VITA 100 Targets 400 GbE, but Remains a Draft Standard

By Sophie Lindqvist

VITA 100 remains in draft, but the proposed embedded-computing architecture targets backplane speeds of up to 400 GbE and eight times the bandwidth of OpenVPX.

Justin Moll, vice president of sales and marketing at Pixus Technologies, wrote that VITA 100 “will potentially provide speeds up to 400 GbE across the backplane.” The sentence states a design target, not measured performance. Moll’s 11-point assessment cites a performance chart and connector figure, but supplies no prototype results, shipment forecast or independent validation.

What does VITA 100 actually change?

The proposal builds on OpenVPX and introduces a new connector and backplane architecture to support higher speed. Because the connector and pinouts differ, VITA 100 will not provide backward compatibility with OpenVPX.

The design is not entirely new. It draws on root concepts from VITA 65, mechanical and cooling approaches from VITA 48, and RF and optical work from VITA 66 and 67. Its slot profiles draw on VITA 65 lessons and Modular Open Systems Approach, or MOSA, profiles, but must accommodate the new connector’s higher pin density.

Will VITA 100 replace OpenVPX?

Moll expects many applications to combine VITA 100 and OpenVPX sections within one backplane or system. He describes the OpenVPX installed base as huge but provides no unit count. His forecast is that “OpenVPX has more than enough capability for most applications and will likely be around in another 30 years.”

Pixus still takes on about one new VME project each year, according to Moll, even though the VME specification is more than 50 years old. That example supports the longevity argument, not VITA 100 adoption. It also comes from one supplier rather than an independent market survey.

When will the specification and systems arrive?

The drafters hope to complete the VITA 100 specification this year, although Moll calls that schedule aggressive. The source gives no publication date, so the target cannot be tied to a specific calendar year.

Moll expects prototype systems to begin emerging in 2027. That estimate covers early prototype hardware. It does not establish a production date, customer qualification schedule or volume-shipment commitment.

Can high-speed modules be cooled?

Moll does not expect enclosures relying only on natural convection for conduction-cooled modules. Airflow across conductive fins can still cool systems, and not every module requires a power-hungry processor.

For hotter systems, Moll identifies three likely cooling approaches:

• Air Flow Through: AFT under VITA 48.5 directs air through the module path.
• Liquid Flow Through: LFT under VITA 48.4 carries coolant through the system.
• Sidewall liquid cooling: Liquid can pass through chassis sidewalls.

Moll presents these as likely options, not as named product deployments. The article identifies no cooling supplier, system model or commercial installation.

Which module sizes are planned?

VITA 100 will retain the 3U and 6U sizes associated with OpenVPX while adding a 4U module. Moll expects heavy processors for AI and quantum computing to require more space than 3U provides.

• 3U: The smaller existing size remains available for modules that do not require more space.
• 4U: A new middle option addresses systems where 6U is too large but 3U cannot provide enough room.
• 6U: The larger size remains necessary for some high-power configurations.

Pixus says some modular enclosures supporting the 4U size already exist. The source names no enclosure supplier, model or production volume.

Does VITA 100 weaken SOSA?

Moll says the opposite. SOSA Consortium members are strongly interested in VITA 100 and are working with joint VITA members to propose slot-profile definitions. The consortium continues to influence OpenVPX, VNX+, SpaceVPX and other VITA standards.

That answer rests on stated participation, not an approved joint specification. The source provides no finalized profile, joint work schedule or independent review of either consortium’s position.

Can developers prototype without waiting?

Moll says prototype development does not require complete customization or long component lead times. Power-and-ground backplanes can power and test boards, and he describes their lead times as short.

The article also discusses rear RTM connectors. They could expose I/O for removal and potentially permit cabled slot-to-slot connections, but that outcome depends on an industry vendor developing suitable cabling solutions. No vendor or cable standard is named.

What carries forward from OpenVPX?

VITA 91 provides the clearest stepping stone. It remains compatible with OpenVPX and offers a theoretical bandwidth of 200 GbE. Its connector uses a double-density approach similar to VITA 100.

Moll also counts VITA 65, 66, 67 and 48 as foundations for VITA 100. The standards path is therefore evolutionary, but systems requiring the new connector cannot make a direct backward-compatible transition from OpenVPX.

Moll has worked at Pixus Technologies since 2012. He is active in VITA and PICMG, previously served as PICMG’s vice president of marketing, and has worked as a keynote speaker and industry commentator.

The near-term contest is coexistence rather than immediate replacement: VITA 100 must validate its 400-GbE target in 2027-era prototypes before OpenVPX users have a clear commercial reason to switch.