World's First: Simultaneous Measurement of Heat and Entropy in a Semiconductor Memory DRAM Cell at the Single-Electron Scale

NTT has successfully measured heat and entropy at the single-electron scale in a room-temperature DRAM cell for the first time in the world, enabling experimental verification of the theoretical limit of minimum energy consumption in information processing.
調査NQ 81/100出典:PR Times

📋 Article Processing Timeline

  • 📰 Published: April 24, 2026 at 00:00
  • 🔍 Collected: April 23, 2026 at 15:32
  • 🤖 AI Analyzed: April 23, 2026 at 16:12 (40 min after Collected)
Key Points of the Announcement:

By using NTT's proprietary nanometer-scale (*1) electronic device capable of detecting changes in the number of electrons one by one, we have succeeded for the first time in the world in measuring heat and entropy (*2) at the single-electron scale in a semiconductor memory device, a DRAM cell, at room temperature.

It has become possible to experimentally verify the theoretical limit of minimum energy consumption in information processing in an actual device operating under realistic conditions, which had been impossible until now.

With this achievement, the energy consumption of information processing circuits can be evaluated based on thermodynamics, and it is expected to be applied to the development of energy-saving information processing devices.

NTT Corporation (Headquarters: Chiyoda-ku, Tokyo; President & CEO: Akira Shimada; hereinafter "NTT") has successfully achieved the world's first simultaneous measurement of heat and entropy at the single-electron level in a semiconductor memory device (DRAM cell) operating at room temperature. This was accomplished using a single-electron detection technology employing NTT's proprietary nanometer-scale electronic device capable of detecting changes in electron numbers individually. Furthermore, through this measurement, we experimentally verified the theoretical limit of minimum energy consumption in information processing. Additionally, we showed that thermal instability when retaining information is a crucial factor causing energy consumption to exceed the theoretical limit, in addition to the conventionally pointed out effects of high-speed processing and peripheral circuits. This research opens the way to evaluating the energy consumption of information processing circuits from the perspective of thermodynamics, and is expected to be applied to the development of energy-saving information processing devices and next-generation memory technologies. This result was published in the U.S. academic journal "Physical Review Letters" on March 20, 2026.

Figure 1: Overview of this research

1. Background
Energy saving in information processing has become an urgent issue against the backdrop of increasing power consumption accompanying the spread of generative AI, and various research and developments are underway. On the other hand, from a more fundamental perspective, a deep relationship between information and thermodynamics has become clear in recent years. For example, in information processing, an "initialization operation" is performed to align scattered information into a certain state. At this time, entropy, which represents the dispersion of information, decreases, and in return, heat generation, that is, energy consumption, occurs. The theoretical minimum value of this energy consumption is known as Landauer's principle (Landauer limit) (*3), and it is an important indicator when considering energy saving in information processing (Figure 2). However, actual electronic devices consume energy far exceeding this limit, and elucidating the cause of this discrepancy has become an important issue for creating energy-saving devices.

Figure 2: Explanation of Landauer limit and a schematic showing the discrepancy with the energy consumption of electronic devices

Conventionally, this discrepancy was thought to be due to the effects of high-speed processing and peripheral circuits, but experimental verification had not been conducted. NTT has previously engaged in research focusing on the relationship between information and thermodynamics (such as demonstrating power generation by Maxwell's demon (*4)), but in this research, we aimed to elucidate the energy-saving limits of information processing in electronic devices. Specifically, in order to eliminate the effects of high-speed processing and peripheral circuits, we focused on the circuit structure of a DRAM cell (Figure 3), which is the minimum unit constituting 1 bit in Dynamic Random Access Memory (DRAM), to verify whether it reaches the Landauer limit when the initialization operation is performed at low speed.

Figure 3: Circuit structure of the DRAM cell focused on in this research

However, there was a technical challenge in this verification. This is because under extreme conditions approaching the Landauer limit, the entropy and heat signals necessary for verification are extremely small and easily buried in noise, and there was no means to measure these in semiconductor devices operating at room temperature.

2. Key Technical Points
NTT solved this problem by applying single-electron detection technology using its proprietary silicon nanodevice. Specifically, using a high-performance detector fabricated by microfabrication technology (Figure 4 (a, b)), we successfully measured the amount of charge stored in the capacitor in units of single electrons (Figure 4 (c)), and evaluated heat and entropy from that information using the following methods.

① Measurement of Heat
When electrons move between the lead and the capacitor associated with information processing, heat (heat generation or heat absorption) is generated according to the potential difference between the two. Therefore, if the potentials of the lead and the capacitor at the moment the electron moves are known, the amount of heat can be calculated. In this study, it became possible to measure heat by calculating the potential of the capacitor from the amount of charge and combining it with the potential of the lead, which is known as the externally applied voltage (Figure 4 (d, e)).

② Measurement of Entropy
Entropy is a physical quantity calculated from the probability distribution of information. The charge amount of the capacitor determining the information of the DRAM cell...