Lines Matching full:worked
16 * (8K-32K worked best for my 256K cpu cache) and how much overhead there is in
