Fujitsu's development of special application chips for AI and HPC applications will use TSMC's 7-nanometer manufacturing process. Artificial Intelligence (AI) and High-Performance Computing (HPC) are new technologies for this year's technology industry. Japan's Fujitsu also targets AI and HPC applications. Developed its own application-specific chip (ASIC), including dedicated DLU deep learning chips specifically designed for AI deep learning, and ARM architecture HPC chips designed for the next-generation Post-K supercomputer, both of which are Will be handed over to foundry leading Taiwan Semiconductor Manufacturing (2330) foundry. Deep learning, one of the AI ​​learning technologies, is an important technique for advancing AI judgment. However, the biggest problem currently encountered in deep learning is to continuously perform calculations on huge amounts of data. In order to shorten the calculation time, Fujitsu has launched the Zinrai Deep Learning System, the world's fastest deep learning server and cloud computing service. It also designs DLU deep learning chips for deep learning and is expected to start shipping this year. Compared to the industry, which uses GPU parallel computing technology for AI deep learning, Fujitsu has developed a DLU multi-core deep learning chip specifically designed for deep learning. At the same time, it uses the principle of parallel computing, but it can be effective. Reduce computational power consumption, the goal is to compare the competitor's deep learning chip, have 10 times performance per Watt performance (Performance per Watt) performance. In addition, Fujitsu will also apply the DLU chip to the super computer "King". Combined with Fujitsu's self-designed Tofu interconnect technology, Fujitsu will be able to build ultra-large-scale neural networks. In the layout of the HPC operation, the Beijing Supercomputer developed by Fujitsu and the Japan Institute of Physical Chemistry has won the world's most powerful supercomputer for two consecutive years in 2011. Fujitsu has developed a commercially available PRIMEHPC series of servers using the Beijing Supercomputer as its architecture, which has received the attention of scientific research organizations. The first customer outside of Japan is the Taiwan Meteorological Bureau. Fujitsu's Beijing Supercomputer currently uses a SPARC-based processor, which was commissioned by TSMC in the past few years. However, the new generation of the Post Beijing Supercomputer will be designed using the ARM architecture for the first time. Fujitsu adopts the ARMv8-A instruction set architecture. The biggest feature is that it can have lower power consumption than other server processors, allowing supercomputers to reduce power consumption more effectively when performing large data operations. It is understood that the first ARM-based processor will use TSMC 7-nanometer production volume.
Automotive Wire Harness
Our wire harnesses are widely used for various parts of automobiles and truck , Like engine valve cover gasket/diesel gasket(DORMAN-FORD), headlamp(GROTE), Intelligent Anti-collision System(Mobileye), audio, rearview mirror, power seat, etc. Also the complete engine cable assembies for electric vehicle.
Related Products:ul wiring assembly,engine gasket cable,cable harness.
Automotive Wire Harness,Auto cable assemblies,OEM Automotive Wire Harness
ETOP WIREHARNESS LIMITED , https://www.wireharness-assembling.com