Optical microscope photographic image of an array of nanophotonic avalanche photodetectors on a silicon chip
On March 4, 2010 IBM scientists unveiled a significant step towards replacing electrical signals that communicate via copper wires between computer chips with tiny silicon circuits that communicate using pulses of light. As reported in the recent issue of the scientific journal Nature, this is an important advancement in changing the way computer chips talk to each other.
What has IBM announced today?
We announced the development of an ultra-high speed, ultra-low power avalanche photodetector which converts faint optical signals into electrical signals. This device is a significant step towards replacing copper wires between the computer chips with tiny silicon circuits that can communicate via pulses of light rather than electrical signals. It is the world’s fastest device capable of receiving optical information signals at 40Gbps (40 billion bits per second) while simultaneously multiply them tenfold. Moreover the device operates with just a 1.5V voltage supply, which is 20 times smaller than previous demonstrations. The IBM device is made of Silicon and Germanium, the materials already widely used in production of microprocessor chips. Moreover it is fabricated using standard processes used in chip manufacturing. Thus, thousands of these devices can be built side-by-side with silicon transistors for high-bandwidth optical communications.
How does the avalanche photodetector work?
First, information encoded onto pulses of light travels down a silicon nanophotonic waveguide on the silicon chip. This waveguide is a nanometer-scale “light pipe” which delivers the pulses of light to the input of the avalanche photodetector. Then, the incoming light pulse frees a few electrons and holes in the photodetector. These electrons and holes in turn free other electrons and holes, and so on, until the original signal is amplified many times. This electronic process is analogous to a snow avalanche on a steep mountain slope. Thus, by utilizing the avalanche multiplication effect, the information encoded on the incoming light pulses is amplified ten-fold and converted into digital electrical pulses. The avalanche multiplication is very important, because it allows the detection of weak light pulses traveling on the silicon chip. In IBM’s device, the avalanche multiplication takes place within just a few tens of nanometers (one thousandth of a millimeter), which enables ultra-fast operation and minimizes signal degradation due to noise.
Why is this a significant advancement?
Optical communications require avalanche photodetectors that are capable of receiving information at very high speed, detecting and amplifying weak pulses of light, and operating at low voltage. Additionally, the photodetectors have to be integrated onto the waveguide and occupy very small footprint so that hundreds of them can be fabricated on the same chip alongside transistors. IBM’s device is the first demonstration which fulfils all of the above requirements. The device operates at 40Gbps while achieving ten-fold multiplication, and requires a voltage of only 1.5V. While conventional avalanche photodetectors are not able to detect fast optical signals because the avalanche builds slowly, IBM’s device utilizes nanoscale effects to obtain very fast multiplication. Also, avalanche multiplication in Germanium photodetectors had been considered a very noisy process that degrades the signals. IBM has “reinvented” Germanium avalanche photodetectors by demonstrating high gain while the multiplication noise is suppressed by 50-70%.
When will this technology be available to the general public?
Currently, optical interconnects are the focus of a big research and development effort. There is a lot to be done before nanophotonic devices to appear on a microprocessor chip, but we have made a lot of progress in demonstrating the various components that are required. Actual integration of photonics devices on the microprocessor chip would probably happen somewhere within the next five to ten years.
However, specialized silicon-based photonic devices integrated with CMOS electronics are already on the market. For example, some companies working components for active cables or telecommunications are making very good progress.
What is the motivation for on-chip optical interconnect networks?
In today’s computer systems, communication between the chips is performed using electrical signals over millions of tiny copper wires. As the number of cores increases in order to handle larger computations with greater speed, an enormous amount of data has to be routed on the chip between each core, and off the chip to nearby memory or another chip. However, a very large amount of power is burned while routing all the data, thus heating of the chip and severely limiting the performance. The copper wiring would simply use up too much power and be incapable of transmitting the enormous amount of information. By using pulses of light rather than electrical signals in wires, as much as 100 times more information can be sent between cores, while using 10 times less power and consequently generating less heat. Hence, optical communications will be critical for next-generation computer systems.
What were the challenges in this project?
The nanophotonic devices have to compact so that a large footprint will not be required when integrating hundreds of them on the same chip. Also, active devices such as modulators and detectors have to be very fast and power efficient. Moreover, the nanophotonic devices have to be fabricated by utilizing standard processes which are used for making transistors. Achieving all of the above is a very exciting challenge.
What are the next steps?
We would like to reduce the dark current and increase the responsivity of the device. Also, we would like to increase the multiplication factor in the photodetectors while maintaining the ultra-low power and ultra-fast performance. There are many technical challenges in making these improvements, but none of them are fundamental barriers. We have already made further progress in these areas. The next step then is a further advanced stage of development wherein we would integrate the devices with CMOS circuitry. The device is the last puzzle in our “nanophotonic toolkit”, and we are currently working on tackling the challenge of integrating all of our nanophotonic devices into computer chips.
IBM’s Silicon Integrated Nanophotonics Project
The ultimate goal of this project is to develop a technology for on-chip integration of ultra-compact nanophotonic circuits for manipulating the light signals, similar to the way electrical signals are manipulated in computer chips. Nanoscale silicon photonics circuits are being developed to enable the integration of complete optical systems on a monolithic semiconductor chip that would eventually allow to overcome severe constraints of today’s mostly copper I/O interconnects