Over the years we've tried many things to help customers look at our products and give them some help in differentiating our software from the other guys. Here's what we found:
- We've been abused with claims that our benchmarks were designed to make us look good.
- We've been accused of staging benchmarks without giving the other guys the same care and attention to detail that we've taken with out tools.
- And we've been criticized for even making benchmarks, since most serious users claimed that they were pointless at best.
Well, for years after that we resisted the impulse to show the world our capabilities and instead relied upon our wise and thoughtful users to review our code and select accordingly. Mostly, this has worked well. But now it seems that one of our erstwhile competitors has begun to make untrue, unfair, and unjust comparisons between their products and our products. They even claimed to use our V6.1 compiler set (from our 32-bit tools release), pretty good trick especially given that we hadn't released it at the time they made their claims! Given those miss-leading and ill advised claims, we too are now reduced to showing you the power and capabilities in our tools.
To set the record straight: The following benchmarks are presented for your amusement, and amazement. Please let us know if anything we've done here is not in accordance with your strict standards of fair play and we'll address those issues for you immediately. Note too that some of our current ads do not yet reflect these current tool results. This is the most current information available anywhere today.
How we do them and measure the results?
Whetstone: Large Model
Whetstone: Small Model
Sieve: Large Model
Sieve: Small Model
Towers of Hanoi: Large Model
Towers of Hanoi: Small Model
Dhrystone: Large Model
Long Operations: Small Model