Intel Corporation

Zhen Fang

Ravishankar Iyer

Srihari Makineni

Xiaowei Jiang

Michael J Espig

David J Harriman

Mahesh U Wagh

Jasmin Ajanovic

Sriram R Vangal

Li Zhao

Donald K Newell

Erik Hallnor

Ravi Iyer

Liqun Cheng

Nitin B Gupte

Dmitri Yurievich Pavlov

Shih-Lien L Lu

Robert S Chappell

Michael W Leddige

Rameshkumar G Illikkal

Nitin Y Borkar

Ramesh Kumar Illikkal

Tong Li

Hemant G Rotithor

Sadagopan Srinivasan

Ryan Carlson

Jeffrey S Wilder

Hongliang Gao

Steven Zhang

Meenakshisundara R Chinthamani

Milind B Kamble

Seung Eun Lee

Ravishankar Iver

An instruction pipeline implemented on a semiconductor chip is described. The semiconductor chip includes an execution unit having the following to execute an interrupt handling instruction. Storage circuitry to hold different sets of micro-ops where each set of micro-ops is to handle a different interrupt. First logic circuitry to execute a set of said sets of micro-ops to handle an interrupt that said set is designed for. Second logic circuitry to return program flow to an invoking program upon said first logic circuitry having handled said interrupt.

Interrupt return instruction with embedded interrupt service functionality

In accordance with some embodiments of the present invention, a branch prediction unit for an embedded controller may be placed in association with the instruction fetch unit instead of the decode stage. In addition, the branch prediction unit may include no branch predictor. Also, the return address stack may be associated with the instruction decode stage and is structurally separate from the branch prediction unit. In some cases, this arrangement reduces the area of the branch prediction unit, as well as power consumption.

Embedded branch prediction unit

An apparatus is described that contains a processing core comprising a CPU core and at least one accelerator coupled to the CPU core. The CPU core comprises a pipeline having a translation look aside buffer. The CPU core comprising logic circuitry to set a lock bit in attribute data of an entry within the translation look-aside buffer entry to lock a page of memory reserved for the accelerator.

Efficient locking of memory pages

Systems and methods of operating a memory controller may provide for receiving a write request from a motion compensation module, wherein the write request includes video data. A compression of the video data may be conducted to obtain compressed data, wherein the compression of the video data is transparent to the motion compensation module. In addition, the compressed data can be stored to one or more memory chips. Moreover, a read request may be received, wherein stored data is retrieved from at least one of the one or more memory chips in response to the request. Additionally, a decompression of the stored data may be conducted to obtain decompressed data.

Dram compression scheme to reduce power consumption in motion compensation and display refresh

Systems and methods may provide for determining whether a memory access request is error-tolerant, and routing the memory access request to a reliable memory region if the memory access request is error-tolerant. Moreover, the memory access request may be routed to an unreliable memory region if the memory access request is error-tolerant. In one example, use of the unreliable memory region enables a reduction in the minimum operating voltage level for a die containing the reliable and unreliable memory regions.

Differentiating cache reliability to reduce minimum on-die voltage

Methods and apparatus for control of On-Die System Fabric (OSF) blocks are described. In one embodiment, a shadow address corresponding to a physical address may be stored in response to a user-level request and a logic circuitry (e.g., present in an OSF) may determine the physical address from the shadow address. Other embodiments are also disclosed.

Control of on-die system fabric blocks

In one embodiment, the present invention includes a method for handling a registration message received from a host processor, where the registration message delegates a poll operation with respect to a device from the host processor to another component. Information from the message may be stored in a poll table, and the component may send a read request to poll the device and report a result of the poll to the host processor based on a state of the device. Other embodiments are described and claimed.

Delegating a poll operation to another device

Methods and apparatus to provide for power consumption reduction in memories (such as cache memories) are described. In one embodiment, a virtual tag is used to determine whether to access a cache way. The virtual tag access and comparison may be performed earlier in the read pipeline than the actual tag access or comparison. In another embodiment, a speculative way hit may be used based on pre-ECC partial tag match to wake up a subset of data arrays. Other embodiments are also described.

Cache memory power reduction techniques

In one embodiment, the present invention includes a method for receiving a memory request from a device coupled to an input/output (IO) interconnect, accessing a mapping table associated with the IO interconnect to determine if an address range including an address of the memory request is coherent, and if so, sending the memory request and a coherency indicator to indicate the coherent state of data at the address, otherwise sending the memory request and the coherency indicator to indicate a non-coherent state. Other embodiments are described and claimed.

Providing address range coherency capability to a device

In one embodiment, a cache memory includes entries each to store a ring level identifier, which may indicate a privilege level of information stored in the entry. This identifier may be used in performing read accesses to the cache memory. As an example, a logic coupled to the cache memory may filter an access to one or more ways of a selected set of the cache memory based at least in part on a current privilege level of a processor and the ring level identifier of the one or more ways. Other embodiments are described and claimed.

Controlling access to a cache memory using privilege level information

In an embodiment of the present invention a method includes: sending request for data to a memory controller; arranging the request for data by order of importance or priority; identifying a source of the request for data; and if the source is an input/output device, masking off P ways in a cache; and allocating ways in filling the cache. Other embodiments are described and claimed.

Searching a shared cache by using search hints and masked ways

In one embodiment, a method includes comparing an occupancy level of a buffer of a port of a router to a threshold, and controlling the port to operate at a first voltage and frequency based at least in part on the comparison, and at least one other port of the router is controlled to operate at a second voltage and frequency. Other embodiments are described and claimed.

Growing community of inventors

Portland, OR, United States of America

Zhen Fang