As artificial intelligence technology rapidly advances, market demand for computing power is growing significantly. Serving as the foundation of computing power, memory chips face challenges with data interaction latency and power consumption, which are bottlenecks constraining computational performance improvements. A new discovery by Chinese scientists offers potential to overcome this hurdle.
Recently, the team led by Professor Zhou Peng and Liu Chunsen at Fudan University achieved a breakthrough in room-temperature single-electron non-volatile quantum memory technology. This accomplishment not only overcomes the physical limitation that quantum effects are typically observable only at extremely low temperatures but also reaches the physical极限 of using a single electron to represent one information bit, drastically reducing memory power consumption by orders of magnitude.
This original "leapfrogging" achievement was published in the journal Science on July 17. Science provided an evaluation, noting it is "promising and potentially high-impact, attracting significant attention in the fields of memory physics and nano-device engineering," and that it "introduces a new theoretical mechanism (the density of states scissors), making engineered manipulation of quantum states possible."
Professor Zhou Peng, in an interview, stated that the team plans to formally establish a startup company in the second half of this year. The initial goal is to complete chip verification based on modifications to existing semiconductor production lines.
Leveraging Existing Processes for Technological Leapfrogging
Previously, the Zhou-Liu team developed the world's fastest 400-picosecond "PoX" flash memory device by deeply understanding and interpreting Maxwell's first equation. This solved the fundamental challenge since the invention of the floating-gate transistor in 1967 where high speed and non-volatility could not be simultaneously achieved. Furthermore, they integrated it into CMOS processes to develop the "CY-01" hybrid architecture full-function flash memory chip, which was hailed by Nature as an "original breakthrough" and selected for the 2025 "Top 10 Scientific Advances in China."
"PoX" signifies the dawn, and "CY-01" represents holding the sword. After "PoX" unlocked the fastest electronic storage speed for humanity, where is the density极限? As an indivisible fundamental particle, the electron is theoretically the ultimate carrier for constructing the smallest data unit—single-electron storage. However, due to its deep involvement with the quantum behavior of fundamental particles, it has long been regarded by the scientific community as a "theoretically feasible but experimentally unobservable" castle in the air, with practical applications considered distant.
The latest Dynamic Random-Access Memory (DRAM) technology requires maintaining 200,000 electrons in a device to independently represent one information bit, limiting the potential for high density. If the charge storage device is likened to a "large reservoir," previous technologies required filling the reservoir to know if it contained water, while the reservoir was constantly leaking. The Fudan University team, by altering the "reservoir" structure, enabled perception of the world "leak-free" with just "one drop of water." This is because when the scale of the charge storage device approaches the physical极限, quantum effects in the microscopic world are dramatically amplified. At this point, storing just "one" electron is sufficient to cause a significant change in the device's state.
Explaining "single-electron storage," Professor Zhou used an analogy: In the past, we needed to fill a "room" with 200,000 people to sense that "someone is there." Now, because the device is extremely微小 and sensitive, even if just one person (one electron) enters, the overall electrical potential of the room undergoes a drastic change and is accurately captured.
Solving the "Room Temperature" Challenge
The team首次 clearly observed non-volatile single-electron storage behavior at room temperature (27°C). Bringing this technology into a room-temperature environment破解了 the biggest bottleneck for its industrialization. This macroscopically observable quantized behavior has long been默认 in the scientific community to appear only under极低温 conditions. For instance, the Nobel Prize-winning research in 2025 on "macroscopic quantum tunneling and energy level quantization in circuits" conducted its key experiments at接近绝对零度的极低温 (around -272°C).
The team not only achieved a leap in the quantum storage window but also introduced颠覆性 innovation at the foundational level,首次 proposing the "density of states scissors" theory. Liu Chunsen explained that by constructing a double Dirac structure, an interval of "zero density of states" that cannot accommodate electrons is introduced into the microscopic world, enabling manipulation of quantum states. Based on this novel "scissors mechanism," the team首次 revealed a previously unseen anomalous quantum storage behavior: using an invisible "quantum scissors" in energy space to precisely "cut out" specific quantum states, making them vanish. This achievement not only pioneers a全新 theoretical system for single-electron quantum storage but also provides a crucial theoretical piece for advancing quantum memory toward engineering applications.
Regarding industrial implementation, the biggest bottleneck currently facing data centers is the energy loss during the memory-compute interaction process (data movement power consumption is often several times that of computation power). This technology can not only provide ultra-low-power memory arrays but also, through Back-End-Of-Line (BEOL) semiconductor integration technology, directly integrate memory cells above computing units, reducing information transmission distance to the hundred-nanometer scale. This excellent "memory-compute integration" characteristic can fundamentally address the energy consumption痛点 of computing centers, empowering the national "East Data West Computing" strategy.
Comments