NL.net proposal

Project name

EOMA68 Eco-Conscious Computing

Website / wiki

https://rhombus-tech.net/nlnet_2019

Please be short and to the point in your answers; focus primarily on the what and how, not so much on the why. Add longer descriptions as attachments (see below). If English isn't your first language, don't worry - our reviewers don't care about spelling errors, only about great ideas. We apologise for the inconvenience of having to submit in English. On the up side, you can be as technical as you need to be (but you don't have to). Do stay concrete. Use plain text in your reply only, if you need any HTML to make your point please include this as attachment.

Abstract: Can you explain the whole project and its expected outcome(s).

The EOMA68 Project is a paradigm shift in computing that puts end-users back in control of their computing devices, respecting their privacy, and almost incidentally brings long-term cost savings that ultimately results in reduced e-waste by encouraging re-use by design through modularity. The full list of benefits are actually extremely comprehensive and required a white paper to properly list (http://rhombus-tech.net/whitepapers/ecocomputing_07sep2015/)

In addition to the ecological benefits (synonymous with cost savings) there are privacy aspects to the project. With the actual computer being "removable" (because it is literally a Computer "Card"), the Computer Card may be taken out and hidden from prying eyes. Interestingly it is mistaken for a mobile phone when viewed by an X-Ray machine (same PCB size, same X-ray signature).

An additional requirement of the project that inspires trust by end-users is that there is a financial incentive for Computer Card Manufacturers to provide full Operating System and boot loader source code, and to permit end-users to run their own Operating Systems.

However the primary reason for applying for this grant is that there is presently an extremely serious problem with the typical processors used in smartphones, tablets, netbooks, laptops and desktop computers (nearly every device in common usage today): Spectre and Meltdown security attacks.

Spectre is basically timing attacks present and undetected for the past 25 years, in almost every modern high-end processor used today. The problem is that it is so prevalent and so recent that new attacks are being discovered every year, and this is hampering processor designers desperately trying to design new processors that have fixed the flaws (only to have new ones discovered during the design phase), where the cost of developing a new processor is literally tens of millions of dollars and can take over two years to get into production.

It turns out that the EOMA68 Project is using lower-cost, low-to-mid-range power-efficient processors that, by virtue of being "in-order", are immune to Spectre attacks. These processors, such as the Allwinner A20, and Ingenic jz4775, are the only types that can be used as a firm fundamental basis to offer end-users guaranteed privacy and immunity from Spectre-style attacks.

Have you been involved with projects or organisations relevant to this project before? And if so, can you tell us a bit about your contributions?

Luke Leighton is an ethical technology specialist who has a consistent 23-year track record of developing code in a real-time transparent (fully libre) fashion, and in managing Software Libre teams. He currently is fulfilling a USD $200,000 successfully-funded crowdfunding campaign: an eco-conscious computing project.

He is also creating a Libre RISC-V SoC that is publishing, to the bedrock, not only the full source code of the bootloader, BIOS and Operating System, but the full HDL of the processor itself (see http://libre-riscv.org)

Requested Amount

EUR $50,000.

Explain what the requested budget will be used for?

The budget will be used to finish prototype Computer Cards and associated Housings. Each prototype Computer Card run requires around USD $2,500 to have 10 samples manufactured, and requires around 2-3 months of design work and planning, per iteration. Those 2-3 months require living expenses to be paid (around USD $1500 per month), so each iteration (revision) of any given Computer Card can cost around USD $7000 a time.

There are 3 Computer Cards that need to be reworked: one uses the Rockchip RK3288, another uses the Ingenic jz4775: it would be good to investigate other processors such as the Allwinner R40.

The focus will be on Cards that have in-order processors, because only in-order processors are fundamentally immune from spectre-style timing attacks. In addition, ONLY processors that are not "DRM Locked" will be used, so that end-users are GUARANTEED to be able to replace the entire bootloader, firmware, kernel and OS, IN FULL.

On the Housings side, there is a 15.6in Laptop Housing, a 7in Tablet Housing, a Mini-Desktop Housing (similar to a NUC), and a Router Housing (which includes an Atheros Libre WIFI chipset). The Laptop Housing is particularly complex, the biggest cost being time (estimated 8-10 months), and approximately USD $4,000 in materials.

Note: Full source code and full 3D CAD files will be provided so that end-users may not only independently audit the products to not have any way of violating privacy, they may also repair and improve the products indefinitely. (Not just "right to repair" but "right to repair, manufacture, improve, upgrade, replace, repurpose and redesign").

Does the project have other funding sources, both past and present?

The project was initially self-funded, for several years. A sponsor (ThinkPenguin) came forward for two years, providing USD $60,000 to complete the first Laptop Housing prototype. This enabled the project to apply for crowdfunding (https://www.crowdsupply.com/eoma68/micro-desktop/) which was successful.

Unfortunately, extremely specialist mid-mount single-supplier components (a USB OTG and extremely small Micro-HDMI connector) went end-of-life several times during the development of the products, requiring several redesigns. The crowdfunding campaign was not set up to cater for Research and Development costs, it was set up to cater for manufacturing. It has taken almost eighteen months to complete the iterations of the pre-production prototypes, leaving barely enough funds to fulfil the campaign promises.

Sponsors, donations and occasional Contract work therefore barely manage to fill in the gaps. Once the first products ship, however (estimated some time in March to April), second batch Crowdfunded funds will become available, allowing the project to begin to move forward again.

Funding for the project will however make things a lot easier, and will reduce the time that it takes to get privacy-respecting products into peoples' hands.

Compare your own project with existing or historical efforts.

There are actually several similar "modular" computing projects. Intel's "Compute" Card claimed to be the "World's! First!" when in fact they are about the 5th.

Interestingly, the EOMA68 Project is the only one that is specifically focussed on respecting end-user privacy and on long-term cost reductions for the end-user, and e-waste reductions.

Intel's Compute Card can already be declared to be a privacy failure because it is single-manufacture. The BIOS is proprietary, the processor contains an NSA spying backdoor co-processor (known as the "Intel Management Engine"), and there is no guarantee that Intel will not pull the plug on OEMs, just like they did with the USB "Compute Stick".

By complete contrast, EOMA68 is actually set up as a Certification Mark that is impartial (FRAND) and may be licensed not only by large Corporations but by small-volume Electronics Engineers (anyone wishing to create a product that is entirely Libre / Open will receive a zero-royalty license as well as huge amounts of assistance and free support).

There is also the kangaroo.cc Computing Project (kangaroo.cc), since shut down. There is also the ICE PC (icecomputer.us) from Taiwan, that seems not to have taken off. We surmise that the fact that, like the Intel Compute Card, it is a single manufacturer, leaves people unwilling to trust its longevity, whereas if EOMA68 supports more than just "Intel Processors" and is independent of any one Manufacturer, it is likely to stick around for a decade or two.

Basically, there really isn't anything that's the same as EOMA68 when you look a bit closer. Even the failed Google Project ARA (a modular smartphone) failed precisely because it was a single company (Google).

What are significant technical challenges you expect to solve during the project, if any?

The most significant challenges faced by the project are actually those of logistics of component supply. Most of those have now been solved: a factory located in Shenzhen where the owner is extremely conscientious.

The primary technical challenges are those of PCB Design. However, again, there are experts that are available to solve this. Some are helping the project out free of charge, due to the nature and benefits of the project, whilst others can be contracted.

Describe the ecosystem of the project, and how you will engage with relevant actors and promote the outcomes?

The project has been running for several years now and has a mailing list (http://lists.phcomp.co.uk/pipermail/arm-netbook/), a wiki (http://rhombus-tech.net) and a Crowdsupply page with over 60 updates (https://www.crowdsupply.com/eoma68/micro-desktop/)

In addition, various milestones have been picked up numerous times on the top techie news sites. During the campaign there were over 70 separate news articles, discussions, and radio interviews, a list of which were recorded at the time, here: http://rhombus-tech.net/crowdsupply/

This trend will continue indefinitely, because it is critical to the project's success that it be fully transparent.

Extra info to be submitted