Re: PCI and DSP bandwidth
- From: Jerry Avins <jya@xxxxxxxx>
- Date: Wed, 10 Sep 2008 12:19:45 -0400
KennyL wrote:
Hi guys,
Thanks for the reply.
The data is being packed together as a single chunk, so there's only 1
interrupt at a time for all the channels. This should in theory mean we'll
be achieving near the max transfer rate. What I'm worried about is the
worst case scenario, does anyone know if the 32MBytes/sec is a documented
minimum rate? Or is this based on hearsay? Also, what other special
circumstances will this occur?
I'm using XP, and the driver was written by that company - however I've
the code to "tweak" the driver for performance.
There's no other cards hooked to the PCI, graphics uses PCIe, and the same
goes for other cards - all PCI express. So no other contention.
What I'm trying to determine is if there's a fundamental issue here and if
the system is able to cope with the worst case scenario. 35 Mbytes/s
requirement, but the worst case rate is only 33Mbytes/sec, which means
there'll be problem?
What else will be going on? If you're writing to disk, how fast can that take data? The whole _system_ needs to be fast enough, not just the _bus_.
Jerry
--
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯
.
- References:
- PCI and DSP bandwidth
- From: KennyL
- Re: PCI and DSP bandwidth
- From: Vladimir Vassilevsky
- Re: PCI and DSP bandwidth
- From: KennyL
- PCI and DSP bandwidth
- Prev by Date: Re: software demoduation
- Next by Date: Re: End Point Effect on prediction
- Previous by thread: Re: PCI and DSP bandwidth
- Next by thread: 802.11 Aggregate Data Rate
- Index(es):
Relevant Pages
|
Loading
