Search
Menu

Optical Computing Breakthrough Addresses Memory Limitations

Facebook X LinkedIn Email
PITTSBURGH, Oct. 25, 2024 — Up to this point, researchers have been limited in developing photonic memory for AI processing — gaining one important attribute, like speed, has meant sacrificing another, like energy usage. An international team of researchers has demonstrated a solution that addresses current limitations of optical memory that have yet to combine nonvolatility, multibit storage, high switching speed, low switching energy, and high endurance in a single platform.

The team includes researchers from the University of Pittsburgh Swanson School of Engineering; the University of California, Santa Barbara (UC Santa Barbara); the University of Cagliari; and the Tokyo Institute of Technology (now the Institute of Science Tokyo).

“The materials we use in developing these cells have been available for decades. However, they have primarily been used for static optical applications, such as on-chip isolators rather than a platform for high performance photonic memory,” said Nathan Youngblood, assistant professor of electrical and computer engineering at the University of Pittsburgh.
A concept image depicting a photonic in-memory computing scheme which leverages the non-reciprocal phase shift in magneto-optical materials. Courtesy of UC Santa Barbara/Brian Long.
A concept image depicting a photonic in-memory computing scheme which leverages the nonreciprocal phase shift in magneto-optical materials. Courtesy of UC Santa Barbara/Brian Long.

The discovery, he said, is a key enabling technology toward a faster, more efficient, and more optical computing architecture that can be directly programmed with CMOS circuitry, which makes it compatible with conventional computer technologies.

“Additionally, our technology showed three orders of magnitude better endurance than other nonvolatile approaches, with 2.4 billion switching cycles and nanosecond speeds,” Youngblood said.

The researchers propose a resonance-based photonic architecture that leverages the nonreciprocal phase shift in magneto-optical materials to implement photonic in-memory computing.


A typical approach to photonic processing is to multiply a rapidly changing optical input vector with a matrix of fixed optical weights. However, encoding these weights on-chip with traditional methods and materials has proven challenging. By using magneto-optic memory cells comprised of heterogeneously integrated cerium-substituted yttrium iron garnet (Ce:YIG) on silicon micro-ring resonators, the cells cause light to propagate bidirectionally, like sprinters running opposite directions on a track.

“It’s like the wind is blowing against one sprinter while helping the other run faster,” said Paulo Pintus, who led the experimental work at UC Santa Barbara. “By applying a magnetic field to the memory cells, we can control the speed of light differently depending on whether the light is flowing clockwise or counterclockwise around the ring resonator. This provides an additional level of control not possible in more conventional nonmagnetic materials.”

The team is currently working to scale up from a single memory cell to a large-scale memory array which can support even more data for computing applications. According to their research, the nonreciprocal magneto-optic memory cell offers an efficient nonvolatile storage solution that could provide unlimited read/write endurance at subnanosecond programming speeds.

“We also believe that future advances of this technology could use different effects to improve the switching efficiency,” said Yuya Shoji, associate professor at the Institute of Science Tokyo, “and that new fabrication techniques with materials other than Ce:YIG and more precise deposition can further advance the potential of nonreciprocal optical computing.”

The research was published in Nature Photonics (www.doi.org/10.1038/s41566-024-01549-1).

Published: October 2024
Glossary
artificial intelligence
The ability of a machine to perform certain complex functions normally associated with human intelligence, such as judgment, pattern recognition, understanding, learning, planning, and problem solving.
resonator
A resonator is a device or system that exhibits resonance, which is a phenomenon that occurs when an external force or stimulus is applied at a specific frequency, causing the system to oscillate with increased amplitude. Resonators are found in various fields and can take different forms depending on the type of waves involved, such as mechanical waves, acoustic waves, electromagnetic waves, or optical waves. Key points about resonators: Resonance: Resonance is a condition where a...
Research & Technologyoptical computingphotonic memoryartificial intelligenceresonatoroptical input vectorring resonatormemory cellmagneto-opticUniversity of PittsburghUniversity of CaliforniaUniversity of CaligariTokyo Institute of TechnologyInstitute of Science TokyoAmericasEuropeAsia-Pacific

We use cookies to improve user experience and analyze our website traffic as stated in our Privacy Policy. By using this website, you agree to the use of cookies unless you have disabled them.