GENiEBEN

du ungezogener moose =))

OK, PM sent. Add a link to this thread so people know what's up.

Can confirm, although I'd prefer selling with board and ram too.

Ok, will do it when I'm back from work.

Please include CPU-Z Memory tab in your screenshots.

Tweaking Windows to improve scores for UCBench is like removing the backseats from a Civic for acceleration improvement.

Of course there is, but there aren't as many gentleman's as expected.

Let's be gentlemen's and share our knowledge for this rather popular benchmark. Since I've started the thread, I will start with my share of tweaks I've been using (partially) for the last 2 years. Operating System * Preferred OS, in order of performance Linux MacOSX Windows NT5 (XP / 2003) Windows NT6 (Vista / 7 / 8 / 8.1) * General OS Tweaks For Windows platforms there are no specific requirements, other than having as few processes and services running. Do not try achieving this by using DIAGNOSTIC MODE, it will generally score lower than in normal mode with startup items disabled etc. Command Line Parameters * Using High Priority for unrar.exe You can achieve this in various ways (up to your general knowledge). Since 99.99% of the users are using Windows for this benchmark, they could use: start /high /b unrar bench test.rar Make sure that launcher.exe was started with Administrator rights for this to work. * Setting unrar.exe intensity unrar bench test.rar -ri# # is a number from 0 to 15 You can also set a sleep time (in milliseconds), which seems to also help unrar bench test.rar -ri# $ is a number representing the amount of milliseconds to sleep. It's up to you to test values, personally I've tried up to 10000. * Changing the Instruction Set It is important NOT to run using the "All Supported" option. General rule is to select the instruction set available before the newest supported. If your cpu supports SSE4, then SSE3 would be the preferred choice. The only exception here would be on CPUs with only SSE2 and i386, where you would definitely want to choose SSE2. Note that it's good practice to also test the latest instruction set, but it will rarely score any better. On AMD processors use XOP if available (Bulldozer and newer). Note that XOP only works under: Windows 7 SP1 / R2 SP2 / 8 / 8.1 OR Linux with kernels newer than 2.6.30 * Changing the Threads used THIS IS NO LONGER ALLOWED, NOTHING TO SEE HERE. ====================================================== UPDATE1: 07/SEP/2014 Thread changing removed.

Let me photoshop that

Yeah, I know.

They don't work, he can set whatever he wants

^ Yes, all good now, have fun.

Press EDIT on the right side of the page, and replace one of those 4 screenshots with a system picture.

My bad, I see what you mean. I've tried editing & recalculating the #2 result but to no avail, you'll have to wait for the bug fix.

All OK, just make sure the UI si also in the screenshot.

Nothing wrong there? You subbed first, you got the points.

Lol, made them long ago.

You just suck at it

It doesn't but it is VERY cpu dependent.

Well for starters he's using an old version that is no longer supported. As for detection, the GPU info is provided by GPU-Z, which always puts the IGP first, so in a IGP+Dedicated the default card used in submission form will be the IGP. Haven't found a workaround yet.

Nice, is this the dv6000?

Please edit OP when upload is done

Please upload the txt file to http://anrieff.net/ucbench/upload.php

Strange, but did you also check out the patch I posted?