← Back to Patent Explorer
US4849926A Expired / Lapsed US ⏱️ 0.0 years remaining

Data processing circuit for calculating either a total sum or a total product of a series of data at a high speed

Assignee: NEC Corp
•
Filing: 1988-03-28
•
Granted: 1989-07-18
•
Expires: 2008-03-28
📄 Original PDF 🌐 Google Patents 🔄 Sync Now

Abstract

Abstract In a data processing circuit for calculating a total sum of a succession of data in a pipeline fashion, an arithmetic unit (28) carries out an addition iteratively on the data to produce intermediate results of addition. Work registers (31, 32), equal in number to the number of the pipeline stages, successively temporarily stores the intermediate results. Specifying circuit (35) cyclically specifies the work registers as specified registers to store the intermediate results in the specified registers in the pipeline fashion. Selector (33) selects each of the work registers as a selected register to supply the arithmetic unit with the intermediate result stored in the selected register. When the arithmetic unit carries out the addition on all the data, contents of the work registers are added to each other by the use of the arithmetic unit to provide an ultimate result representative of the total sum.

Inventors

  • 👤 Kenji Hasegawa

Bibliographic Details

Application Number —
Jurisdiction US
Data Source google_patents
Last Synchronized 2026-11-10 01:41
CPC Classifications
G; G06; G06F; G06F7/00; G06F7/38; G06F7/48; G06F7/50; G06F7/505; G06F7/509; G06F7/5095; G06F7/52; G06F7/523

Claims Analysis

Total Claims: 6
Claim 1
Independent Claim 303 words
A data processing circuit for use in a scalar machine in combination with memory unit to process each of a plurality of instructions through a plurality of pipeline stages in a pipeline fashion, said instructions including an iterative calculation instruction for carrying out an interactive calculation and a scalar instruction for carrying out a scalar calculation, said interactive calculation instruction being used for successively accessing said memory unit to read at least one operand, said iterative calculation being divisible into first through last partial calculations which are sequentially carried out at every pipeline stage by producing a sequence of output results successively appearing from an initial result to a final result during said first through last partial calculations, respectively, said data processing circuit comprising: (a) calculation means operable in response to each of said instructions for selectively carrying out said iterative calculation to produce said output results during said partial calculations, respectively; (b) calculation register means operatively connected to said calculation means for successively and temporarily storing each of the output results of said calculation means when said calculation means is operable in response to said iterative calculation instruction, said calculation register means comprising a plurality of registers operatively connected to said calculation means for defining the pipeline stages and a plurality of work registers cooperating with said plurality of registers and said calculation means and each having a work register function, said work registers storing the output results produced during different ones of said partial calculations, respectively; (c) register administration means for administrating said plurality of the registers and said work registers for cyclically carrying out said successive and temporary storing of said output results through said pipeline stages; and (d) a calculation result register for storing said final result during said last partial calculation in accordance with said iterative calculation instruction.
Claim 2
Dependent Claim 303 words
A data processing circuit for use in a scalar machine in combination with memory unit to process each of a plurality of instructions through a plurality of pipeline stages in a pipeline fashion, said instructions including an iterative calculation instruction for carrying out an interactive calculation and a scalar instruction for carrying out a scalar calculation, said interactive calculation instruction being used for successively accessing said memory unit to read at least one operand, said iterative calculation being divisible into first through last partial calculations which are sequentially carried out at every pipeline stage by producing a sequence of output results successively appearing from an initial result to a final result during said first through last partial calculations, respectively, said data processing circuit comprising: (a) calculation means operable in response to each of said instructions for selectively carrying out said iterative calculation to produce said output results during said partial calculations, respectively; (b) calculation register means operatively connected to said calculation means for successively and temporarily storing each of the output results of said calculation means when said calculation means is operable in response to said iterative calculation instruction, said calculation register means comprising a plurality of registers operatively connected to said calculation means for defining the pipeline stages and a plurality of work registers cooperating with said plurality of registers and said calculation means and each having a work register function, said work registers storing the output results produced during different ones of said partial calculations, respectively; (c) register administration means for administrating said plurality of the registers and said work registers for cyclically carrying out said successive and temporary storing of said output results through said pipeline stages; and (d) a calculation result register for storing said final result during said last partial calculation in accordance with said iterative calculation instruction.
Claim 3
Dependent Claim 50 words
A data processing circuit as claimed in claim 1, further comprising: selector means coupled to said register administration means and said calculation register means for selecting one of said calculation register means under control of said register administration means to make said calculation means successively carry out said iterative calculation.
Claim 4
Dependent Claim 57 words
A data processing circuit as claimed in claim 2, wherein: said register administration means is operable to successively supply a selected one of said calculation register means with a storage command for storing each of said output results and to cyclically supply said selector means with a selection command for selecting one of said calculation register means.
Claim 5
Dependent Claim 31 words
A data processing circuit as claimed in claim 1, wherein said number of work registers is equal to the number of the pipeline stages carried out in said data processing circuit.
Claim 6
Dependent Claim 18 words
A data processing circuit as claimed in claim 1, wherein said iterative calculation instruction is a vector instruction.

References Cited (Backward Citations)

US3673399A Cited
US3875391A Cited
US4128880A Cited
US4298936A Cited
US4484259A Cited
US4651274A Cited
US4546446A Cited
US4405992A Cited
US4490786A Cited
US4589067A Cited

USPTO Annuity & Maintenance Windows

Annuity Stage Window Opens Payment Deadline Surcharge Period Until Estimated Fee Status
3.5 Years (Year 4) 1992-07-17 1993-01-16 1993-07-18 $2000.00 Window Closed / Lapsed
7.5 Years (Year 8) 1996-07-17 1997-01-16 1997-07-18 $3760.00 Window Closed / Lapsed
11.5 Years (Year 12) 2000-07-17 2001-01-16 2001-07-18 $7700.00 Window Closed / Lapsed

Assigned Categories & Tags

No categories assigned to this patent yet.

Assign Category Tag