For additional information
read what "The OLAP Report" says about
TM1.
TM1 enables users throughout the enterprise to share information and collaborate, no matter where they are located. A knowledge worker needs only an understanding of the business, not the structure of the data base to use the power of TM1. Links and relationships between numbers are created automatically, enabling non-technical users to access, view and analyze data from many perspectives. When new data is input, elements added, assumptions changed, or scenarios analyzed, the results are available in real-time. There is no waiting for an answer. Everyone stays on task, focused and moving toward the goal of making the right decision and executing the right plan.
The design of TM1 provides real-time calculations as well as immediate access to incremental updates (daily, weekly, monthly). As soon as data is loaded, it is available for manipulation, calculation and creation of new views. New data elements can be added to a dimension (up to 16 dimensions per cube) without taking servers off-line. New data sets can be added as new, appropriately dimensioned cubes. New rules can be written to integrate the data in inter-cube calculations. Simple spreadsheet commands bring information from any cube on any server (up to eight) into a single, side-by-side view.
The TM1 design departs radically from most OLAP products that achieve fast response time by pre-calculating consolidations and derived values. This leads to "data explosion" with huge data cubes that slow down application development and optimization, make testing challenging and are a difficult to deploy. A typical TM1 model can be up to 1/50-th the size of most other OLAP products.
TM1 stores only the lowest level elements in the model, and does so extremely efficiently. A 100 megabyte input file from a relational database typically requires only 10 megabytes in TM1. Thus, extremely large models can be put into memory, and calculated fast in real-time. TM1 accommodates large models -- the equivalent of several hundred gigabytes in pre-calculated OLAP products.
Other OLAP products achieve fast response times by precalculating and storing consolidations and derived values. This practice leads to "data explosion" -- large cubes of calculated data that require enormous processing power, RAM, and disk storage space. TM1 is unique in that it creates multiple, compact, densely populated data cubes that are tens or even hundreds of times smaller than cubes stored by other OLAP products.
To achieve such compact storage of cubes, TM1 stores only the lowest level elements in the model, using extremely efficient, proprietary algorithms. With these algorithms, a data point requires only 10 bytes of storage. Therefore, a 100 megabyte input file from a relational database typically requires only 10 megabytes in TM1 (assuming approximately100 bytes per input record). As a result, extremely large models can be put into memory and calculated quickly. Thus, TM1 can accommodate much larger models than would be possible with OLAP products that use a precalculated approach.
TM1 real-time, calculation-on-demand design also provides immediate access to incremental updates (daily, weekly, monthly). As soon as the data is loaded, it is available for manipulation, calculation and creation of new views. New data elements can be added to a dimension (up to 16 per cube) while the servers are still up. New data sets can be added as new, appropriately dimensioned cubes. And new rules can be written to integrate the data in inter-cube calculations. Simple spreadsheet commands bring information from any cube (up to 128) on any server (up to eight) side-by-side into a single view.
A TM1 application can scale from an enterprise-wide implementation to a single-user, standalone development. The same engine runs on a laptop and multiple, multi-user NT or UNIX servers. Moving up is as easy to moving files from the PC's directory to the server's. TM1-based solutions can grow in the enterprise as they prove their value. Therefore the risk is much lower than most OLAP alternatives.
The multi-user environment allows Professional IT controls, including security and access rights to cell level by user, automatic audit trails that detail every change to the database, such as the identity of the user, automated back-up, recovery and roll-back.
A TM1 application can scale from an enterprise-wide WAN implementation at hundreds of desks down to a single-user standalone for application development. The same engine runs on a notebook computer and multiple, multi-user NT or UNIX servers. Users have unparalleled flexibility in where and when they can work. Solutions based in TM1 can begin with a single user stand-alone and grow out across a workgroup, a division or the entire enterprise. Moving up is as easy as moving files from the PC's directory to the server's. The only difference in the TM1 multi-user environment is 'professional' IT controls: security and access rights to cell level by user, automatic audit trails that detail every change to the database including the identity of the user; and automated back-up, recovery and roll-back. So, as the system phases in across the enterprise, the added complexities can be incorporated in a controlled and structured manner.
Another important benefit of the multi-cube architecture is the ability to distribute processing at multiple regional cubes (servers) that feed a central, higher-level cube (consolidation) at corporate headquarters. The regional cubes can be broken down further into dimensions such as "years". All of this enables TM1 applications to be more compact and more understandable. In addition, if you elect to store calculations - which is an option - there is much less database explosion.
TM1 applications are easy to change, extend and enhance. A new cube can be built to handle pricing options, for example, and it can be related to an existing application without having to redesign it. This approach makes systems much easier to maintain and prevents the need for a complete rewrite when business needs change.
TM1 multi-cube architecture enables distributed processing at multiple regional cubes (servers), which feed a central, higher-level cube at corporate headquarters. The regional cubes can be broken down further by "years" or other parameters. As a result, TM1 applications are more compact and easier to understand.
TM1 offers an extremely flexible and productive development environment. If you know how to use Excel or Lotus, you can develop TM1 applications very rapidly. The multi-cube architecture gives the developer - whether an end-user or a professional - much more freedom and flexibility in developing applications. And the applications - can be up to 100 times smaller than in a single-cube architecture.
The multi-cube architecture also allows the developer to attack the central requirements of an application, simplifying the concepts initially, and then building in the complexities from a solid base. This technique ensures that there are plenty of feed back loops in the development process and delivers a better system faster. It also reduces risk - fundamental to every businessman, accountant and IS professional!
You can start small (even with a single user) to prove the concept, then phase the roll-out to the user population. As the system rolls out, the added complexities required by different user groups can be incorporated in a controlled and structured manner.
TM1 is elegantly simple by design. The multi-cube architecture also allows the developer to attack the central requirements of an application, simplifying the concepts initially, and then building in the complexities on a solid base. The system features automatic tuning and optimum calculation path. Built-in feed back loops can be built into the development process, delivering a better system faster. This also reduces risk in development and use -- a fundamental benefit to every manager, accountant and IS professional!
Applications created in TM1 are easy to change, extend and enhance. For example, you can build a new cube to handle pricing options, then related it to an existing application without having to redesign the application. This approach makes systems much easier to maintain, and prevents the need for a complete re-write when business needs change.