By Contributors: Kevin Cornell CEO and President AmbiCom Holdings, Inc. [OTCQB: ABHI] Bravatek Solutions, Inc. [OTCQB: BVTK] ?Dr. Thomas A. Cellucci, PhD, MBA Chairman & CEO
November 8, 2017
Automated IT Equipment Optimization: Reducing Energy Consumption and Maximizing Capital Investment Introduction As our society continues to rapidly “computerize” every aspect of business, government, and our personal life. The immediate requirements, as well as the collateralized consequences of these requirements are immense: the increased costs of investment in hardware, software, and infrastructure; the associated, ongoing IT management costs; as well as the hidden costs of massive energy demand to power systems and to responsibly dispose of used equipment. Imperative to any discussion of increased computerization adoption is the measurement of the effectiveness of this computing base compared to the energy cost to support it. Also, relevant to this energy issue are capacity and performance. For IT environments already rife with complexity— heterogeneous mixes of computer servers, storage, networks, devices, and applications—continuous efficiency, capacity, and performance optimization is nearly impossible. IT departments are faced with infrastructure and spending inefficiency as a norm, and public and private entities alike, are confronted with energy inefficiency—wasteful use consumption—as a corollary, and a longer-term effect. We now have capabilities available to government to assist them reduce expense and increase output to their computer systems.
Associated Policy Research The United States Congress began to address server use and demands with Public Law 109-431, passed on December 20, 2006. The law mandated the study and promotion of energy-efficient computer server use in the U.S., producing the resulting Report to Congress on Server and Data Center Energy Efficiency. One of the key metrics highlighted in the report is “Active Server Utilization,” which measures the efficiency of the use of the servers involved, rather than just the linear measurement of cost or amount of energy consumed. Just as the automobile industry used the “Miles Per Gallon” measurement in the 1970s to inform the public and enforce public policy toward a meaningful improvement in transportation industry energy efficiency, Active Server Utilization serves as a touchstone to measure and drive the adoption of more efficient computing resources in society. Heightened policy advocacy and greater strategic implementation of more efficient server use will go a long way toward minimizing energy consumption waste and maximizing infrastructure investments. In response to Public Law 109-431, on August 2, 2007, a research team led by Dr. Jonathan Toomey published Report to Congress on Server and Data Center Energy Efficiency. Since the original publication, the report has been updated several times, most recently in June 2016.
Important Trend Detected Given current trends, computers and data centers could account for up to 50% of total electrical usage in the US within just a few decades. Such a large, attributable source of energy demand outlined ahead requires a clear strategy for addressing the challenge and reducing associated consumption. Vital Analysis Component The report’s initial and enduring analysis focuses on drawing a correlation between electrical usage and processing power, to see if steps may be taken to gain efficiencies in power consumption, without negatively impacting server performance, reliability, features, or overall cost. A vital component of the analysis addresses Active Server Utilization—that is, the amount of output achieved by a server versus its actual capacity. Over the duration of report updates (more than 10 years), the Active Server Utilization percentages have remained relatively steady, as evidenced in Table 1. November 8, 2017 ?Source: Report to Congress on Server and Data Center Energy Efficiency, 2016.
Variety of Active Server Utilizations “Hyperscale” server installations—at such sites as Google, Facebook, and eBay—can dedicate specific servers for specific tasks and, therefore, configure server usage accordingly. Such configurations and resulting server utilization are reflected in the current Hyperscale Active Server Utilization average of 45 percent (45%). Administrators of internal corporate servers and service provider servers (i.e., Cloud or SaaS offering servers) do not always have the luxury of managing server utilization on a prioritized or task-specific basis enterprise-wide. Generally, the non-hyperscale class of servers chronically suffer from low utilization. This low utilization causes multiple consequences led by: 1) server sprawl, 2) high capital cost, and 3) continuous hardware replacement. Although a company may only be gaining 10% of server output for a certain task or set of tasks, the company still is tasked with housing 100% of the server, powering all its components, cooling the server rooms, and, most importantly, paying for 100% of the capital and runtime cost of the server.
Key Recommendation Dr. Koomey, the aforementioned report research- team leader stated: “It’s actually a battle for using your capacity that would otherwise be idle. If you put it in those terms, in financial terms, then I think people would get it.” In other words, Dr. Koomey recognized that full and efficient server capacity utilization is a bit of an albatross for the large majority of IT environments. Final Extent of Utilization Worldwide server shipments in 2016 were $52B. With only 15 percent (15%) utilization, $44B in server investment was made with no corresponding return on that investment. Though one would not want server utilization greater than 50% for management purposes, that still leaves $22B in investments with no real benefit. If one is addressing solely the US government portion of this annual server investment, then the investment amount wasted is about $220M per year.
Low Active Server Utilization Causation All servers are shipped with default settings (e.g., memory, disk, and network) as their lowest common denominators. They are shipped with these settings so that they will support the widest variety of applications, peripherals, and interfaces. Though the baseline setting makes sense from an administrative view, it robs server performance in favor of versatility and compatibility. Certainly, one can evaluate and adjust manufacturer settings to improve server performance, capacity, and efficiency. However, IT organizations, big and small, generally are not tasked with or interested in server setting changes unless the out-of-the-box or down-the-road performance is extremely poor. Additionally, even if technicians were fully skilled in all architecture and design aspects, individually tuning a server to maximize its utilization can take up to 2 to 3 days (16 to 24 hours) of dedicated effort. This finite resource, along with the absence of server production accessibility, is needed each time the server needs to be re-tuned. Most IT organizations have their “Check Server” light permanently switched ON. Figure 1: Check Server. ?November 8, 2017
The Solution It’s possible to maintain and improve server performance without manual effort and measurement and tuning processes can be automated. AmbiCom Holdings, Inc., for example, has created an application that automatically addresses computer system optimization and, within the application, has developed key intelligent algorithms that automatically apply to all computer hardware. AmbiCom’s application automatically evaluates a server’s workload and assets and then adjusts the manufacturer’s supplied settings to maximize server performance. On average, the application produces a 30% improvement in overall server capacity after a single, two-minute application of tuning algorithms. In areas such as network throughput, the application consistently produces improvements of 300%. And, if a server can increase its capacity by 30%, then, not only can the capital budget for servers be reduced by 30% but the power required to run these unnecessary servers is no longer needed. The AmbiCom solution runs a periodic process that automatically tests and optimizes server utilization in order to maximize use and capacity. By fine- tuning server performance, energy consumption is better managed. IT environments are able to maximize infrastructure spends and minimize correlating waste. Figure 2: Automation Process.
Proven Results In order to attest to the positive impact of AmbiCom’s application, they have worked with a number of leading, physical and virtual server manufacturers. To date, the AmbiCom application has been certified as delivering the cited improvement gains by IBM, Microsoft, Citrix, VMware, and a long list of enterprise customers. For More Information, Contact: Kevin Cornell kcornell@ambicom.com Phone: (831) 252-1685 November 8, 2017
Contributor Profiles --Kevin Cornell CEO and President Mr. Cornell has more than 40 years of experience in the computer industry and held senior management positions at Silicon Valley notables such as Oracle, Borland, MIPS, Symphoniq, and Finisar. Although his career has primarily involved running the business aspects of these companies, Mr. Cornell has been instrumental in creating numerous ‘firsts’ in the computer industry. As well, Mr. Cornell has four issued patents in fiber optics and security. See https://www.linkedin.com/in/kevinscottcornell/ for more information.
--The Hon., Thomas A. Cellucci, PhD, MBA Dr. Tom Cellucci has been a senior executive in both the private and public sectors for over 36 years. He served as the US Government’s first-ever Chief Commercialization Officer, working for both President Bush and President Obama and helps out President Trump when he can... Tom has authored or co-authored 25 scholarly books and over 197 high-tech business articles. Cellucci earned a PhD in Physical Chemistry from the University of Pennsylvania (1984), an MBA from Rutgers University (1991) and a BS in Chemistry from Fordham University (1980). He is on 21 Boards and Chairman of the Board of Trustees of Virginia International University near Washington, D.C. He holds an endowed Chair at a prestigious university in Kazakhstan, as well as taught at Harvard Business School, Princeton University and the University of Pennsylvania. ???November 8, 2017
News 
Market Data 
Discover 
