Hi Tom,
in the application that I intend to use, the most "user perceptible"
part is this index creating. So this is as good a test as any.
What I would like is an easyer to use or a universally accepted program.
I am willing to make a speciffic test to exactly discover what is
specifically
Hello Alain,
> I also tested with a real-world .DBF file creating a complex index and
> differences are much smaller then with a big single write/read.
> Can someone point me to a good benchmark, so that I can do more testings?
depends on what you want to measure.
I consider both PRIME95 and Sp
Hi all,
Thanks for all the feedback. I spent all day making tests, the data is
very confusing, I will make more tests and some kind of report on friday.
Just a few points: emm386 is not responsible for most of the slow-down,
most of the time umbpci behaves the same (?) and different combinati
Hello Alain,
> I just made some tests with UDMA2 and had some unexpected results:
> Why is FreeDOS so much slower (60% no UDMA, 14% with UDMA)???
this may depend on
file placement and free space fragmentation as
well, where MSDOS uses an alghorithm other then freedos
what EXACTLY is RAWSPE
At 10:15 PM 6/7/2005 -0300, Alain wrote:
I just made some tests with UDMA2 and had some unexpected results:
Why is FreeDOS so much slower (60% no UDMA, 14% with UDMA)???
Is this consistent with other experiments?
Virtual 8086 mode hurts disk I/O throughput. Not really sure why all the
deta
On Tue, 07 Jun 2005 22:15:25 -0300, you wrote:
Hi Alain,
>Why is FreeDOS so much slower (60% no UDMA, 14% with UDMA)???
>Is this consistent with other experiments?
Try skipping EMM386 ... or try UMBPCI
>FreeDOS kernel= , lbacache 22sep2004 with tickle, Himem/emm386 v2.01
>MSDOS from Win98se, ca
Alain schreef:
FreeDOSMS-DOS7.10
No UDMA, no cacheR=1.4 W=1.4R=2.3 W=2.8
No UDMA, cacheR=1.3 W=1.4R=2.3 W=2.7
UDMA2, no cacheR=6.3 W=4.4R=54.2 W=16.0
UDMA2, cacheR=5.5 W=4.4R=39.5 W=15.7
FreeDOS kernel= , lbacac
I just made some tests with UDMA2 and had some unexpected results:
Why is FreeDOS so much slower (60% no UDMA, 14% with UDMA)???
Is this consistent with other experiments?
Please, I would like to here comments on this ;-)
UDMA 7.0 and UDMA2 2.5 had the same performance.
On instalation it says m
