Unlock The Universe Count H4PC

Ok... My own post got me thinking...

Obviously unlocking the Universe count on H4PC for nothing, doesn't help HES's business. So here is an idea.

HES sells a USB dongle or simply a hardware device that plugs in via usb to a user built, rack mounted Linux or Windows based PC. The dongle/device unlocks either unlimited Artnet/aCN universes or a scalable amount of universes based on the dongle license purchased. The unlocked H4PC machine can be run as an H4PC server with all the processes running, or just the DP/multiple DP processes to allow the rackmounted box to serve as a networkable, (Enter Tounge In Cheek) "Hog DMX Mothership".

Whatcha think?

JB

I think thats how it currently works?
If you plug in a single universe widget, you get a single universe output of art net.
If you plug in a super widget you get 4 universes of art net...

At least thats what I remember hearing....

It is to a max of 16 universes per dp8000, 12 per FB4 and 8 for RH4 and PC4.

But I believe with the latest hardware on a very strong PC that one machine might be able to drive way more than 16 Artnet/aCN per processor with the latest hyper threaded I7s and 8 core amd chips.

Perfect example is this dedicated H4pc I'm using with an 8 core amd. Yesterday just for shits and giggles I ran every single perameter on 8 Universes in a simultaneous effect with cross fades and the performance manager on windows 7 showed the Dp8k process using 16 threads at less than 12% with no dmx latency.

With that kinda overhead we are talking about being able to theoretically run ~ 50 universes from a single box. Considering that I'm not a HES software engineer, my estimates are just theoretical from what I'm seeing, but it would be a nice starting point for a discussion.

I get the idea, but any show that has that much gear to drive can certainly afford a dedicated control system, that IMHO is much more reliable than an "off the shelf PC".

I think HES/FPS has done a nice job at balancing the number of universes you can control directly from the computer before you step up to dedicated processing (DP-8000).

I'd much rather (x4) DP-8000s to get 64 universes for example than 16 Super widgets all hooked up with individual USB cables and power supplies to (theoretically) unlock the same output capacity.....at a certain point it just gets messy and turns into a "franken-console" with far too many potential points of failure....

My $0.02

Actually Marty, I was just envisioning Artnet out. On that large of a show, you would probably be running Artnet and sprinkling nodes around, so no need for widgets. One cat 5 cable out of one box into a switch is a lot less messy than a slew of dp8000s on hog net and then an artnet out on every dp8000 into the first or many Artnet switches. Much more that can go wrong there.

IMHO. The mothership seems like a more robust system than a dp8000 for every 16 Universes. Plus your not dealing with multiple network devices that can possibly conflict.

My idea was not a lower budget idea for a large production, but more or less the exact opposite. The hardware that HES has in the DP8000 is probably the same components you would use to put a custom box together and with the custom box you can probably make the PC more stable.

Nice idea Jeff. Although to get all those artnet universes working from the mothership you would still need to have a widget attached for each artnet universe. There's that mess again.

And personally I would rather deal with setting up a nice network of DPs than hoping that Windows doesn't spontaneously unload one or more widget drivers.

--my $0.02

Mothership fails = all outputs fails
1 DP8000 fails = 16 lines fails

I never would use this PC-mothership for a complex setup....

Gideon,
The USB dongle would be the only thing needed to unlock all of the artnet. So the widgets wouldn't be needed.

And Marc,

You are correct about not having a backup, so instead of having just one, maybe you would run two mother ships on the same net # so one is always idle waiting to be the failover. Two Motherships is still cleaner than 8 DPs.

And if I remember correctly from school years ago.. And someone correct me if I am wrong.. The more points of failure you have, the higher the chance of failure happening, right?

I appreciate all the replies and input.

I still think it might be a great idea...

;)

If you wanted failover for both systems, one based off the DP's and then One based on the MotherShip (Remember tounge in cheek as I already hate that name)... for lets say 50 Lines... You would need...

2 x Motherships(1 failover and 1 primary (Which only uses 2 IP's)), 2 switches (1 Artnet and 1 Hognet) and 4 cat5 jumpers (from Motherships to switches)... lets just leave the consoles out of the equation for now, since both situations will be desk to hognet switch with one cat5 jumper ..


8 x DP's (4 primary and 4 failover (Which now uses 8 seperate IP's)), 2 switches (1 Artnet and 1 Hognet) and now 16 cat5 jumpers (from DP's to switches)...

With the Mothership idea, we have 8 points of failure.... with 2 IP's

With The DP's we now have 26 points of failure.... with 8 IP's



Don't get me wrong.. I am not in any way shape or form saying that DP's are a bad idea... Hell.. Everyone uses the idea of the DMX processor limited to around 8-16 lines of artnet, whether it's HES, Vari-Lite, MA, ETC... etc. But I think they limited themselves, since at the time of design, the hardware was not fast enough to drive more Lines reliably..

JB