Tootfinder

STM32MP2 XC7K160T PCIe test board progress! 66 unrouted nets in the power supply, but I think it'll fit.
There's something for everybody in this board:
* 0201 passives? Check (under the SoC)
* Lots of FPGA? XC7K160T is decently sized
* Lots of CPU? Dual A35s, a M33, a M0 , and a M4
* High speed? PCIe gen2 and a SFP that is theoretically capable of 8 GT/s but I'm gonna try pushing the FPGA a bit and see if I can make it do 10Gbase-R with acceptable BER…

Late-night TV hosts Jimmy Kimmel and Stephen Colbert criticized FCC Chair Brendan Carr for the FCC's new guidance on "equal time" rules for political candidates (Siladitya Ray/Forbes)
https://www.forbes.com/sites/siladit…

Kimmel And Colbert Say FCC ‘Equal Time’ Rule Is A Bid To ‘Stifle’ Them
The late-night hosts questioned the relevance of the old rule in an era where broadcast TV viewership is declining and labeled it as a bid to stifle their speech.

Tired. Broken pipe section *and* failed pump *and* dodgy wiring replaced. So far no leaks in evidence, but it needs a good shakedown before I will trust my inner wall with it. In the words of the Hšvamšl:

468 nets unrouted. Still have to do the FPGA flash, the power supply, bulk caps for the FPGA, and a few other odds and ends but the entire SoC subsystem is basically routed.
So at this point I can already say definitively that routing the STM32MP2 in TFBGA436 on an OSHPark 4-layer board is extremely doable as long as you aren't trying to max the IOs. Using the DDR may or may not be possible (I'm not trying here) but the internal RAM and HyperRAM are enough for a lot of applicat…