Intel Corporation

Motorola Corporation

National Semiconductor Corporation

Via-Cyrix, Inc.

Nicholas G Samra

Andrew Huang

Stephan Jourdan

Jacob Doweck

Jeffrey A Lohman

Ram Gummadi

Namratha R Jaisimha

Glenn J Hinton

David J Sager

Alan Beecher Kyker

Belliappa Manavattira Kuttanna

Bradley Gene Burgess

Bryan P Black

M Clair Webb

Murali S Chinnakonda

Betty Y Kikuta

Stephen J Jourdan

Jacqueline S Nelson

Disclosed are a multi-die processor apparatus and system. Processor logic to execute one or more instructions is allocated among two or more face-to-faces stacked dice. The processor includes a conductive interface between the stacked dice to facilitate die-to-die communication.

Multi-die processor

Method, apparatus and system embodiments provide support for multiple SoEMT software threads on multiple SMT logical thread contexts. A sleep state mechanism maintains a current value of an element of architecture state for each physical thread. The current value corresponds to an active virtual thread currently running on the physical thread. The sleep state mechanism also maintains sleep values of the architecture state element for each inactive thread. The active and inactive values may be maintained in a cross-bar configuration. Upon a read of the architecture state element, simplified mux logic selects among the current values to provide the current value for the appropriate active thread. Upon a thread switch, control logic associated with the sleep state mechanism swaps the active state value for the current thread with the inactive state value for the new thread.

Sleep state mechanism for virtual multithreading

Methods and apparatus are disclosed to control power consumption within a processor. An example apparatus disclosed herein includes logic to identify at least one instruction type and to initialize a counter value corresponding to a maximum number of instructions to be performed, the maximum number being at least partially dependent upon the identified at least one instruction type. The example apparatus also includes processing logic to be enabled or disabled based, at least in part, on the counter value.

Methods and apparatus to monitor instruction types and control power consumption within a processor

Method, apparatus and system embodiments provide support for multiple SoEMT software threads on multiple SMT logical thread contexts. A thread translation table maintains physical-to-virtual thread translation information in order to provide such information to structures within a processor that utilize virtual thread information. By associating a thread translation table with such structures, a processor that supports simultaneous multithreading (SMT) may be easily retrofitted to support switch-on-event multithreading on the SMT logical processors.

Virtual multithreading translation mechanism including retrofit capability

A method, apparatus, and system are provided for a multi-threaded virtual state mechanism. According to one embodiment, active thread state of a first active thread is received using a virtual state mechanism, and virtual thread state is generated in accordance with the active thread state of the first active thread, and the virtual thread state corresponding to the first active thread is forwarded to state update logic.

System for forming a critical update loop to continuously reload active thread state from a register storing thread state until another active thread is detected

Fusing a load micro-operation (uop) together with an arithmetic uop. Intra-instruction fusing can increase cache memory storage efficiency and computer instruction processing bandwidth within a microprocessor without incurring significant computer system cost. Uops are fused, stored in a cache memory, un-fused, executed in parallel, and retired in order to optimized cost and performance.

Fusing load and alu operations

A method and apparatus for dual-target register allocation is described, intended to enable the efficient mapping/renaming of registers associated with instructions within a pipelined microprocessor architecture.

Dual-target block register allocation

Methods and apparatus are disclosed to control power consumption within a processor. An example processor disclosed herein comprises an instruction retirement unit; a first set of functional blocks to process a first set of instructions having a first instruction type; a second set of functional blocks to process a second set of instructions having a second instruction type; and a controller to enable the first set of functional blocks to process an instruction allocated to the instruction retirement unit if the type of the instruction is the first type, and to disable the first set of functional blocks after the instruction is retired by the instruction retirement unit.

Methods and apparatus to control functional blocks within a processor

A stack pointer update technique in which the stack pointer is updated without executing micro-operations to add or subtract a stack pointer value. The stack pointer update technique is also described to reset the stack pointer to a predetermined value without executing micro-operations to add or subtract stack a stack pointer value.

Updating stack pointer based on instruction bit indicator without executing an update microinstruction

Fusing micro-operations (uops) together. Intra-instruction fusing can increase cache memory storage efficiency and computer instruction processing bandwidth within a microprocessor without incurring significant computer system cost. Uops are fused, stored in cache memory, un-fused, executed in parallel, and retired in order to optimize cost and performance.

Intra-instruction fusion

For use in a processor having a first number of decode units for decoding an ordered stream of floating point instructions, a floating point unit (FPU) for receiving decoded ones of the floating point instructions and a method of processing the decoded ones of the floating point instructions. In one embodiment, the FPU includes: (1) a second number of floating point pipelines that execute the floating point instructions, the second number being at least one and less than the first number, the floating point pipeline having a load unit, an execution core and a store unit, (2) a floating point checkpoint buffer, coupled to the decode units, that queues the decoded ones of the floating point instructions for allocation to the floating point pipelines and (3) a floating point register file, coupled to and cooperable with the floating point checkpoint buffer, that preserves states of the execution core to allow the floating point pipelines to execute the floating point instructions out of order.

Pipelined, superscalar floating point unit having out-of-order execution capability and processor employing the same

Disclosed herein is a method for matching dependency coordinates and an efficient apparatus for performing the dependency coordinate matching very quickly. A plurality of buffers to store instructions is set forth. Each storage location of a buffer corresponds to a particular pair of dependency coordinates. Dependency matching logic receives the dependency coordinates for a buffered instruction and scheduling information pertaining to dispatched instructions. The dependency matching logic indicates whether a dependency precludes scheduling of the corresponding buffered instruction. Dependency checking logic produces a ready signal for the buffered instruction when no such dependency is indicated by the dependency matching logic.

Method and apparatus for fast dependency coordinate matching

An apparatus, system and method for accelerating exchange (XCHG) instructions in a processor using a register alias table (RAT) data array and a content addressable memory (CAM) to handle register renaming. The RAT has at least one read port, at least one write port, and a plurality of address entries. The CAM has at least one read address, at least one write address, and a plurality of swap addresses. A plurality of logical register numbers are used as CAM input addresses to the RAT, and the operation of the CAM is completed in a first phase and a second phase of a clock cycle. The logical register numbers that match a pair of input swap addresses are interchanged.

Method and apparatus for accelerating exchange or swap instructions using a register alias table (RAT) and content addressable memory (CAM) with logical register numbers as input addresses

The present invention is directed to a system and method for implementing a re-ordered instruction cache. In one embodiment, groups or &ldquo;packets&rdquo; of instructions with specific packet sizes are formed. Each of packets includes two or more positions. The two or more positions are defined such that they support one or more different types of instructions. Each of the positions are also correlated to a subset of the specialized execution units of the processor. Given a specific packet size and definitions for each of the positions, each of the instructions are re-ordered according to instruction type and loaded into the instruction cache in the new order.

Growing community of inventors

Austin, TX, United States of America

Nicholas G Samra