This should be no problem from the pure hardware specs. Most notebooks with lower power can handle this (e.g. the Depeche Mode World Tour with 2 x notebooks and MADIface).

You need at least Windows XP (see driver support).

A new and powerful computer plus PCI express is a good - but more expensive - choice. If you want to upgrade your system after some time, you need a board with at least one PCI slot for the HDSP MADI. This could be a problem, if you want to use something special, like a Mac Pro.

If you choose the HDSPe MADI you have a future proof solution. Differences of the HDSPe MADI to the PCI MADI card:
    * PCI Express interface
    * Sample rates up to 192 kHz
    * Even lower latencies
    * Enhanced TotalMix
    * Monitor output with low-latency converter technology of ADI-8 QS
    * Direct support of HDSP TCO

The MADI format is nothing special, but a very powerful digital multitrack connection with some useful enhancements, done by RME (like MIDI over MADI and remote control). For the desktop computer or notebook it´s just an audio card with more I/Os than usual.

best regards
Knut

Is this the official stance on this?  Everytime I hear someone ask about PCI vs PCIe and RME's "Even lower latency" advertising copy, RME's response has always been the latency will be identical unless some other PCI or PCIe limitation is present in the system.

I think this advertising copy should be better qualified/quantified if you are speaking of PCIe bus latency that won't have any real-world effect on bottom-line ASIO latency.  It is incredibly misleading unless you are actually talking about ASIO latency IMNSHO.  I understand the Advertising dept likes to make things sound as good as possible in the product literature and advertising copy - but concise unambiguous info is always the best bet as far as the consumer is concerned (if you want an informed consumer)...

Thanks for any clarification and possible elaboration of this lingering point (elaboration in the literature and ad copy - not just on this forum .

So is this talking about ASIO latency the end-user will appreciate, or PCIe bus related latency that the end-user would never even know about (as RME seems to say here again and again)?  "Open-ended/leading" ad copy is dangerous at best, and deceptive at worst...

Words matter, and "accidentally" misleading the consumers is something that needs to be resolved now that the "accident" has been brought to light.  If this is intentional marketing ambiguity - then shame on you!

The fact that you mention that the HDSPe series don't use a PCI>PCIe Bridge Chip is sufficient - no need to add the leading "Even Lower Latencies" copy unless you are actually talking about ASIO (and you aren't).  Anytime the generic term "Latency" is tossed around when talking about a LOW LATENCY AUDIO INTERFACE - it's assumed "Latency" means "Usable Audio Latency" unless otherwise specified - and I think that is blatantly obvious.  Not everyone that buys an RME card is a PC guru (I'd wager most aren't, and have no idea how ASIO latency is different from PCIe bus latency).

Clarification is welcomed - and I hope any clarifications make it into the Ad Copy and product literature to keep the consumers ACCURATELY informed w/o misleading anyone.

Can you tell these kind of advertising practices strike a nerve with me? ed:  The time to save face (and learn for these mistakes) is now.  Water under the bridge or a mammoth flood on the horizon - the ball is in your court.  I know RME will do the right thing and keep their customers accurately informed...