http://www.infobright.org/Learn-More/ICE_IEE_Comparison/ says:

Description     ICE     IEE
...
Superior query performance   ✓    50-500% Faster
...
Infobright Loader supported   Single-Threaded (up to 55GB/hr) ASCII only   Multi-Threaded (up to 300GB/hr) ASCII / Binary (up to 100% faster than ASCII)

Where this page says “50-500% Faster” and “100% faster”, I think it should say “more work” rather than “faster”.

If a query is 50% faster in IEE than ICE, and it takes 100 seconds in ICE, it takes 50 seconds in IEE, and you can run 2 such queries in IEE in 100 seconds. Thus, when something is 50% faster you can run 100% more in the same time.

If a query is 90% faster in IEE than ICE, and it takes 100 seconds in ICE, it takes 10 seconds in IEE, and you can run 10 such queries in IEE in 100 seconds. Thus, when something is 90% faster you can run 1,000% more in the same time.

If a query is 99% faster in IEE than ICE, and it takes 100 seconds in ICE, it takes 1 second in IEE, and you can run 100 such queries in IEE in 100 seconds. Thus, when something is 99% faster you can run 10,000% more in the same time.

If a query is 99.9% faster in IEE than ICE, and it takes 100 seconds in ICE, it takes 0.1 second in IEE, and you can run 1,000 such queries in IEE in 100 seconds. Thus, when something is 99.9% faster you can run 100,000% more in the same time.

If a query is 100% faster in IEE than ICE, and it takes 100 seconds in ICE, it takes 0 seconds in IEE, and you can run an infinite number of such queries in IEE in any amount of time.

Queries and loads cannot be more than 100% faster. However, they can accomplish 500% more work in the same amount of time.