Back to results
US 8,799,577 B2Granted
Gather and scatter operations in multi-level memory hierarchy
AssigneeIntel Corporation
InventorsChristopher J. Hughes, Yen-Kuang Chen, Changkyu Kim, Daehyun Kim, Victor W. Lee, Anthony-Trung D. Nguyen, Nadathur Rajagopalan Satish
Life of this patent
Estimating the term…
Abstract
Methods and apparatus relating to gather or scatter operations in a multi-level cache are described. In some embodiments, a logic may determine whether to perform gather or scatter operations at a first memory or a second memory, based in part on a relative performance of performing the gather or scatter operations at the first memory and the second memory. Other embodiments are also described and claimed.
Claims20
An apparatus comprising:
- a first level cache of a multiple level memory hierarchy;
- and
- a level controller logic to determine whether to steer gather or scatter operations to the first level cache based on one or more locality predictions for the gather or scatter operations at the first level cache, wherein gather or scatter operations are steered to a lower level of the multiple level memory hierarchy than the first level cache based on one or more locality predictions for gather or scatter data of little locality.
A method comprising:
- determining whether to perform gather or scatter operations at a first level of a multiple level memory hierarchy, based in part on one or more locality predictions for gather or scatter data at the first level.
A system comprising:
- a main system memory to store a gather or scatter instruction;
- and
- a processor to fetch the stored instruction, the processor to access, responsive to the stored instruction, a first level cache of a multiple level memory hierarchy when a level controller logic determines to steer operations of the gather or scatter instruction to the first level cache based on one or more locality predictions, wherein operations of the gather or scatter instruction are steered to a lower level of the multiple level memory hierarchy than the first level cache based on one or more locality predictions for gather or scatter data of little locality.
Dependencies are inferred from the claim text.