Since the advent of large language models (LLMs), artificial intelligence (AI) has been utilized in various scenarios, making AI a familiar presence.
The infrastructure technology supporting this AI utilization is a hardware system capable of advanced and large-scale parallel computation, known as AI accelerators. Currently, many AI accelerators are installed in data centers, but the server/storage/network systems traditionally used in data centers are not designed specifically for the operation of AI accelerators.
Therefore, to install AI accelerators in data centers and provide services that meet the demand for AI, many technical challenges need to be addressed. Among these, cooling of the equipment is considered the most critical issue, and the development of next-generation cooling methods for AI accelerator systems is underway. Among these, immersion cooling systems are said to be the most promising technology due to their significant cooling effects.
This article introduces our company's efforts regarding passive component products for AI accelerator systems using this immersion cooling method.

The technology that delivers the metaverse will build on the technology for virtual reality, using more sensors, and more different kinds of sensors. TDK has one of the richest sensor portfolios in the industry, extensive software expertise, and excels at sensor fusion.

The NTC thermistor is a thermally sensitive resistor whose resistance decreases rapidly as the temperature rises. This property can be utilized in various applications such as temperature sensors and thermal protection devices to protect circuits from overheating.
By mounting the NTC thermistor in close proximity to the heat source, it can accurately sense the temperature of the heat source. However, in some cases, such as when there are constraints on the size of the board or the pattern layout, it may need to be mounted in a location away from the heat source.
In this article, considering such conditions, we used the LEDs on the LED flash circuit board as the heat source and simulated heat generation to check the temperature difference between the LEDs and the NTC thermistors caused by the different mounting positions. We also checked the effect of circuit board thickness.

Service robots are widely used in various situations such as logistics / transport, security, cleaning, housework, hobbies / entertainment and care / assistance. Since service robots are used in a wide variety of situations, it is necessary to fulfil many requirements such as smooth communication with humans (HMI), ensuring safety and responding to changes in the real time. In service robots, the role of a sensor is very important. In this article, we will take a vacuum cleaner robots as an example.

Find the optimal chip NTC thermistors for you by commercial grade and automotive grade based on the characteristics of the product (for use with electrically conductive adhesives, copper plated product for embedding, etc.), application, and external dimensions.