The Secret Cambridge Integration Master Plan

(just between you and me)
Ameer Janabi, Founder & CEO, Cambridge Integration Ltd, 2025

As you likely know, the first product of Cambridge Integration Ltd is not a complete converter or module, but the Standard Cell—a universal, physics-optimised power-electronics building block.
Some readers may not realise that this is not intended to be a packaging product on its own. Rather, it is the foundation upon which an entirely new method of designing and manufacturing power converters will be built.

Before explaining the broader plan, it is worth briefly stating why the Standard Cell exists at all.

Power electronics has a persistent bottleneck: packaging limits device capability.
A SiC MOSFET can theoretically switch in tens of nanoseconds, yet practical converters often operate hundreds of nanoseconds slower due to:

• parasitic inductance in module or PCB interconnects,

• imbalanced current sharing across parallel devices,

• thermal gradients and localised hot spots,

• layout constraints that force slow edge rates to avoid over-voltage stress.

These limitations propagate upward.
A slow switch means higher switching losses, which means larger heatsinks, larger magnetics, lower power density, and eventually higher system cost.

The Standard Cell was designed to remove these constraints from the bottom up.

1. Begin With the Fundamental Unit: The Standard Cell

The Standard Cell achieves ultra-low stray inductance by controlling every geometric parameter that contributes to switching loop area. In practical terms, inductance is reduced enough that gate resistors can be minimised or removed without risking voltage overshoot. This allows SiC devices to switch at speeds far closer to their intrinsic device limits, significantly reducing energy loss per transition. Thermally, the device is mounted on a high-conductivity substrate structure with exceptionally low junction-to-case impedance. Heat spreads rapidly and uniformly, allowing devices to operate at higher switching frequencies (or higher current) without thermal derating. In short: The Standard Cell provides electrical and thermal performance that conventional module packaging cannot match.

2. Use the Standard Cell to Enable True Converter Scalability

Traditional converters require redesign for every change in power level.
Parallel MOSFETs rarely share current evenly due to asymmetry in interconnect inductance and thermal gradients. As a result, a 1 kW converter and a 1 MW converter often have almost nothing in common structurally or electrically.

The Standard Cell resolves this.

Each cell is designed to:

• share current uniformly,

• maintain identical switching conditions across devices,

• maintain consistent thermal conduction paths,

• tile seamlessly into larger converter assemblies.

The result is a fully scalable architecture: converters from sub-kilowatt to multi-megawatt can be constructed using the same block with predictable behaviour.
You no longer redesign the converter — you scale it.

3. Standardise Manufacturing Until Converter Assembly Becomes Routine

The long-term value of the Standard Cell is not only performance — it is manufacturability.

Once the cell is reliable at volume, converter production reduces to:

• automated pick-and-place,

• reflow or sintering,

• final assembly and testing.

Large, specialised packaging lines are unnecessary.
Custom high-power module tooling is unnecessary.
Converter production becomes repeatable, modular, and automated.

The industry shifts from handcrafted layouts to cell-based assembly.

4. Automate Converter Design With Physics-Constrained AI

With a universal building block, the design space collapses.
Every converter shares the same:

• current paths,

• switching environments,

• thermal stack,

• mechanical interface.

This opens the door to true AI-driven converter synthesis.

Instead of spending months tuning layout, PCB parasitics, thermal vias, gate inductance, loop area, snubbers, and magnetics placement, an AI system can:

• configure Standard Cell arrangements,

• generate optimised layouts,

• predict thermal maps in seconds,

• size magnetics based on known switching conditions,

• score designs for efficiency, EMI, and manufacturability.

A process that once required expert engineers and weeks of iteration can be reduced to minutes.

This is not theoretical — it is the natural outcome when the building block is fixed and the physics is consistent.

5. Expand Into All Power Domains Using One Architecture

With the Standard Cell, a single architecture can support:

• onboard chargers and traction inverters,

• aviation power modules,

• rail and marine propulsion,

• grid-connected converters,

• high-reliability space systems.

Instead of bespoke converters for each sector, we use the same physics-proven block in every power range.

The cost of innovation drops.
Time-to-market shrinks.
Reliability increases across the entire power spectrum.

6. And a Final Note… It All Started in Cambridge

Just as Tesla hinted that its long-term vision was already written, our direction is clear:

• Build the Standard Cell.

• Use it to make scalable converters.

• Standardise manufacturing until it is routine.

• Automate converter design with AI.

• Deploy one architecture across all power levels.

And quietly remember that this approach — merging materials, integration physics, and scalable architecture — began in Cambridge, where technology has repeatedly been rewritten from first principles.

Don’t tell anyone.