Introduction to Maya - Modeling Fundamentals Vol 1
This course will look at the fundamentals of modeling in Maya with an emphasis on creating good topology. We'll look at what makes a good model in Maya and why objects are modeled in the way they are.
# 1 24-05-2003 , 07:22 PM
tuddel's Avatar
Registered User
Join Date: Jan 2003
Location: ph
Posts: 36

expert help here!

hi! i just benchmark my maya 4.5 (linux and windows)

for windows (win2k sp3 detonator 40.72)

for linux (rh 7.3 maya4_5-4.5-90 xfree 4.2.0 (come with redhat 7.3) nvidia glx 4363) (kernel i686 bigmem)

hardware pentium 3 933 1280M(pc-133) quadro2 mxr tuv4x (bios pnp off) (4x agp all optimization in bios) (both os in 1 harddisk sperate partition, NTFS and ext3fs) no swapfiles.

well i used the rendertest.ma (found at highend3d) to benchmark my system, i made a couple of benchmark in command line and batch renders(gui) it seems that win2k is faster than a linux system with a default kernel mem, then i tried to recompile my kernel (cause i couldn't believe) but still no lack win2k is still faster (about 10 - 30) sec (command , gui respectedly)

anyone can give some nice suggestion about this?

# 2 25-05-2003 , 11:11 AM
BabyDuck's Avatar
Subscriber
Join Date: Nov 2002
Location: USA
Posts: 1,170
what about if maya is just better optimized for windows than it is for linux. if i were a|w i would optimize the software that most of the users use - not sure if there is more windows than linux users - just an assumption on my side.

# 3 25-05-2003 , 12:36 PM
badbunny's Avatar
Subscriber
Join Date: Sep 2002
Location: Lincoln, UK
Posts: 79
what was the total rendertime on the system - I mean is 10-30 secs a very small percentage ?

Also, are the times in Win2k and Linux repeatable ?

I don't know if its using the hard disk much but the hard disk is not a uniform speed across the platter (I know you said swaps were off) - is your Linux partition in the slower area? Although to be honest I wouldn't have though this would have made a massive difference unless its thrashing the hard disk around a lot though - on some computations at work we can get hours of difference on a twenty hour moderately hard disk intensive analysis depending on how full the hard disk is and where its writing to.

Clutching at straws really....

# 4 26-05-2003 , 09:14 PM
tuddel's Avatar
Registered User
Join Date: Jan 2003
Location: ph
Posts: 36
cli win2k fastest 3:01 mins, 3:05 (repeated many times)
gui win2k 3:06 - 3:11 (repeated many times)

cli no x linux 3:13 - 3:17 (repeated many times)
x kde - 3:21 - 3:40 (repeated many times)

i'm planning for a rebenchmarking again, i have already formatted my drives, and some little tweaking, and im all set, and maybe i'll post the render logs here, hope i really made a mistakes, cause i know linux must be better againt windows when it comes to rendering, cause linux can be optimized and scalable,(renderfarms) , and i have read some articles in the net that claiming maya linux is better than win(sort of like that) " Thanks in part to Maya, Linux is becoming the technical desktop of choice for the film industry. "

https://www.linuxjournal.com/article.php?sid=5235

but i can't really make a conclusion, haven't tried it on a different hardware, but for now i'll just rebenchmark again my newly format setup system system, and i'm hoping that this time linux will win,

and guys, any good maya tweaks or system tweaks that i can add, before i make the test, (will render tom. evening.) thankx


Last edited by tuddel; 26-05-2003 at 09:19 PM.
# 5 26-05-2003 , 09:29 PM
badbunny's Avatar
Subscriber
Join Date: Sep 2002
Location: Lincoln, UK
Posts: 79
Linux is better than Windows at some things, but in terms of how quick a program can do something, as long as the software isn't really badly written for one operating system, it shouldn't really be that different with the same hardware, coming down to just processing power.

We use finite element engineering simulation software at work (lots of cpu, LOTS of memory, lots of hard disk thrashing) and I've been doing lots of benchmarking between Linux and Win2k versions of software on the same hardware. To be honest I've found little difference between run times on the two, with some cases the Win2k being quicker. The only thing is, the Linux machine is still capable of doing other things and is fairly responsive, whereas the Win2k machine is completely unusable while its doing the computation.


Last edited by badbunny; 26-05-2003 at 09:32 PM.
Posting Rules Forum Rules
You may not post new threads | You may not post replies | You may not post attachments | You may not edit your posts | BB code is On | Smilies are On | [IMG] code is On | HTML code is Off

Similar Threads