Although using ultrahigh-intensity x-ray free-electron laser light to study the structure of matter was inconceivable until recently, today it is transforming how we visualize the atomic world. With far shorter wavelengths and higher intensities than other lasers, XFEL light enables researchers to observe and manipulate objects on an unrivaled scale, benefiting fields from medicine and drug discovery to nanotechnology. The advent of lasers in 1960 fundamentally changed optical sciences and technologies, thanks to the unprecedented high intensity, high degree of coherence and narrow pulse width of laser light. Since that time, tremendous efforts have been made to attain shorter-wavelength lasers in the hard x-ray region, with the expectation that similar fundamental changes that are emerging in the infrared, visible and ultraviolet spectral regions will occur in x-ray sciences and technologies. One of the biggest advantages of the shorter wavelength lies in the fact that the ultimate spatial resolution attainable in observation with light is determined by the wavelength. With shorter-wavelength x-rays, we can resolve subnanometer structures such as atoms and molecules. Although x-ray lasers cannot be obtained with the same technologies as longer-wavelength lasers, it is becoming evident that accelerator-based free-electron lasers in the self-amplified spontaneous emission (SASE) scheme can generate coherent electromagnetic radiation in the x-ray region. The SASE x-ray free-electron laser consists of an electron linear accelerator (linac) and a long undulator. A high-energy, high-density, low-emittance electron bunch is alternatively deflected in a periodic magnetic field of the undulator and emits quasi-monochromatic x-rays. The energy of these is determined by the electron energy, the magnetic field strength and the magnetic period. The interaction between the electromagnetic field of the emitted x-rays with the electron bunch during travel through the long undulator eventually aligns the electrons with the period of the x-ray’s wavelength. The SASE principle is that the aligned electrons move coherently in the magnetic field of the undulator to emit coherent x-rays. SACLA is pronounced “sakura” in Japanese; sakura is the word for “cherry blossom.” SASE-XFEL facility construction projects were discussed in the US and Europe around 2000 and later materialized as LCLS and Euro-XFEL projects. Both proposals were based on a high-energy linac combined with a long undulator, in which the magnet arrays are placed outside the vacuum chamber through which high-energy electron beams pass. The chamber limits the minimum gap between magnetic poles of the undulator, leading to a longer spatial period for the electrons running in the vacuum chamber to feel the strength of the magnetic field. Hard x-ray generation with such undulators requires a higher-energy electron beam because the wavelength of the undulator radiation is proportional to the spatial period length of the undulator magnet and inversely proportional to the square of the electron energy. Consequently, the total lengths of the facilities are as long as 2.2 km at LCLS and 3.4 km at Euro-XFEL. While SASE-XFEL facilities were being discussed in the US and Europe, we were commissioning an 8-GeV electron-storage ring for synchrotron radiation, SPring-8, in Harima, Japan – one of the three large-scale synchrotron radiation sources at the time, along with the European Synchrotron Radiation Facility in Grenoble, France, and the Argonne National Laboratory’s Advanced Photon Source in Illinois. We constructed a 1-km beamline to obtain high spatial coherence in the hard x-ray region, and a 27-m-long undulator to obtain high brilliance from the x-ray beam. Technologies developed for these constructions are directly related to those necessary for XFEL construction. In addition, we have been developing an in-vacuum undulator technology in which undulator magnet arrays are placed inside the ultrahigh-vacuum chamber for the electron beam path. Because the in-vacuum technology removes the constraint on the minimum gap between magnetic poles, we reduced the spatial period length of the undulator magnets to generate sufficient field along the electron beam path. The Japanese XFEL project arose from a question: What happens when we use an in-vacuum undulator instead of the conventional out-of-vacuum type for SASE-XFEL? The short-period, in-vacuum undulator would reduce the necessary electron beam energy for hard x-ray generation, considerably reducing the length of the linac. In addition, adopting higher-frequency accelerator tubes (5712 MHz) for the C-band instead of the conventional 2856 MHz S-band tubes increases the acceleration gradient more than 35 MV/m, resulting in further reduction of the linac length. However, because lower-energy operation of XFEL required a higher-quality electron beam, we did not use a laser-radio-frequency electron gun as was used in LCLS and Euro-XFEL but rather a combination of a classical thermionic electron gun with a single CeB6 crystal as a cathode and a velocity-bunching system consisting of radio-frequency cavities with different frequencies. Thus, the concept of the SPring-8 Compact SASE Source was developed. Based on this concept, we designed an SASE XFEL with 6-GeV linac to emit 0.1-nm-wavelength coherent electromagnetic radiation. However, because many users wanted to use a shorter wavelength and because the linac may be used as an excellent quality injector for SPring-8, the energy of linac was changed to 8 GeV. Even with this enhancement, the total length of the facility is 700 m, less than one-third the length of LCLS or Euro-XFEL. The linear building on the left houses SACLA, the brand-new XFEL facility, while the circular building on the right houses SPring-8, which has a diameter of 500 m. Electron beam commissioning at SACLA began in March 2011. Courtesy of RIKEN Harima Institute. The construction project of the XFEL was launched in fiscal 2006 as one of five Key Technologies of National Importance in Japan to be completed in fiscal 2010. At the end of that period, all of the hardware – including an 8-GeV linac, two undulator lines, photon beamlines and optics, and end stations with four tandem radiation shielding hutches – was completed. The facility was named SACLA, or SPring-8 Angstrom Compact Free-Electron Laser. Unlike with LCLS or Euro-XFEL, SPring-8 and SACLA are collocated. A new building was set up for XFEL and SPring-8 x-ray beams to intersect at the sample position in the building. The relaxation process of the material can be observed with SPring-8 x-rays after an instantaneous impact with the XFEL beam. An electron beam transport from the XFEL linac to the SPring-8 storage ring also was built to use the XFEL linac as the injector of SPring-8. During the electron beam commissioning, we had an historic earthquake and tsunami in the northeast region of Japan, followed by severe damage to the nuclear power plants in Fukushima. Although we felt some minor effects of the quake, such as belated component deliveries from companies in the northeast region, no direct effects were observed, primarily because we are more than 1000 km away from the quake’s center. Meet the author Dr. Tetsuya Ishikawa is the director at RIKEN Harima Institute in Harima, Japan; e-mail: ishikawa@spring8.or.jp. Editor’s Note: The whole world paused in early March when an earthquake and resulting tsunami caused widespread damage and nuclear crisis in Japan. Some research institutions in the country were affected, and some companies experienced slowdowns or production stoppages. But some good news has come out of Japan this spring as well: The world’s second x-ray free-electron laser (XFEL) has entered the stage in Harima, one year behind the first one at Linac Coherent Light Source (LCLS) at SLAC National Accelerator Laboratory in Menlo Park, Calif. The construction was launched in 2006 as part of the Key Technologies of National Importance by the Japanese government, and the launch is a promise of Japan’s bright technologicalfuture, even in the face of large-scale natural disasters.