Blog Page


Research Bits: Nov. 21 – SemiEngineering

Graphene heater for phase-change switches; silicon defects for quantum computing; flash defects for more storage.
Researchers from the University of Washington, Stanford University, Charles Stark Draper Laboratory, University of Maryland, and Massachusetts Institute of Technology designed an energy-efficient, silicon-based non-volatile switch that manipulates light through the use of a phase-change material and graphene heater.
Aiming to reduce the power consumption of data centers, the “set and forget” switch is capable of maintaining a connection without any additional energy.
The non-volatile phase-change material can be put in a state by heating it, and it remains in that state until it receives another heat pulse, when it reverts back to its original state.
“This platform really pushes the limits of energy efficiency,” said Arka Majumdar, a UW associate professor of electrical and computer engineering and physics, and faculty member at the UW Institute for Nano-Engineered Systems and the Institute for Molecular & Engineering Sciences. “Compared with what is currently being used in data centers to control photonic circuits, this technology would greatly reduce the energy needs of data centers, making them more sustainable and environmentally friendly.”
Doped silicon has been proposed as a way to heat the phase-change material, but the researchers said that method results in wasted heat as the entire 220nm thick doped silicon layer has to be heated to transform only 10 nm of phase-change material. “We realized we had to figure out how to reduce the volume that needed to be heated in order to boost the efficiency of the switches,” said Zhuoran (Roger) Fang, a UW Ph.D. student in electrical and computer engineering.
Instead, they turned to an un-doped 220nm silicon layer to propagate light and introduced a layer of graphene between the silicon and phase-change material to conduct electricity. “This design eliminates wasted energy by directing all heat generated by the graphene to go towards changing the phase-change material. In fact, the switching energy density of this setup is only 8.7 attojoules (aJ)/nm3, a 70-fold reduction compared to the widely used doped silicon heaters, the current state-of-the-art. This is also within one order of magnitude of the fundamental limit of switching energy density (1.2 aJ/nm3),” the researchers note.
The team’s graphene-based heater could reliably switch the state of the phase-change material more than 1,000 cycles. “Even 1,000 is not enough,” said Majumdar. “Practically speaking, we need about a billion cycles endurance, which we are currently working on.”
They also plan to show that the switches can be used for optical routing of information through a network of devices, and investigate applying the technology to silicon nitride for routing single photons for quantum computing. “The ability to be able to tune the optical properties of a material with just an atomically thin heater is a game-changer,” said Majumdar. “The exceptional performance of our system in terms of energy efficiency and reliability is really unheard of and could help advance both information technology and quantum computing.”
Researchers from Simon Fraser University found that a specific luminescent defect in silicon, called T centers, can provide a ‘photonic link’ between qubits that could aid in construction massively scalable quantum computers and the quantum internet.
The researchers said this was the first measurement of any single spin in silicon to be performed with only optical measurements.
“An emitter like the T center that combines high-performance spin qubits and optical photon generation is ideal to make scalable, distributed, quantum computers, because they can handle the processing and the communications together, rather than needing to interface two different quantum technologies, one for processing and one for communications,” said Stephanie Simmons, Canada Research Chair in Silicon Quantum Technologies at Simon Fraser University.
Additionally, T centers emit light at the same wavelength that fiber communications and telecom networking equipment use.
“With T centers, you can build quantum processors that inherently communicate with other processors,” Simmons said. “When your silicon qubit can communicate by emitting photons (light) in the same band used in data centers and fiber networks, you get these same benefits for connecting the millions of qubits needed for quantum computing.”
Simmons also noted the benefits of using silicon. “By finding a way to create quantum computing processors in silicon, you can take advantage of all of the years of development, knowledge, and infrastructure used to manufacture conventional computers, rather than creating a whole new industry for quantum manufacturing. This represents an almost insurmountable competitive advantage in the international race for a quantum computer”.
Engineers from Pohang University of Science and Technology (POSTECH) and Samsung Electronics developed a flash memory with increased data storage by intentionally generating defects. In addition to increasing capacity, the method could have benefits for neural network processing.
The research team used a strong plasma bombardment process during the deposition of the data storage layer to generate artificial defect sites in the flash memory device. They found that more electrons could be stored in the generated defects, increasing the total trap density by 64%, dramatically increasing the amount of data storage compared to conventional flash memory.
In addition, the memory demonstrated multiple levels of data when electrons are gradually filled in the data storage layer with many defects. The multilevel flash memory developed by the team was able to reliably distinguish eight data levels, which can be used as weights, potentially improving inference accuracy and reliability.

(Note: This name will be displayed publicly)

(This will not be displayed publicly)

How prepared the EDA community is to address upcoming challenges isn’t clear.
Details on more than $500B in new investments by nearly 50 companies; what’s behind the expansion frenzy, why now, and challenges ahead.
More factors need to be weighed than just process scaling; leadership can vary month-to-month and by application.
113 startups raise $3.5B; batteries, AI, and new architectures top the list.
New memory approaches and challenges in scaling CMOS point to radical changes — and potentially huge improvements — in semiconductor designs.
Changes are steady in the memory hierarchy, but how and where that memory is accessed is having a big impact.
But that doesn’t mean it’s going to be mainstream anytime soon.
Exponential increase is not sustainable. But where is it all going?
Challenges continue to mount, but most of them appear surmountable with enough investment.
ML/DL is increasing design complexity at the edge, but it’s also adding new options for improving power and performance.
113 startups raise $3.5B; batteries, AI, and new architectures top the list.
To significantly reduce the power being consumed by machine learning will take more than optimization, it will take some fundamental rethinking.