Showing headlines posted by deadline

Time series analysis (financial data, web logs, etc.) works best with row oriented databases. A traditional database consists of tables with data ordered by row (row-oriented). When data are organized by rows, data look-up can be done quickly by using a key-value to identify the row and then its associated data. When data are organized by column (column-oriented), operations on columns are extremely fast because there is no key-value look-up. This approach is used for time-series analysis. If you are a kdb+ user or are intrigued with time-series analysis, there are resources waiting for you. First, there is a free version of kdb+ (32 bit) available for download. The free version has all the functionality of the 64-bit version and can be used for commercial, non-commercial, or educational purposes.

Of course it all depends, but deciding to use the Cloud for Linux HPC is not as simple as it may seem.

As Linux High Performance Computing (HPC) continues to grow in academic, government, and commercial sectors, one might wonder just who are the current movers and shakers of the Linux HPC industry and community? These days those pushing the HPC market and the technology are not necessarily pushing top performance, but rather expanding and improving HPC application spaces. There is new steering group being developed by Intersect360 that is inviting HPC users to help shape the future of the market and community Maybe you know someone who should apply. In fact, it could be you! Let's make sure Linux gets represented!

Many Linux/Unix programmers are not aware that a battle over Bash scripts is currently raging in a US (Kansas) courtroom this week.

Programming GPUs under Linux is not easy. It often requires massive re-programming existing applications. There are some tools that can help in this process.

Watson, a 2880 core IBM Linux cluster computer will attempt to beat two human experts on the USA game show Jeopardy. This is more than a chess match because in Jeopardy you must supply the question based on an answer. It is a measure of how far Artificial Intelligence has progressed.

When the first Linux clusters were constructed, Ethernet was one of the few choices for an interconnect. Of course there were more expensive and custom ways to connect computers, but Ethernet was the first network technology supported by Linux. How will 10GigE fair in the Linux HPC sector?

The WhisperingWulf is a fully functional totally quiet personal HPC cluster. This project was originally started as a quest for a small and silent HTPC (Home Theater Personal Computer).

With the availability high performance interconnects like InfiniBand, one has to ask why do people still use Ethernet for Linux HPC clusters? The answer may not be what you think!

There continues to be plenty of news in the GP-GPU (General Purpose - Graphical Processing Unit) sector. In case you missed the previews, the standard video card now can be used as a data parallel computer right on your desktop!

Everyone knows applications drive the HPC (High Performance Computing) boat. It is one thing to run benchmarks and burn-in programs, but when it is time for production work, applications take over. Fortunately, there are many applications that can take advantage of clusters.

If you are interested in HPC clustering and don't know where to begin, you are in the right place. All to often HPC beginners jump right into building hardware without understanding the nuances of HPC clusters.

Recently the following was posted to the Beowulf Mailing list: I would like to know what server has the best performance for HPC systems between The Dell Poweredge 1950 (Xeon) And 1435SC (Opteron). Please send me suggestions...

The response was probably not what the poster expected. From his specifications, almost certainly the Opteron would work well for a variety of reasons. If you think you know about buying a Linux performance, learn why you should not believe the answer.

Software continues to be one of the largest challenges to the Linux parallel computing market. When considering parallel and multi-core computing, questions about software are often the most difficult. To help set expectations and ensure a successful project, the important questions to ask (and some answers!) about parallel software are presented.

Can you draw a line with a single point? Sure you can, but you can draw the line anywhere and conclude anything. Measuring network performance on Linux clusters suffers from a similar issue. Often single network performance numbers are quoted when discussing clusters. This article looks beyond single measurements and the subsequent wrong conclusions they can invite. Often times real application data contradicts the conclusions based on single point of performance data.

Read this article and become a cluster design expert! Use a new tool from the aggregate.org to determine price and performance before you buy! Get a handle on everything from Ethernet cables, to GFLOPS, to power and cooling.

High performance interconnects used for Linux HPC clusters are often compared using artificial benchmarks. The important question is how well these artificial benchmarks translate into real-world performance. This article attempts to answer this question buy taking a look at the real-world performance for Mellanox for Qlogic Interconnects.

A hard to swallow conclusion from the table of cluster. In the first two installments of this series I have pushing a sales pitch of sorts. If you believe some of the things I talked about, large clusters will break, applications will need to tolerate failure and be easy to write, then you may agree that dynamic programming algorithms are one method to solve these problem. The next question is, how do implement these ideas?

A recent panel at the SC06 show in Tampa discussed Commercialization of Open source. The panel include Beowulf creators Tom Sterling and Don Becker. In addition, Mike Fitzmaurice founder of the Baltimore Area Beowulf Users Group added his thoughts. You can hear the entire session on ClusterCast -- a new podcast site for clustering enthusiasts.

A discussion of dynamic parallel execution is presented. The article mentions some possible alternatives to MPI, offers dynamic execution ideas, and develops some solutions.