UnRAR-crack is a free RAR password cracker. It employs various techniques to maximize cracking speed (e.g. using SSE2, SSSE3, etc. to test more passwords at once, and using multithreading to take advantage of multi-core/hyperthreaded processors).

 

UnRAR-crack benchmark uses the password cracking core of UnRAR-crack to test the performance of your system. It works by launching several thousand password checks and measuring the average testing speed, in passwords per second. It repeats the same procedure for different number of threads (up to the max your system can utilize), and using different crypto functions (SSE2-optimized, SSSE3-optimized, etc.). Using it, you can measure the SMP boost, the SSE unit performance, etc.

 

The whole benchmark takes a minute or two. After finishing, the program generates a report file. This file can be uploaded to the result database so you can compare your machine with your peers.

 

(~ http://anrieff.net/ucbench2011/)

 

Just came across this benchmark - looks pretty cool as it reminds me of the simplicity of PiFast .

 

I've set up a beta-testing competition for this benchmark. Not much limitations: just a screenshot and a verification link needed. All is explained here:

 

To test the performance of your machine, follow these steps:

 

1. Go to http://anrieff.net/ucbench2011/download.php and get a build for your OS (Windows, Linux and Mac OS X supported);
   
2. Extract the .ZIP file somewhere and run the launcher;
   
3. There you should fill in your "username" (e.g. the your nick from the forum) and motherboard model. The program automatically detects CPU type and speed, but you can override its guesses if you find they are incorrect . You also can append more info in the large box (e.g. the full PC specs);
   
4. Hit "Run". The test takes a minute or two.
   As always, it's best to turn off any powersaving features. Same for turbo modes, etc. CPU speed should be constant throughout the test;
   
5. When ready, the test generates a file like report_XXX_X.txt in the directory where the launcher resides. To upload it to the result database, use the following link: http://anrieff.net/ucbench2011/upload.php. The database itself is here.
   

 

Have phun! The test depends mainly on CPU. Memory and mobo don't play much.

 

Link to the competition: http://hwbot.org/competition/ucbench2011

 

This competition is set up to allow community members to beta-test this new benchmark. Just like with Unigine Heaven, it's meant to uncover bugs or other problems. If you come across a problem, don't hesitate to post in the forums .

 

Have fun!

downloading.......

will this benchmark be eligible for points in the future?

It's for that that there is a challenge, to test if it will be OK

cpu speed is not detected correctly...huge diff of 70mh while detecting cpu speed

http://hwbot.org/community/submission/2110777_slammer95_ucbench_2011_beta_core_2_e8600_3.33ghz_389.2_mpt_score?new=true

 

this is mine

 

25 Mhz difference from cpu-z and benchmark in cpu speed.

 

cpu isn't detected correctly, it says only wolfdale core due duo (6M)

 

 

I hope it will be useful for you

Edited January 21, 201114 yr by slammer95

Well this is mine: http://anrieff.net/ucbench2011/viewitem.php?id=134

10mhz+ cpu speed at benchmark window.

 

But which is the final result.?

Well this is mine: http://anrieff.net/ucbench2011/viewitem.php?id=134

10mhz+ cpu speed at benchmark window.

 

But which is the final result.?

 

the way I understand this benchmark 264.7 p/s would be your final result.

http://hwbot.org/img/examples/UCBench%202011%20%28beta%29.jpg

missing

 

I beat MM for once!

http://hwbot.org/competition/ucbench2011?tab=stage%202

Hm, okay, let's see how it can be fixed.

 

So, what do you guys think about the bench?

nice one....good for determining CPU performance...need a little brush up on the cpu detection engine though...the launcher gives weird ratings at times.

btw.....if i run an 4 core "daneb" as "heka" core...will it be valid?

It's CPU usage is bizarre, at the halfway point it was only using three cores, then six, then 9, then 12 by the end.

I was positively surprised to see that it did work just fine on my old Pentium I. I did not take very long to run either. :-)

I wonder how cheat-proof it is though? I'd rather see that you add the distributed.net client benchmark to HWBot though (you can benchmark multiple OSes, CPU architectures and GPUs with a single benchmark and it is as simple as pifast). :-)

@above

link please....

@above

link please....

 

Client software:

http://www.distributed.net/Download_clients

 

Tables with benchmarking data:

http://n1cgi.distributed.net/speed/

 

Run the client with the prameter "-benchmark" from a command prompt/terminal window.

100MHz difference for CPU clock reading. Otherwise, good bench. Took "fun" to run it.