US 12,443,394 B2
Dynamic directional rounding
Alex Fit-Florea, Los Altos Hills, CA (US); Boris Ginsburg, Mountain View, CA (US); Pooya Davoodi, Santa Clara, CA (US); and Amir Gholaminejad, San Jose, CA (US)
Assigned to NVIDIA Corporation, Santa Clara, CA (US)
Filed by NVIDIA Corporation, Santa Clara, CA (US)
Filed on Feb. 1, 2021, as Appl. No. 17/163,855.
Application 17/163,855 is a continuation of application No. 16/200,325, filed on Nov. 26, 2018, granted, now 10,908,878.
Prior Publication US 2021/0232366 A1, Jul. 29, 2021
This patent is subject to a terminal disclaimer.
Int. Cl. G06F 7/499 (2006.01); G06F 7/483 (2006.01)
CPC G06F 7/49957 (2013.01) [G06F 7/483 (2013.01)] 18 Claims
OG exemplary drawing
 
1. A method, comprising:
performing, using a processor, an operation using two or more floating point operands specified by a program instruction to generate a floating point result; and
rounding, using a rounding unit, the floating point result based, at least in part, on a value from a sign of one of the two or more floating point operands independent of a sign of the floating point result,
wherein the program instruction specifies which operand of the two or more floating operands is used to round the floating point result.