China
Africa
Explore chapters and articles related to this topic
Internet of Things (IoT)
Published in Prashant Ranjan, Ram Shringar Rao, Krishna Kumar, Pankaj Sharma, Wireless Communication, 2023
Prashant Ranjan, Ram Shringar Rao, Krishna Kumar, Pankaj Sharma
The management service delivers the data that can be accessed by analytics, access controls, process modelling and system management. One of the critical aspects of this layer is to communicate and interact with objects and devices to provide information such as incidents or weather, traffic data and current location. Some of these aspects include the filtering, routing, and normalization of periodic data. Any of the material includes responding to urgent emergencies, such as attending to the health conditions of the patient. The rule engines help to formulate the real-time decision logics. The rule engines present a very responsive IoT system by prompt interactive and automated processes. Data analytics tools extract relevant information from the enormous amount of raw data and are processed without a time lag. In-memory analytics reduces the time required for data query, making decision making faster, and allows this enormous amount of data to be cached in RAM without storing it in physical disks. Streaming analytics helps in the analysis of data, which is data-in-motion and is required to be carried out in real-time so that the decisions can be made without any latency.
Intelligent Edge
Published in Haishi Bai, Zen of Cloud, 2019
A scalable storage service has two characteristics. First, it's horizontally scalable. Data storage involves I/O operations, which are much slower than in-memory operations. And the maximum IOPs per disk are constrained by many factors such as disk design, bus bandwidth, and OS limitations. To overcome the bottleneck, data is sliced and sent to multiple nodes for parallel processing. Second, its compute plane is separate from its data plane. As data volume increases, the demand on data plane increases much faster than the stress on compute plane. If compute plane and data plane are bundled together, scaling often leads to over-stressed data layers and under-utilized compute layers. If the two planes are separate, they can be scaled separately so that the data plane can grow as needed independent of the compute plane.
Renewable digital transformation and cybersecurity management
Published in Henry K. H. Wang, Renewable Energy Management in Emerging Economies, 2020
One key digital transformation area is in data process and decision making. Traditionally, many of the traditional data processes used by companies have required staff to move data from one corporate database to another. It could take even longer if these data have to cross many functional silos, which is normally the case in most companies. Then they will need to process the data and generate appropriate reports for management to review and to make appropriate decisions. New advanced in-memory computing digital technologies could significantly accelerate these processes and reduce the processing time greatly.
A novel low-latency DRAM based on the bitline-discharge rate
Published in International Journal of Electronics, 2018
Memory wall represents one of the main challenges faced by computer architects. This is due to the increased performance gap between the fast processor and the slow memory. In this paper, a novel readout scheme that depends on the difference between the bitline-discharging rates in cases of ‘1’ and ‘0’ readings was presented in order to reduce the memory latency. The proposed scheme was analysed quantitatively and compared to the conventional scheme. In order to minimise the power-delay and the energy-delay products, one must use an optimum value for the precharge-voltage level that depends on the probabilities of occurrence of ‘1’ and ‘0.’ This point was investigated quantitatively. The proposed scheme was verified by simulation using the 45 nm CMOS technology and showed 44 and 7.7% reductions in the average read-access and cycle times, respectively, compared to the conventional scheme. It was shown that a power saving can be achieved if the probability of occurrence of ‘0’ storage exceeds 66.7%. This lower limit can be alleviated, however, at the expense of a smaller saving in the access time.