Whether it’s multi-core approaches such as the Tile64 or FPGA-based approaches such as the OpenSPARC reported on recently, evidence of the dematerialization trend is everywhere. I recently came upon a site that is at the center of this

In those early days of my career, hardware design was a real man’s game. We designed big boxes with loud fans that roared as if boasting of its impressiveness. Then came ASICs where all of sudden your innovation was miniaturized into something only an inch across. Today, chips are disappearing altogether and the real design work is in IP— making chips and systems are simply manufacturing steps.

The focus of an engineer today is either in creating IP, or assembling others IP into dream fulfilling subsystems. The power and influence of the engineer simply keeps expanding, being able to create larger and larger works from the work of others.

Design & Resuse 

This site should be helpful to people creating solutions requiring devices that interact with the internet.

Via comments on an entry about Fujitsu’s 3D image recognition chip, I came upon this Cadalyst article on visualization which like the commenter, points out the advantages of integrating CAD and traditional design visualization tools

The ability to turn a design drawing into a visualization that mimics reality is an invaluable tool for troubleshooting a design, convincing a nervous client or helping to promote a design firm’s capabilities. This segment of the CAD software industry continues to grow and evolve, with rendered visualizations becoming more sophisticated.

While this trend is indeed a positive one, it is fundamentally constrained by an exclusively designer-centric perspective. This view understandably stems from the historical hardware constraints which made customer/end user access to the CAD data prohibitively expensive. However, as can be seen in the high end, proprietary CAD/PLM offerings of Dassault, this tradeoff is not an inherent requirement. In a rhythmeering environment integration is needed for manufacturing, maintenance, supply chain, marketing analysis - throughout the complete product lifecycle. In order to achieve this level of deep integration, the underlying information models for CAD have to become features of the operating system and eventually the hardware. Open source platforms like Croquet point the way.

Software used in today’s conventional milling machines is helping nanotech researchers make progress in nanomanufacturing:

The new technique suggests that the nanotechnology factories of the future might not operate so differently from existing manufacturing plants.

“If you can take prototyping and nanomanufacturing to a level that leverages what engineers know how to do, then you are ahead of the game,” Clark said. “Most engineers with conventional training don’t think about nanoscale manipulation. But if you want to leverage a workforce that’s already in place, how do you set up the future of manufacturing in a language that engineers already use to communicate? That’s what we’re focused on doing here.”

Duke News: Automation of Nanotech Manufacturing May Be Ahead

This should prove to be very helpful because there’s no getting around the strange and often counter-intuitive aspects of the nanoscale realm where thermal and quantum fluctuations make moving molecules from place to place like walking in a hurricane. Similar to the proven patterns used in software virtual machines, the Duke researchers are abstracting out the unfamiliar/counter-intuitive and substituting the more familiar. This is yet another sign of dematerialization and virtualization so I expect it will bear fruit.

In a nutshell, “rapid manufacturing” is poised for an unprecedented explosion of growth in the next 3 to 5 years. To see why this potential exists, it’s necessary to examine a broad set of shaping factors. If only a single segment is explored, significant growth looks to be much further out but when one takes into account the converging sources of influences and innovation at work, a different perspective emerges. in this regard it is helpful to examine some other patterns of technology evolution.
Initially PC’s were no match for mainframes when it came to raw processing power, but their accessibility(price and learning curve) enabled people to do things they simply couldn’t do before. In the process, people pushed the limits of PC’s and accelerated the demand for reducing their limits. They also at the same transformed the design and operation of mainframes - Linux and Java are significant contributors to renewed interest in mainframes. During the early days of PC’s many people didn’t see the potential for rapid growth because PC database programs could only manage a fraction of the data that mainframes did. These skeptics didn’t realize that departments and groups within departments did not need the capacity of a mainframe for many important tasks. They didn’t recognize how big an impact spreadsheets would have or what it would mean to empower thousands of developers previously unable to create solutions because they couldn’t afford the necessary equipment. A similar pattern unfolded for the web, although desktop publishing is probably more relevant to the subject of desktop manufacturing.
I found via reBang to an excellent, but narrowly focused review of a Design News feature on Rapid Manufacturing’s Role in the Factory of the Future. The discussion is valuable but assumes that traditional high production volume factories will continue to dominate the manufacturing landscape forever and ignores overlapping influences. Like mainframes and printing presses, high production volume factories will be with us for a long time, their fall from dominance will happen faster then most people think and they will be significantly transformed by the emerging paradigm. How will this happen? Services such as Xardas and Ponoko are starting to give people the very powerful experience of “holding ideas in their hands” and providing engineers with insights into new forms of fabrication. With 3D printer prices dropping into the consumer electronics range, the number of people and organizations able to fabricate goods from their computers will grow rapidly. Architects, landscapers and engineering entrepreneurs will find immediate uses for these but many folks especially those lacking professional design and manufacturing experience will be frustrated. Parts will break or won’t come out right, but through Supplier Source and other online sources connections to professionals will be found. It’s not hard to envision Google figuring out a fabrication tie-in to it’s 3D Warehouse. All of this activity will expand the base of experiences and provide valuable feedback for engineers and designers. It will also drive demand for higher end 3D fabrication machines, as well as CNC machines.

At some point I expect that Fed-Ex/Kinkos will probably throw their hat in the ring and some distributed manufacturing network startup with have a huge IPO. Perhaps more significantly, a new type of product or service that hasn’t been thought of yet will emerge(think Lotus 1-2-3 or Amazon). One source in the Design News article put the widespread use of direct digital manufacturing 20 years out but by then nanotechnology will have already started having a significant impact. Desktop manufacturing is being driven by exponentially growing factors it’s just always hard to see it in the early stages. I think Ray Kurzweil has it exactly right

Although technology grows in the exponential domain, we humans live in a linear world. So technological trends are not noticed as small levels of technological power are doubled. Then seemingly out of nowhere, a technology explodes into view. For example, when the Internet went from 20,000 to 80,000 nodes over a two year period during the 1980s, this progress remained hidden from the general public. A decade later, when it went from 20 million to 80 million nodes in the same amount of time, the impact was rather conspicuous.

Ray Kurzweil: The Law of Accelerating Returns