GPUs And Their Usage In Computing

Find and summarise ONE resource that discusses the impacts of legacy systems on organisations.  The resource and your summary should clearly identify problems with maintaining legacy systems.

What is a GPU?

GPU uses and how it differs from CPU.

Medium. (2018). GPU vs CPU Computing: What to choose? – altumea – Medium. [online] Available at: https://medium.com/altumea/gpu-vs-cpu-computing-what-to-choose-a9788a2370c4 [Accessed 27 Sep. 2018]. 

A GPU (Graphic’s processing unit) is a type of microprocessor that is used in the graphics display and perform some computational tasks. It has a number of cores compared to the CPU. GPU main purpose is video rendering. Video rendering involves mathematical operations which are best performed by a GPU. A GPU has many processing cores. The core tend to operate simultaneously. Each core is purpose to perform the video rendering by performing the arithmetic operations. Parallelism enable the GPU to perform the rendering of 3D graphics. The GPU is used for security purposes in encryption and decryption. The encryption need more processing power. The CPU can also perform encryption but GPU performs the encryption in a better way. A GPU has a number of difference from the CPU. The CPU carries out program instructions from various applications whereas the GPU is specifically for rendering 2D and 3D images that are displayed on a computer. Secondly the CPU performs multiple task but the GPU specializes of image rendering. Finally the CPU is considered the brain of a computer but the GPU complements the CPU for performing complex tasks. The CPU are more flexible that GPUs (Medium, 2018). 

Artificial Intelligence and their capabilities in operating systems.

Globalme. (2018). A Complete Speech Recognition Technology Overview. [online] Available at: https://www.globalme.net/blog/the-present-future-of-speech-recognition [Accessed 27 Sep. 2018]. 

There are a number of intricacies of AI in operating systems such as multitasking, recognize speech, people recognition and detection of antivirus, telling jokes, having reminders and read and write information. The greatest AI capabilities in the operating systems is speech recognition (Globalme, 2018). The operating systems are have some programs installed such as Siri in mac, Cortana in windows, Alexa and Google Assistant. This programs help in the AI to enhance the interaction between the computers and the users. Today, the AI programs are set to recognize the owner’s voice so that if commanded to perform a task they use the voice recognition to perform the tasks requested by a user. One can request for a joke from a computer and the computer will share a joke which the user may find funny.  The speech recognition helps in a number of operations like calling, texting or emailing your contact, getting directions, playing music, setting alarms, searching for nearby restaurants, adjusting the setting, creating a calendar entry, launching a program or an application and many more activities. Voice recognition is simplify the interaction of users in the operating system. It is convenient for driver while on the roads too. There is a lot of competition among the existing voice recognition apps. More advancement are launched periodically. 

Kennedy, C. (2018). Key differences between AWS and Microsoft Azure - Managed Data Center News. [online] Managed Data Center News. Available at: https://resource.onlinetech.com/key-differences-between-aws-and-microsoft-azure/ [Accessed 27 Sep. 2018].  

Netflix uses AWS for some features such as keeping truck of the Netflix users, keeping track of the content they watch, keeping track what the Netflix users click and finally tracking the massive catalog content that is stored in the cloud. Netflix uses AWS for the backend infrastructure, storing, streaming the web series.

How do GPUs differ from CPUs?

Microsoft azure offers infrastructure as service that is provided instantly over the internet. The payment is done only when one uses the service. The resources are offered as separate service components. One can rent a particular resource for as long as the person needs it. The infrastructure is managed by cloud computing service provider, while a person or company buys, installs and manage the software installed. There are a number of differences between AWS and Microsoft Azure. Based on the compute power, AWS users configure their own VMs or opt to choose pre-configured machines images. The AWS users choose the size, power, number of virtual machines. For Azure, the users choose a Virtual Hard Disk (VHD) to create a VM. VHD is preconfigured by Microsoft or the Azure users. Based on storage, AWS has temporary storage that is allocated when an instance is started and destroyed once the instance is terminated while, Azure offer a temporary storage through D drive (Kennedy,2018) 

IPV6 and IPV4

System, O., Chen, A. and Chen, A. (2018). Difference Between IPv4 and IPv6 (With Comparison Chart) - TechDifferences. [online] Tech Differences. Available at: https://techdifferences.com/difference-between-ipv4-and-ipv6.html [Accessed 27 Sep. 2018]. 

IPv6 is also referred to as IPng. It uses 128-bit addresses which increase the number of possible addresses compared to the IPv4. The addresses are more complex as they consist of hexadecimal numbers. The complexity of IPv6 leads to a big range of IP address to uses and enhance security as well. The long extension for the IPv6 is intended for the growth of internet and network addresses.  IPv6  has some features which include: requires IPSec support, the addresses are 128 bits, they don’t require manual configuration or DHCP, uses PTR, uses multicast neighbor solicitation, uses multicast listener discovery to manage the membership in a local subnet and uses ICMPv6 router solicitation.

There are a number of differences between IPv6 and IPv4. Based on address configuration, IPv4 supports manual and DHCP configuration, while, IPv6 supports auto-configuration. End-to-end connection integrity is achievable on IPv6 unlike on the IPv4. The security feature are inbuilt in IPv6 but security on IPv4 is dependent on the application. IPv4 has 32 bits while IPv6 has 128 bits. In IPv4 security is not provided but in IPv6 the security is provided. The address representation in IP v6 is in hexadecimal while in IPv4 is represented in decimal (System, Chen and Chen, 2018).  

Popli, N. and Girdhar, A. (2017). WannaCry Malware Analysis. MERI-Journal of Management & IT, 10(2).

Trautman, L. and Ormerod, P. (2018). Wannacry, Ransomware, and the Emerging Threat to Corporations. SSRN Electronic Journal. 

WannaCry ransomware consists of a number of components. The components include the application that encrypts and decrypts data, the file containing the encryption keys and a copy of Tor. The malware arrives on the infected computer in form of a dropper. When the malware is launched it tries to access the hardcore URL. If does not manage to get to the hard core URL, then it accesses the encrypt files in various file formats such as MKV or MP3. The malware then makes the files inaccessible as it displays a ransom demanding notice to decrypt the files (Popli and Girdhar, 2017).

Use cases for GPUs

WannaCry exploits the Windows operating system. The malware lies in the implementation of Server Message Block (SMB). The SMB has nodes that could be tricked by crafted packets in executing the arbitrary code. WannaCry used EternalBlue to infect computers and began spreading rapidly. The malware tries to access the URL. If it succeeds it shuts itself down. Once the files are encrypted by the encrypting application, there is little remedy to restore the file (Trautman and Ormerod, 2018). Incase people fail to back up their files they end up experiencing data loss. Data loss is a big threat to all companies as well as individuals. 

SDxCentral. (2018). What is OpenFlow? Definition and How it Relates to SDN. [online] Available at: https://www.sdxcentral.com/sdn/definitions/what-is-openflow/ [Accessed 27 Sep. 2018]. 

Tourrilhes, J., Sharma, P., Banerjee, S. and Pettit, J. (2014). SDN and OpenFlow Evolution: A Standards Perspective. Computer, 47(11), pp.22-29. 

OpenFlow is a protocol that allows a network to tell the switches where to send the packets. Each switch has a software that directs it on what to do. In OpenFlow the decisions are centralized to ensure that the network are programmed independent of the data center and the switches. Usually an OpenFlow switch separates the control path and the data path. The data path is found on the switch and a separate controller will make high-level routing decisions (SDxCentral, 2018).  The switch and the controller communicate via OpenFlow protocol.  The methodology of this communication is called SDN. The SDN attends to allow a more effective and efficient use of the network resources compared to the traditional networks. OpenFlow networks makes it possible to place the VMs anywhere in a data center to enhance stranded computing capacity.  The OpenFlow is a standard of the SDN. The SDN controllers are tasked with federating between SDN Controller domains, using application interfaces such as OpenFlow (Tourrilhes et al., 2014). Through the OpenFlow Controllers, the SDN pushes the changes to the switch flow table in order to allow network administrations to separate the traffic and control flow for an efficiency performance. OpenFlow allows programmability, centralized intelligence and abstraction.  

Trinisys.com. (2018). Understanding the Organizational Costs of Legacy Systems | Trinisys.com. [online] Available at: https://www.trinisys.com/2018/07/understanding-the-organizational-costs-of-legacy-systems/ [Accessed 27 Sep. 2018]. 

A legacy system is a system that is still in use but it is obsolete. The data in the system cannot be changed to the latest format and the program cannot be upgraded.

The legacy systems are slow and therefore make the operation process in an organization to be slow. They are prone to failure as there are installed in obsolete hardware. The maintenance for the legacy systems is expensive as it is difficult for the IT personnel to detect any problems in the systems. The IT personnel may also waste a lot of time in trying to understand the legacy systems because they lack documentation and few people can tell about all the functionalities of these systems. The legacy systems also lack better interface to allow them to integrate with other applications. This I suite expensive as the organization may end up getting a totally new software that will help in the operation and integrate with other applications.  The organization also incurs a cost of training the users of the system on how to use it as it does not have functionality documentation.  A legacy system are open to security breaches and attacks. In case company behind the systems does not release security breaches nor updates to protect the system from threats (Trinisys.com, 2018). 

How GPUs complement CPUs

What Cambridge Analytica did controversial

Meyer, R. (2018). The Cambridge Analytica Scandal. [online] The Atlantic. Available at: https://www.theatlantic.com/technology/archive/2018/03/the-cambridge-analytica-scandal-in-three-paragraphs/556046/ [Accessed 27 Sep. 2018]. 

Cambridge Analytica is a major data company. Some leaders such as President Trump misused the company for personal gain. He gain access to private information of more than 50 million users. The tools used could identify the personality of American voters and he could influence the behavior of those voters. Cambridge has been offering to entrap politicians. The data that was accessed includes users ‘identities, friends and the likes. The information was to be used to target audiences with digital ads.  Apart from accessing Facebook users, the data was matched with other records to construct personality profiles on many American voters. Cambridge Analytica classified voters with personalities of openness, conscientiousness, extraversion, agreeableness and neuroticism. The main aim of getting around 50 million users’ data was to identify the personalities and influence the behavior using psychographic modelling techniques. The psychographic tools was meant to target online ad and but the Brexit campaign. The psychographic tools built with the Facebook data had a lot of hidden information regarding the entire process and the goal. Cambridge acted as a vote profiling company. Cambridge Analytica was misused during the campaigns since it was technology re-shaped in the United Kingdom and United States of America (Meyer, 2018).  

Performance testing tool.

Anon, (2018). [online] Available at: https://www.softwaretestingclass.com/performance-testing-using-webload/ [Accessed 27 Sep. 2018].

Radview.com. (2018). Load Testing & Website Performance Testing Tools - RadView. [online] Available at: https://www.radview.com/ [Accessed 27 Sep. 2018]. 

WebLOAD is the best load testing tool. WebLOAD provides a comprehensive and robust option in largescale performance testing. It includes extensive scripting capabilities, detailed analytics and is built-in integration with APM solutions and open source platforms.  The integrated development environment tends to provide a proxy-based recorder to load all the web activities. Recording enable one to determine the web page resource. WebLOAD provides other ways to edit and enhance the scripts such as HTTP configuration, parameterization and building blocks that can be dropped in the script. WebLOAD has an automatic correlation of Dynamic Values. The WebLOAD correlation replaces the values that are unique in each test (Anon, 2018). There are quite a number of dynamic values within the URLs, cookies and headers. WebLOAD had a number of protocols such as HTTP, HTTPS. Rest API, AJAX, FTP, SMTP, TCP and many more. Unique testing is used to determine the flexibility of the Java objects.  WebLOAD has Selenium scripts that are used to measure the real user experience.  WebLOAD has a smooth integration that is used with Amazon EC2 where it can automatically start and stop hosts. Finally, this tool generates reports and depict the areas with violations. The results shows the performance test results (Radview.com, 2018).  

Australian IT business that failed.

Coyne, A. and Coyne, A. (2018). Victoria looks for new operator to run Myki. [online] iTnews. Available at: https://www.itnews.com.au/news/victoria-looks-for-new-operator-to-run-myki-401131 [Accessed 27 Sep. 2018]. 

Victorian Myki Smart Card was the most costly Smart Card system for use in the public system. By the time the app got to the production level it did cost an approximate of 1.4 billion dollars. The system was introduced to replace Metcard system. In the year 2015, a repot was done by Victorian auditor saying the system was not successful due to poor initial planning of the original scope. There are a number of reasons that lead to the failure of the costly public system.  Poor time planning is one of the factor. The time that was take to develop and implement the system was nine years as opposed in the expected and planned time span of two years.  The system was also out of budget as it was assumed that t would use 944 million dollars but 1.4 billion dollars was spent in the project implementation.  Report shows that between the year 2010 and 2014 there were too many complaints about Myki. The government renegotiated the contract in 2010 and the NTT Data bought out Kamco FOR 1.2 billion dollars. NTT Data deal expired and Myki started to seek for an operator to take over the reins. The new operator was expected to deliver maximum value for money for the Victorian Taxpayers (Coyne and Coyne, 2018).