Phase-change random-access memory (PCRAM) is an emerging ultrahigh capacity nonvolatile memory technology. It relies on an electric-pulse-induced phase transition to SET (crystallize) or RESET (amorphize) the phase-change material in the memory cell. There are at least two orders of magnitude difference in electrical resistance between the crystalline and amorphous states. This techonology has been developed using many novel structures and layer designs for high-density/high-speed operation. Up to now, the composition Ge 2 Sb 2 Te 5 (Ge-ST for short) and its forms modified with nitrogen, oxygen, silicon, and antimony, among others, have been extensively adopted in PCRAMs. Indeed, the Ge-ST materials possess a wellrecognized high crystallization speed, high resistance ratio, and thermal stability. However, Ge-ST compositions have a relatively high ($620 8C) melting temperature (T m ), which tends to require a high RESET current (I RESET ), and easily induces thermal cross-talk problems in high-density memory arrays. They also have a quite low ($160 8C) crystallization temperature (T c ), which leads to stability problems with reported data retention for ten years at 85-120 8C. In this paper, we describe the use of Sb-rich GaTeSb materials. Sb-rich thin films, such as a doped SbTe composition, have been elucidated to possess a growth-dominated crystallization behavior instead of a nucleation dominated one. This allows even thin films to crystallize at higher speeds. In earlier studies on phase-change optical media, a group of GaTeSb materials was disclosed. Among them, the compound Ga 2 Te 3 Sb 5 (Ga-TS for short) was found to possess superior thermal properties: T c ¼ 228 8C and T m ¼ 563 8C, and was identified as an extremely fast phase transition media in our earlier studies of rewritable optical disks. As another comparison, the electrical resistivity of the crystalline films of Ga-TS and Ge-ST are 650 and 300 mV cm, respectively. This indicates that Ga-TS is potentially more easily RESET than Ge-ST. In this paper, we demonstrate measured results of test cells for phase-change memory (PCM) using Ga-TS.Ga 2 Te 3 Sb 5 and Ge 2 Sb 2 Te 5 were deposited, respectively, into PCM test cells of the size 200 nm by 200 nm. We present here the results for Ga 2 Te 3 Sb 5 test cells; those for Ge 2 Sb 2 Te 5 are available in the Supporting Information, for comparison. The crosssectional transmission electron microscopy (TEM) image in Figure depicts the test-cell structure, which includes the bottom electrode, Ga-TS (the active material), and the upper electrode. An extraneous interfacial layer and voids exist between the active material and the bottom electrode, because of our imperfect fabrication processing. This leads to a highly resistive current path in the test circuit, which requires an initial punch through before measurement. Figure shows resistance-voltage (R-V ) and resistance-programming current (R-I ) curves, and endurance-test results of the test cells. To evaluate the dynamic resistance and programming current through the phase-change layer, a sensing resistor is utilized. It is a reference necessary to determine the relationship between the resistance and programming current and voltage across the test cell. The programming current of the test cell increased with an increasing applied electric field, as shown in Figure . In Figure the voltage across the PCM test cell fluctuates depending on the ratio of the dynamic resistance and the fixed resistance of the sensing resistor. Therefore, the voltage across the test cell does not always increase with an increment in the applied electric field. Figure COMMUNICATION www.advmat.de