Efficient communication between integrated circuits within an electronic component is a major concern for designers of electronic systems. The 'Interconnect,' as it is called in the industry, is considered to be a major bottleneck in the design of high performance electronic systems, both in terms of speed, and in terms of its contribution to the total energy consumption of such systems. Be it on a mobile device, where data communication may take place between the processor and memory, or in a server farm where communication takes place between devices in different racks, such high speed interconnects can contribute up to 40% of the total energy consumption. Taking into account that data centers alone consume 200 billion kilowatt hours annually worldwide, any decrease in the energy consumption of the interconnect has a meaningful impact on the society at large. At the heart of today?s interconnects lies ?differential signaling,? a method for transmitting information between components in an electronic device. This method uses two wires to communicate one bit by putting a physical quantity (voltage or current) of opposite polarities on the two wires, and reading the information as the polarity of the difference. The method provides for excellent noise immunity, but because of the limits on the number of pins on a chip, it needs to transmit the data at double the speed. This typically leads to a higher energy consumption, especially at high speeds. In the course of our research within the ERC funded Advanced Grant 'ECCSciEng' we have discovered a completely new class of signaling methods which have better noise immunity, require far fewer wires, and consume far less energy than differential signaling. The method has the potential to drastically reduce the power consumption of electronic devices. The goal of this proposal is to explore this new class of signaling methods further, and build prototypes to convince to adopt them into their products.
Call for proposal
See other projects for this call