Patent US4725973A successfully refreshed from Google Patents.
← Back to Patent Explorer
US4725973A Expired / Lapsed US ⏱️ 0.0 years remaining

Vector processor

Assignee: Hitachi Ltd
•
Filing: 1983-10-25
•
Granted: 1988-02-16
•
Expires: 2003-10-25
📄 Original PDF 🌐 Google Patents 🔄 Sync Now

Abstract

Abstract A vector processor for executing vector instructions comprises a plurality of vector registers and a plurality of pipeline arithmetic logic units. The vector registers are constructed with a circuit which operates in a speed equal to 2n times as fast as the processing speed of the pipeline arithmetic logic units. Either the read or the write operation from or to the vector registers are carried out in the time obtained by a processing cycle of each of the pipeline arithmetic logic units multiplied by n/2.

Inventors

  • 👤 Tsuguo Matsuura
  • 👤 Toshihiko Odaka

Bibliographic Details

Application Number —
Jurisdiction US
Data Source google_patents
Last Synchronized 2026-10-29 12:44
CPC Classifications
G; G06; G06F; G06F15/00; G06F15/76; G06F15/80; G06F15/8053; G06F15/8092

Claims Analysis

Total Claims: 13
Claim 1
Independent Claim 116 words
A vector processor for executing vector instructions comprising: timing means for providing a timing signal; a plurality of pipeline-type arithmetic logic units, each connected to receive said timing signal, to receive data in a cycle pitch and providing an arithmetic data result in said cycle pitch; a plurality of vector registers each connected to receive said timing signal for executing 2n (n is an integer not less than one) accessing operations in succession during one cycle pitch with respect to a selected vector register; and data transfer means connected between said plurality of pipeline-type arithmetic logic units and said plurality of vector registers for transferring data between said pipeline-type arithmetic logic units and said vector registers.
Claim 2
Dependent Claim 116 words
A vector processor for executing vector instructions comprising: timing means for providing a timing signal; a plurality of pipeline-type arithmetic logic units, each connected to receive said timing signal, to receive data in a cycle pitch and providing an arithmetic data result in said cycle pitch; a plurality of vector registers each connected to receive said timing signal for executing 2n (n is an integer not less than one) accessing operations in succession during one cycle pitch with respect to a selected vector register; and data transfer means connected between said plurality of pipeline-type arithmetic logic units and said plurality of vector registers for transferring data between said pipeline-type arithmetic logic units and said vector registers.
Claim 3
Dependent Claim 83 words
A vector processor according to claim 1, each of said a plurality of vector registers comprising n write data registers, n read data registers, 2n read/write address registers, and selector means for selecting one of said write data and read data registers and one of said read/write address registers, wherein desired ones of said registers are sequentially selected by use of said selector so that n write operations and n read operations are enabled on the same vector register in one cycle pitch.
Claim 4
Dependent Claim 22 words
A vector processor according to claim 2, wherein each of said vector registers is connected to said selector with two signal lines.
Claim 5
Dependent Claim 172 words
A vector processor for executing vector instructions comprising: timing means for providing a timing signal 2n times during each of a plurality of successive operation cycles (where n is an integer not less than one); a plurality of vector registers, each connected to receive said timing signals and including vector storage means having a plurality of addressable storage locations and read/write control means responsive to said timing signal for executing 2n accessing operations in succession with respect to the storage locations of said vector storage means of a selected vector register during each operation cycle; a plurality of pipeline-type arithmetic logic units, each connected to receive said timing signals and data from selected ones of said plurality of vector registers in an operating cycle for providing an arithmetic data result in one operating cycle; and data transfer means connected between said plurality of pipeline-type arithmetic logic units and said plurality of vector registers for transferring data between said pipeline-type arithmetic logic units and said vector registers under control of said timing signal.
Claim 6
Dependent Claim 172 words
A vector processor for executing vector instructions comprising: timing means for providing a timing signal 2n times during each of a plurality of successive operation cycles (where n is an integer not less than one); a plurality of vector registers, each connected to receive said timing signals and including vector storage means having a plurality of addressable storage locations and read/write control means responsive to said timing signal for executing 2n accessing operations in succession with respect to the storage locations of said vector storage means of a selected vector register during each operation cycle; a plurality of pipeline-type arithmetic logic units, each connected to receive said timing signals and data from selected ones of said plurality of vector registers in an operating cycle for providing an arithmetic data result in one operating cycle; and data transfer means connected between said plurality of pipeline-type arithmetic logic units and said plurality of vector registers for transferring data between said pipeline-type arithmetic logic units and said vector registers under control of said timing signal.
Claim 7
Dependent Claim 88 words
A vector processor according to claim 4, wherein said read/write control means in each of said plurality of vector registers includes n write data registers, n read data registers, n read address registers, n write address registers, and selector means responsive to said timing signal for selectively connecting said write data registers, said read address registers and said write address registers to said vector storage means to sequentially access said addressable storage location so that n write operations and n read operations are performed during one operating cycle.
Claim 8
Dependent Claim 60 words
A vector processor according to claim 5, wherein said selector means includes a first selector connecting said write data registers to said vector storage means, a second selector connecting said read address registers and said write address registers to said vector storage means, and an n bit counter responsive to said timing signal for controlling said first and second selectors.
Claim 9
Dependent Claim 28 words
A vector processor according to claim 4, wherein n=2, and each of said vector registers is connected to said data transfer means by way of two signal lines.
Claim 10
Dependent Claim 47 words
A vector processor according to claim 4, wherein said data transfer means includes a first selector connecting said outputs of said vector registers to inputs of said arithmetic logic units, and a second selector connecting outputs of said arithmetic logic units to inputs of said vector registers.
Claim 11
Dependent Claim 119 words
A vector processor according to claim 4, wherein said read/write control means in each of said plurality of vector register includes n write data registers, n read data registers, n read address registers, n write address registers, (where n is an integer not less than 2) and selector means responsive to said timing signals for selectively connecting said write data registers, said read address registers and said write address registers to said vector storage means to sequentially access a plurality of addressable storage locations of a selected vector storage means so that both a data write operation and a data read operation is effected with respect to each of said sequentially accessed addressable storage locations in one operating cycle.
Claim 12
Dependent Claim 172 words
A vector processor for executing vector instructions comprising: timing means for providing a timing signal a plurality of times during each of a plurality of successive operation cycles; a plurality of vector registers, each connected to receive said timing signals and each including vector storage means having a plurality of addressable storage locations and read/write control means responsive to said timing signals for executing a data write operation and a data read operation in succession with respect to the same storage location of a selected one of said vector storage means during each operation cycle; a plurality of pipeline-type arithmetic logic units, each connected to receive said timing signals and data from selected ones of said plurality of vector registers in an operating cycle for providing an arithmetic data result in one operating cycle; and data transfer means connected between said plurality of pipeline-type arithmetic logic units and said plurality of vector registers for transferring data between said pipeline-type arithmetic logic units and said vector registers under control of said timing signal.
Claim 13
Dependent Claim 172 words
A vector processor for executing vector instructions comprising: timing means for providing a timing signal a plurality of times during each of a plurality of successive operation cycles; a plurality of vector registers, each connected to receive said timing signals and each including vector storage means having a plurality of addressable storage locations and read/write control means responsive to said timing signals for executing a data write operation and a data read operation in succession with respect to the same storage location of a selected one of said vector storage means during each operation cycle; a plurality of pipeline-type arithmetic logic units, each connected to receive said timing signals and data from selected ones of said plurality of vector registers in an operating cycle for providing an arithmetic data result in one operating cycle; and data transfer means connected between said plurality of pipeline-type arithmetic logic units and said plurality of vector registers for transferring data between said pipeline-type arithmetic logic units and said vector registers under control of said timing signal.

References Cited (Backward Citations)

US3541516A Cited
US4127899A Cited
US4270181A Cited
US4193127A Cited
US4302818A Cited
US4400768A Cited
US4447891A Cited
US4435765A Cited
US4525796A Cited
US4490786A Cited
US4489381A Cited

USPTO Annuity & Maintenance Windows

Annuity Stage Window Opens Payment Deadline Surcharge Period Until Estimated Fee Status
3.5 Years (Year 4) 1991-02-15 1991-08-17 1992-02-16 $2000.00 Window Closed / Lapsed
7.5 Years (Year 8) 1995-02-15 1995-08-17 1996-02-16 $3760.00 Window Closed / Lapsed
11.5 Years (Year 12) 1999-02-15 1999-08-17 2000-02-16 $7700.00 Window Closed / Lapsed

Assigned Categories & Tags

No categories assigned to this patent yet.

Assign Category Tag