Renesas RZ family board design

The silicon family we have the deepest production experience on — vision AI, industrial HMI and TSN networking, across modules and carriers.

RZ/V2LRZ/V2MRZ/V2MARZ/G2LRZ/G3ERZ/N2LDRP-AITSN

Fit

When this is the right choice

Choose it when

  • You need AI inference at the edge with a power budget that rules out a GPU
  • Industrial temperature range and long-term availability matter
  • You need TSN or industrial Ethernet natively rather than bolted on
  • Vision processing is central and DRP-AI fits the model you are running

Look elsewhere when

  • Your model needs more inference throughput than DRP-AI delivers — look at Jetson-class
  • Your team already has deep i.MX or Sitara ecosystem investment
  • You need a part with a broader community ecosystem for rapid prototyping

Engineering

What’s actually hard about it

The parts that decide whether it works on the first spin, and the parts that cost a respin when they are wrong.

DDR on a constrained footprint

Fly-by topology and write levelling inside Renesas tolerances, on a module-sized board with no routing room.

Power sequencing exactly to spec

The RZ family is specific about power-up ordering. Out of order, it does not boot — and it looks like a dead part.

Thermal under inference load

DRP-AI running is not the idle datasheet figure. Junction temperature has to be budgeted at real worst-case ambient.

Camera pipeline design

MIPI CSI-2 lane matching and a clean reference for the ISP, which is where most vision bring-up problems originate.

What we have built on it

Boards around RZ/V2L, RZ/V2M, RZ/V2MA, RZ/G2L and RZ/N2L across SMARC, OSM and LGA form factors, with matching carriers and evaluation platforms — including DDR fly-by routing and MIPI CSI-2 camera pipelines on this silicon.

Got a board to design — or one that won’t boot?

You talk to the engineer who would do the work. Reply within one business day, and we’ll sign your NDA before you go into detail.