu) but also did hide the rendered image which has to be shown mandatorily for the result to be valid. Sorry that you are unable to read rules and to understand that one cpu is one cpu even if it has eight cores - maybe you learn from it. Happy new year nontheless^^

Only one, the winner gets it

Stage 3 was not removed because of your score but because of tests and several scores made by different guys. Too dangerous to let it stay

The 750 ti does not support sli and you submit results that are beaten by single cards and locked i3 at vantage for example for sli to grab cups? Very funny

The superposition stage will get deleted and I will add a UH Extreme stage instead. For this stage I prolong the contest to 20th of january

If you use an external hoster you can enter the pictures as thumbnails which lead to the big versions in a new tab example

Please note that push is allowed every 24 hours (marketplace rules) thanks

All downloads work for me, they lead to website of piblisher or direct download at 3dm03, but I need to use the normal left click and cannot download with right click. Same as on hwbot directly, also the 03 being shown as shortened url problem is same. All works nontheless

I do not understand the question about the download - if I click the ocesports download link at cheapazz it leads to unigine site directly for download link, so there is no difference for me. Does it lead you somewhere else?

The benchmark was tested with a couple of different GPUs, but you can´t test new benchmarks on all different cards available. For this competition, hwbot did not determine the benchmarks used but the community did at a special preparations thread so you are asking the wrong person for "if this was chucked in"^^. The benchmarks we use normally are tested and bugs are known unless we make special stuff like the realbench beta comps in the past to evaluate new options

I might exchange stage 3 if we cannot find a solution that either shows these runs are tweaked or bugged. It is always risky to use newer not much tested benchmark on rarely used hardware. This decision will be made saturday/sunday most likely because everyone will feel bad if we have to rely on chance.

We might have a problem, maybe you can give me an explanation for results like these http://hwbot.org/submission/3672434_sergeyr_gpupi_for_cpu___1b_2x_xeon_x5690_46sec_718ms/ in comparison with this http://hwbot.org/submission/3410802_sparkey247_gpupi_for_cpu___1b_core_i7_extreme_980x_4min_29sec_662ms and this http://hwbot.org/submission/3668752_splave_gpupi_for_cpu___1b_core_i9_7920x_1min_26sec_607ms/ and especially these http://hwbot.org/submission/2860344_skulstation_gpupi_for_cpu___1b_xeon_x5690_4min_47sec_284ms http://hwbot.org/submission/3626243_skulstation_gpupi_for_cpu___1b_2x_xeon_e5645_59sec_189ms I think there definitely is a more than weird scaling and results of multi cpu xeons are real dodgy

The only changes made at GA were that I removed more results, and mainly because the raymarch bug was happening more often than usual. On the "bugged" runs, so how to see these? minimum and maximum fps are completely in line for the results. We cannot block results because of efficiency being too good if we do not have more hints. This is a quite new benchmark and it has not been run on lower end gpus much. What is even more hilarious, the community discussed the benchmark choice and hardware for months and obviously no one did in depth testing.

To answer some questions, GPUPI 3.0 is ok for me and we stay with extreme at superposition like agreed on the thread, too late to change this I think. The result tab shows min and max fps at least

There is no information if the site will return and when

The aquamark file is the problem, it has 5,01MB - the screenshot of 3d2000 you sent is easily submittable if you want to use the result

I checked and tried to upload it and got same result like you, 5MB is huge file above the limit. This happens sometimes, mostly if you bench and screen at high resolution or something goes wrong at normal settings

You can use both backgrounds, I found the new version with Galax sponsorship and improvized.

George Storm I heard

10KB is much too small for a valid aquamark file, but 2MB is too big for rig picture or discrete screenshot. Try to downsize this a bit. P.S. you can send the file to michael@hwbot.org so I can check it

I do not remember rtls, but I used a normal test pattern for subs like at this screen