kernel(): 1. Compute sum of pairwise product at respective index, while within bounds. 2. Shift to the next component, by a stride of total no. of threads (4). 3. Store per-thread sum in shared cache ...

Sire: Moro Tap, gr/ro, 2010. Raced 3 yrs, 19 sts, 3 wins, $161,816. 3rd Louisville H (G3T), Centaur S (T). Lifetime: 5 crops, 123 foals, 72 rnrs (59%), 42 wnrs (34% ...