Photonic communication platform
Described herein are photonic communication platforms that can overcome the memory bottleneck problem, thereby enabling scaling of memory capacity and bandwidth well beyond what is possible with conventional computing systems. Some embodiments provide photonic communication platforms that involve use of photonic modules. Each photonic module includes programmable photonic circuits for placing the module in optical communication with other modules based on the needs of a particular application. The architecture developed by the inventors relies on the use of common photomask sets (or at least one common photomask) to fabricate multiple photonic modules in a single wafer. Photonic modules in multiple wafers can be linked together into a communication platform using optical or electronic means.
1. A device comprising:
a first computing system and a second computing system, each of the first and second computing systems comprising:
a plurality of processing units configured to transmit and receive signals;
a plurality of waveguides connecting the plurality of processing units into a photonic network supporting propagation of the signals among the plurality of processing units; and
a plurality of optical switches configured to selectively place subsets of the plurality of processing units in communication with each other;
a controller configured to control the plurality of optical switches; and
a fiber connecting the first computing system with the second computing system, wherein the fiber supports propagation of the signals from the first computing system to the second computing system.
2. The device of claim 1 , wherein the plurality of processing units and the plurality of waveguides are 3D-stacked.
3. The device of claim 1 , wherein the first computing system further comprises a first substrate comprising the plurality of optical waveguides and a second substrate comprising a first processing unit of the plurality of processing units, wherein the first substrate and the second substrate are bonded to one another.
4. The device of claim 1 , wherein each of the first and second computing systems further comprises an optical transmitter (TX) coupled to a first optical switch of the plurality of optical switches, and an optical receiver (RX) coupled to the first optical switch.
5. The device of claim 4 , wherein the first switch is configured to arbitrate between the optical TX and the optical RX.
6. The device of claim 1 , wherein the controller of the first computing system is further configured to:
at a first time, program the plurality of optical switches of the first computing system to form a first optical communication path coupling together a first subset of the plurality of processing units of the first computing system, and
at a second time subsequent to the first time, program the optical switches of the first computing system to form a second optical communication path coupling together a second subset of the plurality of processing units of the first computing system.
7. The device of claim 1 , wherein the fiber is edge coupled to both the first and second computing systems.
8. The device of claim 1 , wherein each of the first and second computing systems further comprises an out-of-plane coupler, wherein the fiber couples the out-of-plane couplers of the first and second computing systems to one another.
9. The device of claim 1 , wherein the plurality of processing units are configured to implement an artificial neural network.
10. The device of claim 1 , wherein each of the first and second computing systems further comprises a plurality of dies, each of the plurality of dies comprising a respective processing unit of the plurality of processing units.
11. The device of claim 1 , wherein the first computing system further comprises a laser die coupled to at least one of the plurality of waveguides.
12. The device of claim 1 , wherein the first computing system further comprises a memory unit, wherein the plurality of waveguides further connect the plurality of processing units to the memory unit, and wherein the plurality of optical switches are further configured to selectively place a first processing unit of the plurality of processing units to the memory unit.
13. The device of claim 12 , wherein the memory unit comprises a stacked memory unit.
14. The device of claim 1 , wherein the plurality of optical switches comprise Mach Zehnder interferometers.