Real time program language accelerator
AssigneeRazim Technology, Inc.
InventorsYair Raz, Arik Paran
Life of this patent
Estimating the term…
Abstract
A computer system for accelerated processing of stack oriented interpretive language instructions comprising a translator which establishes correlations between address values and core operations, a processing accelerator core for performing core operations, a cache memory for storing operand data, a control logic device for reading and writing operand data and for determining core operations corresponding with address values and an arithmetic logic unit for receiving data and performing core operations, being specified solely by address values.
Claims25
A computer system providing accelerated processing of stack oriented interpretive language instructions, said system comprising:
- a native processing unit for executing native language instructions;
- a translator for translating the interpretive language instructions to native language instructions, for generating address values associated with selected ones of said native language instructions, and for establishing a correlation between each of said address values and a corresponding core operation;
- a dedicated processing accelerator core communicatively coupled with said processing unit, responsive to address values received from said processing unit, and operative to perform core operations, said core including,
- a dedicated hardware stack cache memory device having a plurality of memory address locations addressable by said address values, said locations for storing operand data upon which said core operations are to be performed,
- core control logic responsive to said address values and read/write control signals provided by said native processing unit, operative to control writing and reading of said operand data to and from said cache memory device, and being further operative to determine core operations corresponding with said address values based solely on said correlation established by said translator, and
- an arithmetic logic unit communicatively coupled with said core control logic for receiving said operand data from corresponding ones of said address locations of said cache memory device, and being operative to perform core operations on said operand data, each of said core operations being specified solely by said address values.
In a computer system including a native processing unit for executing native language instructions, and a processing accelerator core coupled for communication with the processing unit, the core including a stack cache memory device having a plurality of memory address locations, control logic operative to control writing and reading of operand data to and from the locations of the cache memory device, and an arithmetic logic unit operative to perform core operations on said operand data, a method of providing accelerated processing of stack oriented interpretive language instructions, the method comprising the steps of:
- receiving interpretive language instructions;
- translating said interpretive language instructions to native language instructions;
- generating address values associated with selected ones of said native language instructions, each of said selected instructions and associated address values corresponding with a core operation;
- writing and reading operand data to and from memory address locations of a dedicated hardware stack cache memory device, said locations being addressable by corresponding ones of said address values, said operand data being associated with corresponding ones of a plurality of core operations to be performed on said operand data;
- and
- retrieving said operand data from corresponding ones of the locations using said address values, and using an arithmetic logic unit to perform core operations on said operand data, each of said core operations being specified solely by said corresponding address value.
A method of providing accelerated processing of stack oriented interpretive language instructions, the method comprising the steps of:
- translating interpretive language instructions to native language instructions;
- generating address values associated with selected ones of said native language instructions, each of said address values being associated with corresponding core operation;
- writing and reading operand data to and from storage locations of a dedicated hardware stack cache memory device, said locations being addressable by corresponding ones of said address values;
- and
- retrieving said operand data from corresponding ones of the storage locations using said address values, providing said operand data to an arithmetic logic unit, and using the arithmetic logic unit to perform core operations on said operand data, each of said core operations being specified solely by said corresponding address values used to retrieve said operand data.
An accelerator core for a processing unit, comprising:
- an arithmetic logic unit (ALU) for performing operations on at least one operand provided to the ALU;
- a stack cache memory device having a plurality of storage locations, each location being addressable by an address value specified by a translator that associates the address value with a core operation, the memory device being configured to store at least one operand at a location whose address value corresponds to the operation to be performed on the operand, and to provide at least one operand to the ALU;
- and
- a controller communicatively coupled to the memory device and the ALU, responsive to a memory address value indicating a location of the memory device, operative to access an operand from the location, to provide the operand to the ALU, and to specify the operation to be performed on the operand by the ALU based solely on the address value used to access the operand.
Dependencies are inferred from the claim text.