Patent US5801975A successfully refreshed from Google Patents.
US5801975A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Computer modified to perform inverse discrete cosine transform operations on a one-dimensional matrix of numbers within a minimal number of instruction cycles
Assignee: Advanced Micro Devices Inc
•
Filing: 1996-12-02
•
Granted: 1998-09-01
•
Expires: 2016-12-02
Abstract
Abstract A multimedia extension unit (MEU) is provided for performing various multimedia-type operations. The MEU can be coupled either through a coprocessor bus or a local central processing unit (CPU) bus to a conventional processor. The MEU employs vector registers, a vector arithmetic logic unit (ALU), and an operand routing unit (ORU) to perform a maximum number of the multimedia operations within as few instruction cycles as possible. Complex algorithms are readily performed by arranging operands upon the vector ALU in accordance with the desired algorithm flowgraph. The ORU aligns the operands within partitioned slots or sub-slots of the vector registers using vector instructions unique to the MEU. At the output of the ORU, operand pairs from vector source or destination registers can be easily routed and combined at the vector ALU. The vector instructions employ special load/store instructions in combination with numerous operational instructions to carry out concurrent multimedia operations on the aligned operands.
Inventors
- 👤 John S. Thayer
- 👤 John G. Favor
- 👤 Frederick D. Weber
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-11-09 21:44
CPC Classifications
G; G06; G06F; G06F9/00; G06F9/06; G06F9/30; G06F9/30003; G06F9/30007; G06F9/30021; G06F17/00; G06F17/10; G06F17/14; G06F17/147; G06F9/3001; G06F9/30014; G06F9/30032; G06F9/30036; G06F9/30098; G06F9/30105; G06F9/30109; G06F9/3012; G06F9/3013; G06F9/38; G06F9/3885; G06F9/3887
Claims Analysis
Total Claims: 22
Claim 1
Independent Claim
109 words
A method for concerting a matrix of input values to a matrix of output values within a computer incoming, comprising: loading an incoming signal represented as a matrix of input values into a first source register and a second source register, each of which are partitioned into a plurality of first source slots and second source slots, respectively; changing the order in which the input values are arranged within said second source slots; and performing a plurality of concurrent arithmetic operations on respective pairs of said input values stored within the first and second source slots to convert the matrix of input values to a matrix of output values.
Claim 2
Independent Claim
109 words
A method for concerting a matrix of input values to a matrix of output values within a computer incoming, comprising: loading an incoming signal represented as a matrix of input values into a first source register and a second source register, each of which are partitioned into a plurality of first source slots and second source slots, respectively; changing the order in which the input values are arranged within said second source slots; and performing a plurality of concurrent arithmetic operations on respective pairs of said input values stored within the first and second source slots to convert the matrix of input values to a matrix of output values.
Claim 3
Dependent Claim
24 words
The method as recited in claim 1, wherein said matrix of input values and said matrix of output values each comprise a one-dimensional matrix.
Claim 4
Dependent Claim
15 words
The method as recited in claim 2, wherein said one-dimensional matrix is a 1×8 matrix.
Claim 5
Dependent Claim
44 words
The method as recited in claim 1, wherein said matrix of input values comprise a matrix of values describing pixel elements in a spatial domain, and wherein said matrix of output values comprise a matrix of values describing pixel elements in a frequency domain.
Claim 6
Dependent Claim
27 words
The method as recited in claim 4, wherein said matrix of input values are converted to said matrix of output values using a discrete cosine transform algorithm.
Claim 7
Dependent Claim
44 words
The method as recited in claim 1, wherein said matrix of input values comprise a matrix of values describing pixel elements in a frequency domain, and wherein said matrix of output values comprise a matrix of values describing pixel elements in a spatial domain.
Claim 8
Dependent Claim
28 words
The method as recited in claim 1, wherein said matrix of input values are converted to said matrix of output values using an inverse discrete cosine transform algorithm.
Claim 9
Dependent Claim
27 words
The method as recited in claim 1, wherein said plurality of first source slots comprise eight slots, and said plurality of second source slots comprise eight slots.
Claim 10
Dependent Claim
25 words
The method as recited in claim 8, wherein said plurality of arithmetic operations comprise eight operations performed on up to eight pairs of spatial values.
Claim 11
Dependent Claim
52 words
The method as recited in claim 9, wherein each pair of said eight pairs of spatial values is stored in a pair of slots, whereby one slot of said pair of slots is within said first source slots and the other of said pair of slots is within said second source slots.
Claim 12
Dependent Claim
56 words
The method as recited in claim 1, wherein said changing step comprises moving spatial values of one of each said pairs of input values to a slot position within said second source slots commensurate with a slot position in which the other one of each said pairs of input values reside within said first source slots.
Claim 13
Dependent Claim
20 words
The method as recited in claim 1, wherein said plurality of concurrent operations comprise add, multiply, subtract, and accumulate operations.
Claim 14
Dependent Claim
39 words
The method as recited in claim 1, wherein up to eight of said operations occur during a single instruction cycle, and wherein said matrix of output values consists of 1×8 output values achieved in fewer than seven instruction cycles.
Claim 15
Dependent Claim
162 words
A computer, comprising: an input/output device operably coupled to a microprocessor, wherein the microprocessor includes: an instruction cache adapted for storing coded integer and vector instructions; a decode unit for decoding and routing said vector instructions to a multimedia extension unit; said multimedia extension unit comprising: a first source register and a second source register, each of which are partitioned into a plurality of first source slots and a plurality of second source slots, respectively; an operand routing unit responsive to said vector instructions for changing the order in which a first plurality of input values are arranged within said second source slots; and an arithmetic logic unit responsive to said vector instructions for performing a sequence of concurrent operations on said first plurality of input values rearranged within said second source slots and a second plurality of input values within said first source slots to produce a one-dimensional matrix of output values represented in a domain dissimilar from said input values.
Claim 16
Dependent Claim
162 words
A computer, comprising: an input/output device operably coupled to a microprocessor, wherein the microprocessor includes: an instruction cache adapted for storing coded integer and vector instructions; a decode unit for decoding and routing said vector instructions to a multimedia extension unit; said multimedia extension unit comprising: a first source register and a second source register, each of which are partitioned into a plurality of first source slots and a plurality of second source slots, respectively; an operand routing unit responsive to said vector instructions for changing the order in which a first plurality of input values are arranged within said second source slots; and an arithmetic logic unit responsive to said vector instructions for performing a sequence of concurrent operations on said first plurality of input values rearranged within said second source slots and a second plurality of input values within said first source slots to produce a one-dimensional matrix of output values represented in a domain dissimilar from said input values.
Claim 17
Dependent Claim
22 words
The computer as recited in claim 14, wherein said decode unit is adapted for concurrently dispatching at least two said integer instructions.
Claim 18
Dependent Claim
31 words
The computer as recited in claim 14, wherein said decode unit is adapted for concurrently dispatching at least one of said integer instructions with at least one of said vector instructions.
Claim 19
Dependent Claim
45 words
The computer as recited in claim 14, wherein said input values are converted to said output values by said operations using discrete cosine transform such that said input values are represented in a spatial domain dissimilar from said output values represented in a frequency domain.
Claim 20
Dependent Claim
46 words
The computer as recited in claim 14, wherein said input values are converted to said output values by said operations using inverse discrete cosine transform such that said input values are represented in a frequency domain dissimilar from said output values represented in a spatial domain.
Claim 21
Dependent Claim
53 words
The computer as recited in claim 14, wherein said multimedia extension unit further comprises a destination register partitioned into a plurality of destination slots wherein said destination slots are configured to receive a plurality of intermediate values and to forward said intermediate values to said arithmetic logic unit during an accumulate vector instruction.
Claim 22
Dependent Claim
37 words
The computer as recited in claim 14, wherein up to eight of said operations occur during each vector instruction cycle, and wherein said matrix of 1×8 output values are achieved in fewer than seven vector instruction cycles.
References Cited (Backward Citations)
US5181183A
Cited
US5453945A
Cited
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 2001-08-31 | 2002-03-02 | 2002-09-01 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 2005-08-31 | 2006-03-02 | 2006-09-01 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2009-08-31 | 2010-03-02 | 2010-09-01 | $7700.00 | Window Closed / Lapsed |