I had a simple question about the client:

Since we are testing for 11 primes currently if we find a prime and start sieving for the remaining 10, will we actually see a 9% increase in speed??? Or are there other factors etc.

The reason why I ask is there was a 1-2% speed loss in seive file 300K-20m vs 3m-20m if I'm not mistaken, wondering if this is simply due to the dat file/size number of entries etc.

Also the sieve client seems to only use 25 MB max memory would there be any benifit to higher resource demands etc.