[SI-LIST] EM Simulation, Chipsets & USB
- From: Georg.Ramsch@xxxxxxxxxxx (Georg Ramsch)
- To: si-list@xxxxxxxxxxxxx
- Date: Sat, 29 Jan 2005 09:46:47 +0100
Hello together !
*1)* Got a question related to the computation performance of "cots" (
commercial off the shelf )
mainboards ( MSI, ASUS, Tyan etc. ), which are available in every
computer shop around the corner.
I try to assemble a computer configuration, that is best suited for *SI
/ EMC* simulation, with
a high floating point performance and reliability.
Meanwhile I tried several approaches with different mainboard brands and
I am not content with the assembled solutions up to now .
Therefore I try to collect some information from this forum.
Okay, I know: dual channel memory is better, low CAS latency is better,
speedy cpu needed and so on.
What rests in the fog, is the the role and reliability of the chip sets.
*[* Sometimes I hear the "rumor", that a lot of chipsets tend to skip
data words, when high bandwidth throughput
from cpu to memory or cpu to hard disk happens.
My assuption is, that some blue screens, that I got last time, has the
root in a chipset, where the SSN skips the synchronisation
of the internal PLL. Other thing is, that the high speed I/O bus is a
well known differential one. *]*
Has anybody further information for a comparison of chipsets /
mainboards regarding their reliability and performance?
Which hard disk brands can be regarded as reliable at the moment?
Does a dual cpu MB solution exist, which delivers more than the average
20-30% performance increase, compared with a single cpu? ( assumption:
the SW is capable of using multiple cpus )
( I already looked on some internet pages like "Tom's hardware guide" )
*2)* USB 2.0 : a friend of mine developed a camera interface for quality
assurance and applied a cypress seminconductor USB chip.
The bandwidth of the chip seems to be good ( up to *280 Mbit/s* ), but
he has latency problems at the host computer ( "cots" configuration,
WinXP as OS ).
The application does not seem to require a real time OS.
My first proposal was: try IEEE 1394, the B revision; but for that
decision it was too late.
BIOS settings for USB have been adjusted ( from "full speed" = 12
Mbit/s to "high speed" = "somewhat beyond 30 Mbit/s" )
and useless Windows tasks have been cancelled as far as thay are known.
The whole thing becomes a very time consuming problem at the moment,
the USB 2.0 specification and HW data sheets do not tell all .
Where are the main latency "blocks" in the hardware?
So which "srews" inside Windows have to be tuned , that the application
reads "fluently"? Anybody any idea?
Has anybody of you experience with "strange" effects of USB?
( As there is: computer is turned off, then you pull a USB stick from
his port and the computer boots up again /
or: computer stays in reset because the 5V USB cable power "floats"
via a resistor to the watchdog circuitry and keeps a low active signal
high. )
Thanks and best regards,
Georg
------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
or to administer your membership from a web page, go to:
//www.freelists.org/webpage/si-list
For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
List FAQ wiki page is located at:
http://si-list.org/wiki/wiki.pl?Si-List_FAQ
List technical documents are available at:
http://www.si-list.org
List archives are viewable at:
//www.freelists.org/archives/si-list
or at our remote archives:
http://groups.yahoo.com/group/si-list/messages
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
Other related posts:
