Fundamentally they're taking a similar approach as the Tesla supercharger - a large number of parallel inverters in a centralized "power electronics" enclosure, cabled up to the end-user accessible charging stations. There's a bit of a difference in that the charging stations themselves have a pair of inverters each, which allows them to function independently of the inverter bank.
Unfortunately for Zero, Chargepoint appears to be
settling on a 200V minimum.

They're also settling on a maximum of 400A output from each charging pod regardless of voltage; a 350V car can receive a maximum of 140 kW, about the same as Tesla's Supercharger today. A realistic maximum even for the higher voltage vehicles is likely to be closer to 350 kW delivered to the vehicle due to voltages at low SOC around 800V or less.
However, what's clever about the external, parallel design is that it can accommodate a mix of system voltages and output power requirements. Lower voltage, lower power vehicles can coexist with higher voltage, higher power vehicles.
For example:
Another configuration shows a maximum of 750 kW shared output delivered among four stations, each with two inverter modules and a "power cube" with 16 inverter modules. That's a total of 24 inverter modules.
Hypothetical Scenario 1- 3 low SOC Chevy Bolts pull up and begin charging - say 320V and can accept 250A each (80 kW)
- 1 low SOC Porsche Mission E pulls up just after and begins charging - say 700V and can accept 400A (280 kW)
- each of the 24 inverter modules can supply up to 78A below 400V "constant current" or 32.5 kW above 400V "constant power"
The charging system will work out the following assignments:
- each Bolt supplied by 4 inverter modules, 250A and 80 kW to each vehicle
- Mission E supplied by 9 inverter modules, 400A and 280 kW to the vehicle (total system 520 kW)
- as Mission E system voltage rises and power requirement increases, a 10th and 11th module will be added to the array. Alternatively the system may want to assign all 12 remaining inverter modules to the high voltage car from the beginning.
Hypothetical Scenario 2- 3 low SOC Tesla Model S 100D pull up and begin charging - say 320V and can accept 400A each
- then 1 mid SOC Porsche Mission E pulls up and begins charging - say 800V and can accept 400A
The charging system will work out the following assignments:
- initially each Model S will be serviced by 6 inverter modules, 400A and 128 kW to each vehicle
- when the Mission E arrives, only 6 inverter modules are available to service it. The system will have to distribute the charging inverters in some way.
Option 1: Give the 6 inverter modules to the Mission E, which then receives "only" 195 kW, 61% of the initial request. As the Model S vehicles begin to taper, their inverters can be repurposed for the Mission E. Total system output is 579 kW.
Option 2: Get as close as possible to each vehicle's request. 5 inverter modules can supply 97% of each Model S request; the remaining 9 inverter modules can supply 91% of the Mission E request. These may even work in some timesliced fashion, ie perhaps once a minute rob inverters from one vehicle to service another. Total system output is 667 kW.
Option 2 has the advantage of maximizing system utilization, getting everyone on the road quickly, and ultimately maximizing charged-mile-throughput for the station. I believe Tesla's Supercharger works like this and I assume Chargepoint's system does as well.
