BLOG

Article contributed by Dr. Suresh Chandra Wariyal, Assistant Professor, Amrapali Institute of Technology and Sciences

The concept of Deep learning is unbeaten at dealing with different type of environment and tasks. This resource plays a vital role to accelerate the performance of deep learning. The utilization of hardware is the key aspect when number of application wants easy use of the resource. On the other way we can say that there are varieties of applications and these varieties of application may use different deep learning frameworks to utilize different amounts of resources.

Image recognition and speech recognition are the major aspect of deep learning and now the deep learning approach can be used for general purpose machine learning. The performance of machine learning is accelerated due to deep learning approach. However, it is not easy and not so cheap to deal with complex model and massive training data that occurred due to deep learning.

A number of online services nowadays rely upon machine learning to extract valuable information from data collected from various real time sources. There are different working areas of deep learning where it is implemented such as:-

• The cultural evolution proposes a theory and experimental tests, relating difficulty of learning in deep architectures to culture and language.

• Towards the Poisoning of Deep Learning Algorithms with Back-gradient Optimization.

• Deep Learning Algorithm for Brain Tumor Detection and Analysis Using MR Brain Images which aims is to create deep learning algorithm to detect brain tumor using magnetic resonance brain images and analysis the performance of algorithm based on different values, accuracy, sensitivity, specificity.

• Effective Media Traffic Classification Using Deep Learning in which Traffic classification (TC) is very important as it can provide useful information which can be used in the flexible management of the network.

• SINGA: A Distributed Deep Learning Platform a distributed deep learning system, called SINGA, for training big models over large datasets.

• The Business Impact of Deep Learning. This has broad implications for all organizations that rely on data analysis. It represents the latest development in a general trend towards more automated algorithms, and away from domain specific knowledge.

• Medical image processing play great role in helping the radiologists and facility to patient’s diagnosis.
• Deep neural networks (DNNs) are brain-inspired machine learning methods designed to recognize key patterns from raw data. State-of-the-art DNNs, even the simplest ones, require a huge amount of memory to store and retrieve data during computation.

The implication of this blog will explore the opportunity for various researcher to do their research on any area with the inference of deep learning.

Article contributed by Mr. Hem Chandra Joshi, Assistant Professor, Amrapali Institute of Technology and Sciences

Since in Current networking Situation Data rates are increasing, Protocols are becoming more dynamic and complicated and Protocols are being introduced more rapidly so traditional GP (General-purpose Processor), ASIC (Application Specific Integrated Circuit) are not optimized for networking applications

A network processor is an integrated circuit which includes a feature set specifically targeted at the networking application domain .Network processors are typically software programmable devices and would have generic characteristics  same as general purpose central process units that are usually employed in various types of equipment and products. Network processors have evolved into ICs with specific functions. This evolution has resulted in more flexible and more complex ICs being created. The newer circuits allow a single hardware IC design to undertake  variety of various functions where the appropriate software is installed since they are programmable A Network processor  executes programs to handle packets in a data network, network processors are used in the manufacture of many different types of network equipment like router, software routers and switches, Firewalls, Session border controllers, Intrusion detection devices, Intrusion prevention devices and Network monitoring systems

Some latest network processor

Marvell’s Xelerated

Marvell’s Xelerated family of network processors is AN integral piece of the next-generation network. Designed for carrier LAN platforms for unified fiber access, mobile backhaul, railway LAN, packet-optical transport, as well as  switching systems in cloud computing environments, the Xelerated product modify wealthy services at competitive worth points.

Cisco nPower™ X1

The Internet of Everything would require extremely advanced silicon innovation and it begins today with the launch of the world’s most scalable and programmable network processor, the Cisco nPower™ X1. With quite 4 billion transistors, this highly integrated 400 Gbps throughput single-chip will enable Terabit class solutions

Broadcom hits 100Gbps network processor

Broadcom fourth generation LAN network processor, the BCM88030, achieves full-duplex 100 Gbps performance courtesy of a massively parallel design supported sixty four packet process cores, every running at 1GHz. the company says four of the elements might be enclosed in a line card to attain 400 Gbps performance

Where are network processors used?

A network processor is used in a network traffic manager, which is placed between a network interface and a switch fabric in a switcher/router. The traffic manager decides where, when, and how incoming and outgoing data will be sent next. It strips, adds, and modifies packet headers. It also makes routing and schedule decisions. The traffic manager has interfaces to the network and to the switch fabric. In Figure these are labeled PHY (physical interface) and CSIX (common switch interface) respectively

Early traffic managers were build around a general purpose processor (GPP). The GPP was support by a direct memory access controller (DMAC) and simple I/O devices. Traffic was transfer in protocol data units (PDUs) between memory and the switch fabric or network interface.

This new architecture modified as network speed outpaced processor and bus speed. The switch fabric interface and network interface were integrated into a single application-specific integrated circuit (ASIC) to allow PDUs to be transferred without passing over the system bus

This new design meant that management of individual PDUs was delegated to the ASIC. . The ASIC ran hard-wired network protocols. It passed the majority of traffic through, transferring to the GPP only those PDUs involved in control or signaling, or those that required unusual processing. Network processors are designed to switch the fixed-function ASIC, adding software programmability to wire speed processing

Conclusion and future scope

Network  processor have an important role in networking as we know Configure a router for existing or new network is very tedious job so to avoid manual process again and again we can use programmed  network processor in router so it Can automatically configured

Now a days we are trying to use IPV6 in place of IPV4 so the number of user increasing tremendously so maintain and update configure of router is not so easy if we Use a programmed network processor in router it becomes much easy

References

1– Jianhua Huang “Network processor design” ASIC 5^th international conference (volume 1) on 21-24 Oct 2003.

2– M. Coss and R. Sharp, “The network processor decision” in Bell Labs Technical Journal, vol. 9, no. 1, pp. 177-189, 2004, doi: 10.1002/bltj.20012.

3-Sanchita Vishwa and Rajeshwari M. Hegde “Optimization of network processor microarchitecture” in International Journal of Electrical, Electronics and Data Communication, ISSN(p): 2320-2084, ISSN(e): 2321-2950 Volume-7, Issue-4, Apr.-2019