Blog: Richard Hackathorn

« August 2008 | Main | October 2008 »

September 26, 2008

InfoNOW Aims at Channel Data Management

At the Boulder BI Brain Trust, we heard a presentation from InfoNow about their company background, product offering, and future directions. They offer SaaS for the management of the demand chain (manufacturer to distributor to retailer). See the complete blog here.

Share:

September 12, 2008

Kickfire Kicks TPC-H Benchmark

I went to the TPC-H benchmarks recently and found that Kickfire was at the top of the price/performance list for 100GB and 300GB systems. Using a 3-year system cost, they were under a dollar per query while averaging around 50 thousand queries per hour! Who is Kickfire, I asked myself?

I was briefed by Steve Dille, VP of Marketing, and Beverley Bird, Director of Marketing Communications, for Kickfire. They were founded in 2006, have 50 employees, closed $30M in VC funding, and have been very busy.

As a new vendor in the increasingly crowded data warehouse appliance (DWA) marketplace, the ability to distinguish themselves is critical. The non-distinguishing features are twofold:

First, their product is column-oriented with data compression. Second, they are a ‘true’ appliance by offering an integrated hardware/software package.

The distinguishing features of Kickfire are twofold:

First, they are the only DWA built upon (underneath?) the open source database MySQL. I am increasingly impressed with the technology and community of MySQL. It is gaining creditability within enterprise circles; however, MySQL is not known for its ability to process large data sets and complex queries. Kickfire feels that they have solved those problems and are offering a unique solution to the growing MySQL community.

Second, they process SQL using a proprietary co-processor (like a graphics co-processor) whose machine language is SQL operations (like relational algebra operators). With their aggressive data compression, those SQL operations can be performed in memory. Further, the SQL operations are performed on the compressed data, avoiding the compression/decompression overhead.

A disadvantage is that the data volume is limited to 3 TB with the current nonclustered product. They plan to scale to higher data volumes with an array of similar processor units. On the other hand, I believe that there is huge demand for sub-TB warehouses that are inexpensive and convenient.

The product is in beta with two customers. Availability is mid October. I am going to watch the initial customer reactions on this one!

Share:

LyzaSoft Bridges Enterprise Analytics with Desktop

At the Boulder BI Brain Trust, we heard a presentation from LyzaSoft about their company background, product offering, and future directions. They offer an analysis tool for the desktop. See the complete blog here.

Share:

IBM Optim Covering Enterprise Data Management

A webcast from IBM describes the integration of the Optim product from Princeton Softech, acquired in September 2007. The focus is on the set of tools and procedures for managing the IT architecture across a wide scope of functions. The figure below show this scope as 14 sectors. I was impressed with the breath of functions that much be considered as a whole if a company is to manage properly its enterprise data asset.

Share:

September 3, 2008

Data Driven - New Book by Tom Redman

Tom Redman, better known as the 'Data Doc', has published a new book Data Driven: Profiting from Your Most Important Business Asset through Harvard Business Press.

In contrast to several recent books on the importance of managing with data analytics, Data Driven starts with the IT infrastructure required to maintain consistent data, then focuses on data quality from the executive perspective of the hidden costs of poor data, and finally, explores how to make better decisions through proper data management.

A nice twist is a chapter on content providers who bring packaged data to the marketplace. This is a growing segment that is applicable to every business, since every business collects data that has value to companies in its ecosystem.

There is a chapter on Social Issues, which is great in intent but weak in content. Sad...

The book ends with "what to do over the next one hundred days" advice. If managers are serious about treating data as a business asset, then this chapter lays out the essentials of what to do.

I recommend this book for business executives to orient their thinking about data as a business asset and to realize what tangible actions must be done to make that a reality in their companies. I know that these are old old themes for the IT profession. However, these fundamental themes are oldies but goodies!

Share: