Update: Phiar has shut its doors, according to GigaOm.
A competitor to semiconductors would break a 50-plus year monopoly on everything from processing power to communications. Yet that’s exactly what Phiar says it can do, with a new technology called metal-insulated electronics.
The new technology uses a phenomenon called quantum tunneling to achieve greater speeds and efficiencies than semiconductors. It essentially makes low-energy electrons tunnel directly through the insulating material of the electronics, whereas most semiconductors change the electrical states of electrons to channel them through specific bands. Think of it like pinball; semiconductors shoot the electron around the side, while Phiar’s technology prods it right through the barriers.*
However, there are decades of history behind using semiconductors. Even if it’s possible for Phiar to spark a computing revolution, it won’t happen overnight, or even in a decade. That’s why the company is starting off in two vital areas it thinks it can dominate: Short range communications and flash memory.
Take communications. Imagine, if you can, transferring a movie from your computer to a nearby television in as much time as it would take to slide a DVD into the player.
That’s possible today, using the 60 gigahertz wireless spectrum, but the materials used for the semiconductors, indium phosphate or gallium arsenide, are extremely expensive and difficult to integrate with silicon, the cheap standard material used in semiconductors. Several corporations are at work on modifying semiconductors for 60GHz, but Phiar says its devices will be much cheaper.
The metal-insulated chips used by Phiar will also be less power-hungry than semiconductors, meaning they can be used in devices like cell phones and laptops without quickly draining their power. And metal-insulated electronics can be “grown” on top of silicon. Phiar is currently in a co-development deal with Motorola for the devices, among partners.
For flash memory, Phiar similarly promises to make large strides forward, again in speed but also in miniaturization. There are currently two competing types of flash memory, NAND and NOR; the former being slow but densely made (and thus small), the latter being fast but less dense.
The company says it can bridge the gap between the two technologies, packing more cells into less space but using the innate properties of metal-insulated electronics to break through speed barriers that semiconductors have not yet reached.
Bob Goodman, Phiar’s CEO, told us that the company has a deal with a major flash manufacturer, but couldn’t disclose which one.
We’re talking about a potentially huge story for the computing world here — that is, if the company is successful in spreading its technology. However, it’ll be tough to break the existing relationships around semiconductors — a $200 billion industry — and create a new ecosystem for technology that is still unproven.
Goodman acknowledges that ongoing research in semiconductors could yield a breakthrough to challenge metal-insulated electronics. Everyone is eager to add 60GHz networking, the equivalent of Bluetooth sped up hundreds of times over, to their products. A partnership between IBM and semiconductor maker MediaTek is just one of the many projects chipping away at modifying semiconductors for 60GHz.
Phiar thinks its technology will win out, though, as it’s a physical impossibility for today’s semiconductors to achieve the speed feats of metal-insulated electronics. Mark Siegel, a partner at investor Menlo Ventures, told us it’s unusual to see a technology with as much potential, saying, “It’s very infrequent that we’d invest in something so early, but the technology really is novel.”
He thinks that the company has the potential to raise money through a public offering at some point in the future, which would help it to build production facilities and make the chips on its own.
A semiconductor market analyst we talked to,Vahé Mamikunian of Lux Research, said he thinks Phiar’s chances are good based just on its scientific expertise. “They seem to have a good hold on how physics work at the nanoscale, better than some of the leading companies in the markets they’re looking at,” he said.
Phiar received its first funding, for $9 million, from Menlo Ventures. It’s currently working on raising another round.
*(Continuing from the explanation above: This is just the tip of the diode. Phiar is actually a double take on an older single-insulator technology that’s been used for almost as long as semiconductors, creating their own metal-insulator-insulator-metal model — think of it like a sandwich. Where one insulator tends to raise its resistance to electrons passing through as they gain energy, with two, Phiar can cause the materials to create single-directional “quantum wells” that mimic empty space, allowing the electrons to perform their tunneling trick and pass through. For further information, go get a Ph.D.)
VB's research team is studying web-personalization... Chime in here, and we’ll share the results.