We have to accept the limitations of silicon chips.
Current manufacturing methods may not be able to shrink silicon transistors much smaller than they’ve already become without compromising chip performance, but our appetite for ever-increasing computing power doesn’t seem to have an end. Emerging alternatives show promise as materials to build the next generation of computing hardware. Graphene, one molecule thick and more conductive than any other known material, can move electrons faster in less space than even the tiniest bit of silicon.
A biocomputer made of the molecular “motors” that perform tasks in living cells can also perform parallel calculations faster, cheaper, and with far more energy efficiency than conventional electrical computers. And for long-term storage, synthetic DNA packs so much data into so little space—think of how little DNA it takes to make one person—that every bit of information humans have ever generated could fit in a spoonful of lab-grown goo with room to spare.
Computers are binary.
Conventional computers execute instructions using sequences of electrical pulses: zero (off) and one (on). New computing technologies fracture this binary approach. Qubits, the bits used in quantum computing, operate through fluctuating electrical fields, which can represent a zero, a one, both at once, or some point in between—all at the same time. Theoretically, these computers will be able to solve highly complex problems millions of times faster than the ones we have today.
Neuromorphic technology, meanwhile, is modeled on the human brain, with thousands of electronic neurons connected to many other neurons, all exchanging electrical signals at the same time. While still binary, they can perform calculations faster because they work in parallel, not sequentially.
New kinds of computers are a long way off.
While quantum computers, neuromorphic chips, and other advances may sound like science fiction, they are on their way to becoming commercial products. General Vision has created a neuromorphic chip that can recognize patterns in microseconds and learn in real time; it’s already being used for data and visual pattern recognition applications. Google researchers announced in 2015 that they had developed a new way to protect the quantum entanglement—or stability—of qubits, a feat necessary for quantum computing to work without errors and outperform conventional computers. IBM’s 2015 breakthrough in creating graphene nanotubes is the first step toward building processors that will squeeze enormous computing power onto chips the size of dust particles. And researchers at Sweden’s Lund University have already built an experimental biocomputer, though it’s still only capable of doing simple calculations. As for storage, Microsoft is investigating synthetic DNA for secure long-term storage and has already successfully encoded and recovered 100% of its initial test data. D!
Read more thought provoking articles in the latest issue of the Digitalist Magazine, Executive Quarterly.Comments