MultiPort Hubs vs. Dedicated Bandwidth ?

Just as one, of many, examples ...

Avid and other Video Edit Suites dictate a dedicated firewire port, for off-board HDD/RAID which does not share lane bandwidth with any other traffic or "overhead".

Discussion: ... That was certainly the case with the old PCI bus ... Because of the high density throughput demands for video, streaming fast storage and video I/O options can individually saturate that bus ...

.... Today, corporations like Blackmagic Design are introducing USB3 breakout options which REQUIRE no less than a dedicated USB3 on a (explicitly specified) X58 Reference(Plus) mobo ... and they even have a short list of certified (specific) mobo models.

Furthermore ... Even the latest USB2 devices (media thumb drives and other flash media) are at (or pushing) the extreme maximum limit(s) of USB2 spec bandwidth.

Finally ... Todays mobos may come with a dozen (or more) USB and firewire ports, of various flavors. ... PCIeX1 firewire 400 cards (Rosewill et. al.) are spec'd as Two External plus ONE internal ports on-board.

Pre-amble to "The Question at Hand" : ... I have never been entirely clear as to the sustainable/dedicated/combined bandwidth for these sorts of multi-port controller hubs or how many of which type of "mastering"(or not) device can co-exist, on a given hub, before "saturation-conflicts" occur.

I also have no real way of identifying (with any great ease) which physical ports (holes), on the back (and internal "pin-out" headers) are assigned to which (and however many) hub controllers.

So far ... I KNOW that the "card bracket" ports are strung from ONE (particular) HUB ... and I know that the rear I/O panel ports are connected to (prolly) at least two other HUBs ... and that any front panel ports (if connected internally) are stemming from yet another hub.

While I am speaking (mostly) about USB2/USB3 ... I am also wondering the SAME questions about FireWire 400/800 and multiple Gigabit/CSA ports AND ...

... AND ... I'm also guessing that I can just "think of" shared eSATA ports as another USB device ... but how can I confirm whether this port is connected to the same hub as other USB devices and which will get any "master priority" ??

SO ... [deep breath] ... after "all that" ... I have one "3 word" question : ...

Question ? : "Wassup wid dat?"

Seriously ... I'm sure I will need to consult/peruse WikiPedia, concerning the various standards and, also, ....

... also ... I know that limitations/bandwidth/solutions *may* differ, depending if the hub is attached to PCIe, vs. Native SouthBridge, vs. Non-native attacments (whatever that might mean).

Best Answer Points to whomever can shed the most light ... extra points for concise and simple (comprehensible) explanations.

And, hey ... If you don't know ... Don't even try ... This issue is sticky enough, as it is.

Any bold takers ?

ThankYou (OF COURSE ! )

= Alvin =
2 answers Last reply
More about multiport hubs dedicated bandwidth
  1. These are questions that have lurked in the back of my mind for some time. I don't even know how bandwidth is managed with USB. Is it on demand or dedicated?

    This is mostly just a tag in case you actually get good answers ;)
  2. I belive it is "device mastering" , rather than time-sliced polling ... clueless on that, ATM ... but ... that much *should* be covered in the openning description ... on WikiPedia.
Ask a new question

Read More

New Build Bandwidth Systems