USB-6008 National Instruments DAQ

Chances are, if you're working with LabView then you have their USB-6008 DAQ. It's an I/O module. Its digital outputs are configured as "open drain".  That's not uncommitted open drain, though - they're tied high, each via 4.7kΩ to its 5V. Not good for much at that rate and their "knowledge base" reveals a certain lack of electronics comprehension among users.
NI's suggested "fix ya" is a pull-up, but that doesn't hit it for the "ON = 5V" people or those who need capability greater than its paltry own unassisted. They may have reasons for doing this (keeping the electronically uneducated out of big trouble), but, clearly, the device is certainly limited.

There are better solutions, and this is mine: a high-side driver. It requires two components, a resistor and a PNP transistor.
All with the example build atop the 6008, the circuit diagram is shown beside the module in the following picture:

I used a 2N3906 because its rated collector current is 200mA and it's commonly available. The 750Ω value isn't super-critical (510Ω - 1kΩ will do).
It should be noted that users have to check the "Invert Line" box in the Config tab.
You don't have to limit yourself to the USB 200mA, get an external supply you have the power.

VFD - DG10FI

2017 SEP 24

I bought several VFD tubes (NOS) from an ebay seller (look for old_guy_radiola).

With a datasheet from a similar tube RadioShack once sold, many aeons ago, and a pin diagram for this device that I found on another webpage, I began experimenting almost immediately.

This is the 'final' demo circuit.
The segment anodes are pulled up, externally, with 10K resistors. Each '2003 output shunts the segment anode to which it's connected. CG is the control grid, it's a display enable for mux'ing.
[ i.e. logic input HI → output LO (to Gnd.) → segment Off ]
The filament runs on 1.5V

Basically a static demo at this point.
I will work out an Arduino sketch, for the Update, that will run the numbers (DEC and HEX), do BIN with the horizontal segments and a chase routine. (segment_a, _b, _c, _d, _e, _f, _a, _b,...)

The "A" looks odd because of the segment geometries were likely optimized for presenting numerals (there is a Segment h, a little nub for a more pleasing "4"). Segment c is kind of long. I will see how an "a" looks in comparison.

2017SEP30 - Update
I made the demo (as promised, but less the HEX part, too lazy)

2019MAR23
All Electronics has an excellent Noritake dot-matrix VFD. Accepts data Asynch, SPI, I2C.
This pic shows double-size characters.