Slashdot Mirror
35,000 Linux Benchmarks In a Week
G3ckoG33k writes "Openbenchmarking.org has received 37,027 benchmarks (mainly Linux, and some Macs) in the first week since its inauguration. 241,384 completed tests using 468,344 components from 438 hardware vendors. All results submitted by end users. I guess the hardware support for Linux must become even better thanks to this effort. Yes, the benchmarks are easy to install and run, and you can readily compare your own system anonymously with the results already submitted, using any or all of hundreds of free applications in 47 categories."
First Post Slashdot? I didn't know the kernel was getting that specialised, but you proved it is.
Expert in software patents or patent law? Contribute to the ESP wiki!
My biggest issue with most benchmarks is the underhanded deals between sites/mags and tech companies who provide incentives and free hardware. While I take individual benchmarks with a grain of salt as well for a lot of reasons, I'm far more confident that thousands of people haven't been bought and sold in the way ad supported publications seem to be.
Probably because you can benchmark every mac made in the last three years in a couple days. You don't find many people wondering about which part they should upgrade on their mac.
Also, it's crowd sourced software that runs on OS X, so mac users probably just aren't as motivated.
I'm not sure why they are spending so much time on Linux, when it is used so little for actual desktop work. This site should really set its focus on OS X which certified Unix (unlike Linux which is just a non-standard proprietary clone).
By your same logic, they should also devote more time to SCO products ...
Windows users have nothing like this. Too bad for them.
Windows users have hundreds of different benchmarking tools to choose from, so they are in kind of a better position. 3DMark Vantage + Aida64 is already plenty enough to benchmark more-or-less all components of your system and if you're still not satisfied then throw PCMark in the bunch. Phoronix Test Suite apparently allows you to craft custom tests too, but aside from that it doesn't offer that much functionality that isn't available for Windows users already. Linux users on the other hand have lack of good benchmarking tools and suites and for them Phoronix Test Suite is one of the few and arguably the most full-featured one.
The article summary here intrigued and excited me. I headed to the site, eagerly hoping to compare the performance of my system against the 35,000 submissions, and to submit my own results for the community to share. But when I got there, I was lost and confused. The site is almost completely incomprehensible to me. Navigation is a nightmare. I can't figure out how to see the results for a given system. When I do, by accident, stumble upon a page with some graphs, I can't figure out how to compare these with another system. There seems to be no rhyme nor reason to the hyperlinks, and precious little in the way of explanations and guidance. It's a pity that a site which has the potential to be such a useful resource has been rendered completely unusable by poor organization and UI design.
I've had a quick look, this is still about "bigger numbers are better". It would be nice te have more focus on power usage of modern computers!
Your old blog from 2007 at http://blogs.technet.com/b/jonjah/ tells us right away that you are a shill, a self-admitted MICROSOFT shill at that.
Go get a life.
Here's what it has to say about that:
Lenovo ubuntu 10.10 is a motherboard. This product is available from Lenovo. The Lenovo ubuntu 10.10 has been tested via the Phoronix Test Suite in the configurations listed below.
Let me know when they've sanitized their DB.
So what is the best benchmark (however that is defined)?
Phoronix test suite
The venerable UnixBench BYTE magazine lineage, updated by Yahoo.
Geekbench
I'm not a lawyer, but I play one on the Internet. Blog