Archives of the TeradataForum
Message Posted: Fri, 02 Apr 2004 @ 14:48:07 GMT
<-- Anonymously Posted: Thursday, April 01, 2004 16:14 -->
Exec Summary from Forester website: Teradata quietly withdrew its 10 TB and 3 TB TPC-H decision-support (data warehousing) benchmarks on March 4, 2004, without formal, public comment and will abstain from future publications of TPC-H benchmarks. Teradata asserts the benefits of not competing will outweigh the flak it will take from its competitors, which continue to compete. However, the real loss will be experienced by clients and prospects that will be poorer in information and the innovations driven by the benchmark.
From the article (in my words):
- Teradata quit because it feels that the TPC-H benchmark is artificial and of little value for its customers.
- Vendors have reverse engineered the benchmark and adapted their product so it will place better.
- Although the TPC benchmarks have their value, customers need benchmarks specific to their selection process. Teradata wants to focus on custom benchmarks to help close pending deals (reportedly completes more than 40 per year).
- Although it has withdrawn from the TPC-H, Teradata will remain an active participant in the TPC. Wants to work toward a more real-world benchmark.
- Although it wasn't felt to be a direct factor, the TPC-H required specific configuration and pricing to be detailed. This would have included any discount information and given a buyer's market, this could have been leveraged by the customer in future sales.
- In the previous TPC-H results, Teradata's price/performance was tied with IBM's DB2. By leaving the TPC-H, it leaves the competitors in the position where they will be comparing their product against the last public TPC-H results (apples/oranges instead of apples/apples).
The Withdrawn TPC Results List can be found at:
|Copyright 2016 - All Rights Reserved|
|Last Modified: 23 Jun 2019|