What is the current status of the EOMA Earth Friendly Computing Devices project? On the one hand, I see no updates on the Crowdsupply project page.
20190304/https://www.crowdsupply.com/eoma68/micro-desktop
On the other hand, I do know that there is still being progress made by the same community on the RISC-V project. So, let’s see where that is at.
20190304/https://www.crowdsupply.com/libre-risc-v/m-class
20190304/https://www.crowdsupply.com/libre-risc-v/m-class/updates/microarchitecture-by-osmosis
20190304/https://www.crowdsupply.com/libre-risc-v/m-class/updates/modernising-1960s-computer-technology-learning-from-the-cdc-6600
20190304/https://www.crowdsupply.com/libre-risc-v/m-class/updates/microarchitectural-decisions
So, the most exciting thing that I wanted to point out. There is a architectural processor optimization that was implemented in the CDC 6600 decades ago, yet somehow this was not picked up by modern Intel and ARM processors. Out-of-order processor execution can be implemented either by the Tomosalu Algorithm with a reorder buffer or by scoreboarding. Tomosalu Algorithm is straightforward to implement, but scoreboarding, in the common interpretation of implementation, is energy intensive due to the use of content-addressable memory. However, the two methods can be understood to be functionally identical by the means of implementing a single-bit logical AND decision tree. I haven’t completely$ explained this here (see the articles for details), but the overall implementation and its results is ingenious. It allows you to get powerful out-of-order execution capabilities for a very low power budget.
Now, with the consideration of an extended instruction set for 3D capabilities in mind, there are other processor implementation specifics to watch out for when using many technologies together.
But yes, I think this project is overall exciting, and it is great for me personally to look at how some of my computer architecture knowledge that I learned in college, but is not used in 99% of the relevant job market, can be put to use… but not necessarily to profit. Yes, anyways it’s fun to follow along the project and think of how improvements can be made to the implementation of today’s processors.