Techila

Techila has published a new familiy of tutorial videos. These tutorial videos show step-by-step how to build an autonomous private High-Performance Computing cloud with Techila, in just minutes.

Setting up a flexible private HPC cloud with Techila consists of four steps: Setting up a Techila Server, updating the license of the Server, adding and trusting Workers and creating secure accounts for the End-Users.

Techila Server can be run on a physical or virtualized platform. It is possible to run even in the cloud, which enables building a full-cloud solution. The Techila Worker software supports a full range of operating systems from Windows to Linux and Mac OS X. The Techila Worker too can run on physical, virtualized or cloud platforms. Private, hosted, public or hybrid. For example, in this tutorial, the Worker is running on VMware.

In these Techila tutorial videos you can see, how you can do this all from scratch to ready-for-computing in just minutes.

Today, businesses, research and product development projects rely heavily on number crunching in its many forms, like optimization, modeling, simulations and data analysis. Cloud services have unforeseen amounts of cost-efficient computing power available, which could meet the end-users needs. Cloud computing is conceptually relatively simple but still hard to implement and manage on the application level.

According to the Total Economic Impact Analysis by Forrester Research application development to launch ready in the cloud does not happen over night. Because of the cost of development, developers want to be careful when selecting their platform. Launch-ready application code requires optimization for the cloud environment, which falls into two categories: Minimization of cloud resource consumption and to be able to predict cloud resource consumption, and implementation of scalability and robustness that performs well in the cloud environment. Using a proven middleware, which provides scalability and shielding from the complexity could save a lot of time and money here.

According to one ISV, they “developed their first platform in four years . . . re-architected for Azure over a nine-month period” and “reused a lot of code because it was directly portable.”

Will going with a cloud create vendor dependence?

For the last decade, IT managers have been working to enable solutions without vendor lock-in. When moving to the cloud era, they do not want to take a step backward. Thomas Erl, CEO of Arcitura Education Inc. expresses a caution in a Forbes article that going with a cloud solution means buying into the specific protocols, which create a vendor dependence.

The Techila middleware enables going with a cloud without taking a step backward. Techila enables moving workloads to the cloud and across different clouds without a vendor lock-in. With Techila you can move your Workload, for example, to Amazon EC2 or Windows Azure knowing that you can move it in back to on-premise or to another cloud platform in the future without any the impact. Techila makes it possible to do the application integration to clouds in a universal way.

Techila offers customers a middleware plug-in, which enables their applications to securely harness the infinite capacity of the cloud, public or private. This makes Techila the missing link between applications and the computing capacity of the clouds. Without a cloud computing vendor lock-in.

MacRumors community tells about a brilliant example of what easy-to-integrate computing capacity can enable: Digital video recorders produce enormous amounts of content. Encoding this content on a regular Mac HTPC can take a long time.

Distributed encoding and sharing the workload across several computers can speed up the process dramatically. Setting up a private media encoding cloud with the Techila middleware is easy. In this case, an one-evening project (including the integration of the existing application). Powering the encoder with the flexibly scalable capacity of the cloud was an easy project.

If the operators would provide faster uplinks, this could be a brilliant hybrid cloud use case for anyone archiving a digital media library.

What will be your application for the Techila middleware?

The European Commission has started putting a great focus on building and growing the high-performance computing market. Conventional HPC uses supercomputers to solve advanced computation problems and is a key enabler of ground-breaking innovation that creates high wage jobs and allows Europe’s competitiveness to be maintained.

The world of high-performance computing is in the middle of a radical change and more focus will be laid on software rather than on hardware. An example of this trend is cloud computing that promises to facilitate the high performance computing market. The cloud computing power is supplied from large data centres in the network, and offers economies of scale as the cost per server is drastically lowered based on a “pay per use” principle.

The Techila middleware is a key element in enabling this change. Techila enables producing HPC from the cloud and provides an easy integration and secure use, without the complexity.

Skipper, Kowalski, Private and Rico have been enlisted by the Council on Competitiveness in a video:

“From Medicine to Consumer Products, from Energy to Aerospace, High-Performance computing is accelerating breakthroughs to improve our lives and our competitiveness.

The Public and Private sectors must continue together to harness the full potential of High-Performance Computing as an innovation enabler. High-Performance Computing makes high-performance business and high-performance lives.”

GARCH models have been applied to option pricing with a good success. The main advantage of GARCH models is that the volatility dynamics can be characterized in imaginative ways. In fact, the family of GARCH models amounts to an infinite state space setups.

However, there is a drawback: In most GARCH option pricing models, no closed-form analytical solution for the option price is available and the price is available only through Monte Carlo simulation. This makes the model calibration computationally expensive, even if we use variance reduction techniques.

In his Global Derivatives Trading & Risk Management presentation, Associate Professor Juho Kanniainen shows how the performance of GARCH models can be assesed using distributed computing. With the accuracy used in the presentation, the calibration of the leverage model could take about 800 – 1200 hours of CPU time. With Techila, the total time will be reduced to 3-6 hours.