US 7,467,286 B2
Executing partial-width packed data instructions
Mohammad Abdallah, Folsom, Calif. (US); James Coke, Shingle Springs, Calif. (US); Vladimir Pentkovski, Folsom, Calif. (US); Patrice Roussel, Portland, Oreg. (US); and Shreekant S. Thakkar, Portland, Oreg. (US)
Assigned to Intel Corporation, Santa Clara, Calif. (US)
Filed on May 09, 2005, as Appl. No. 11/126,049.
Application 11/126049 is a continuation of application No. 09/852217, filed on May 08, 2001, granted, now 6,970,994.
Application 09/852217 is a continuation of application No. 09/053127, filed on Mar. 31, 1998, granted, now 6,230,253.
Prior Publication US 2005/0216706 A1, Sep. 29, 2005
This patent is subject to a terminal disclaimer.
Int. Cl. H04Q 3/00 (2006.01)
U.S. Cl. 712—2  [712/5; 712/20; 712/22; 712/23; 712/27] 29 Claims
OG exemplary drawing
 
1. A processor-implemented method comprising:
receiving a single scalar packed data instruction, the scalar packed data instruction specifying locations in a 128-bit logical register file of a first 128-bit packed data operand and a second 128-bit packed data operand, each of the 128-bit packed data operands including a low-order segment and a high-order segment, and each of the segments including two 32-bit single precision floating point data elements; and
storing, as a result of executing the single scalar packed data instruction within the processor, in a storage area specified by the scalar packed data instruction, a 128-bit packed data result operand, the 128-bit packed data result operand including a single data element that is a result of an operation selected from an add operation and a multiply operation performed on a single pair of corresponding data elements of the first and second 128-bit packed data operands, and the 128-bit packed data result operand including a plurality of other data elements having values that are not results of either add operations or multiply operations performed on pairs of corresponding data elements of the first and second 128-bit packed data operands.