Google's self-driving cars involved in 11 crashes Google comes clean on the number of accidents its driverless cars have been in over the past six years. If you're wondering why you've never heard of NOR flash, here's one reason: NAND flash is a $40B industry; NOR flash is a $2B niche. It also costs significantly more than NAND, so if designers can replace NOR with NAND, they do. But there is a steady and growing market for NOR. NOR's big advantage is that it is a lot more rugged than NAND.
Every digital computing system requires non-volatile memory to retain the primary or necessary programs and data, permanently, in order to avoid data loss on occasions of power failure or system shutdown. This has led to the invention of non-volatile memories. NOR flash is a type of nonvolatile semiconductor memory widely used in electronic devices. The NOR flash memory was first introduced by Intel in 1988. In NOR flash memory, the individual memory cells are connected in parallel configuration, which enables fast random access to the content stored on the memory cells.
TORONTO – Jedec's new xSPI standard for non-volatile memory (NVM) devices is aimed meeting the demands of instant-on applications while also maintaining performance standards for NOR flash sitting outside of the SoC. Adesto Technologies, known for its small, ultra-low power NVM products, is claiming to be first out the gate, having been working on products and collaborating on the standard for the past three years. In a telephone interview with EE Times, Adesto Chief Techology Officer Gideon Intrater said its eXecute-in-Place (XiP) EcoXiP product family takes advantage of the new Jedec standards to give customers such as system developers and controller designers assured compatibility with controllers and peripheral devices that should accelerate adoption. Developed by a task force comprised of representatives from most NOR flash device manufacturers and several PC and microcontroller companies, the xSPI standard establishes mechanical, electrical and transactional guidelines for developing high-throughput octal devices. Although Jedec has been a defining organization for NAND flash, said Intrater, until recently NOR flash has been somewhat of a "wild west" in terms of standards, which has led to divergent products for enabling communications between host controllers and memories and confusion for controller designers.
SANTA CLARA, Calif. and DENVER, Colo., Nov. 19, 2019 -- Marvell today announced the availability of NVIDIA GPU support on its ThunderX family of Arm-based server processors. Following NVIDIA's June announcement to bring CUDA to the Arm architecture, Marvell has collaborated with NVIDIA to port its CUDA-X AI and HPC libraries, GPU-accelerated AI frameworks and software development tools to the ThunderX platform. The computational performance and memory bandwidth of ThunderX2, Marvell's latest 64-bit Armv8-A based server processor, combined with the parallel processing capabilities of NVIDIA GPUs provide a compelling path to energy-efficient exascale computing. Artificial intelligence (AI) and machine learning (ML) continue to become essential technology components to data center server requirements at the cloud and network edge. To address these evolving AI and ML workloads, as well as the most challenging and complex problems in science and research, supercomputers need processors that are optimized to provide cutting-edge throughput, application latency and power.
The past year has been inundated by a wave of consolidation among chipmakers. Throughout 2019, billions of dollars changed hands as industry giants attempted to fill gaps in their product lines and bolster their positions in some of the world's most competitive markets including data center connectivity, artificial intelligence (AI), and 5G networking. Here's a rundown of 2019's five hungriest hippos and the silicon morsels they gobbled up or tossed aside. Intel closed out 2019 by betting big on AI with the purchase of AI startup Habana Labs for $2 billion. The purchase is aimed at cementing Intel's position in the highly competitive data center processor space, which the company predicts will be worth more than $25 billion by 2024.