This is the text I wrote in 2016, shortly before we incorporated Hyperganic. It describes the founding idea of Hyperganic and our quest to evolve the way we design objects in the world of industrial 3D printers. Today, with our Algorithmic Engineering approach, using Hyperganic Core, we are demonstrating the power of moving engineering under a software paradigm. In my talk at the Singularity University Summit in 2017, I emphasized the impact on our industry.

Until the end of World War II, computers were incredibly complex machines that filled entire buildings, built using many mechanical and electrical components. IBM’s Thomas Watson estimated in 1943 that the worldwide market for computers would be a handful of computers — because computers took years to design and build, and were therefore extremely exclusive and expensive machines.

Then, in 1945, the transistor was invented and finally, computers became electronic machines that required less space. But building and designing them was still a very complex effort.

In 1958/59, the first integrated circuits were demonstrated, that combined multiple electronic elements on a wafer instead of discrete individual components. The potential was there to print entire computers on a chip.

In 1971, the first microprocessor was introduced, the Intel 4004 — essentially the first printed computer. But the 4004 and its successors were still designed by hand and its schematics manually drawn graphically. With the 4004, the mass production of computers was possible, but the way the blueprints were created had not really evolved much from the early days of computing.

In 1980, Carver Mead and Lynn Conway at XEROX PARC introduced a publication called “Introduction to VLSI [Very Large Scale Integrated] Systems” in which they described a fundamental new paradigm for chip design. Instead of following the traditional way of graphically describing all the elements and connections on a microprocessor, they advocated the use of abstract specification languages to describe the functionality of the chip — and then let the software design the actual physical layout.

This was a dramatic change in the way chips were designed and created the entire space of Electronic Design Automation (EDA) tools, which are a key driver of Moore’s Law. EDA tools allowed chip designers to focus on the high-level functionality and removed the error-prone tedium of manual layout work. The complexity of chip structures started to skyrocket and in conjunction with the progress in the field of lithographic printing processes for chips, made the revolution in the computer industry possible.

In order for 3D printing to live up to its full potential, we need to evolve the paradigm for designing objects and machinery in a very similar way. Currently, CAD/CAM tools require the designer to essentially describe every step of the construction process manually, similar to the graphical layout work that the first chip designers did. Automation and generative features in the existing tools are at an absolute minimum and a far cry from the breakthrough which Mead and Conway described in “Introduction to VLSI Systems”.

If we want to advance to the next level of complexity and move industrial manufacturing under Moore’s Law, we need to dramatically reduce the effort to design complex objects and machinery through computer-generated and computer-simulated design tools. These new tools must use very abstract descriptions of design intent to then automatically create the blueprints of the physical objects to be printed.

The Additive Manufacturing machines are now at the level where they are starting to reliably produce end-use parts that can be used in a final product. We are therefore at a very similar stage that we were, when Intel introduced the first 4004 microchip and the design tools started to evolve.

The time is ripe to move the process and the entire software ecosystem for 3D printing to the next level.