SlideShare ist ein Scribd-Unternehmen logo
1 von 40
Downloaden Sie, um offline zu lesen
IBM’S DEEP BLUE
CHESS
GRANDMASTER CHIPS
DESIGN PHILOSOPHY, GENERAL ARCHITECTURE,
AND PERFORMANCE OF THE CHESS CHIPS
 Ujitha Iroshan
 Nadeeshaan Gunasinghe
 Dulanga Sashika
Presenters
 Introduction
 History
 Origin of IBM’s Chess Grandmaster
 Task and Design Philosophy
Journey
 General Architecture
 Common chess machine Overview
 System Configuration
 Chip Overview
 Move Generator
 Evaluation Function
 Smart Move stack
 Search Control
 Performance
Journey Cont..
IBM Deep Blue World chess champion
Garry Kasparov in 1997
What is Deep Blue ?
Vs
Creating the first world champion class chess
computer is one of the oldest challenge in
computer science.
 Blitz chip won against grand master in
shortened time chess game in 1977
 Cray Blitz computer is defeated by
International master David Levy in shortened
time chess game in 1984
History
 PC programmed in Blitz won against 2nd rank chess
player in 1988 under faster time controls.
 Deep blue early version(Deep thought) was defeated
by the Champion Garry Kasparov in 1996 under
regular time control.
 Finally in 1997 Deep Blue won against the World
Champion Garry Kasparov under regular time control
History Ctd...
Winning a match against the human World
chess champion under regulation time control
The game had to play out no faster than three
minutes per move
Task description of Deep Blue
Regular time control
 Standard chess tournament time control
Faster time control (Shortened time)
 Limited time for entire match and for the each
moves
Task Description Ctd...
 Integrate level
 Integrate the maximally possible amount of
software-modifiable chess knowledge on to the
chess chip.
 Computational Speed
 Assign weights to positional features
Design Philosophy
Common chess machine Overview
 Follows the design outlined by Claud Shanon in
1949
 Chess 4.5 is the earliest and the first
successful chess machine design
 Today Chess Machines in one way or another
way, are the grandchildren of this Chess 4.5
General Architecture
Choosing the next move
 Shannon’s basic design uses a min-max tree
 The evaluation function is designed from the
computer player’s viewpoint
Common chess machine Overview Ctd...
 Assign numerical values for evaluation
function
■ +1- Computer Player Wins
■ -1- Computer Player Loses
■ 0- Drawn position
Choosing the next move Ctd..
Can regular Min-Max search be used in a
Chess Game?
Common chess machine Overview cont..
 New developments in Chess 4.5
 A hardware move generator
 Quiescence search (Capture Search) and
Alpha beta Pruning algorithm
Common chess machine Overview cont..
 1997 version of DB included 480 chess chips
 each chess chip could search 2 to 2.5 million
chess positions/s
 More software work could increase the speed.
But applied software work in chess knowledge
area
System Configuration
 Search goes parallelly in two levels
 One over IBM RS/6000 SP switching network
 The other over Micro Channel bus inside a
workstation node
 Partitioning the search provided a great deal
of design flexibility and increased the search
speed
 SW handled a 1% of the search, but the ⅔
search depth was reduced by SW
System Configuration
Block diagram of the Chess Chip
Chip Overview
The chip’s cycle time
lies between 40 and
50 nanoseconds in a
0.6-micron, three-
metal
layer, 5-V CMOS
process. The
average number
of cycles per position
searched is about
10;
power dissipation is
about one watt
 Each chess Chip works as a chess machine
 Why Chess chips rather than software
algorithm for the competing?
 Provided design possibilities unacceptable by SW
 Time-scaling factor dropped from instruction cycle
time to gate delay
 Only thing required was reexamine of the time
complexity of the competing algorithm
Chip Overview
Inside of a chess chip
Chess
Chip
Smart
Move
stack
Evaluation
Function
Search
Move
generator
 What is move generator?
 Used to decide the next best move of a
chess piece
 An 8x8 array of combinational logic
 Move generation is controlled by a
hardwired finite state machine
Move Generator
Why Deep Blue’s move generator special?
 Generate check and check evasion Moves
 Generates attacking moves
 Hardware pruning of irrelevant chess moves
Move Generator Ctd..
A chip cell
The combinational logic array is a silicon chess
board
Each has four major components
 Find victim transmitter
 Find Attacker Transmitter
 Receiver
 Distributed arbiter
 Four-bit piece register
Move Generator Ctd..
Find Victim and Find Attacker
Move Generator Ctd..
As described earlier, move generator computes
all the moves implicitly for a single move.
What if SW is used for this task?...
Move Generator Ctd..
Generating moves from a partially
generated position
 Need to mask the already searched moves
 Belle design used 2-bit disable-stack for
masking
 What is used in Deep Blue?
Move Generator Ctd..
Other moves generated by Move generator
 checking moves
 check-evasion moves
 Detect Hung Pieces
 Measures the forcefulness of checks
Move Generator Ctd..
Evaluation Function
What is Evaluation Function?
 Evaluation function is a function used by game-
playing programs to estimate the value or
goodness of a position in the minimax and related
algorithms.
Evaluation Function of “Deep Blue”
 Contain about 66,000 gates not including RAMs
and ROMs.
 All sub blocks of the evaluation function divided
into two sections
• Lower sub blocks provide fast evaluation
• Upper sub blocks provide slow evaluation
Fast Evaluation
The piece placement table
computes a position’s incremental
evaluation.
Uses three RAMs.
The Game Phase Control
 Contains a piece count register
 Maintain an XORed piece-location register
 The piece-count register addresses game phase
control RAMs
 The game phase control RAMs produce several
control values
 simply a bonus or penalty to be added to the
evaluation based on the material left on the
board.
 The most important game phase control value is
the king safety relevance
 tells the chess chip how to adjust the penalty for
bad pawn structure and the bonus for passed
pawns
The king-and-pawn array
mainly detects the “pawn can run” condition
The Endgame logic and ROMs
Mainly recognizes unusual endgame conditions
This is the single most complicated element
on the chip, occupying close to half of the
chip core.
Slow evaluation has a three-stage pipeline
 8 X 1 systolic array
 40-plus synchronous RAMs
 adder tree
Computes values for chess concepts such as,
 square control, pins, x-rays, king safety, pawn
structure, passed pawns, ray control, outposts,
pawn majority, rook on the 7th, blockade,
restraint and etc.
Slow evaluation
This doesn’t exist in the chess chip’s older
release.
Repetition detector
 Contains a 32-entry circular buffer of the last 32
plies of moves.
 Use hardware content addressable memory
algorithm
 Does this detects just simple repetition?
 Time complexity is O(n)
Smart move stack
implements a minimum-window alpha-beta
search algorithm
This search can only tell us whether the position
searched is better or worse than a single test
value.
Efficiency- wise, minimum-window-based search
seems about the same as the regular alpha-beta
search
Search Control
In early 1997, used a single chip running at
70% clock speed
A hardware bug reduced the chip to 7% to
14% of its regular speed
Deep Blue Jr. against the commercial chess
programs played 40 games and chess chip
score 95% against PC program
Performance
The 1997 version of Deep Blue only played six
games
Deep blue won the match by the score of 3.5 to
2.5
Kasparov is rated around 2815, which placed
Deep Blue’s performance at about 2875
Thank You
Questions?

Weitere ähnliche Inhalte

Was ist angesagt?

Blockchain and distributed ledgers
Blockchain and distributed ledgersBlockchain and distributed ledgers
Blockchain and distributed ledgersRobin Teigland
 
IRJET- A Decentralized Voting Application using Blockchain Technology
IRJET- 	 A Decentralized Voting Application using Blockchain TechnologyIRJET- 	 A Decentralized Voting Application using Blockchain Technology
IRJET- A Decentralized Voting Application using Blockchain TechnologyIRJET Journal
 
Blockchain 101 - public, tokenized blockchains
Blockchain 101 - public, tokenized blockchainsBlockchain 101 - public, tokenized blockchains
Blockchain 101 - public, tokenized blockchainsBrett Colbert
 
Decentralised ethereum blockchain voting application
Decentralised ethereum blockchain voting applicationDecentralised ethereum blockchain voting application
Decentralised ethereum blockchain voting applicationRakesh Ranjan
 
Game development history
Game development historyGame development history
Game development historylitoon dev
 
License Plate Recognition
License Plate RecognitionLicense Plate Recognition
License Plate RecognitionAmr Rashed
 
MPI DevCon Hsinchu City 2017: MIPI C-PHY/D-PHY Dual Mode Subsystem Performanc...
MPI DevCon Hsinchu City 2017: MIPI C-PHY/D-PHY Dual Mode Subsystem Performanc...MPI DevCon Hsinchu City 2017: MIPI C-PHY/D-PHY Dual Mode Subsystem Performanc...
MPI DevCon Hsinchu City 2017: MIPI C-PHY/D-PHY Dual Mode Subsystem Performanc...MIPI Alliance
 
Cricket technologies
Cricket technologiesCricket technologies
Cricket technologiesUsman Khalid
 
Consensus Algorithms.pptx
Consensus Algorithms.pptxConsensus Algorithms.pptx
Consensus Algorithms.pptxRajapriya82
 
Overview of Blockchain Consensus Mechanisms
Overview of Blockchain Consensus MechanismsOverview of Blockchain Consensus Mechanisms
Overview of Blockchain Consensus MechanismsJohannes Ahlmann
 
Gaming console technology 2017 ppt
Gaming console technology 2017 ppt Gaming console technology 2017 ppt
Gaming console technology 2017 ppt keshav kumar
 
Memory forensics
Memory forensicsMemory forensics
Memory forensicsSunil Kumar
 

Was ist angesagt? (20)

Blockchain and distributed ledgers
Blockchain and distributed ledgersBlockchain and distributed ledgers
Blockchain and distributed ledgers
 
Blockchain technology
Blockchain technologyBlockchain technology
Blockchain technology
 
IRJET- A Decentralized Voting Application using Blockchain Technology
IRJET- 	 A Decentralized Voting Application using Blockchain TechnologyIRJET- 	 A Decentralized Voting Application using Blockchain Technology
IRJET- A Decentralized Voting Application using Blockchain Technology
 
Blockchain
BlockchainBlockchain
Blockchain
 
Blockchain 101 - public, tokenized blockchains
Blockchain 101 - public, tokenized blockchainsBlockchain 101 - public, tokenized blockchains
Blockchain 101 - public, tokenized blockchains
 
Computer chess
Computer chessComputer chess
Computer chess
 
Decentralised ethereum blockchain voting application
Decentralised ethereum blockchain voting applicationDecentralised ethereum blockchain voting application
Decentralised ethereum blockchain voting application
 
Game development history
Game development historyGame development history
Game development history
 
License Plate Recognition
License Plate RecognitionLicense Plate Recognition
License Plate Recognition
 
MPI DevCon Hsinchu City 2017: MIPI C-PHY/D-PHY Dual Mode Subsystem Performanc...
MPI DevCon Hsinchu City 2017: MIPI C-PHY/D-PHY Dual Mode Subsystem Performanc...MPI DevCon Hsinchu City 2017: MIPI C-PHY/D-PHY Dual Mode Subsystem Performanc...
MPI DevCon Hsinchu City 2017: MIPI C-PHY/D-PHY Dual Mode Subsystem Performanc...
 
Cricket technologies
Cricket technologiesCricket technologies
Cricket technologies
 
Consensus Algorithms.pptx
Consensus Algorithms.pptxConsensus Algorithms.pptx
Consensus Algorithms.pptx
 
Connect (4|n)
Connect (4|n)Connect (4|n)
Connect (4|n)
 
CLOUD GAMING
CLOUD GAMING CLOUD GAMING
CLOUD GAMING
 
Cryptocurrency in real world
Cryptocurrency in real worldCryptocurrency in real world
Cryptocurrency in real world
 
Overview of Blockchain Consensus Mechanisms
Overview of Blockchain Consensus MechanismsOverview of Blockchain Consensus Mechanisms
Overview of Blockchain Consensus Mechanisms
 
Blockchain
Blockchain Blockchain
Blockchain
 
Gaming console technology 2017 ppt
Gaming console technology 2017 ppt Gaming console technology 2017 ppt
Gaming console technology 2017 ppt
 
Blockchain and Banking
Blockchain and BankingBlockchain and Banking
Blockchain and Banking
 
Memory forensics
Memory forensicsMemory forensics
Memory forensics
 

Ähnlich wie IBM Deep Blue's Chess Chip Design

Chess Engine Programming
Chess Engine ProgrammingChess Engine Programming
Chess Engine ProgrammingArno Huetter
 
Topic - 6 (Game Playing).ppt
Topic - 6 (Game Playing).pptTopic - 6 (Game Playing).ppt
Topic - 6 (Game Playing).pptSabrinaShanta2
 
Artificial Intelligence gaming techniques
Artificial Intelligence gaming techniquesArtificial Intelligence gaming techniques
Artificial Intelligence gaming techniquesSomnathMore3
 
Alpha go 16110226_김영우
Alpha go 16110226_김영우Alpha go 16110226_김영우
Alpha go 16110226_김영우영우 김
 
International journal of engineering issues vol 2015 - no 2 - paper1
International journal of engineering issues   vol 2015 - no 2 - paper1International journal of engineering issues   vol 2015 - no 2 - paper1
International journal of engineering issues vol 2015 - no 2 - paper1sophiabelthome
 
Better Chess through Genetic Programming
Better Chess through Genetic ProgrammingBetter Chess through Genetic Programming
Better Chess through Genetic Programmingcaswenson
 
Lecture 02: Layered Architecture of Game Engine | GAMES104 - Modern Game Engi...
Lecture 02: Layered Architecture of Game Engine | GAMES104 - Modern Game Engi...Lecture 02: Layered Architecture of Game Engine | GAMES104 - Modern Game Engi...
Lecture 02: Layered Architecture of Game Engine | GAMES104 - Modern Game Engi...Piccolo Engine
 
Architectural Analysis of Game Machines
Architectural Analysis of Game MachinesArchitectural Analysis of Game Machines
Architectural Analysis of Game MachinesPraveen AP
 
CptS 440/ 540 AI.pptx
CptS 440/ 540 AI.pptxCptS 440/ 540 AI.pptx
CptS 440/ 540 AI.pptxDrDejaVu2
 
Lunch session: Quantum Computing
Lunch session: Quantum ComputingLunch session: Quantum Computing
Lunch session: Quantum ComputingRolf Huisman
 
Visibility Optimization for Games
Visibility Optimization for GamesVisibility Optimization for Games
Visibility Optimization for GamesUmbra
 
Visibility Optimization for Games
Visibility Optimization for GamesVisibility Optimization for Games
Visibility Optimization for GamesSampo Lappalainen
 

Ähnlich wie IBM Deep Blue's Chess Chip Design (20)

Chess Engine
Chess EngineChess Engine
Chess Engine
 
Computer Chess 2004
Computer Chess 2004Computer Chess 2004
Computer Chess 2004
 
chess-180927202627.pdf
chess-180927202627.pdfchess-180927202627.pdf
chess-180927202627.pdf
 
Chess Engine Programming
Chess Engine ProgrammingChess Engine Programming
Chess Engine Programming
 
Topic - 6 (Game Playing).ppt
Topic - 6 (Game Playing).pptTopic - 6 (Game Playing).ppt
Topic - 6 (Game Playing).ppt
 
Artificial Intelligence gaming techniques
Artificial Intelligence gaming techniquesArtificial Intelligence gaming techniques
Artificial Intelligence gaming techniques
 
AlphaGo and AlphaGo Zero
AlphaGo and AlphaGo ZeroAlphaGo and AlphaGo Zero
AlphaGo and AlphaGo Zero
 
Alpha go 16110226_김영우
Alpha go 16110226_김영우Alpha go 16110226_김영우
Alpha go 16110226_김영우
 
International journal of engineering issues vol 2015 - no 2 - paper1
International journal of engineering issues   vol 2015 - no 2 - paper1International journal of engineering issues   vol 2015 - no 2 - paper1
International journal of engineering issues vol 2015 - no 2 - paper1
 
Games.4
Games.4Games.4
Games.4
 
Better Chess through Genetic Programming
Better Chess through Genetic ProgrammingBetter Chess through Genetic Programming
Better Chess through Genetic Programming
 
Lecture 02: Layered Architecture of Game Engine | GAMES104 - Modern Game Engi...
Lecture 02: Layered Architecture of Game Engine | GAMES104 - Modern Game Engi...Lecture 02: Layered Architecture of Game Engine | GAMES104 - Modern Game Engi...
Lecture 02: Layered Architecture of Game Engine | GAMES104 - Modern Game Engi...
 
Architectural Analysis of Game Machines
Architectural Analysis of Game MachinesArchitectural Analysis of Game Machines
Architectural Analysis of Game Machines
 
CptS 440/ 540 AI.pptx
CptS 440/ 540 AI.pptxCptS 440/ 540 AI.pptx
CptS 440/ 540 AI.pptx
 
Lunch session: Quantum Computing
Lunch session: Quantum ComputingLunch session: Quantum Computing
Lunch session: Quantum Computing
 
l3.pptx
l3.pptxl3.pptx
l3.pptx
 
484-7.ppt
484-7.ppt484-7.ppt
484-7.ppt
 
Visibility Optimization for Games
Visibility Optimization for GamesVisibility Optimization for Games
Visibility Optimization for Games
 
Visibility Optimization for Games
Visibility Optimization for GamesVisibility Optimization for Games
Visibility Optimization for Games
 
Videogame Optimization
Videogame OptimizationVideogame Optimization
Videogame Optimization
 

Kürzlich hochgeladen

ISYU TUNGKOL SA SEKSWLADIDA (ISSUE ABOUT SEXUALITY
ISYU TUNGKOL SA SEKSWLADIDA (ISSUE ABOUT SEXUALITYISYU TUNGKOL SA SEKSWLADIDA (ISSUE ABOUT SEXUALITY
ISYU TUNGKOL SA SEKSWLADIDA (ISSUE ABOUT SEXUALITYKayeClaireEstoconing
 
Choosing the Right CBSE School A Comprehensive Guide for Parents
Choosing the Right CBSE School A Comprehensive Guide for ParentsChoosing the Right CBSE School A Comprehensive Guide for Parents
Choosing the Right CBSE School A Comprehensive Guide for Parentsnavabharathschool99
 
HỌC TỐT TIẾNG ANH 11 THEO CHƯƠNG TRÌNH GLOBAL SUCCESS ĐÁP ÁN CHI TIẾT - CẢ NĂ...
HỌC TỐT TIẾNG ANH 11 THEO CHƯƠNG TRÌNH GLOBAL SUCCESS ĐÁP ÁN CHI TIẾT - CẢ NĂ...HỌC TỐT TIẾNG ANH 11 THEO CHƯƠNG TRÌNH GLOBAL SUCCESS ĐÁP ÁN CHI TIẾT - CẢ NĂ...
HỌC TỐT TIẾNG ANH 11 THEO CHƯƠNG TRÌNH GLOBAL SUCCESS ĐÁP ÁN CHI TIẾT - CẢ NĂ...Nguyen Thanh Tu Collection
 
ECONOMIC CONTEXT - PAPER 1 Q3: NEWSPAPERS.pptx
ECONOMIC CONTEXT - PAPER 1 Q3: NEWSPAPERS.pptxECONOMIC CONTEXT - PAPER 1 Q3: NEWSPAPERS.pptx
ECONOMIC CONTEXT - PAPER 1 Q3: NEWSPAPERS.pptxiammrhaywood
 
Science 7 Quarter 4 Module 2: Natural Resources.pptx
Science 7 Quarter 4 Module 2: Natural Resources.pptxScience 7 Quarter 4 Module 2: Natural Resources.pptx
Science 7 Quarter 4 Module 2: Natural Resources.pptxMaryGraceBautista27
 
How to do quick user assign in kanban in Odoo 17 ERP
How to do quick user assign in kanban in Odoo 17 ERPHow to do quick user assign in kanban in Odoo 17 ERP
How to do quick user assign in kanban in Odoo 17 ERPCeline George
 
Field Attribute Index Feature in Odoo 17
Field Attribute Index Feature in Odoo 17Field Attribute Index Feature in Odoo 17
Field Attribute Index Feature in Odoo 17Celine George
 
How to Add Barcode on PDF Report in Odoo 17
How to Add Barcode on PDF Report in Odoo 17How to Add Barcode on PDF Report in Odoo 17
How to Add Barcode on PDF Report in Odoo 17Celine George
 
ACC 2024 Chronicles. Cardiology. Exam.pdf
ACC 2024 Chronicles. Cardiology. Exam.pdfACC 2024 Chronicles. Cardiology. Exam.pdf
ACC 2024 Chronicles. Cardiology. Exam.pdfSpandanaRallapalli
 
Culture Uniformity or Diversity IN SOCIOLOGY.pptx
Culture Uniformity or Diversity IN SOCIOLOGY.pptxCulture Uniformity or Diversity IN SOCIOLOGY.pptx
Culture Uniformity or Diversity IN SOCIOLOGY.pptxPoojaSen20
 
Procuring digital preservation CAN be quick and painless with our new dynamic...
Procuring digital preservation CAN be quick and painless with our new dynamic...Procuring digital preservation CAN be quick and painless with our new dynamic...
Procuring digital preservation CAN be quick and painless with our new dynamic...Jisc
 
4.18.24 Movement Legacies, Reflection, and Review.pptx
4.18.24 Movement Legacies, Reflection, and Review.pptx4.18.24 Movement Legacies, Reflection, and Review.pptx
4.18.24 Movement Legacies, Reflection, and Review.pptxmary850239
 
Proudly South Africa powerpoint Thorisha.pptx
Proudly South Africa powerpoint Thorisha.pptxProudly South Africa powerpoint Thorisha.pptx
Proudly South Africa powerpoint Thorisha.pptxthorishapillay1
 
What is Model Inheritance in Odoo 17 ERP
What is Model Inheritance in Odoo 17 ERPWhat is Model Inheritance in Odoo 17 ERP
What is Model Inheritance in Odoo 17 ERPCeline George
 
ENGLISH 7_Q4_LESSON 2_ Employing a Variety of Strategies for Effective Interp...
ENGLISH 7_Q4_LESSON 2_ Employing a Variety of Strategies for Effective Interp...ENGLISH 7_Q4_LESSON 2_ Employing a Variety of Strategies for Effective Interp...
ENGLISH 7_Q4_LESSON 2_ Employing a Variety of Strategies for Effective Interp...JhezDiaz1
 
Visit to a blind student's school🧑‍🦯🧑‍🦯(community medicine)
Visit to a blind student's school🧑‍🦯🧑‍🦯(community medicine)Visit to a blind student's school🧑‍🦯🧑‍🦯(community medicine)
Visit to a blind student's school🧑‍🦯🧑‍🦯(community medicine)lakshayb543
 
Influencing policy (training slides from Fast Track Impact)
Influencing policy (training slides from Fast Track Impact)Influencing policy (training slides from Fast Track Impact)
Influencing policy (training slides from Fast Track Impact)Mark Reed
 
AUDIENCE THEORY -CULTIVATION THEORY - GERBNER.pptx
AUDIENCE THEORY -CULTIVATION THEORY -  GERBNER.pptxAUDIENCE THEORY -CULTIVATION THEORY -  GERBNER.pptx
AUDIENCE THEORY -CULTIVATION THEORY - GERBNER.pptxiammrhaywood
 

Kürzlich hochgeladen (20)

ISYU TUNGKOL SA SEKSWLADIDA (ISSUE ABOUT SEXUALITY
ISYU TUNGKOL SA SEKSWLADIDA (ISSUE ABOUT SEXUALITYISYU TUNGKOL SA SEKSWLADIDA (ISSUE ABOUT SEXUALITY
ISYU TUNGKOL SA SEKSWLADIDA (ISSUE ABOUT SEXUALITY
 
Choosing the Right CBSE School A Comprehensive Guide for Parents
Choosing the Right CBSE School A Comprehensive Guide for ParentsChoosing the Right CBSE School A Comprehensive Guide for Parents
Choosing the Right CBSE School A Comprehensive Guide for Parents
 
HỌC TỐT TIẾNG ANH 11 THEO CHƯƠNG TRÌNH GLOBAL SUCCESS ĐÁP ÁN CHI TIẾT - CẢ NĂ...
HỌC TỐT TIẾNG ANH 11 THEO CHƯƠNG TRÌNH GLOBAL SUCCESS ĐÁP ÁN CHI TIẾT - CẢ NĂ...HỌC TỐT TIẾNG ANH 11 THEO CHƯƠNG TRÌNH GLOBAL SUCCESS ĐÁP ÁN CHI TIẾT - CẢ NĂ...
HỌC TỐT TIẾNG ANH 11 THEO CHƯƠNG TRÌNH GLOBAL SUCCESS ĐÁP ÁN CHI TIẾT - CẢ NĂ...
 
ECONOMIC CONTEXT - PAPER 1 Q3: NEWSPAPERS.pptx
ECONOMIC CONTEXT - PAPER 1 Q3: NEWSPAPERS.pptxECONOMIC CONTEXT - PAPER 1 Q3: NEWSPAPERS.pptx
ECONOMIC CONTEXT - PAPER 1 Q3: NEWSPAPERS.pptx
 
Science 7 Quarter 4 Module 2: Natural Resources.pptx
Science 7 Quarter 4 Module 2: Natural Resources.pptxScience 7 Quarter 4 Module 2: Natural Resources.pptx
Science 7 Quarter 4 Module 2: Natural Resources.pptx
 
How to do quick user assign in kanban in Odoo 17 ERP
How to do quick user assign in kanban in Odoo 17 ERPHow to do quick user assign in kanban in Odoo 17 ERP
How to do quick user assign in kanban in Odoo 17 ERP
 
Field Attribute Index Feature in Odoo 17
Field Attribute Index Feature in Odoo 17Field Attribute Index Feature in Odoo 17
Field Attribute Index Feature in Odoo 17
 
How to Add Barcode on PDF Report in Odoo 17
How to Add Barcode on PDF Report in Odoo 17How to Add Barcode on PDF Report in Odoo 17
How to Add Barcode on PDF Report in Odoo 17
 
ACC 2024 Chronicles. Cardiology. Exam.pdf
ACC 2024 Chronicles. Cardiology. Exam.pdfACC 2024 Chronicles. Cardiology. Exam.pdf
ACC 2024 Chronicles. Cardiology. Exam.pdf
 
Culture Uniformity or Diversity IN SOCIOLOGY.pptx
Culture Uniformity or Diversity IN SOCIOLOGY.pptxCulture Uniformity or Diversity IN SOCIOLOGY.pptx
Culture Uniformity or Diversity IN SOCIOLOGY.pptx
 
Procuring digital preservation CAN be quick and painless with our new dynamic...
Procuring digital preservation CAN be quick and painless with our new dynamic...Procuring digital preservation CAN be quick and painless with our new dynamic...
Procuring digital preservation CAN be quick and painless with our new dynamic...
 
4.18.24 Movement Legacies, Reflection, and Review.pptx
4.18.24 Movement Legacies, Reflection, and Review.pptx4.18.24 Movement Legacies, Reflection, and Review.pptx
4.18.24 Movement Legacies, Reflection, and Review.pptx
 
Proudly South Africa powerpoint Thorisha.pptx
Proudly South Africa powerpoint Thorisha.pptxProudly South Africa powerpoint Thorisha.pptx
Proudly South Africa powerpoint Thorisha.pptx
 
What is Model Inheritance in Odoo 17 ERP
What is Model Inheritance in Odoo 17 ERPWhat is Model Inheritance in Odoo 17 ERP
What is Model Inheritance in Odoo 17 ERP
 
ENGLISH 7_Q4_LESSON 2_ Employing a Variety of Strategies for Effective Interp...
ENGLISH 7_Q4_LESSON 2_ Employing a Variety of Strategies for Effective Interp...ENGLISH 7_Q4_LESSON 2_ Employing a Variety of Strategies for Effective Interp...
ENGLISH 7_Q4_LESSON 2_ Employing a Variety of Strategies for Effective Interp...
 
Visit to a blind student's school🧑‍🦯🧑‍🦯(community medicine)
Visit to a blind student's school🧑‍🦯🧑‍🦯(community medicine)Visit to a blind student's school🧑‍🦯🧑‍🦯(community medicine)
Visit to a blind student's school🧑‍🦯🧑‍🦯(community medicine)
 
Influencing policy (training slides from Fast Track Impact)
Influencing policy (training slides from Fast Track Impact)Influencing policy (training slides from Fast Track Impact)
Influencing policy (training slides from Fast Track Impact)
 
FINALS_OF_LEFT_ON_C'N_EL_DORADO_2024.pptx
FINALS_OF_LEFT_ON_C'N_EL_DORADO_2024.pptxFINALS_OF_LEFT_ON_C'N_EL_DORADO_2024.pptx
FINALS_OF_LEFT_ON_C'N_EL_DORADO_2024.pptx
 
YOUVE_GOT_EMAIL_PRELIMS_EL_DORADO_2024.pptx
YOUVE_GOT_EMAIL_PRELIMS_EL_DORADO_2024.pptxYOUVE_GOT_EMAIL_PRELIMS_EL_DORADO_2024.pptx
YOUVE_GOT_EMAIL_PRELIMS_EL_DORADO_2024.pptx
 
AUDIENCE THEORY -CULTIVATION THEORY - GERBNER.pptx
AUDIENCE THEORY -CULTIVATION THEORY -  GERBNER.pptxAUDIENCE THEORY -CULTIVATION THEORY -  GERBNER.pptx
AUDIENCE THEORY -CULTIVATION THEORY - GERBNER.pptx
 

IBM Deep Blue's Chess Chip Design

  • 1. IBM’S DEEP BLUE CHESS GRANDMASTER CHIPS DESIGN PHILOSOPHY, GENERAL ARCHITECTURE, AND PERFORMANCE OF THE CHESS CHIPS
  • 2.  Ujitha Iroshan  Nadeeshaan Gunasinghe  Dulanga Sashika Presenters
  • 3.  Introduction  History  Origin of IBM’s Chess Grandmaster  Task and Design Philosophy Journey
  • 4.  General Architecture  Common chess machine Overview  System Configuration  Chip Overview  Move Generator  Evaluation Function  Smart Move stack  Search Control  Performance Journey Cont..
  • 5. IBM Deep Blue World chess champion Garry Kasparov in 1997 What is Deep Blue ? Vs
  • 6. Creating the first world champion class chess computer is one of the oldest challenge in computer science.  Blitz chip won against grand master in shortened time chess game in 1977  Cray Blitz computer is defeated by International master David Levy in shortened time chess game in 1984 History
  • 7.  PC programmed in Blitz won against 2nd rank chess player in 1988 under faster time controls.  Deep blue early version(Deep thought) was defeated by the Champion Garry Kasparov in 1996 under regular time control.  Finally in 1997 Deep Blue won against the World Champion Garry Kasparov under regular time control History Ctd...
  • 8. Winning a match against the human World chess champion under regulation time control The game had to play out no faster than three minutes per move Task description of Deep Blue
  • 9. Regular time control  Standard chess tournament time control Faster time control (Shortened time)  Limited time for entire match and for the each moves Task Description Ctd...
  • 10.  Integrate level  Integrate the maximally possible amount of software-modifiable chess knowledge on to the chess chip.  Computational Speed  Assign weights to positional features Design Philosophy
  • 11. Common chess machine Overview  Follows the design outlined by Claud Shanon in 1949  Chess 4.5 is the earliest and the first successful chess machine design  Today Chess Machines in one way or another way, are the grandchildren of this Chess 4.5 General Architecture
  • 12. Choosing the next move  Shannon’s basic design uses a min-max tree  The evaluation function is designed from the computer player’s viewpoint Common chess machine Overview Ctd...
  • 13.  Assign numerical values for evaluation function ■ +1- Computer Player Wins ■ -1- Computer Player Loses ■ 0- Drawn position Choosing the next move Ctd..
  • 14. Can regular Min-Max search be used in a Chess Game? Common chess machine Overview cont..
  • 15.  New developments in Chess 4.5  A hardware move generator  Quiescence search (Capture Search) and Alpha beta Pruning algorithm Common chess machine Overview cont..
  • 16.  1997 version of DB included 480 chess chips  each chess chip could search 2 to 2.5 million chess positions/s  More software work could increase the speed. But applied software work in chess knowledge area System Configuration
  • 17.  Search goes parallelly in two levels  One over IBM RS/6000 SP switching network  The other over Micro Channel bus inside a workstation node  Partitioning the search provided a great deal of design flexibility and increased the search speed  SW handled a 1% of the search, but the ⅔ search depth was reduced by SW System Configuration
  • 18. Block diagram of the Chess Chip Chip Overview The chip’s cycle time lies between 40 and 50 nanoseconds in a 0.6-micron, three- metal layer, 5-V CMOS process. The average number of cycles per position searched is about 10; power dissipation is about one watt
  • 19.  Each chess Chip works as a chess machine  Why Chess chips rather than software algorithm for the competing?  Provided design possibilities unacceptable by SW  Time-scaling factor dropped from instruction cycle time to gate delay  Only thing required was reexamine of the time complexity of the competing algorithm Chip Overview
  • 20. Inside of a chess chip Chess Chip Smart Move stack Evaluation Function Search Move generator
  • 21.  What is move generator?  Used to decide the next best move of a chess piece  An 8x8 array of combinational logic  Move generation is controlled by a hardwired finite state machine Move Generator
  • 22. Why Deep Blue’s move generator special?  Generate check and check evasion Moves  Generates attacking moves  Hardware pruning of irrelevant chess moves Move Generator Ctd..
  • 23. A chip cell The combinational logic array is a silicon chess board Each has four major components  Find victim transmitter  Find Attacker Transmitter  Receiver  Distributed arbiter  Four-bit piece register Move Generator Ctd..
  • 24. Find Victim and Find Attacker Move Generator Ctd..
  • 25. As described earlier, move generator computes all the moves implicitly for a single move. What if SW is used for this task?... Move Generator Ctd..
  • 26. Generating moves from a partially generated position  Need to mask the already searched moves  Belle design used 2-bit disable-stack for masking  What is used in Deep Blue? Move Generator Ctd..
  • 27. Other moves generated by Move generator  checking moves  check-evasion moves  Detect Hung Pieces  Measures the forcefulness of checks Move Generator Ctd..
  • 28. Evaluation Function What is Evaluation Function?  Evaluation function is a function used by game- playing programs to estimate the value or goodness of a position in the minimax and related algorithms. Evaluation Function of “Deep Blue”  Contain about 66,000 gates not including RAMs and ROMs.  All sub blocks of the evaluation function divided into two sections • Lower sub blocks provide fast evaluation • Upper sub blocks provide slow evaluation
  • 30. The piece placement table computes a position’s incremental evaluation. Uses three RAMs.
  • 31. The Game Phase Control  Contains a piece count register  Maintain an XORed piece-location register  The piece-count register addresses game phase control RAMs  The game phase control RAMs produce several control values  simply a bonus or penalty to be added to the evaluation based on the material left on the board.  The most important game phase control value is the king safety relevance  tells the chess chip how to adjust the penalty for bad pawn structure and the bonus for passed pawns
  • 32. The king-and-pawn array mainly detects the “pawn can run” condition
  • 33. The Endgame logic and ROMs Mainly recognizes unusual endgame conditions
  • 34. This is the single most complicated element on the chip, occupying close to half of the chip core. Slow evaluation has a three-stage pipeline  8 X 1 systolic array  40-plus synchronous RAMs  adder tree Computes values for chess concepts such as,  square control, pins, x-rays, king safety, pawn structure, passed pawns, ray control, outposts, pawn majority, rook on the 7th, blockade, restraint and etc. Slow evaluation
  • 35. This doesn’t exist in the chess chip’s older release. Repetition detector  Contains a 32-entry circular buffer of the last 32 plies of moves.  Use hardware content addressable memory algorithm  Does this detects just simple repetition?  Time complexity is O(n) Smart move stack
  • 36. implements a minimum-window alpha-beta search algorithm This search can only tell us whether the position searched is better or worse than a single test value. Efficiency- wise, minimum-window-based search seems about the same as the regular alpha-beta search Search Control
  • 37. In early 1997, used a single chip running at 70% clock speed A hardware bug reduced the chip to 7% to 14% of its regular speed Deep Blue Jr. against the commercial chess programs played 40 games and chess chip score 95% against PC program Performance
  • 38. The 1997 version of Deep Blue only played six games Deep blue won the match by the score of 3.5 to 2.5 Kasparov is rated around 2815, which placed Deep Blue’s performance at about 2875