| A. Marquez, K. Theobald, X. Tang, G. Gao, "A Superstrand Architecture," CAPSL Technical Memo 14, Dec. 1997. |
....units. Stalling a thread because of memory latency does not stall the entire processor. In 22 the HTMT, multithreading was chosen as the main approach to manage latencies in accessing memory and in communication between processing nodes. 4.2. 3 Stranding HTMT is a superstrand architecture [17] which exploits the notion of a strand. A strand is a block of instructions grouped together by the programmer (or compiler) to become a scheduling quantum of execution at the SPELL level. Strands are not ordered sequentially, but are scheduled according to data and control dependences: they ....
A. Marquez, K. Theobald, X. Tang, G. Gao, "A Superstrand Architecture," CAPSL Technical Memo 14, Dec. 1997.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC