VITA 100 Draft Targets 400 GbE Backplanes: Eleven Myths Sorted From Facts

By Amara Osei

VITA 100 aims to push up to 400 GbE (Gigabit Ethernet) across the backplane — roughly eight times the bandwidth of today's OpenVPX — and it will need a new connector to get there. That much is settled. The rest, according to Justin Moll, Vice President of Sales & Marketing at Pixus Technologies, is surrounded by what he calls spinning myths. The specification is still in Draft stage, which never slows the rumor mill.

Moll laid out eleven of those myths, and his corrections sketch a clearer picture of the architecture than most datasheets manage.

What is VITA 100, exactly?

It's a next-generation embedded computing architecture built on OpenVPX. The much higher signaling speeds forced a new connector, and the new connector forced a new architecture. That's the physics of it: pins that happily carry 10- or 40-gigabit traffic don't behave at 400.

The draft leverages a great deal of existing VITA work rather than starting from a blank sheet. Pinouts differ, but the root concepts come from VITA 65 (the OpenVPX standard itself), the mechanicals and cooling from VITA 48, and the RF/optical provisions from VITA 66/67. So no, it isn't a completely new design standard. It's an aggressive recombination.

Does VITA 100 replace OpenVPX?

No — and the installed base is the reason. OpenVPX, Moll argues, has more than enough capability for most applications and "will likely be around in another 30 years." His evidence: Pixus still does a new VME project roughly every year, and that specification is over 50 years old. Legacy architectures die slowly in embedded computing. Hybrid systems pairing an OpenVPX segment with a VITA 100 segment seem the likely outcome.

Backwards compatibility, though, is off the table. The new connector interface rules it out.

When will hardware actually appear?

The working groups hope to complete the specification this year; Moll concedes that timeline may be aggressive. Prototype systems will likely start to emerge in 2027. Anyone budgeting a program around 400 GbE backplanes should pencil that in pencil, not ink.

Are the slot profiles just recycled VITA 65?

Not quite. VITA 100 will leverage SOSA-aligned slot profiles (SOSA being the Sensor Open Systems Architecture alignment the defense market increasingly demands), but the profiles will take lessons from existing pinouts while maximizing the high pin density of the new connectors.

Can you cool an 8x-bandwidth system?

Yes, with effort. Moll doesn't expect natural-convection enclosures for conduction-cooled modules at these speeds. But airflow over conductive fins remains viable — not every module in a system carries a power-hungry processor. Hotter systems will lean on:

• Air Flow Through cooling (AFT, VITA 48.5 and others)
• Liquid Flow Through (LFT, VITA 48.4)
• Liquid circulated through the chassis sidewalls

The takeaway: thermal engineering gets harder, not impossible.

Same 3U and 6U form factors?

VITA 100 will use 3U and 6U (the module heights, in rack units, familiar from OpenVPX). But AI and quantum computing processors are heavy-duty parts, and many need more than 3U. 6U works but is simply too much space for many applications. So the specification introduces a 4U module as a third option. Some modular enclosures already support that size — a small but telling sign that the ecosystem is preparing.

Does VITA 100 hurt SOSA?

On the contrary. Many SOSA Consortium members are strongly interested in VITA 100 and are working with joint VITA members to suggest slot-profile definitions for the specification. The same members continue to influence OpenVPX, VNX+, SpaceVPX, and other VITA standards. If anything, the alignment effort is widening.

Can you prototype without customization pain?

Power and ground backplanes — which let boards power up and be tested — are always available with short lead times. There are also discussions of rear-transition module (RTM) connectors that could pull I/O out the back and potentially allow cabled slot-to-slot connections, if an industry vendor develops the cabling solutions. That's a conditional, and Moll leaves it as one.

Is there a stepping stone?

Yes: VITA 91. That architecture is OpenVPX-compatible and in use today for a theoretical 200 GbE bandwidth. Its connector scheme uses a similar double-density connector approach to VITA 100, making it a stepping stone in both bandwidth and density. Programs that can't wait for a draft standard to ratify have a halfway house.

The verdict

VITA 100 is an incremental revolution — new connector, new density, borrowed bones. The performance gap over OpenVPX is real, at up to 8x bandwidth. The schedule is the soft part.

The question program managers should now be asking isn't whether to adopt VITA 100. It's whether their SOSA-aligned slot-profile requirements, cooling budgets, and delivery dates can tolerate a specification that still says "Draft" on the cover.