Induction accelerator is an old but novel accelerator. Its concept had been discussed in the doctoral dissertation of Wideroe in 1927. From a historical point of view, it should have been natural that the evolution of accelerator starts from an electrostatic accelerator governed by Gauss Law, second induction accelerators governed by Faraday Induction Law follow, and last RF accelerators dominated by a full set of Maxwell Equation join the race. However, the actual history is not the case. After WWII, RF and microwave accelerators drastically evolved. The first induction accelerator was realized as a betatron in 1941.

In 1950's, an induction linac had been proposed and independently developed by Vekslar in former USSR and by Christofilos in US. The induction linac was actually employed as a driver of high intensity electron beam acceleration. In 1986, LLNL demonstrated giga watt microwave/mm wave FELs driven by ETA (4.5 MeV, 3 kA induction linac). However, performance of these induction linacs in rep-rate was limited to less than 100 Hz due to limited capability of the power supply driving the induction cell (1 to 1 pulse transformer).

In 2000, the concept of induction synchrotron has been proposed at KEK, where a conventional RF cavity is replaced by an induction cell, claiming notable benefits such as no band-width limitation and a large freedom of beam handling. Just after this proposal, R&D works on a switching power supply to drive the induction cell, which had to be operated with a few kV at 1 MHz, was initiated there. Fortunately, Si-MOSFETs of 700 V and 20 A were becoming available as a switching device at that time. In addition, Finemet was just on the market as a low loss magnetic core material at the high rep-rate operation. The concept was fully demonstrated using the KEK 12 GeV proton synchrotron in 2006. Since then, basic studies on a fast cycling induction synchrotron or induction microtron have been conducted there, demonstrating various beam handling techniques that are never realized in a conventional RF synchrotron.

As one of notable applications based on these novel circular induction accelerators, the next hadron therapy system, which allows 3D continuous tracking irradiation on a moving target, is under development in a world-wide collaboration. In addition, an induction microtron system for high energy giant cluster ions will provide a unique opportunity to realize an unknown stage in heavy ion mutation technology and to create an extreme non-equilibrium state in materials associated with penetration of a high energy giant cluster ion. At last, it is noted that a giant cluster ion driven inertial fusion system is in our scope. The third scheme that is original distinguished from the exiting induction linac scheme in US and RF accelerator scheme in EU is under design by a recently established forum integrating the quantum beam fusion society in Japan. The driver consists of the giant cluster ion sources, the induction microtrons as injector, the permanent magnet stacking rings, the two-way multiplex induction synchrotron as a main driver taking a role of bunch compression after acceleration, and the drift compressor, and final focusing systems.

We have fabricated Si tetramer (Si₄) atom switches on Si (111)-7×7 surfaces using atom manipulation at room temperature. Then, we have demonstrated on Si₄ that 1) tunneling current-induced downward atom switching by scanning tunneling microscope, 2) attractive atomic force-induced upward atom switching by atomic force microscope, and 3) tuning of atom switching by using both current and force at same time.

The geometric phase, generally discarded as a global phase, allows universal holonomic gating of an ideal logical qubit, which we call a geometric spin qubit, defined in the degenerate subspace of the triplet spin qutrit. We here experimentally demonstrate non-adiabatic and non-abelian holonomic quantum gates over the geometric spin qubit on an electron or nitrogen nucleus. We manipulate purely the geometric phase with a polarized microwave in a nitrogen-vacancy center in diamond under a zero-magnetic field at room temperature. We also demonstrate a two-qubit holonomic gate to show universality by manipulating the electron-nucleus entanglement.

The doubly magic features of the very unstable nucleus located far from the stability line, ⁷⁸Ni, were investigated experimentally. To accomplish the measurement of this short-lived isotope, a new target and detector system was developed to measure the de-excitation gamma rays with a better Doppler reconstruction following proton-knockout reactions. While the high energy of the first excited state of ⁷⁸Ni corroborates its closed-shell nature, a second state was found at a similar excitation energy. This observation suggests a possible shape-coexisting nature in this nucleus and questions the nuclear shell robustness for isotopes further away from stability, where the nucleosynthesis is proposed to occur.

Graph partitioning as an inference problem has been an important topic in multiple fields of science. In this article, we derive a performance limit called the algorithmic detectability limit on graph partitioning using a technique developed in statistical physics. This limit is a phase transition point beyond which an algorithm completely loses the ability to identify the group structure that is assumed in a random graph model.