Truebench: Test your system and provide results!

[oshunluvr]

Vinny and I have done this and the developer has allowed me to post here about it.

Truebench is a small program that runs a benchmark test on your CPU. Simply download the program, making it executable, and run it with nothing else running. Save the results by redirecting them to a file and email it to the developer for posting on the comparison blog. He wants "vanilla" results - no overclocking!

As a challenge: I'm currently at the top of the list with Vinny right behind me. Try and beat us!

[wizard10000]

Let's just say that my first-generation i7 laptop is a workhorse but it's not particularly efficient. landed between midrange i3 and low-end i5

I did send Kiran results, though. Neat project.

[oshunluvr]

I'm on there three times (so far ) The Q6600, the i3-4160T and the top-o-the-heap i7-6700K

[vinnywright]

I'm on there three times (so far ) The Q6600, the i3-4160T and the top-o-the-heap i7-6700K

told you you'd take it .

cool little benchmark !

VINNY

[whatthefunk]

My i5-2400 is #7. Just submitted it so its not on the list yet.

[wizard10000]

Mine is #14 and the Xeon directly above it is the server hosting Kiran's website. He sent me a note and said

Thanks a lot Allan.

I soon update your results. And mention your Hardware Installation (sample) which corresponds to this CPU model.

Wow this is a Q3-2010 CPU, which means may be the laptop is 2011/2010 model.

https://ark.intel.com/products/49024...cache-1_73-GHz

Quite Strange: I just now benchmarked a Xeon based server (my inmotionhosting) in which this website is hosted:
http://localhost/toffee-website/true...eon-e5-2670-v3
Which has the 12-core (and dual cpu) installed server which gave me 141,365,038 which is pretty close to yours 145,995,213.

This shows how far Intel need to scale down cpu performance and compromise so that they can accommodate 12-cores
on a single xeon cpu and make sure it works stable and never overheats

Intel(R) Core(TM) i7 CPU Q 740 @ 1.73GHz
>> Is a good sample. I am impressed

cheers, Kiran

[wizard10000]

Got my Christmas laptop today and emailed results in already, but...

Intel(R) Core(TM) i7-4800MQ CPU @ 2.70GHz
8 cores/threads

TrueBench-11 :: Release: 10-Sep-2016 :: (c)2016 :: Author: Kiran Kankipati

Starting TrueBench system benchmark test.
It may take several minutes/hours to complete depending upon your system.

Elapsed Duration:
in Seconds: 79
in Milliseconds: 78857
in Microseconds: 78857277

Looks like for the short term I'm #3 now

[Qqmike]

I get really, really bored when I don't have something useful to do. So ...

I didn't do too bad, on my mainstream, vanilla, on-a-budget, home-built 2015 PC:

mike@mike-desktop:~$ cd /home/mike/Downloads
mike@mike-desktop:~/Downloads$ chmod +x truebench_x86_64
mike@mike-desktop:~/Downloads$ ./truebench_x86_64

TrueBench-11 :: Release: 10-Sep-2016 :: (c)2016 :: Author: Kiran Kankipati

Starting TrueBench system benchmark test.
It may take several minutes/hours to complete depending upon your system.

Elapsed Duration:
in Seconds: 92
in Milliseconds: 91976
in Microseconds: 91976961
--------------------------------------------------------
CPU Info:
Intel(R) Core(TM) i5-4570 CPU @ 3.20GHz
4 cores/threads

--------------------------------------------------------
dummy_result=63
dummy_result2=63

Didn't see where he asked to be emailed with results, so I sent them to his contact email address in a text file.

[vinnywright]

Didn't see where he asked to be emailed with results, so I sent them to his contact email address in a text file.

thats correct

VINNY

[acheron]

This PC.

TrueBench-11 :: Release: 10-Sep-2016 :: (c)2016 :: Author: Kiran Kankipati

Starting TrueBench system benchmark test.
It may take several minutes/hours to complete depending upon your system.

Elapsed Duration:
in Seconds: 87
in Milliseconds: 86247
in Microseconds: 86247616
--------------------------------------------------------
CPU Info:
Intel(R) Core(TM) i5-4460 CPU @ 3.20GHz
4 cores/threads

[vinnywright]

This PC.

And an 8 core Xeon server, which shows the scaling down previously mentioned.

? ,,,,what do you mean by "scaling down" ,,,,,,,,, I'm slow .

VINNY

[oshunluvr]

From post #6 on this thread:

This shows how far Intel need to scale down cpu performance and compromise so that they can accommodate 12-cores
on a single xeon cpu and make sure it works stable and never overheats

[acheron]

From post #6 on this thread:

Indeed.

I got rid of my Xeon result though, as they were not using the full resources I thought there were.

My main machine is currently 5th in the table, which is not bad for something bog standard.

[wizard10000]

...I got rid of my Xeon result though, as they were not using the full resources I thought there were.

Yeah. My understanding is Kiran is only benchmarking a single core. I still think that's fascinating, in that my first-generation i7 got whupped by an i3

Also, a single core on my 4th generation i7 appears to be about twice as fast as on my first-generation i7.

They have made strides in power management, though - I've never seen my old laptop below 50°C except on a cold boot; the new (to me) one idles about ten degrees cooler and recovers more quickly from thermal load than the i7-740QM did.

Speaking of performance I was talking in another thread about tweaking intel_pstate. I finally landed on a solution that I haven't implemented yet but is the most elegant solution I've seen - I'm currently using laptop-mode-tools to switch CPU governors based on whether the laptop is on mains power or battery; apparently I can get rid of laptop-mode-tools and do what I need to do with a couple of udev rules.

Found this while I was googling - /etc/udev/rules.d/50-scaling-governor.rules

SUBSYSTEM=="power_supply", ATTR{type}=="Mains", ATTR{online}=="1", RUN+="/usr/bin/cpufreq-set -g performance -r"
SUBSYSTEM=="power_supply", ATTR{type}=="Mains", ATTR{online}=="0", RUN+="/usr/bin/cpufreq-set -g powersave -r"

which is exactly what I was looking for, and since udev is already running we don't need another utility to switch governors

edit: Well, the udev rule worked as expected but the kernel flipped the governor back to performance after being in powersave for about one second. Grumble.

[vinnywright]

Yeah. My understanding is Kiran is only benchmarking a single core.

yup ,,,single core , 1 set of instructions to carry out and timed ,

VINNY