Massman

Updated leaderboard

114-ish with 38x ratio seems accurate with ambient cooling.

Regarding the AMD/NVIDIA split up. The reason why @Strong Island's idea resonates with me is because it's a separation which is more closer to how people segment graphics cards in real life. No one evaluates or buys a card based on the SP or ROP count. Mainly because it's not something that is really comparable across vendors and architectures. It would be like using a metric like pipeline length for processors (or any other metric related to architecture design). People do separate AMD and Nvidia in their segmentation. I agree it's less 'clean' or 'technical', but in this case I would prefer a more real-world segmentation. Just my opinion though, @Christian Ney has final word on this matter.

Alright, I'll have a look at the changelogs. At IDF I saw version v6.3 on display so maybe the next version will fix this.

That is strange indeed. Last question: what version of XTU are you using?

Welcome to the light side @wiLLius

Seems so, yea. I'm seeing similarities with free-to-play games. Some players enjoy grinding the game and putting in lots of hours to collect all the goals/achievements/skins. Others don't have the time to grind away and prefer to just spend their hard-earned money. I guess the binning is nice for those who want to have a reasonably competitive chip without spending time to bin yourself. In that respect, more binning sources should be better for prices

Hm, it shows 4 active cores in the submission link and the score is also in line with expectations. It's possible it shows 2 active cores in the XTU monitoring as the benchmark workload is ramping up. What do you see in task manager during the benchmark?

Previous TITAN: vendors had branded cards. Current TITAN: vendors have no branded cards. It's only interesting to try to aim for OC marketing if you have branded cards (even if it's just a sticker). Hmmm ... maybe vendor globals is something to consider.

Hi @eightyeightkeys, do you have a link to your XTU submission?

Why is the 4th category not 96-127? For future-proofness should we also prepare 128-159, 160-191 and 192+ ? Looks good, but I'd like to hear a more technical point of view on the segmentation per ROP count.

Ships from China, where Streacom is located

It's not impossible, but this will create a LOT of new global rankings. I just did a quick check with 3DMark03 and without verifying if all the graphics cards have their SP filled out correctly, there would be ~80 global rankings (instead of the 4 we have now). We'd also have to find a solution to group the older architectures that don't have the ROP/SP layout.

Just a quick message: it's available! Home - Open Benchtable

You can't opt out telemetry?!

One of those projects I'm extremely proud to be part of, but also will be happy when it's wrapped up

T-10 T-6 T-4

Is it still like that? I switched servers just now. Could be interesting to see hear if you still have issues

With R7, the following items affect this particular issue: Slope for rank 1 to 5: top-5 has higher points than the rest Points weight: depending on the weight, the points can be lower than the maximum of 49 Base principle: the points are based on the result in relation to the top results, meaning equal score is equal points The discussion on whether an equal score should get equal points is a really fundamental one and I can see valid arguments for both points of view. The right or wrong decision depends on your point of view, so it's difficult to make a 'correct' choice. Below two screenshots relevant to the XTU 2xCPU issue: Global points distribution: http://snaptest.hwbot.org/2016/08/11/2d8522ce7a6a205810d06ef0cd8c85661ca57f78.png Hardware points distribution: http://snaptest.hwbot.org/2016/08/11/cad7fc0877a808ebef7186903ab7d2a6cf9f252d.png A score of 742 in 2xCPU gives you (GL+HW) 166.7 points in R7 compared to 197 points in R6. Ranking 10th gives you 98.7 points in R7 compared to 197 points in R6.

From a community point of view: when Christian Ney has concluded that the revision can go up. From a developer point of view: earliest in September, after the final round of adjustments so that it only has to be updated once. From a business point of view: as soon as possible. This revision will very likely reduce our server costs by quite a bit. The way things are going, I think the new revision should be up and running by the time we are to enjoy Bristol Ridge, Summit Ridge and Kaby Lake. Our current point adjustment options are: Toggle Benchmark Points Toggle Global Points Toggle Hardware Points Adjust overall benchmark weight

Hey guys, Just got a new version of XTU: v6.1.2.9. Can anyone give this a spin to see if the bugged-high scores are still happening? http://hwbotdownloads.s3.amazonaws.com/downloads/benchmarks/Intel%20XTU/XTU-Setup_6.1.2.9.exe

Can you make a screenshot of the submission page?

You certainly have a valid point and we'll try to adapt the rules for competitions in the future to also accommodate for the Core i5 users. For this competition we will leave things as it is, however, as the competition is already on-going. In any case, I appreciate the feedback

Ah, try: (no https)

What format for the youtube URL video are you using?