Re: VB6 to VB.NET, comments please
"Dick Grier" <dick_grierNOSP AM@.msn.comwrot e in message
news:uvhKJqBuHH A.3400@TK2MSFTN GP03.phx.gbl...
My USB drivers are from Jungo. I originally started with the net Jungo
drivers. Now that was slooow. I got up to 0.5 BPM rate. Keep in mind that
the normal use of USB is sending/recieving large streams of data. To a
printer or from a scanner etc.
Whereas I'm getting small 24 byte bursts fairly rapidly. The printers and
scanners don't really care that a little extra time is used by net. I do.
As a side note, my proprietary USB device is 100% assembly language using a
188 type processor and a NetChip USB chip. It's fast.
Galen
"Dick Grier" <dick_grierNOSP AM@.msn.comwrot e in message
news:uvhKJqBuHH A.3400@TK2MSFTN GP03.phx.gbl...
Hi,
>
It's interesting to note that I develope on a Pent D 940 dual-core at
3.2GHz. But our program runs at high BMP rates on older Win98 machines
with
way less power and speed.
<<
>
You have to really work hard with threading to make the dual-cores perform
(the one that I have is only a 2 GHz notebook, and it is about 30-35%
slower than the Hyperthreaded 3.2 GHz notebook it is replacing -- Vista
looks nice, but looks aren't everything). The OS (and .NET) have added so
much that gets in the way of raw performance that we have to use as much
skill as we can muster to avoid the bottlenecks. And, COM (and STAThread)
marshaling is a big bottleneck.
>
Perhaps you can investigate a native dll (or even .NET) interface to your
USB hardware. Getting away from the ActiveX interface might provide a
substantial speed improvement.
>
Dick
>
--
Richard Grier, MVP
Hard & Software
Author of Visual Basic Programmer's Guide to Serial Communications, Fourth
Edition,
ISBN 1-890422-28-2 (391 pages, includes CD-ROM). July 2004, Revised March
2006.
See www.hardandsoftware.net for details and contact information.
>
>
>>>
3.2GHz. But our program runs at high BMP rates on older Win98 machines
with
way less power and speed.
<<
>
You have to really work hard with threading to make the dual-cores perform
(the one that I have is only a 2 GHz notebook, and it is about 30-35%
slower than the Hyperthreaded 3.2 GHz notebook it is replacing -- Vista
looks nice, but looks aren't everything). The OS (and .NET) have added so
much that gets in the way of raw performance that we have to use as much
skill as we can muster to avoid the bottlenecks. And, COM (and STAThread)
marshaling is a big bottleneck.
>
Perhaps you can investigate a native dll (or even .NET) interface to your
USB hardware. Getting away from the ActiveX interface might provide a
substantial speed improvement.
>
Dick
>
--
Richard Grier, MVP
Hard & Software
Author of Visual Basic Programmer's Guide to Serial Communications, Fourth
Edition,
ISBN 1-890422-28-2 (391 pages, includes CD-ROM). July 2004, Revised March
2006.
See www.hardandsoftware.net for details and contact information.
>
drivers. Now that was slooow. I got up to 0.5 BPM rate. Keep in mind that
the normal use of USB is sending/recieving large streams of data. To a
printer or from a scanner etc.
Whereas I'm getting small 24 byte bursts fairly rapidly. The printers and
scanners don't really care that a little extra time is used by net. I do.
As a side note, my proprietary USB device is 100% assembly language using a
188 type processor and a NetChip USB chip. It's fast.
Galen
Comment