STM32 Availability for new designs

[FJB2069]

I have several eval boards and was interested changing to ST platform prior to pandemic. Now looking at getting back to it, but checking with Mouser, Digikey, etc.. and nearly all of the STM32F4... parts I would be interested in using are still out of stock with no expected delivery dates!

What is the outlook ST Vs Microchip availability in the next year?

What is the general consensus on new product development with ST?

[Tesla DeLorean]

>>What is the general consensus on new product development with ST?

Order parts early in the process.

Use a foot-print that is common across multiple parts/families.

Consider SoM for larger/complex boards if you plan for 3-5+ year product lifespan.

I would tend to recommend making design choices around what you can order.

There are a wide range of MCU options these days, if the design can be split across multiple, simpler, ICs consider that. Yield per wafer scales a lot better with smaller die.

Look if you can externalize FLASH usage, there are broad options with XIP and QUAD/OCTO-SPI parts.

FPGA?

RISC-V?

[AScha.3]

RISC-V?

Clive, you are sooo bad...

:)

[Tesla DeLorean]

I'm playing from a stacked deck, I've built products / tools around MIPS, MIPS-X, SPARC, V850, and was involved in Acorn equipment in the early 1980's.

RISC-V has got some significant traction, I expect that to accelerate into 2023

[Pavel A.]

> RISC-V has got some significant traction

Because it is kind of open source. License free, by contrast with ARM.

[Uwe Bonnes]

F4 is quite old, I would not use for a new design. G4 uses smaller feature size and more advanced peripherals. As this will give more chips per wafer, supply for those families with more chips per wafer can be expected to be better. E.g. Farnell has 2 F4 but 5 G4 types in stock.

[FJB2069]

Thanks, I did not know that.