Here are the times I clocked, looks like it's kudos to the G4
for the optimized write.
Mandrake Linux 10 on AMD 2GHz (J2SE 1.4.2_05)
Block Size [4096]
Loop Size [65536]
Standard Stream IO [1712]
Standard Random IO RWD [414516]
Standard Random IO RWS [409571]
Solaris 9 on SPARC
Here are the results of my tests on Randy's program on both Linux and
Windows:
Processor: AMD Duron 700 MHz
Hard drive Rpms: 5400
-
Linux computer
Libranet 2.8.1 (Debian-based distribution)
File system: Reiser
No explicit hdparm tuning performed
Randy
I have a Linux box that I should be able to run your program on if you
want to send it to me.
Daniel
Randolph Kahle wrote:
I have been investigating the performance of Java IO
on different operating systems. The results are so
different that I am concerned I am making some sort of
Randy,
I have a Redhat 9 Linux machine with a 1.8 Ghz P4 that I could run your
tests.
Eddie Dimond
Daniel Casey wrote:
Randy
I have a Linux box that I should be able to run your program on if you
want to send it to me.
Daniel
Randolph Kahle wrote:
I have been investigating the performance of
I'd like to run your tests on my systems too. I have a similar
set of tests but they test straight read/write channel I/O
as compared to I/OStreams. You're welcome to it.
Jeffrey Peacock
[EMAIL PROTECTED]
Eddie Dimond wrote:
Randy,
I have a Redhat 9 Linux machine with a 1.8 Ghz P4 that I could