IBM tips nanophotonic switch for on-chip optical network

Mar. 17, 2008 – IBM says it has developed a silicon broadband optical switch, a key component to enable on-chip optical interconnects. The device helps “direct traffic” of electrical signals, converted into light pulses so optical messages can efficiently get form one processor core to another.

An increasing trend in microelectronics is increasing parallelism in computation by mulithreading, multichip systems, and most recently with multicore chips — but the latter works with simultaneous data transmission/receipt of each core. Copper wiring in today’s multicore chips would use too much power and not transmit enough information to enable massively multicore processors. But IBM says its work will help enable transmission of data encoded as pulses of light, whereby up to 100x more information can be sent between cores, using 10x less power (and thus generating less heat).

The new switch has several attractive characteristics for use in on-chip applications, according to IBM in a statement. It’s compact; 2000 can fit in a 1 sq. mm space. It also can route “a huge amount of data” by switching many different light wavelengths simultaneously — each carrying up to 40Gb/sec of data, for a total of 1Tb/sec. And the optical switch can operate “with a realistic on-chip environment” with a temperature-drift tolerant operation, meaning it can work even where “hot spot” temperatures on the chip can move around depending upon the chip functionality at a given time.

“This new development is a critical addition in the quest to build an on-chip optical network,” stated Yurii Vlasov, manager of silicon nanophotonics at IBM’s TJ Watson Research Center. “In view of all the progress that this field has seen for the last few years it looks that our vision for on-chip optical networks is becoming more and more realistic.”

The report on IBM’s work is published in an advance online publication of the April 2008 issue of the journal Nature Photonics.

POST A COMMENT

Easily post a comment below using your Linkedin, Twitter, Google or Facebook account. Comments won't automatically be posted to your social media accounts unless you select to share.