WO2018044332A1 - Method and device for determining branch history for branch prediction - Google Patents
Method and device for determining branch history for branch prediction Download PDFInfo
- Publication number
- WO2018044332A1 WO2018044332A1 PCT/US2016/051040 US2016051040W WO2018044332A1 WO 2018044332 A1 WO2018044332 A1 WO 2018044332A1 US 2016051040 W US2016051040 W US 2016051040W WO 2018044332 A1 WO2018044332 A1 WO 2018044332A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- branch
- bits
- current global
- branch history
- history
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Ceased
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/30—Arrangements for executing machine instructions, e.g. instruction decode
- G06F9/38—Concurrent instruction execution, e.g. pipeline or look ahead
- G06F9/3836—Instruction issuing, e.g. dynamic instruction scheduling or out of order instruction execution
- G06F9/3842—Speculative instruction execution
- G06F9/3844—Speculative instruction execution using dynamic branch prediction, e.g. using branch history tables
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F12/00—Accessing, addressing or allocating within memory systems or architectures
- G06F12/02—Addressing or allocation; Relocation
- G06F12/08—Addressing or allocation; Relocation in hierarchically structured memory systems, e.g. virtual memory systems
- G06F12/0802—Addressing of a memory level in which the access to the desired data or data block requires associative addressing means, e.g. caches
- G06F12/0864—Addressing of a memory level in which the access to the desired data or data block requires associative addressing means, e.g. caches using pseudo-associative means, e.g. set-associative or hashing
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/30—Arrangements for executing machine instructions, e.g. instruction decode
- G06F9/38—Concurrent instruction execution, e.g. pipeline or look ahead
- G06F9/3802—Instruction prefetching
- G06F9/3804—Instruction prefetching for branches, e.g. hedging, branch folding
- G06F9/3806—Instruction prefetching for branches, e.g. hedging, branch folding using address prediction, e.g. return stack, branch history buffer
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/30—Arrangements for executing machine instructions, e.g. instruction decode
- G06F9/38—Concurrent instruction execution, e.g. pipeline or look ahead
- G06F9/3802—Instruction prefetching
- G06F9/3814—Implementation provisions of instruction buffers, e.g. prefetch buffer; banks
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/30—Arrangements for executing machine instructions, e.g. instruction decode
- G06F9/38—Concurrent instruction execution, e.g. pipeline or look ahead
- G06F9/3836—Instruction issuing, e.g. dynamic instruction scheduling or out of order instruction execution
- G06F9/3842—Speculative instruction execution
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/30—Arrangements for executing machine instructions, e.g. instruction decode
- G06F9/38—Concurrent instruction execution, e.g. pipeline or look ahead
- G06F9/3836—Instruction issuing, e.g. dynamic instruction scheduling or out of order instruction execution
- G06F9/3842—Speculative instruction execution
- G06F9/3848—Speculative instruction execution using hybrid branch prediction, e.g. selection between prediction techniques
Definitions
- a branch predictor is a digital circuit that tries to guess which way a branch, such as an if-then-else structure in a program or code, will go before this is known for sure.
- a branch predictor implements a method of branch prediction. Branch prediction improves the flow in an instruction pipeline and plays a role in achieving high effective performance in many microprocessor architectures.
- Branch prediction is based on branch history, that is, a record of how often a given branch is taken and not taken.
- branch history that is, a record of how often a given branch is taken and not taken.
- branch histories of multiple branches are stored in one set of bits.
- FIG. 2 shows a traditional type of processing device and method used to update a global branch history.
- Circuitry determines whether or not a branch has been found (step 225). When a branch has not been found, a current global history is unchanged (step 220). When a branch is found, the current global branch history is shifted by one bit (step 215; shown as a left shift) and the oldest history bit (bit 3) is lost. The rightmost bit in the current global branch history is replaced by a bit indicating whether or not the branch has been taken (step 230) - for example, a 0 for not taken and a 1 for taken. In this manner, the current global branch history is updated. The current branch history is used in a branch predictor to predict whether a branch will be taken or not taken in the future.
- FIG. 1 is a block diagram of an example device in which one or more disclosed aspects may be implemented
- FIG. 2 shows a processing device configured to update a global branch history
- FIG 3 shows another processing device configured to update a global branch history
- FIG 4 shows another processing device configured to update a global branch history
- FIG 5 shows a method for determining and updating a global branch history
- FIG 6 shows another method for determining and updating a global branch history.
- the disclosed method and device improves timing and avoids loss of previous branch predictor training when a new branch is discovered.
- the disclosed method and device also provide benefit of being able to differentiate between identical taken/not taken (T/NT) patterns at different code locations. If the computer code being executed contains discovered NT branches, the disclosed method and device is able to track longer sequences than traditional global history updating methods and devices.
- the method includes shifting first bits of a branch signature into a current global branch history and performing a bitwise exclusive-or (XOR) function on second bits of the branch signature and shifted bits of the current global branch history.
- the second bits are distinct from the first bits.
- the branch signature includes bits that identify a particular branch whose branch history is being updated.
- the shifting includes discarding one or more of the oldest bits in the current global branch history.
- the shifting and the performing of the bitwise XOR function are executed when a branch identified by the branch signature is taken.
- the current global branch history is not changed when a branch identified by the branch signature is not taken.
- the method also includes shifting one or more newest bits in the current global branch history by a first number of places, shifting one or more oldest bits in the current global branch history by a second number of places different from the first number of places, and performing a bitwise XOR function between bits shifted by the second number of places and newer bits in the current global branch history.
- These parts of the method are executed when a branch identified by the branch signature is taken, and the current global branch history does not change when a branch identified by the branch signature is not taken.
- the second number of places is less than the first number of places (e.g., the first number of places is three and the second number of places is one).
- the method is implemented such that the updated current global branch history is included in a first level of a two-level adaptive predictor.
- the processing device configured to determine and update global branch history for branch prediction, includes shift logic configured to shift bits representing a current global branch history, a register configured to store the current global branch history, decision circuitry configured to determine whether or not a branch is taken, and a first plurality of exclusive-or (XOR) gates.
- shift logic configured to shift bits representing a current global branch history
- a register configured to store the current global branch history
- decision circuitry configured to determine whether or not a branch is taken
- XOR exclusive-or
- the determining and updating of global branch history include the shift logic shifting bits representing a current global branch history and storing one or more of the shifted bits in the register, the register receiving and storing first bits of a branch signature, the first plurality of XOR gates performing a bitwise XOR function on second bits of the branch signature and shifted bits of the current global branch history, and storing a result of the bitwise XOR function in the register.
- the processing device is configured to determine and update global branch history when the decision circuitry determines that a branch identified by the branch signature has been taken. In an implementation, the processing device is configured to not change the current global branch history when the decision circuitry determines that a branch identified by the branch signature is not taken.
- the processing device is further configured to execute compression of branch history information and branch signature information.
- the processing device is configured to shift one or more of the newest bits in the current global branch history by a first number of places, and shift one or more of the oldest bits in the current global branch history by a second number of places different from the first number of places.
- the processing device further includes a second plurality of XOR gates configured to perform a bitwise XOR function between bits shifted by the second number of places and newer bits in the current global branch history.
- This implementation of the processing device is also configured to execute compression of branch history information and branch signature information when the decision circuitry determines that a branch identified by the branch signature has been taken. The processing device does not change the current global branch history when the decision circuitry determines that a branch identified by the branch signature is not taken.
- the second number of places may be less than the first number of places.
- the first number of places may be three and the second number of places may be one.
- Another particular implementation may include multiple levels of compression. As one example, newer bits in the current global branch history may be shifted by three places, older bits by two places, and still older bits by one place. For effective compression, the amount of shifting tends to decrease as the historical age of the bit increases.
- the processing device includes a two-level adaptive predictor apparatus configured to receive the updated current global branch history in a first level.
- Figure 1 is a block diagram of an example device 100 in which one or more disclosed embodiments are implemented.
- the device 100 include a computer, a gaming device, a handheld device, a set-top box, a television, a mobile phone, or a tablet computer.
- the device 100 includes a processor 102, a memory 104, a storage 106, one or more input devices 108, and one or more output devices 110.
- the device 100 includes an input driver 112 and an output driver 114. It is understood that the device 100 may include additional components not shown in Figure 1.
- the processor 102 includes a central processing unit (CPU), a graphics processing unit (GPU), a CPU and GPU located on the same die, or one or more processor cores, wherein each processor core may be a CPU or a GPU.
- the memory 104 is located on the same die as the processor 102, or the memory 104 is located separately from the processor 102.
- the memory 104 includes a volatile or nonvolatile memory, for example, random access memory (RAM), dynamic RAM, or a cache.
- the storage 106 includes a fixed or removable storage, for example, a hard disk drive, a solid state drive, an optical disk, or a flash drive.
- the input devices 108 include a keyboard, a keypad, a touch screen, a touch pad, a detector, a microphone, an accelerometer, a gyroscope, a biometric scanner, or a network connection (e.g., a wireless local area network card for transmission and/or reception of wireless IEEE 802 signals).
- the output devices 110 include a display, a speaker, a printer, a haptic feedback device, one or more lights, an antenna, or a network connection (e.g., a wireless local area network card for transmission and/or reception of wireless IEEE 802 signals).
- the input driver 112 communicates with the processor 102 and the input devices 108, and permits the processor 102 to receive input from the input devices 108.
- the output driver 114 communicates with the processor 102 and the output devices 110, and permits the processor 102 to send output to the output devices 110. It is noted that the input driver 112 and the output driver 114 are optional components, and that the device 100 will operate in the same manner if the input driver 112 and the output driver 114 are not present.
- FIG. 3 shows an example implementation of an electronic processing device 300 configured to determine and update global branch history for branch prediction. Specifics of electronic processing device 300 such as exact numbers of bits are not to be considered as limiting.
- a current global branch history is stored as a bit sequence in a register
- the current global branch history is shown as having eight bits, but this number is not to be considered hmiting.
- decision circuitry 325 determines whether the branch has been taken or not taken. If the branch has not been taken, the current global history is not changed 320 and is re-stored in register 310. If the branch has been taken, the current global branch history is updated, as in the following example, which is not to be considered limiting.
- the updating begins with shift logic 315 shifting bits representing a current global branch history, and storing the shifted bits in register 310. As shown in FIG. 3, for example, bits 0 through 5 are shifted to the left by two places. In this process, older bits 6 and 7 of this branch history are discarded. One or more of the shifted bits, such as bits 2-5, are stored directly into register 310. Register 310 receives and stores first bits S[l] and S[0] of a branch signature, which identifies a particular branch whose branch history is being updated.
- a branch signature could be as simple as the least significant bits of a linear address, or it could be as complicated as (a first linear address portion) XOR (a second linear address portion) XOR (bits representing branch type) XOR (bits representing other attributes of the branch). If a 4 bit branch signature is needed, for example, the 12 least significant bits of a linear address (LINAD) can be XOR'ed with four bits representing branch type, as follows:
- XOR gates 330 performing a bitwise XOR function on second bits of the branch signature - such as S[3] and S[2] - and shifted bits of the current global branch history, such as bits 0 and 1. Bits resulting from this performance of bitwise XOR function are stored in register 310, thereby updating the current global branch history stored in register 310.
- FIG. 4 shows an alternative example implementation of an electronic processing device 400 configured to determine and update global branch history for branch prediction. Specifics of electronic processing device 400 such as exact numbers of bits are not to be considered as limiting.
- the implementation of FIG. 4 includes compression of branch history information and branch signature information through variable shifting and additional XOR functions.
- a current global branch history is stored as a bit sequence in a register 410.
- the current global branch history is shown as having 19 bits, but this number is not to be considered hmiting.
- the current global history is not changed 420 and is re-stored in register 410. If the branch has been taken, the current global branch history is updated, as in the following example, which is not to be considered limiting.
- the updating begins with shift logic 415 shifting bits representing a current global branch history, and storing the shifted bits in register 410.
- one or more of the newest bits of the current global branch history are shifted by a first number of places, and one or more of the oldest bits are shifted by a second number of places, the first number of places and the second number of places being not equal. In particular, the second number of places is less than the first number of places.
- bits 0 through 7 are shifted to the left by three places, while bits 10 through 16 are shifted to the left by one place.
- One or more of the shifted bits are stored directly into register 410. In the example of FIG. 4, these bits are 3-7, 15, 16, and 17. In this process, oldest bit 18 of this branch history is discarded.
- register 410 receives and stores first bits S[2], S[l] and S[0] of a branch signature. Similar to FIG. 3, updating of current global branch history continues with a first plurality of XOR gates 430 performing a bitwise XOR function on second bits of the branch signature - S[5], S[4] and S[3] in the example of FIG. 4 - and shifted bits of the current global branch history, bits 0, 1, and 2 in the example of FIG. 4. Bits resulting from this performance of the bitwise XOR function are stored in register 410 as part of the updated current branch history. [0039] In contrast to FIG. 3, the device of FIG. 4 performs a compression using a second plurality of XOR gates 433.
- the XOR gates 433 are configured to perform a bitwise XOR function between current global branch history bits shifted by the second number of places and newer bits of the current global branch history. Bits resulting from this performance of bitwise XOR function are stored in register 410 to complete updating of current global branch history.
- the following XOR functions are performed: 10 XOR 5, 11 XOR 6, 12 XOR 7, 13 XOR 8, and 14 XOR 9.
- global branch history stored in register 410 is passed to a branch predictor 435 that predicts a likelihood that a branch will be taken using the current global branch history.
- the branch predictor 435 is a two-level adaptive predictor apparatus that is configured to receive updated current global branch history in its first level. Alternative implementations may include other types of branch predictors.
- FIG. 5 shows a method 500 of determining global branch history for branch prediction. The exact order of steps shown is not necessarily to be construed as limiting.
- a branch is encountered (step 510).
- a determination is made whether or not this branch is taken (step 515).
- a current global branch history is not changed (step 520) and the method returns to wait for a next encountered branch (step 510).
- the current global branch history is updated. Updating begins by shifting first bits of a branch signature into a current global branch history (step 525). The shifting includes discarding one or more of the oldest bits in the current global branch history. A bit-wise XOR function is performed on second bits of the branch signature and shifted bits of the current global branch history (step 530). The current global branch history is now updated and stored in a register, as described above. The method returns to wait for a next encountered branch (step 510).
- FIG. 6 shows another implementation 600 of a method of determining global branch history for branch prediction.
- the exact order of steps shown is not necessarily to be construed as hmiting.
- the implementation shown in FIG. 6 includes compression of branch history information and branch signature information through variable shifting and additional XOR functions.
- a branch is encountered (step 610) and a determination is made (step 615) whether or not this branch has been taken. If it has not been taken then, as in FIG. 5, the current global branch history is not changed (step 620) and the method returns to wait for the next branch to be encountered (step 610).
- step 625 When the branch is taken, one or more the newest bits in the current global branch history are shifted by a first number of places (step 625). First bits of a branch signature are shifted into the current global branch history (step 627). An exclusive XOR function is performed on second bits of the branch signature and bits of the current global branch history shifted by the first number of places (step 630).
- One or more of the oldest bits in the current global branch history are shifted by a second number of places different from the first number of places (step 635).
- the second number of places is less than the first number of places.
- the first number of places is three and the second number of places is one.
- a bitwise XOR function is performed between bits of the current global branch history shifted by the second number of places and newer bits in the current global branch history (step 640). One or more of the shifted bits of the current global branch history become bits of the updated branch history without further processing.
- the current global branch history is now updated.
- the updated global branch history is stored in a register, as described above.
- the method returns to wait for a next encountered branch (step 610).
- Suitable processors include, by way of example, a general purpose processor, a special purpose processor, a conventional processor, a digital signal processor (DSP), a plurality of microprocessors, one or more microprocessors in association with a DSP core, a controller, a microcontroller, Application Specific Integrated Circuits (ASICs), Field Programmable Gate Arrays (FPGAs) circuits, any other type of integrated circuit (IC), and/or a state machine.
- DSP digital signal processor
- ASICs Application Specific Integrated Circuits
- FPGAs Field Programmable Gate Arrays
- Such processors can be manufactured by configuring a manufacturing process using the results of processed hardware description language (HDL) instructions and other intermediary data including netlists (such instructions capable of being stored on a computer readable media). The results of such processing can be maskworks that are then used in a semiconductor manufacturing process to manufacture a processor which implements aspects of the embodiments.
- HDL hardware description language
- non-transitory computer-readable storage mediums include a read only memory (ROM), a random access memory (RAM), a register, cache memory, semiconductor memory devices, magnetic media such as internal hard disks and removable disks, magneto-optical media, and optical media such as CD-ROM disks, and digital versatile disks (DVDs).
- ROM read only memory
- RAM random access memory
- register cache memory
- semiconductor memory devices magnetic media such as internal hard disks and removable disks, magneto-optical media, and optical media such as CD-ROM disks, and digital versatile disks (DVDs).
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Software Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Advance Control (AREA)
Abstract
Description
Claims
Priority Applications (4)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2019532917A JP6850888B2 (en) | 2016-08-30 | 2016-09-09 | Methods and devices for determining branch history for branch prediction |
| CN201680088627.4A CN109643236B (en) | 2016-08-30 | 2016-09-09 | Method and apparatus for determining branch history for branch prediction |
| KR1020197005417A KR102719424B1 (en) | 2016-08-30 | 2016-09-09 | Method and device for determining branch history for branch prediction |
| EP16203526.5A EP3291082B1 (en) | 2016-08-30 | 2016-12-12 | Method and device for determining branch history for branch prediction |
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| US15/252,168 | 2016-08-30 | ||
| US15/252,168 US10698691B2 (en) | 2016-08-30 | 2016-08-30 | Method and device for determining branch prediction history for branch prediction by partially combining shifted branch prediction history with branch signature |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2018044332A1 true WO2018044332A1 (en) | 2018-03-08 |
Family
ID=61242679
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/US2016/051040 Ceased WO2018044332A1 (en) | 2016-08-30 | 2016-09-09 | Method and device for determining branch history for branch prediction |
Country Status (5)
| Country | Link |
|---|---|
| US (1) | US10698691B2 (en) |
| JP (1) | JP6850888B2 (en) |
| KR (1) | KR102719424B1 (en) |
| CN (1) | CN109643236B (en) |
| WO (1) | WO2018044332A1 (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10521236B2 (en) * | 2018-03-29 | 2019-12-31 | Intel Corporation | Branch prediction based on coherence operations in processors |
| US11086629B2 (en) * | 2018-11-09 | 2021-08-10 | Arm Limited | Misprediction of predicted taken branches in a data processing apparatus |
| CN113721985B (en) * | 2021-11-02 | 2022-02-08 | 超验信息科技(长沙)有限公司 | RISC-V vector register grouping setting method, device and electronic equipment |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6721877B1 (en) * | 2000-05-25 | 2004-04-13 | Advanced Micro Devices, Inc. | Branch predictor that selects between predictions based on stored prediction selector and branch predictor index generation |
| US7000096B1 (en) * | 2000-08-03 | 2006-02-14 | International Business Machines Corporation | Branch prediction circuits and methods and systems using the same |
| US20070150712A1 (en) * | 2005-12-22 | 2007-06-28 | Sun Microsystems, Inc. | Deferred branch history update scheme |
| US20090125707A1 (en) * | 2007-11-13 | 2009-05-14 | Applied Micro Circuits Corporation | System and method for speculative global history prediction updating |
| US20090198984A1 (en) * | 2008-01-31 | 2009-08-06 | Loschke Jon A | Global History Branch Prediction Updating Responsive to Taken Branches |
Family Cites Families (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7797521B2 (en) * | 2007-04-12 | 2010-09-14 | International Business Machines Corporation | Method, system, and computer program product for path-correlated indirect address predictions |
| US8782384B2 (en) | 2007-12-20 | 2014-07-15 | Advanced Micro Devices, Inc. | Branch history with polymorphic indirect branch information |
| US8181005B2 (en) * | 2008-09-05 | 2012-05-15 | Advanced Micro Devices, Inc. | Hybrid branch prediction device with sparse and dense prediction caches |
| CN101826002B (en) * | 2010-04-16 | 2012-12-26 | 浙江大学 | Hardware realization method of recording branch predictor |
| US9524166B2 (en) * | 2013-07-23 | 2016-12-20 | Globalfoundries Inc. | Tracking long GHV in high performance out-of-order superscalar processors |
| US9858081B2 (en) * | 2013-08-12 | 2018-01-02 | International Business Machines Corporation | Global branch prediction using branch and fetch group history |
| US10209992B2 (en) * | 2014-04-25 | 2019-02-19 | Avago Technologies International Sales Pte. Limited | System and method for branch prediction using two branch history tables and presetting a global branch history register |
| US9983878B2 (en) | 2014-05-15 | 2018-05-29 | International Business Machines Corporation | Branch prediction using multiple versions of history data |
-
2016
- 2016-08-30 US US15/252,168 patent/US10698691B2/en active Active
- 2016-09-09 KR KR1020197005417A patent/KR102719424B1/en active Active
- 2016-09-09 JP JP2019532917A patent/JP6850888B2/en active Active
- 2016-09-09 WO PCT/US2016/051040 patent/WO2018044332A1/en not_active Ceased
- 2016-09-09 CN CN201680088627.4A patent/CN109643236B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6721877B1 (en) * | 2000-05-25 | 2004-04-13 | Advanced Micro Devices, Inc. | Branch predictor that selects between predictions based on stored prediction selector and branch predictor index generation |
| US7000096B1 (en) * | 2000-08-03 | 2006-02-14 | International Business Machines Corporation | Branch prediction circuits and methods and systems using the same |
| US20070150712A1 (en) * | 2005-12-22 | 2007-06-28 | Sun Microsystems, Inc. | Deferred branch history update scheme |
| US20090125707A1 (en) * | 2007-11-13 | 2009-05-14 | Applied Micro Circuits Corporation | System and method for speculative global history prediction updating |
| US20090198984A1 (en) * | 2008-01-31 | 2009-08-06 | Loschke Jon A | Global History Branch Prediction Updating Responsive to Taken Branches |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109643236A (en) | 2019-04-16 |
| US10698691B2 (en) | 2020-06-30 |
| CN109643236B (en) | 2024-11-29 |
| KR102719424B1 (en) | 2024-10-21 |
| US20180060074A1 (en) | 2018-03-01 |
| JP2019526872A (en) | 2019-09-19 |
| JP6850888B2 (en) | 2021-03-31 |
| KR20190039159A (en) | 2019-04-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109643237B (en) | Branch target buffer compression | |
| CN105320519B (en) | Conditional branch prediction using long history | |
| KR101643675B1 (en) | Instruction processing circuits, and the convergence of conditional write instructions having opposite conditions in the associated processor systems, methods, and computer-readable media | |
| JP7523361B2 (en) | Low latency synchronization for the operation cache and instruction cache for fetching and decoding instructions | |
| US8788797B2 (en) | Combined level 1 and level 2 branch predictor | |
| KR20180127379A (en) | Providing load address predictions using address prediction tables based on load path history in processor-based systems | |
| US20210149676A1 (en) | Branch Prediction Method, Branch Prediction Unit and Processor Core | |
| US11860787B2 (en) | Cache miss predictor | |
| JP2005100403A (en) | Branch predictor and implementation method for reducing power consumption | |
| US10698691B2 (en) | Method and device for determining branch prediction history for branch prediction by partially combining shifted branch prediction history with branch signature | |
| WO2014025815A1 (en) | Fusing flag-producing and flag-consuming instructions in instruction processing circuits, and related processor systems, methods, and computer-readable media | |
| CN105607888A (en) | Data bit width conversion method and device | |
| US9858077B2 (en) | Issuing instructions to execution pipelines based on register-associated preferences, and related instruction processing circuits, processor systems, methods, and computer-readable media | |
| US9778934B2 (en) | Power efficient pattern history table fetch in branch predictor | |
| CN116505959A (en) | Decoding method and related equipment based on BCH code | |
| US20170083333A1 (en) | Branch target instruction cache (btic) to store a conditional branch instruction | |
| EP3291082B1 (en) | Method and device for determining branch history for branch prediction | |
| US20160054979A1 (en) | Data storage method, ternary inner product operation circuit, semiconductor device including the same, and ternary inner product arithmetic processing program | |
| CN105027074A (en) | Prime number generation | |
| US20220237164A1 (en) | System and method for multiplexer tree indexing | |
| US20120079240A1 (en) | Reduced-Level Shift Overflow Detection | |
| CN113227970B (en) | Instruction tightly coupled memory and instruction cache access prediction | |
| US10411731B1 (en) | Device and method of compressing data using tiered data compression | |
| US20120011343A1 (en) | Data processing apparatus and method | |
| KR101528249B1 (en) | Dynamic branch predictor using global branch history and embedded system including the dynamic branch predictor |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 16915403 Country of ref document: EP Kind code of ref document: A1 |
|
| ENP | Entry into the national phase |
Ref document number: 20197005417 Country of ref document: KR Kind code of ref document: A |
|
| ENP | Entry into the national phase |
Ref document number: 2019532917 Country of ref document: JP Kind code of ref document: A |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| 122 | Ep: pct application non-entry in european phase |
Ref document number: 16915403 Country of ref document: EP Kind code of ref document: A1 |