Dual Core Linux Performance: Two Penguins are Better than One
by Kristopher Kubicki on July 1, 2005 5:55 AM EST- Posted in
- Linux
Multitasking Scenario 4: DVD Burning
Thanks to NeroLINUX, we can now run the first benchmark we did in reverse. We will be paying particular attention to the buffer for NeroLINUX. If the burn continually empties the buffer, then the system is playing catch-up to the burner. We will count the number of times the buffer dies and graph this separately. We used a NEC 3520A burner and a 4.5GB ISO image for the burn.
- Open FireFox 1.0.4 and load all 5 web pages.
- Open XMMS and start playing a Nine Inch Nails CD ripped to Ogg
- Open Thunderbird for news
- Login to our news server and start downloading headers for our subscribed news groups
- Begin an ISO burn with NeroLINUX and start timing
There seems to be a very linear correlation between the number of buffer dumps and the speed at which the discs burned. We really only saw the buffer even come in danger of decreasing at the beginning of the burn and when Thunderbird took a little too long to process something coming in from the news server.
Even though the processors we used in this analysis are more than capable, the additional tasks of FireFox and Thunderbird proved just a tad too much for the chips to handle at all once. This benchmark is a great illustration of where dual core (or dual socket) goes so wonderfully well on the desktop. The additional overhead to run XMMS, FireFox and the other applications is not very substantial, but when FireFox decides to chew up resources while the only CPU is managing the write buffer on a DVD burn, the results can be quite dramatic.
Facebook
Twitter
RSS
69 Comments
View All Comments
sweatshopking - Saturday, May 9, 2009 - link
maybe we should take the linux tab off the top since nobody has updated since '05....PrincessNybor - Sunday, June 21, 2009 - link
I was just going to say the same thing. If Anandtech isn't going to cover Linux, then just remove the tab and call it a day. If you are going to even pretend to make Linux a priority, then perhaps at least ONE article per YEAR would be a nice way to start. No articles since 2005 is just embarrassing.Ryan offered that Ubuntu article well over a year ago, and has offered nothing but excuses since then. I realize you are all busy, and I don't fault him for anything besides not coming out and saying that Linux isn't a priority at Anandtech.
rossmcdonald - Thursday, September 11, 2008 - link
Didn't AT talk about doing a linux for a month article many months ago?Milleman - Sunday, September 14, 2008 - link
Yepp! He/they did!What happened to that one? They've probably lost it down the drain somewhere. Or is M$ paying them to not favor any other O/S now...? Somehow it seems that AT lost all interrest in Linux. Bad move...
forkd - Monday, November 17, 2008 - link
I agree, Anandtech wasn't a bad source of linux info 'til this article stopped everything.I'm not really upset but I am a little disappointed.
At some point they should take the Linux tab off of the website.
Years ago I subscribed to consumerreports.com for information on products I was considering on purchasing. I quickly realized that most of the articles on things i looked for were out dated by years.
Of course I never paid for my anandtech subscription so I can't complain about them leaning toward more profitable press.
Milleman - Sunday, June 15, 2008 - link
Hello??Is this the latest Linux article from Anand, made 3 years ago...?
Come on...
fstratzero - Saturday, March 22, 2008 - link
I'd say testing linux is freaking hard to do. I'm a Gentoo linux user.And having complied tons of kernels over the years, performance can range greatly in what options you use in your kernel. I found that the CFQ scheduler can hurt performance since it tries to spread out cpu usage evenly between all processes.
That and also with linux you could assign a single process to a single core. Alternatively you could "nice" the application ie lower priority and still have a really usable system.
People in our forums will compile stuff on one cpu and use the other for playing games while having tons of crap running in the background, and report their seemingly invincible performance feeling.
Although I don't know much about Suse, I'd say this review is pretty well done for people that don't "get under the hood." Yet if you do get your hands a little wet with nice and taskset, you can do just about anything and hardly feel a slow down.
orbatos - Sunday, November 5, 2006 - link
Why no Shake vs. Shake, sure 3.5 isn't available, but 3.01(?) would be fine. Where's GIMP on windows? NewTek's LightWave® 3D 7.5b != ScreamerNet 7.5b , Why not go Blender vs. Blender or Maya vs. Maya? Who uses Mozilla Composer as an alternative to Dreamweaver as opposed to the myriad of other available web dev software? Outlook? what about Thunderbird on Windows? XMMS vs. iTunes doesn't make sense, they're different types of applications, try Rhythmbox. WinZip vs. Gzip doesn't make sense, try comparing a GUI tool. MS Office vs. OO.o, at least try the windows versions! If you're comparing workalikes as opposed to the same software, don't compare Nero vs. Nero, the Linux version is terrible.Suggestion: WaveLab is a do-all of audio manipulation, not the type of software that has direct analog under Linux, try looking at suites of software that can be used to the same effect all through a similar interface (e.g. JACK). Antivirus? ClamAV sounds decent, did you even look?
hojit - Tuesday, July 12, 2005 - link
Has any one pointed out the mod15 error using the sil3112/3114 and the seagate hd's the hardware in the article seems to suggest a conflict on the amd system. I had this same issue and it killed all my speeds for things like transcoding and any thing hd intencive.MarcusAsleep - Saturday, July 9, 2005 - link
Howdy!I was just wondering what kernel and kernel settings you were using. Since the article was on multi-tasking it seems especially relevent since the development of low-latency/preemptable options in recent years!
Any clue?
Mark.