Green Computing: A Dual Technology for Cloud Computing and HPC
Mehdi Sheikhalishahi
Lucio Grandinetti
NOV 13, 2012 09:21 AM
A+ A A-

Climate change and global warming are viewed by many as the two most challenging problems facing the Earth. Green IT and in particular, green computing, are two ways the information and communications technology community is working to address those problems.

With the explosive growth of Internet-enabled cloud computing and high-performance computing centers, IT's energy consumption and sustainability impact are expected to continue climbing well into the future. Efforts are underway in both industry and academia, however, to address it.

As demand for computing and communication continue to grow, servers, networks, and data centers will consume more and more energy. For example, IT resources in the US now consume more than 1.5 percent of total electricity consumption. Power consumption of US data centers in 2006 was 1.5 percent of the total energy consumed, and at a cost of more than US $4.5 billion.

With the expected 30-fold increase in data traffic over the next decade, the overall power consumption of data centers and networks will become an issue of vital importance to the IT and telecommunications industries.

The Dual Nature of Computing

While computing and information technologies consume energy, they also enable productivity enhancements and directly contribute to energy efficiency. Some even argue that computing and information technologies are vital, significant and critical in moving towards a low-carbon future. In this article, we envision the duality of green computing with technological trends in other fields of computing such as HPC and cloud computing, business, and economy.

Green computing is a challenge for the future of HPC. For example, to reach exascale computing, we need huge amounts of energy to operate an exascale system. On the other hand, HPC provides solutions for green computing and climate change. Complicated processes of new sources of energy, need exascale computing for modeling and simulation.

In addition, these contemporary technologies are moving towards intelligent computation in order to optimize resource and energy consumption without losing performance. intelligent computations are done with the techniques and mechanisms of new computing technologies such as hardware and software co-design, application profiling, and virtual machine consolidation to optimize resource consumption, or pay-as-you-go business models to reduce costs.

Green computing solutions will drive the development of HPC and cloud computing. On the other hand, HPC and cloud computing solutions will drive the development of green computing. Therefore, we may say these technologies are dual to each other, thus we envision green computing as a dual technology for HPC and cloud computing.

Cloud Delivers IT on the Fly

In the cloud computing world, IT capabilities are delivered on the fly and on-demand through the Internet when the need arises, instead of drawing from desktop computers. Many design and architectural patterns are emerging around cloud computing that make it difficult to fit everything into a prefect definition.

In a formal definition, we denote cloud computing as an extreme specialization (hyperspecialization) and at the same time a general-purpose model of IT to digest the flux of future IT. More specifically, we connote the real definition of cloud computing as the convergence of the following essential and ideal characteristics of various distributed computing technologies:

  1. infinity: large scale data centers;
  2. outsourcing: remote, over the Internet;
  3. utility: pay per use;
  4. economy of scale;
  5. self-service: self-provisioning, on the fly;
  6. multi-tenancy
  7. on-demand;
  8. elasticity: scalability, autoscaling;
  9. abilities: availability, reliability, scalability, sustainability, etc.

The nature and the anatomy of cloud computing is totally green. From the users' point of view, cloud computing makes their IT life easier. They can access IT services without spending too much effort and energy, compared to the other IT models that are dedicated. On the other hand, from the scientific point of view, the characteristics of cloud computing and deployment models make it a ubiquitous green IT paradigm.

Economy of Scale

Cloud providers purchase data center infrastructure equipment such as hardware, network, and bandwidth much cheaper than a regular business. The construction and operation of extremely large-scale, commodity-computer data-centers at low-cost locations is the key necessary enabler of cloud computing. The cost of electricity, network bandwidth, operations, software, and hardware is decreased in the factors of 5-7 at very large economies of scale. User self-provisioning is one the greatest benefits of the cloud. With that, you have the ability to get applications up and running in a fraction of the time you would need in a conventional scenario.

In a cloud, users prepare their resources on the fly by themselves. On the other hand, infrastructure is adapted to the applications. For instance, Amazon Web Services makes it possible for anyone with an Internet connection and a credit card to access the same kind of world-class computing systems that Amazon uses to run its US $34 billion per-year retail operation. This cloud's characteristic significantly reduces time to solutions and access to services, thus it is energy efficient and green.

Multi-Tenancy

In traditional data centers, computing systems suffer from underutilization of computing power and networking bandwidth. Multi-tenant is a business model that provides a secure, exclusive virtualized computing environment in which servers, databases, and other resources are shared by multiple user companies in a cloud environment.

Public cloud is hosted, operated, and managed by a third-party vendor from one or more data centers. Since the services are offered to multiple customers with the aforementioned characteristics, it is multiple tenants over a common infrastructure. Public cloud providers optimize energy consumption as a way to offer competitive prices. This is Green IT at a low level powered by market rules.

With virtualization and multi-tenant feature, the resources in the cloud are not devoted to specific usages and users. At one time, a cloud resource can be used for an application by a user, and at another time it can be used for another diverse application by the same user or another user.

Virtualization and shared hosting technologies coupled with multicore servers are the enablers of cloud infrastructures to support a large number of disparate applications running simultaneously on multicore servers. Moreover, these technologies enable VM consolidation, infrastructure-adapted-to-applications (intelligent computation), and other resource optimization techniques.

This multi-usage feature of cloud addresses energy efficiency and green computing challenges by significantly improving resource utilization and minimizing resource waste.

High-Performance Computing

HPC is the use of advanced parallel processing systems (usually, above a teraflop or 1012 floating-point operations per second) for running complicated and huge processes quickly, efficiently, and reliably. The energy (power), cooling, and data center design are the three most prominent challenges of future HPC systems. Power has become the pre-eminent design constraint for future HPC systems. Moreover, the energy cost becomes an increasingly important factor.

We believe green and performance objectives converge to the same point. In this direction, HPC provides solutions for green computing. HPC is used to provide energy, and vice-versa energy is required to operate HPC systems. In particular, as exascale systems are emerging, they would need huge amounts of electricity to sustain. HPC systems of today need about 10MW power requirement. For instance, Sequoia, the IBM BlueGene/Q system installed at the US Department of Energy's Lawrence Livermore National Laboratory, is on the top of the Top 500 list with 16.32 petaflop/s. It is also one of the most energy efficient systems on the list with 7890kW power requirement. In addition, the target for the future exascale systems' power requirement is 20MW.

ExxonMobil predicts the outlook for energy enabled with supercomputing. ExxonMobil's global energy outlook projects through 2040. The projections indicate that, at that time, the world's population will be 8 billion, roughly 25 percent higher than today. Along with this population rise will be continuing economic growth. This combination of population and economic growth will increase energy demand by over 50 percent since 2000.

HPC drives the process of finding new sources of energy. Effective technology solutions to energy challenges rely on modeling complicated processes which will lead to a strong need for supercomputing. Two examples of the supercomputing need in the oil business are seismic approaches for finding petroleum and petroleum reservoir fluid-flow modeling (also known as "reservoir simulation").

HPC and exascale systems are the main tools to solve climate change challenges. HPC will become an even more critical resource to help the broader research community develop solutions to potential environmental problems. On the electricity grid front, power grid applications are exploiting HPC and networking. Technological and policy changes make this an urgent priority.

From the new hardware technologies point of view, Nvidia evolutionary systems,(i.e. GPUs, flash technology, and special-purpose hardware systems) drive the development of green and energy efficient architecture and systems. From the processing technologies point of view, Nvidia evolutionary systems, (i.e. GPUs), will be among the main building blocks of future exascale systems. Accelerator-based supercomputers now occupy the top eight slots of the most recent the Green500 list, so we observe that heterogeneous computing is extensively examined as a means for achieving computing system energy efficiency.

Hardware and software co-design (intelligent computation) is another key solution for green computing and energy efficiency. For instance, GreenFlash is a radically new approach to application-driven hardware and software co-design inspired by design principles from the consumer electronics marketplace. The GreenFlash system addresses some aspects of green computing and exascale system for specific applications. In addition, it significantly reduces cost and accelerates the development cycle of exascale systems. The application that is the target of Green Flash is "The Earth's Atmosphere at Kilometer Scales."

Economy and Business

Green computing and climate change have a direct and significant impact on economy and business. More and more companies consider green policy and climate change risks as part of their business policy and strategy. One of the main reasons is that companies with long-term climate change risks and opportunities in their business policy will gain strategic advantage over their competitors.

In addition, investors fund businesses and companies that consider the environmental policies and the environmental risks as part of their business strategies. For instance, there were floods in Thailand last year with the total cost of US $15-20 billion that had big impact on the sales and services of companies. The floods had negative impact on the automotive supply industry such as the automaker Daimler. In addition, the floods resulted in shortages of critical IT components and materials of Hewlett-Packard and Dell. Last but not the least, insurance companies have to take into account climate change risks in their risk analysis and calculations. They supply investors with information about potential environmental risks, costs and benefits. Similarly, a healthy economy contributes to the development of green computing and climate change solutions.

Mehdi Sheikhalishahi is a research fellow at University of Calabria and Lucio Grandinetti is a professor at University of Calabria.

FIRST
PREV
NEXT
LAST
Page(s):
[%= name %]
[%= createDate %]
[%= comment %]
Share this:
Please login to enter a comment:
 

Computing Now Blogs
Business Intelligence
by Keith Peterson
Cloud Computing
A Cloud Blog: by Irena Bojanova
The Clear Cloud: by STC Cloud Computing
Careers
Computing Careers: by Lori Cameron
Display Technologies
Enterprise Solutions
Enterprise Thinking: by Josh Greenbaum
Healthcare Technologies
The Doctor Is In: Dr. Keith W. Vrbicky
Heterogeneous Systems
Hot Topics
NealNotes: by Neal Leavitt
Industry Trends
Internet Of Things
Sensing IoT: by Irena Bojanova

 

RESET