Monday MPW-2 Update - Aug 29, 2022
Jeff
Hello everyone,
This is the first, of what will be an ongoing series of Monday updates, from Efabless to the community about MPW-2 related work. We expect this will focus on technical developments and engagement with the community such as MPW-2 deliveries, progress on our debugging and the latest and greatest on how you will get the most out of your designs.
If you want to get even more up to date information, you can follow along with the Efabless team working on these issues in the slack channel at #mpw-two-silicon. Please direct your questions to myself or Matt Venn (@jeffdi or @Matt Venn on Slack) to allow the team to focus on the delivery.
As mentioned last week we will be hosting Office Hours three times a week to help as you bring up your project. This will commence with shipment of hardware. In the meantime, if you have questions, please let myself or Matt know.
We will also host a general webinar walking through MPW-2 status and results.
MPW-2 Silicon Testing
I/O Configuration
Analog IO found in the Caravan analog harness have been identified as not affected by GPIO configuration issues due to their fixed setup. Users who are only using these I/O can expect full functionality.
GPIO pins 37 to 30 have been successfully switched between management and user space control and digital input vs output.
Work continues to enable the remaining 30 pins (0 thru 29).
Harness Functionality
A CPU stress test of the management core has been successfully run consisting of arithmetic, recursion and sort functions working when configured using the DFFRAM.
Automation
An automated remote test harness has been built and test automation is being developed to accelerate test coverage across multiple parts.
MPW-2 Hardware
MPW-2 WCSP packaged parts are now at Efabless office in California.
Efabless will be keeping 10 samples of each ICs. This will allow us to better assist users when they ask questions about their ICs.
Breakout board fabrication and assembly is starting. Still waiting to get confirmed delivery dates back from the vendor.
We are continuing to evaluate the final specification for the test harness. One option uses a hardware setup based on the configuration Sylvain Munaut developed for MPW-1. We are continuing to review with Sylvain.
A second option uses an off-the-shelf ST Nucleo development board. The block diagram for this option has been defined and the schematic is under development. We are expecting this to be completed tomorrow and plan to build and test this harness in parallel with the first option being considered.
MPW-2 Simulation
Review of the design flow continues. Several issues have been identified with respect to extraction and timing analysis. Issues are being created for the respective components.
The design and implementation for the GPIO configuration logic continues to be examined. Several paths have been identified as potential contributors to GPIO configuration issues observed during bench testing.
Proprietary licenses have been secured and are being set up for simulation and timing analysis in addition to the current open-source tools.
Other Analysis
All user projects on MPW-2 are being reviewed to understand the potential impact of GPIO configuration issues. We expect to have further information on this as part of next week's update.
Regards,
Jeff DiCorpo