text
stringlengths
30
4k
source
stringlengths
60
201
6 6.189 IAP 2007 MIT Cell Chip Courtesy of International Business Machines Corporation. Unauthorized use not permitted. Michael Perrone © Copyrights by IBM Corp. and by other(s) 2007 17 6.189 IAP 2007 MIT Cell Features ● Heterogeneous multicore system architecture „ Power Processor Element for control tas...
https://ocw.mit.edu/courses/6-189-multicore-programming-primer-january-iap-2007/56d952b2cf5c52c89a4018bd081ac92b_lec2cell.pdf
2.0.2) 2-Way hardware multithreaded L1 : 32KB I ; 32KB D L2 : 512KB Coherent load / store VMX-32 Realtime Controls – – Software / hardware managed TLB – Bandwidth / Resource Reservation – Mediated Interrupts Locking L2 Cache & TLB ● Element Interconnect Bus (EIB): „ Four 16 byte data rings supporting mult...
https://ocw.mit.edu/courses/6-189-multicore-programming-primer-january-iap-2007/56d952b2cf5c52c89a4018bd081ac92b_lec2cell.pdf
controllable using PPE MMIO DMA 1,2,4,8,16,128 -> 16Kbyte transfers for I/O access Two queues for DMA commands: Proxy & SPU L o c a l S t o r e A U C S P U M F C L o c a l S t o r e A U C S P U M F C N N 96 Byte/Cycle N N NCU Power Core (PPE) L2 Cache N N MFC AUC SPU Local Store MFC AUC SPU Local Store ...
https://ocw.mit.edu/courses/6-189-multicore-programming-primer-january-iap-2007/56d952b2cf5c52c89a4018bd081ac92b_lec2cell.pdf
Two configurable interfaces (60GB/s @ 5Gbps) – Configurable number of bytes – Coherent (BIF) and / or I/O (IOIFx) protocols Supports two virtual channels per interface Supports multiple system configurations L o c a l S t o r e A U C S P U M F C N L o c a l S t o r e A U C S P U M F C N 25 GB/sec XDR DRAM MIC...
https://ocw.mit.edu/courses/6-189-multicore-programming-primer-january-iap-2007/56d952b2cf5c52c89a4018bd081ac92b_lec2cell.pdf
From Coherent Interconnect From IOIF0 or IOIF1 – Interrupt Priority Level Control Interrupt Generation Ports for IPI Duplicated for each PPE hardware thread ● „ I/O Bus Master Translation (IOT) Translates Bus Addresses to System Real Addresses Two Level Translation – „ I/O Segments (256 MB) I/O Pages (4K, ...
https://ocw.mit.edu/courses/6-189-multicore-programming-primer-january-iap-2007/56d952b2cf5c52c89a4018bd081ac92b_lec2cell.pdf
AP 2007 MIT Courtesy of International Business Machines Corporation. Unauthorized use not permitted. 6.189 IAP 2007 Lecture 2 Cell Performance Characteristics Michael Perrone © Copyrights by IBM Corp. and by other(s) 2007 23 6.189 IAP 2007 MIT Why Cell Processor Is So Fast?...
https://ocw.mit.edu/courses/6-189-multicore-programming-primer-january-iap-2007/56d952b2cf5c52c89a4018bd081ac92b_lec2cell.pdf
™ 64 X2 2.4 GHz Intel Pentium D® 3.2 GHz PowerPC® 970MP 2.5 GHz Cell Broadband EngineTM 3.2 GHz Michael Perrone © Copyrights by IBM Corp. and by other(s) 2007 25 6.189 IAP 2007 MIT Cell BE Performance ● BE can outperform a P4/SSE2 at same clock rate by 3 to 18x (assuming linear scaling) in vario...
https://ocw.mit.edu/courses/6-189-multicore-programming-primer-january-iap-2007/56d952b2cf5c52c89a4018bd081ac92b_lec2cell.pdf
v) 200 fps (IA32) 290 fps (per SPE) BE Perf Advantage 8x 8x 2x 6x 12x 15x 14x 10x 6x 18x 12x 12x Michael Perrone © Copyrights by IBM Corp. and by other(s) 2007 26 6.189 IAP 2007 MIT Michael Perrone © Copyrights by IBM Corp. and by other(s) 2007 1 Key Performance Characteristics ● Cell's performa...
https://ocw.mit.edu/courses/6-189-multicore-programming-primer-january-iap-2007/56d952b2cf5c52c89a4018bd081ac92b_lec2cell.pdf
s by IBM Corp. and by other(s) 2007 29 6.189 IAP 2007 MIT Cell Application Affinity – Target Industry Sectors ƒƒ Petroleum Industry Petroleum Industry ƒƒ Seismic computing Seismic computing ƒƒ Reservoir Modeling, Reservoir Modeling, …… ƒƒ Aerospace & Defense Aerospace & Defense ƒƒ Signal & Im Signal & I age Proce...
https://ocw.mit.edu/courses/6-189-multicore-programming-primer-january-iap-2007/56d952b2cf5c52c89a4018bd081ac92b_lec2cell.pdf
2007 Lecture 2 Cell Software Environment Michael Perrone © Copyrights by IBM Corp. and by other(s) 2007 31 6.189 IAP 2007 MIT Cell Software Environment Programmer Experience Development Environment Development Tools Stack Code Dev Tools Samples Workloads Demos Debug Tools SPE Management Lib Applic...
https://ocw.mit.edu/courses/6-189-multicore-programming-primer-january-iap-2007/56d952b2cf5c52c89a4018bd081ac92b_lec2cell.pdf
s by IBM Corp. and by other(s) 2007 35 6.189 IAP 2007 MIT Cell Simulator Debugging Environment Execution Environment Michael Perrone © Copyrights by IBM Corp. and by other(s) 2007 36 6.189 IAP 2007 MIT Linux on CBE ● Provided as patched to the 2.6.15 PPC64 Kernel „ Added heterogeneous lwp/thread model – SP...
https://ocw.mit.edu/courses/6-189-multicore-programming-primer-january-iap-2007/56d952b2cf5c52c89a4018bd081ac92b_lec2cell.pdf
GNU Libs (glibc,etc) ILP32 Processes Library (64-bit) SPE Object Loader Services std. PPC64 elf interp 64-bit GNU Libs (glibc) LP64 Processes System Call Interface exec Loader File System Framework Device Framework Network Framework Streams Framework SPU Management Framework Misc format bin SPU Ob...
https://ocw.mit.edu/courses/6-189-multicore-programming-primer-january-iap-2007/56d952b2cf5c52c89a4018bd081ac92b_lec2cell.pdf
9 6.189 IAP 2007 MIT Optimized SPE and Multimedia Extension Libraries ● Standard SPE C library subset „ „ optimized SPE C99 functions including stdlib c lib, math and etc. subset of POSIX.1 Functions – PPE assisted Execution Environment ● Audio resample - resampling audio signals ● FFT - 1D and 2D fft functio...
https://ocw.mit.edu/courses/6-189-multicore-programming-primer-january-iap-2007/56d952b2cf5c52c89a4018bd081ac92b_lec2cell.pdf
● VSE_subdiv - variable sharpness subdivision Execution Environment algorithm Michael Perrone © Copyrights by IBM Corp. and by other(s) 2007 42 6.189 IAP 2007 MIT Bringup Workloads / Demos ● Numerous code samples provided to demonstrate system design constructs ● Complex workloads and demos used to evaluat...
https://ocw.mit.edu/courses/6-189-multicore-programming-primer-january-iap-2007/56d952b2cf5c52c89a4018bd081ac92b_lec2cell.pdf
„ Three modes of debugging SPU threads – Standalone SPE debugging – Attach to SPE thread • Thread ID output when SPU_DEBUG_START=1 Michael Perrone © Copyrights by IBM Corp. and by other(s) 2007 45 6.189 IAP 2007 MIT SPE Performance Tools (executables) ● Static analysis (spu_timing) „ Annotates assembly sour...
https://ocw.mit.edu/courses/6-189-multicore-programming-primer-january-iap-2007/56d952b2cf5c52c89a4018bd081ac92b_lec2cell.pdf
: data + code + stack ● Communication „ DMA and Bus bandwidth – DMA granularity – 128 bytes – DMA bandwidth among LS and System memory „ Traffic control – Exploit computational complexity and data locality to lower data traffic requirement „ Shared memory / Message passing abstraction overhead „ Synchronizatio...
https://ocw.mit.edu/courses/6-189-multicore-programming-primer-january-iap-2007/56d952b2cf5c52c89a4018bd081ac92b_lec2cell.pdf
) ● Chassis „ „ „ Standard IBM BladeCenter form factor with: 7 Blades (for 2 slots each) with full performance – 2 switches (1Gb Ethernet) with 4 external ports each – Updated Management Module Firmware. External Infiniband Switches with optional FC ports ● Typical Configuration (available today from E&TS) „ ...
https://ocw.mit.edu/courses/6-189-multicore-programming-primer-january-iap-2007/56d952b2cf5c52c89a4018bd081ac92b_lec2cell.pdf
. IBM may not make these offerings available in other countries, and the information is subject to change without notice. Consult your local IBM business contact for information on the IBM offerings available in your area. In no event will IBM be liable for damages arising directly or indirectly from any use of the...
https://ocw.mit.edu/courses/6-189-multicore-programming-primer-january-iap-2007/56d952b2cf5c52c89a4018bd081ac92b_lec2cell.pdf
system dependent and may not be available on Linux. For more information, please check: http://www.ibm.com/systems/p/software/whitepapers/linux_overview.html Any performance data contained in this document was determined in a controlled environment. Actual results may vary significantly and are dependent on many...
https://ocw.mit.edu/courses/6-189-multicore-programming-primer-january-iap-2007/56d952b2cf5c52c89a4018bd081ac92b_lec2cell.pdf
trademarks of Rambus, Inc. UNIX is a registered trademark in the United States, other countries or both. Linux is a trademark of Linus Torvalds in the United States, other countries or both. Fedora is a trademark of Redhat, Inc. Microsoft, Windows, Windows NT and the Windows logo are trademarks of Microsoft Cor...
https://ocw.mit.edu/courses/6-189-multicore-programming-primer-january-iap-2007/56d952b2cf5c52c89a4018bd081ac92b_lec2cell.pdf
ation, life support, or other hazardous uses where malfunction could result in death, bodily injury, or catastrophic property damage. The information contained in this document does not affect or change IBM product specifications or warranties. Nothing in this document shall operate as an express or implied license o...
https://ocw.mit.edu/courses/6-189-multicore-programming-primer-january-iap-2007/56d952b2cf5c52c89a4018bd081ac92b_lec2cell.pdf
-late ● VMX-like SIMD dataflow „ Broad set of operations (8 / 16 / 32 Byte) „ Graphics SP-Float IEEE DP-Float „ CHANNEL DMA SMM SBI B E B ATO RTB 14.5mm2 (90nm SOI) ● Unified register file „ 128 entry x 128 bit ● 256KB Local Store „ Combined I & D „ 16B/cycle L/S bandwidth „ 128B/cycle DMA bandwidth M...
https://ocw.mit.edu/courses/6-189-multicore-programming-primer-january-iap-2007/56d952b2cf5c52c89a4018bd081ac92b_lec2cell.pdf
„ „ DMA/DMA-list Branch hint ● VMX-like SIMD dataflow „ Graphics SP-Float „ No saturate arith, some byte „ IEEE DP-Float (BlueGene-like) ● Unified register file ● 128 entry x 128 bit „ 256KB Local Store „ Combined I & D „ „ 16B/cycle L/S bandwidth 128B/cycle DMA bandwidth ● Memory Flow Control (MFC) LS L...
https://ocw.mit.edu/courses/6-189-multicore-programming-primer-january-iap-2007/56d952b2cf5c52c89a4018bd081ac92b_lec2cell.pdf
Enqueue DMA Command Michael Perrone © Copyrights by IBM Corp. and by other(s) 2007 61 6.189 IAP 2007 MIT SPE Block Diagram Floating-Point Unit Fixed-Point Unit Permute Unit Load-Store Unit Branch Unit Channel Unit Result Forwarding and Staging Register File Local Store (256kB) Single Port SRAM I...
https://ocw.mit.edu/courses/6-189-multicore-programming-primer-january-iap-2007/56d952b2cf5c52c89a4018bd081ac92b_lec2cell.pdf
8 PPE-side Command Queue entries 16 SPU-side Command Queue entries „ ● MMU similar to PowerPC MMU 8 SLBs, 256 TLBs 4K, 64K, 1M, 16M page sizes Software/HW page table walk PT/SLB misses interrupt PPE „ „ „ Atomic Cache Facility ● „ „ 4 cache lines for atomic updates 2 cache lines for cast out/MMU reload ● Up to...
https://ocw.mit.edu/courses/6-189-multicore-programming-primer-january-iap-2007/56d952b2cf5c52c89a4018bd081ac92b_lec2cell.pdf
Next Program Counter SPU Execution Status SPU Privileged Control SPU Channel Counter Initialize SPU Channel Data Initialize SPU Signal Notification Control SPU Decrementer Status & Control MFC DMA Control MFC Context Save / Restore Registers SLB Management Registers 4K Physical Page Boundary 4K Physical Page...
https://ocw.mit.edu/courses/6-189-multicore-programming-primer-january-iap-2007/56d952b2cf5c52c89a4018bd081ac92b_lec2cell.pdf
Decrementer Count (Channel 8) 16 Entry MFC Command Queue Interface (Channels 16-21) DMA Tag Group Query Mask (Channel 22) Request Tag Status Update (Channel 23) Immediate Conditional - ALL Conditional - ANY Read DMA Tag Group Status (Channel 24) DMA List Stall and Notify Tag Status (Channel 25) DMA List Stall ...
https://ocw.mit.edu/courses/6-189-multicore-programming-primer-january-iap-2007/56d952b2cf5c52c89a4018bd081ac92b_lec2cell.pdf
Data cache region touch (DMA Get hint) sdcrtst - Data cache region touch for store (DMA Put hint) sdcrz - Data cache region zero sdcrs - Data cache region store sdcrf - Data cache region flush Synchronization Commands Lockline (Atomic Update) Commands: getllar - DMA 128 bytes from EA to LS and set Reservation p...
https://ocw.mit.edu/courses/6-189-multicore-programming-primer-january-iap-2007/56d952b2cf5c52c89a4018bd081ac92b_lec2cell.pdf
96B/cycle peak bandwidth „ Over 100 outstanding requests Courtesy of International Business Machines Corporation. Unauthorized use not permitted. Michael Perrone © Copyrights by IBM Corp. and by other(s) 2007 70 6.189 IAP 2007 MIT Element Interconnect Bus – Command Topology ● “Address Concentrator” tree struc...
https://ocw.mit.edu/courses/6-189-multicore-programming-primer-january-iap-2007/56d952b2cf5c52c89a4018bd081ac92b_lec2cell.pdf
Data Arb 16B 16B 16B 16B 16B 16B 16B 16B 16B 16B 16B 16B MIC SPE0 SPE2 SPE4 SPE6 BIF/IOIF0 Michael Perrone © Copyrights by IBM Corp. and by other(s) 2007 72 6.189 IAP 2007 MIT Internal Bandwidth Capability ● Each EIB Bus data port supports 25.6GBytes/sec* in each direction ● The EIB Command Bus s...
https://ocw.mit.edu/courses/6-189-multicore-programming-primer-january-iap-2007/56d952b2cf5c52c89a4018bd081ac92b_lec2cell.pdf
Controller Controller Controller Controller Controller Controller Controller Controller Controller Controller Ramp 5Ramp 5 Ramp 7 Ramp 4 Ramp 4 Ramp 8 3 Ramp Ramp 3 Ramp 9 2 Ramp Ramp 2 Ramp 10 Ramp 1 Ramp 1 Ramp 11 0 Ramp Ramp 0 MICMICPPE SPE0SSPE0PE1 SPE2SSPE2PE3 SPE4SSPE4PE...
https://ocw.mit.edu/courses/6-189-multicore-programming-primer-january-iap-2007/56d952b2cf5c52c89a4018bd081ac92b_lec2cell.pdf
20.110/5.60 Fall 2005 Lecture #5 page 1 Thermochemistry Much of thermochemistry is based on finding “easy” paths to calculate changes in enthalpy, i.e. understanding how to work with thermodynamic cycles. • Goal: To predict H∆ for every reaction, even if it cannot be carried out in the laboratory The heat of...
https://ocw.mit.edu/courses/20-110j-thermodynamics-of-biomolecular-systems-fall-2005/56de76a110080853b28006a5ce0b0620_l05.pdf
G. Bawendi, Robert W. Field 20.110/5.60 Fall 2005 Lecture #5 page 2 • ∆ H f (° 298.15 K ) : The heat of formation is the heat of reaction to create 1 mole of that compound from its constituent elements in their most stable forms. Example (T = 298.15 K) ½ H2 (g,T,1 bar) +...
https://ocw.mit.edu/courses/20-110j-thermodynamics-of-biomolecular-systems-fall-2005/56de76a110080853b28006a5ce0b0620_l05.pdf
• to thermochemistry] Reactants ∆ rxH Products − ∆∑ ν i ° H , reactants f i ( i ) Elements °∆∑ H ν i f i ( , products i ) 20.110J / 2.772J / 5.601JThermodynamics of Biomolecular SystemsInstructors: Linda G. Griffith, Kimberly Hamad-Schifferli, Moungi G. Bawendi, Robert W. Field ...
https://ocw.mit.edu/courses/20-110j-thermodynamics-of-biomolecular-systems-fall-2005/56de76a110080853b28006a5ce0b0620_l05.pdf
H f ,CH 4 = ∆ ° H f ,CO 2 ∆ H IV = H 2 HO 2 − − H H 2 O 2 H 2 = ∆ 2 ° H f ,H O 2 ∆ H rx = ∆ 2 ° H f ,H O 2 + ∆ ° H f ,CO 2 ∴ In general, − ∆ H° 4f ,CH ∆ H rx = ∑ ν i i ∆ ° H f i , ( products ) − ∆ ° H f i , ( reactants ) ∑ ν i i ν ≡ stoichiometric coefficient • H∆ at constant p and for reversible pV process is = p...
https://ocw.mit.edu/courses/20-110j-thermodynamics-of-biomolecular-systems-fall-2005/56de76a110080853b28006a5ce0b0620_l05.pdf
calorimeter adiabatic constant p React. (T1) + Cal. (T1) ∆HI adiabatic consta nt p ∆ rxH → constant p Prod. (T1) + Cal. (T1) ∆ HII constant p Prod. (T2) + Cal. (T2) I ) ∆ IH React. (T1) + Cal. (T1) adiabatic = constant p Prod. (T2) + Cal. (T2) II) ∆ IIH Prod. (T2) + Cal. (T2) = constant p Prod. (T1) + Cal....
https://ocw.mit.edu/courses/20-110j-thermodynamics-of-biomolecular-systems-fall-2005/56de76a110080853b28006a5ce0b0620_l05.pdf
∆ H II ∴ ∆ H rx ( T 1 ) = − T 2 ∫ T 1 + . C od Cal dT Pr . ( p ) • Constant volume (when gases involved) reaction calorimeter adiabatic constant V R eact. (T1) + Cal. (T1) ∆UI adiabatic consta nt V ∆ rxU → constant V Prod. (T1) + Cal. (T1) ∆U II constant V Prod. (T2) + Cal. (T2) I ) ∆ IU React. (T1) + Cal. (T1) ...
https://ocw.mit.edu/courses/20-110j-thermodynamics-of-biomolecular-systems-fall-2005/56de76a110080853b28006a5ce0b0620_l05.pdf
= 0 0 I I (II) Purpose is to m rom T back to T. f 2 1 easure heat qV needed to take prod. + cal. q V = T ∫ 1 T 2 C od C Pr . + V ( ) . al dT = ∆ U II ∴ ∆ U rx ( T 1 ) = − T 2 ∫ VT 1 ( + . C od Cal dT Pr . ) ∆ ( ) Now use H U pV + = or ∆ ) H U pV + = ∆ ( Assume only significant contribution to ∆ pV is from gases...
https://ocw.mit.edu/courses/20-110j-thermodynamics-of-biomolecular-systems-fall-2005/56de76a110080853b28006a5ce0b0620_l05.pdf
2.772J / 5.601JThermodynamics of Biomolecular SystemsInstructors: Linda G. Griffith, Kimberly Hamad-Schifferli, Moungi G. Bawendi, Robert W. Field 20.110/5.60 Fall 2005 Lecture #5 page 7 Temperature dependence of ∆Hrx uppose know ∆Hrx at some ...
https://ocw.mit.edu/courses/20-110j-thermodynamics-of-biomolecular-systems-fall-2005/56de76a110080853b28006a5ce0b0620_l05.pdf
C prod C react dT ⎣ p .) .) ⎤ ⎦ − ( ( p ) T 2 ∫ + ∆ T 1 CdT p 20.110J / 2.772J / 5.601JThermodynamics of Biomolecular SystemsInstructors: Linda G. Griffith, Kimberly Hamad-Schifferli, Moungi G. Bawendi, Robert W. Field
https://ocw.mit.edu/courses/20-110j-thermodynamics-of-biomolecular-systems-fall-2005/56de76a110080853b28006a5ce0b0620_l05.pdf
5.8 The Harmonic Oscillator To illustrate the beauty and efficiency in describing the dynamics of a quan- tum system using the dirac notation and operator algebra, we reconsider the one-dimensional harmonic oscillator discussed in section 4.4.2 and described by the Hamiltonian operator with H = p2 2m + 1 2 K x2, [x, p] =...
https://ocw.mit.edu/courses/6-974-fundamentals-of-photonics-quantum-electronics-spring-2006/575470ae5ba3cafa4fd36e35b0d29e17_harmonic_oscil.pdf
) and the Hamiltonian operator can be rewritten in terms of the new operators a and a+ as ¡ ¢ ¡ ¢ H = ~ω0 2 a+a + aa+ = ~ω0 ¡ a+a+ µ ¢ . 1 2 ¶ We introduce the operator N = a+a, (5.138) (5.139) (5.140) which is a hermitian operator. Up to an additive constant 1/2 and a scaling factor ~ω0 equal to the energy of one quan...
https://ocw.mit.edu/courses/6-974-fundamentals-of-photonics-quantum-electronics-spring-2006/575470ae5ba3cafa4fd36e35b0d29e17_harmonic_oscil.pdf
that contains one less quantum is a new eigenstate to N with eigenvalue Nn (5.143) (5.144) (5.145) | | − n i | − ¡ | a n i | = C n | − , 1 i (5.146) where C is a yet undetermined constant. This constant follows from the normalization of this state and being an eigenvector to the number operator. a+a n | h n C = | | i |...
https://ocw.mit.edu/courses/6-974-fundamentals-of-photonics-quantum-electronics-spring-2006/575470ae5ba3cafa4fd36e35b0d29e17_harmonic_oscil.pdf
with i energy ~ω0/2 we can generate the n-th energy eigenstate by n-fold application of the creation operator a+ and proper normalization n + 1 i n i | = √n + 1 | | . = n i | 1 (n + 1)! n a+ , 0 i | ¡ ¢ with and p H n i | = En , n i | En = ~ω0 n+ µ 1 2 . ¶ (5.155) (5.156) (5.157) 5.8.2 Matrix Representation We can expr...
https://ocw.mit.edu/courses/6-974-fundamentals-of-photonics-quantum-electronics-spring-2006/575470ae5ba3cafa4fd36e35b0d29e17_harmonic_oscil.pdf
MASSACHUSETTS INSTITUTE OF TECHNOLOGY SLOAN SCHOOL OF MANAGEMENT 15.565 Integrating Information Systems: Technology, Strategy, and Organizational Factors 15.578 Global Information Systems: Communications & Connectivity Among Information Systems Spring 2002 Lecture 9 NETWORK PROTOCOLS COMPLEXITY OF COMMUNICATION N...
https://ocw.mit.edu/courses/15-565j-integrating-esystems-global-information-systems-spring-2002/57599d12f997dc9b58d783ce08d5353b_lecture09.pdf
IEEE 5 ISO REFERENCE MODEL FOR OSI 7. APPLICATION: APPLICATION DEPENDENT (E.G., USER PROGRAM) 6. PRESENTATION: COMPRESSION AND CONVERSIONS (E.G., LIBRARY) 5. SESSION: PROCESS-TO-PROCESS (E.G., OS SOFTWARE) 4. TRANSPORT: HOST-TO-HOST (E.G., OS SOFTWARE) 3. NETWORK: ROUTING (E.G., DEVICE DRIVER) 2. DATA: RELIABLE ...
https://ocw.mit.edu/courses/15-565j-integrating-esystems-global-information-systems-spring-2002/57599d12f997dc9b58d783ce08d5353b_lecture09.pdf
TELEPHONE • T1 = 1.544M bps (USA & CANADA) OR 2.048M (ELSE) – SHARED CABLE (ETHERNET) • 10-100M bps (Typical) • CSMA/CD (CARRIER SENSE MULTIPLE ACCESS/COLLISION DETECT) – SATELLITE • 5-10 CHANNELS, EACH 50M bps • UP-LINK & DOWN-LINK = 270 MILLISECONDS • VSAT – FIBER-OPTIC • 100M - 10G bps (Typical) • INTERNE...
https://ocw.mit.edu/courses/15-565j-integrating-esystems-global-information-systems-spring-2002/57599d12f997dc9b58d783ce08d5353b_lecture09.pdf
NEXT (ACK RECEIVED UP TO) 10 DYNAMIC CHANNEL SHARING • SATELLITE – MULTIPLEX (TIME OR FREQUENCY) – “SLOTTED” ALOHA (PRE-DATES ETHERNET) LA • EACH “SLOT”, EITHER SEND OR NO SEND • IF SEND, LISTEN FOR COLLISION (270 MS) • IF COLLIDE, TRY AGAIN LATER -- BUT WHEN? • “BEST ATTAINABLE UTILIZATION” ABOUT 37% SF • ...
https://ocw.mit.edu/courses/15-565j-integrating-esystems-global-information-systems-spring-2002/57599d12f997dc9b58d783ce08d5353b_lecture09.pdf
= COUNTRY NET 10 = NETWORK OPERATOR CHOICE (E.G., 5 = HOST #, 5 = USER #) -- -- -- 13 SYNCHRONIZATION AND MULTIPLEXING ISSUES • SYNCHRONIZATION ISSUES: -- -- -- UNEXPECTED MESSAGE RECEIVED MULTIPLE PACKETS (DUE TO TIME-OUT & RETRANSMIT) CLOSING CONNECTIONS (E.G., TWO ARMY DIVISION PROBLEM: “YOU ATTACK WHEN YOU...
https://ocw.mit.edu/courses/15-565j-integrating-esystems-global-information-systems-spring-2002/57599d12f997dc9b58d783ce08d5353b_lecture09.pdf
Transmission Control Transmission control Network services Internet Protocol Network Access Path control Transport Data link control Data link control Physical Physical Physical Approximate correspondences between the various networks. 16 TCP/IP Protocol Architecture Application layer Provides communicat...
https://ocw.mit.edu/courses/15-565j-integrating-esystems-global-information-systems-spring-2002/57599d12f997dc9b58d783ce08d5353b_lecture09.pdf
Lecture 8 8.321 Quantum Theory I, Fall 2017 40 Lecture 8 (Oct. 2, 2017) 8.1 General Time Dependent Hamiltonians The Schr¨odinger equation dictates that quantum states evolve in time according to i(cid:126) d d t | (t)(cid:105) = H(t)|ψ(t) ψ (cid:105) . (8.1) In the last class, we saw that if the Hamiltonian is independ...
https://ocw.mit.edu/courses/8-321-quantum-theory-i-fall-2017/579cb61c4eaa494b9adb90830d921099_MIT8_321F17_lec8.pdf
, U (t, t ) = e−iH(t−t0)/(cid:126) 0 . (8.2) (8.3) (8.4) (8.5) (8.6) (8.7) 2. If [H(t), H(t(cid:48))] = 0 for all t, t(cid:48), then we can simultaneously diagonalize the Hamiltonian at all times, meaning we can choose a basis of states that are eigenstates of H(t) for all time (the associated eigenvalues may change as...
https://ocw.mit.edu/courses/8-321-quantum-theory-i-fall-2017/579cb61c4eaa494b9adb90830d921099_MIT8_321F17_lec8.pdf
We could choose each time interval to have a different length, but the ultimate result will not be affected by our choice. We take N large, so that ∆t is small. Then, (cid:1) U (ti+1, ti) = 1 − H(ti)∆t + O (∆t)2 . (cid:0) i (cid:126) If we are ignoring terms of order (∆t)2, then we can write U (t i+1 , t ) ≈ e−iH(ti)∆t/(...
https://ocw.mit.edu/courses/8-321-quantum-theory-i-fall-2017/579cb61c4eaa494b9adb90830d921099_MIT8_321F17_lec8.pdf
48) (cid:48) (cid:1) (cid:19)(cid:21) . (8.14) be the product in Eq. (8.13) in the This is called a time-ordered exponential, and is defined to limit N → ∞, ∆t → 0 with N ∆t held constant. The operator T is called the time-ordering operator ; it reorders the operators in its argument so that they are time-ordered (disre...
https://ocw.mit.edu/courses/8-321-quantum-theory-i-fall-2017/579cb61c4eaa494b9adb90830d921099_MIT8_321F17_lec8.pdf
(8.15), we have U (t, t0) = 1 − ˆ t t0 i (cid:126) (cid:1) dt(cid:48) H t(cid:48) U t(cid:48) t0 . (cid:0) (cid:1) (cid:0) , (8.17) This is an expression in terms of the Hamiltonian and with t(cid:48) ≤ t. We can similarly write time -ev olution operator at times t(cid:48) U (cid:0) t(cid:48), t0 (cid:1) = 1 − ˆ t(cid:...
https://ocw.mit.edu/courses/8-321-quantum-theory-i-fall-2017/579cb61c4eaa494b9adb90830d921099_MIT8_321F17_lec8.pdf
1 d 2 · · · t t0 t0 t0 · · · dtn H(t1) H(tn) + · · · . Note that the operators in each integrand are time-ordered. This observation allows us to write Eq. (8.19) in such a way that the limits of integration are not so complicated, using ˆ t(cid:48) ˆ t dt(cid:48) t0 t 0 dt(cid:48)(cid:48) H (cid:0)t(cid:48) H (cid:1) (...
https://ocw.mit.edu/courses/8-321-quantum-theory-i-fall-2017/579cb61c4eaa494b9adb90830d921099_MIT8_321F17_lec8.pdf
:18) U (t, t0) = T exp − ˆ t t0 i (cid:126) dt(cid:48) H t (cid:0) (cid:48)(cid:1) (cid:19)(cid:21) . (8.22) 8.2 Interaction Picture We have seen the Schr¨odinger picture and the Heisenberg picture; now we will discuss a third picture, due to Dirac, called the interaction picture. This is a mixed picture that is useful...
https://ocw.mit.edu/courses/8-321-quantum-theory-i-fall-2017/579cb61c4eaa494b9adb90830d921099_MIT8_321F17_lec8.pdf
.26) (8.27) (8.28) The interaction picture is useful when we fully understand the dynamics of H0, and V is a weak probe. We can then use this approach to determine how quickly our system absorbs energy from the weak probe. In the interaction picture, we are hiding the time evolution due to the system we understand, so ...
https://ocw.mit.edu/courses/8-321-quantum-theory-i-fall-2017/579cb61c4eaa494b9adb90830d921099_MIT8_321F17_lec8.pdf
I dt = V U = V U0UI , i(cid:126) dU I dt = U −1 0 V U0UI = VI(t)UI(t) . (8.30) (8.31) (8.32) (8.33) (8.34) (8.35) (8.36) MIT OpenCourseWare https://ocw.mit.edu 8.321 Quantum Theory I Fall 2017 For information about citing these materials or our Terms of Use, visit: https://ocw.mit.edu/terms.
https://ocw.mit.edu/courses/8-321-quantum-theory-i-fall-2017/579cb61c4eaa494b9adb90830d921099_MIT8_321F17_lec8.pdf
(cid:0)(cid:1)(cid:2)(cid:3)(cid:4)(cid:5) Last modi(cid:0)ed(cid:1) September (cid:2)(cid:3)(cid:4) (cid:2)(cid:5)(cid:5)(cid:6) Many(cid:0)body phenomena in condensed matter and atomic physics (cid:0) Lecture (cid:1)(cid:2) Vortices(cid:3) super(cid:4)uidity(cid:2) Trapped gases(cid:2) BEC at (cid:5)nite temperature(...
https://ocw.mit.edu/courses/8-514-strongly-correlated-systems-in-condensed-matter-physics-fall-2003/57fb8bd5b6d291ec7eda1bc1c920a403_lec6.pdf
(cid:0)(cid:1)(cid:3) Super(cid:4)uidity(cid:1) Vortices(cid:1) Let us consider the circulation of velocity in a super(cid:18)ow(cid:2) It follows from the relation b e t ween the velocity and the phase(cid:1) Eq(cid:2) (cid:6)(cid:19)(cid:7)(cid:1) that the circulation around any contour C obeys (cid:22) (cid:4) d (ci...
https://ocw.mit.edu/courses/8-514-strongly-correlated-systems-in-condensed-matter-physics-fall-2003/57fb8bd5b6d291ec7eda1bc1c920a403_lec6.pdf
is the origin of super(cid:18)uidity(cid:2) The only way to eliminate a super(cid:18)ow is to produce excitations with discrete vorticity and then remove them (cid:6)along with the vorticity(cid:7) from the system(cid:2) Also mention the Landau criterion for super(cid:18)uidity(cid:25) The quasiparticle energy (cid:8) ...
https://ocw.mit.edu/courses/8-514-strongly-correlated-systems-in-condensed-matter-physics-fall-2003/57fb8bd5b6d291ec7eda1bc1c920a403_lec6.pdf
generated in the (cid:18)ow e v en at velocities below v (cid:1) one cannot c use it to predict the actual value of critical velocity(cid:2) The observed critical velocities are system(cid:0)dependent (cid:6)non(cid:0)universal(cid:7) and are typically several orders of magntude below v c estimated from quasiparticle d...
https://ocw.mit.edu/courses/8-514-strongly-correlated-systems-in-condensed-matter-physics-fall-2003/57fb8bd5b6d291ec7eda1bc1c920a403_lec6.pdf
(cid:11) (cid:4) (cid:6)(cid:11)(cid:11)(cid:7) (cid:4) p (cid:1)m 1 For weakly interacting Bose gas(cid:0) the critical velocity is equal to Bogoliubov s o u n d v elocity(cid:1) (cid:9) ...
https://ocw.mit.edu/courses/8-514-strongly-correlated-systems-in-condensed-matter-physics-fall-2003/57fb8bd5b6d291ec7eda1bc1c920a403_lec6.pdf
circulation(cid:1) l (cid:4) (cid:11)(cid:1) we expect (cid:0)(cid:6)(cid:9) (cid:7) (cid:4) pne (cid:1) and i(cid:0) (cid:0)(cid:6)(cid:9) (cid:11) (cid:7) (cid:9)e (cid:2) Thus one can take a trial function of the form i(cid:0) (cid:5) (cid:6) (cid:7) (cid:8) (cid:0)(cid:6) (cid:7) (cid:4) n e (cid:6)(cid:11)(cid:9)(...
https://ocw.mit.edu/courses/8-514-strongly-correlated-systems-in-condensed-matter-physics-fall-2003/57fb8bd5b6d291ec7eda1bc1c920a403_lec6.pdf
a vortex located on the symmetry axis of the jar of radius b and height L(cid:1) the kinetic energy of the (cid:18)ow is L b L b (cid:1) (cid:1) (cid:11) (cid:8)h d(cid:9) (cid:8)h b (cid:1) E (cid:4) n(cid:6)(cid:9)(cid:7) mv (cid:6)(cid:9)(cid:7)(cid:9)(cid:7)(cid:9)d(cid:9)dz (cid:4) (cid:7) n dz (cid:4) (cid:7) n l...
https://ocw.mit.edu/courses/8-514-strongly-correlated-systems-in-condensed-matter-physics-fall-2003/57fb8bd5b6d291ec7eda1bc1c920a403_lec6.pdf
out to b e approximately (cid:7)n ln (cid:11)(cid:12)(cid:16)(cid:19)(cid:16)L(cid:1) which is smaller than our m kinetic energy estimate (cid:6)(cid:11)(cid:14)(cid:7)(cid:2) The energy of the vortex in a jar rotating with velocity (cid:27) i s E (cid:6)(cid:27)(cid:7) (cid:4) E (cid:27)M (cid:1) where v v M is the an...
https://ocw.mit.edu/courses/8-514-strongly-correlated-systems-in-condensed-matter-physics-fall-2003/57fb8bd5b6d291ec7eda1bc1c920a403_lec6.pdf
:1) which means that it is easier (cid:6)(cid:3)(cid:7) to produce vortices in a larger jar(cid:2) c (cid:6)(cid:3)(cid:7) If the rotation velocity is larger than (cid:27) and keeps increasing(cid:1) one can reach t h e n e x t c critical value (cid:27) at which the second vortex appears(cid:1) and then(cid:1) at some ...
https://ocw.mit.edu/courses/8-514-strongly-correlated-systems-in-condensed-matter-physics-fall-2003/57fb8bd5b6d291ec7eda1bc1c920a403_lec6.pdf
Of course(cid:1) since N is integer(cid:1) in reality the numb e r of vortices increases discretely(cid:1) in steps(cid:1) on average following the proportionality relation (cid:6)(cid:11)(cid:19)(cid:7)(cid:2) (cid:0)(cid:1)(cid:5) Trapped gases(cid:1) Bose condensation of con(cid:3)ned gasses di(cid:28)ers somewhat f...
https://ocw.mit.edu/courses/8-514-strongly-correlated-systems-in-condensed-matter-physics-fall-2003/57fb8bd5b6d291ec7eda1bc1c920a403_lec6.pdf
to the fact that the lowest energy quantum state in which atoms condense is p e a k ed at the trap center and has spatial extent much less than the size of thermal cloud at temperatures slightly above T (cid:2) BEC In the experiments on BEC in trapped gasses(cid:1) atoms are con(cid:3)ned by a magnetic trap(cid:1) whic...
https://ocw.mit.edu/courses/8-514-strongly-correlated-systems-in-condensed-matter-physics-fall-2003/57fb8bd5b6d291ec7eda1bc1c920a403_lec6.pdf
e r of atoms is large(cid:2) In the presence of (cid:5) (cid:2) (cid:9) interactions(cid:1) on can easily reach the limit when the interaction energy per particle is much larger than the level spacing in the trap(cid:1) (cid:1)n (cid:8)h(cid:13) (cid:2) For that(cid:1) the number of atoms should exceed N (cid:4) h(cid:...
https://ocw.mit.edu/courses/8-514-strongly-correlated-systems-in-condensed-matter-physics-fall-2003/57fb8bd5b6d291ec7eda1bc1c920a403_lec6.pdf
)x(cid:7) dx (cid:6)(cid:11)(cid:21)(cid:7) (cid:1) h (cid:8) (cid:11) (cid:1) (cid:1) (cid:8) (cid:0) (cid:9)m jr j (cid:0) j j j j (cid:1) (cid:9) Z with the particle numb e r N (cid:4) (cid:0) dx being (cid:3)xed by a c hemical potential (cid:5)(cid:2) (cid:1) Let us argue that one can discard the gradient term in t...
https://ocw.mit.edu/courses/8-514-strongly-correlated-systems-in-condensed-matter-physics-fall-2003/57fb8bd5b6d291ec7eda1bc1c920a403_lec6.pdf
1)N(cid:6)m(cid:13) (cid:7) (cid:2) At large (cid:9) (cid:3)(cid:3)(cid:9) (cid:1) 2 2 (cid:9) ; h (cid:2) m(cid:2) (cid:1) (cid:1) N N (cid:1) the value R is much larger than l that satis(cid:3)es l (cid:4) l (cid:2) Hence the c (cid:2) (cid:1)m (cid:1) (cid:2) (cid:2) (cid:10) 2 2 (cid:2)h m(cid:2) (cid:1) (cid:1) ;...
https://ocw.mit.edu/courses/8-514-strongly-correlated-systems-in-condensed-matter-physics-fall-2003/57fb8bd5b6d291ec7eda1bc1c920a403_lec6.pdf
a local density approximation(cid:1) i(cid:2)e(cid:2) treating each small part of the BEC cloud as a uniform system(cid:2) For the latter(cid:1) as we already know(cid:1) (cid:16) ...
https://ocw.mit.edu/courses/8-514-strongly-correlated-systems-in-condensed-matter-physics-fall-2003/57fb8bd5b6d291ec7eda1bc1c920a403_lec6.pdf
density distribution n(cid:6)x(cid:7) so that the term (cid:1)n in Eq(cid:2)(cid:6)(cid:11)(cid:23)(cid:7) compensates the potential U (cid:6)x(cid:7) variation in space(cid:1) which gives n(cid:6)x(cid:7) (cid:4) (cid:6)(cid:11)(cid:26)(cid:7) (cid:0) (cid:6)(cid:5) U (cid:6)x(cid:7)(cid:7) (cid:6)(cid:1) (cid:3) U (c...
https://ocw.mit.edu/courses/8-514-strongly-correlated-systems-in-condensed-matter-physics-fall-2003/57fb8bd5b6d291ec7eda1bc1c920a403_lec6.pdf
3)(cid:3)(cid:9) R (cid:4) N (cid:6)(cid:9)(cid:9)(cid:7) (cid:0) (cid:1) (cid:16)(cid:7)m (cid:13) (cid:1) For large N N (cid:1) the radius R is much larger than the BEC healing length (cid:11) (cid:4) h(cid:6) (cid:8) (cid:1)nm c p estimated for typical density n (cid:4) N(cid:6) R (cid:1) which determines the scale ...
https://ocw.mit.edu/courses/8-514-strongly-correlated-systems-in-condensed-matter-physics-fall-2003/57fb8bd5b6d291ec7eda1bc1c920a403_lec6.pdf
BEC coexists with the broad thermal distribution(cid:2) As temperature goes down and becomes very small(cid:1) the thermal component in the density distribution disappears(cid:1) and one obtains the zero(cid:0)temperature state (cid:6)(cid:9)(cid:15)(cid:7)(cid:2) (cid:0)(cid:1)(cid:6) Finite e(cid:7)ects(cid:8) quasip...
https://ocw.mit.edu/courses/8-514-strongly-correlated-systems-in-condensed-matter-physics-fall-2003/57fb8bd5b6d291ec7eda1bc1c920a403_lec6.pdf
Golden Rule for transition rate(cid:25) W (cid:4) f i (cid:16) (cid:6)E E (cid:7) (cid:6)(cid:9)(cid:16)(cid:7) i f int f i ! (cid:9)(cid:7) (cid:1) (cid:8)h jh jH j ij (cid:0) f X In a normal Bose gas(cid:1) at T (cid:10) T (cid:1) the rate of scattering out of the state i is BEC j i out (cid:25) (cid:4) M f f (cid:6)...
https://ocw.mit.edu/courses/8-514-strongly-correlated-systems-in-condensed-matter-physics-fall-2003/57fb8bd5b6d291ec7eda1bc1c920a403_lec6.pdf
Features(cid:25) (cid:6)(cid:9)(cid:21)(cid:7) The rate vanishes in equilibrium(cid:1) since (cid:11) (cid:10) f (cid:4) e f (cid:1) etc(cid:2) j j df i (cid:6) (cid:7) j (cid:3) dt For near(cid:0)equilibrium distribution(cid:1) (cid:4) (cid:6)f f (cid:7) w ith (cid:4) (cid:7)a nv the classical dt (cid:8) (cid:8) i T i...
https://ocw.mit.edu/courses/8-514-strongly-correlated-systems-in-condensed-matter-physics-fall-2003/57fb8bd5b6d291ec7eda1bc1c920a403_lec6.pdf
:1) (cid:4) M (cid:6)f f (cid:6)(cid:11) (cid:10) f (cid:7) (cid:6)(cid:11) (cid:10) f (cid:7)(cid:6)(cid:11) (cid:10) f (cid:7)f (cid:7) (cid:16) (cid:6)(cid:8) (cid:10) (cid:8) (cid:8) (cid:7) (cid:6)(cid:14)(cid:11)(cid:7) ij(cid:5)m i j m i j m i j m dt h(cid:8) (cid:0) j j (cid:0) (cid:0) ij(cid:5)m X Due to the p...
https://ocw.mit.edu/courses/8-514-strongly-correlated-systems-in-condensed-matter-physics-fall-2003/57fb8bd5b6d291ec7eda1bc1c920a403_lec6.pdf
6)(cid:16)(cid:17)(cid:8) c h (cid:7)T T T (cid:9) (cid:9) (cid:4) (cid:6) (cid:2) f (cid:6)(cid:2) (cid:8) (cid:7) (cid:4) p p n (cid:5) (cid:5)(cid:3)(cid:1) (cid:7) (cid:14) (cid:0) Z (cid:6)(cid:9)(cid:7)(cid:8) h (cid:7) (cid:7) BEC (cid:9) BEC (cid:9)(cid:6)T (cid:6)T (cid:7) T T (cid:15) T (cid:7) where T (cid:4...
https://ocw.mit.edu/courses/8-514-strongly-correlated-systems-in-condensed-matter-physics-fall-2003/57fb8bd5b6d291ec7eda1bc1c920a403_lec6.pdf
:1) a collective mode that appears in the two(cid:0)(cid:18)uid regime(cid:2) In this mode(cid:1) the relative fraction of the normal and super(cid:18)uid component oscillates and can propagate in a sound(cid:0)like fashion(cid:2) We consider a uniform system(cid:1) in the absence of external potential(cid:2) The total...
https://ocw.mit.edu/courses/8-514-strongly-correlated-systems-in-condensed-matter-physics-fall-2003/57fb8bd5b6d291ec7eda1bc1c920a403_lec6.pdf
For super(cid:18)uid velocity(cid:1) o n e can write m(cid:2) (cid:4) (cid:5) (cid:6)(cid:14)(cid:17)(cid:7) v t s (cid:0)r This relation(cid:1) derived above from the Gross(cid:0)Pitaevskii equation(cid:1) is in fact very general(cid:1) and is true for any super(cid:18)uid(cid:2) It follows from the relation between t...
https://ocw.mit.edu/courses/8-514-strongly-correlated-systems-in-condensed-matter-physics-fall-2003/57fb8bd5b6d291ec7eda1bc1c920a403_lec6.pdf
9) From Eqs(cid:2)(cid:6)(cid:14)(cid:19)(cid:7)(cid:1)(cid:6)(cid:14)(cid:17)(cid:7)(cid:1)(cid:6)(cid:14)(cid:14)(cid:7)(cid:1) combined with (cid:4) (cid:9) (cid:10) (cid:9) (cid:1) obtain j v v s s n n (cid:2) (cid:6) (cid:7) (cid:4) (cid:18) T (cid:6)(cid:14)(cid:21)(cid:7) v v t n s (cid:9) (cid:0) (cid:0) r (cid...
https://ocw.mit.edu/courses/8-514-strongly-correlated-systems-in-condensed-matter-physics-fall-2003/57fb8bd5b6d291ec7eda1bc1c920a403_lec6.pdf
with (cid:1) (cid:1) (cid:1) (cid:2) p (cid:9) (cid:18) T (cid:2) p T s (cid:1) (cid:1) c (cid:4) (cid:3) c (cid:4) (cid:3) c (cid:4) (cid:6)(cid:16)(cid:16)(cid:7) (cid:3) (cid:1) (cid:5) (cid:0) (cid:1) (cid:0) (cid:1) (cid:9) C (cid:9)C (cid:2)(cid:9) (cid:2)T n T (cid:10) (cid:20) (cid:20) The constants c and c are...
https://ocw.mit.edu/courses/8-514-strongly-correlated-systems-in-condensed-matter-physics-fall-2003/57fb8bd5b6d291ec7eda1bc1c920a403_lec6.pdf
16)(cid:17)(cid:7) (cid:0) (cid:1) (cid:0) (cid:1) (cid:0) (cid:1) (cid:2)T (cid:2)V (cid:2)(cid:9) (cid:0) (cid:10) T T The I a n d I I sound velocities(cid:1) obtained from Eq(cid:2)(cid:6)(cid:16)(cid:14)(cid:7)(cid:1) are (cid:1) (cid:1) (cid:1) (cid:1) (cid:11) (cid:11) (cid:1) (cid:1) (cid:1) (cid:1) (cid:1) (cid...
https://ocw.mit.edu/courses/8-514-strongly-correlated-systems-in-condensed-matter-physics-fall-2003/57fb8bd5b6d291ec7eda1bc1c920a403_lec6.pdf
ov quasiparticles at low energies(cid:1) (cid:9) (cid:7) u (cid:4) (cid:1)n(cid:6)m(cid:3) (cid:6)(cid:16)(cid:21)(cid:7) I while the I I sound velocity i s (cid:14) times lower(cid:1) p q u (cid:4) u (cid:6) (cid:14) (cid:4) (cid:1)n(cid:6)(cid:14)m(cid:3) (cid:6)(cid:16)(cid:23)(cid:7) II I p The velocity u decreases...
https://ocw.mit.edu/courses/8-514-strongly-correlated-systems-in-condensed-matter-physics-fall-2003/57fb8bd5b6d291ec7eda1bc1c920a403_lec6.pdf
18.409 An Algorithmist’s Toolkit October 6, 2009 Lecturer: Jonathan Kelner Scribe: Alessandro Chiesa (2009) Lecture 8 1 Administrivia You should probably know that • the first problem set (due October 15) is posted on the class website, and • its hints are also posted there. Also, today in class there was a maj...
https://ocw.mit.edu/courses/18-409-topics-in-theoretical-computer-science-an-algorithmists-toolkit-fall-2009/5823d396bfe2ae2c5bea9c431601276e_MIT18_409F09_scribe8.pdf
in some graph G, we would like know if v is contained in a cluster, i.e. a subset of vertices that defines a cut with low conductance. However, we want the running time of our algorithm to depend on the cluster size, and not on the size of the graph. Last time we mentioned that a good example of a problem of this sor...
https://ocw.mit.edu/courses/18-409-topics-in-theoretical-computer-science-an-algorithmists-toolkit-fall-2009/5823d396bfe2ae2c5bea9c431601276e_MIT18_409F09_scribe8.pdf
that the random walk is at w; then take the vertices with the k largest probability masses as a possible cut. Repeat this until you get a good cut or you reach a predetermined limit. 2.3 Obstacles We need a bound that says that our general strategy works, and that is why we proved the Lov´asz-Simonovits theorem. H...
https://ocw.mit.edu/courses/18-409-topics-in-theoretical-computer-science-an-algorithmists-toolkit-fall-2009/5823d396bfe2ae2c5bea9c431601276e_MIT18_409F09_scribe8.pdf
time t, if x ≡ w∈W � v∈V dv � � � � � � w∈W p t(w) − π(w)� ≤ min � � � � �√ √ � x, 2m − x � 1 1 − φ(W )2 2 �t . Note that in the last lecture we stated a slightly weaker form of the theorem, where the conductance ϕ(W ) of the cut (W, W ) was replaced by the conductance φ(G) of the whole graph. Ne...
https://ocw.mit.edu/courses/18-409-topics-in-theoretical-computer-science-an-algorithmists-toolkit-fall-2009/5823d396bfe2ae2c5bea9c431601276e_MIT18_409F09_scribe8.pdf
discuss this approach in the following section. (Note that for all of this to work we still need to prove a partial converse. Indeed, one can show that if there exists a cut C of conductance φ2, then at least |C|/2 of its vertices will give a cut of conductance φ, otherwise the random walk would mix too quickly.) ...
https://ocw.mit.edu/courses/18-409-topics-in-theoretical-computer-science-an-algorithmists-toolkit-fall-2009/5823d396bfe2ae2c5bea9c431601276e_MIT18_409F09_scribe8.pdf
we denote this stationary distribution (which depends on the parameters s and α) by prα(s) and call it the PageRank vector ; note that prα(s) is a vector in Rn, where n = |V |. Moreover, it is easy to see that the stationary distribution prα(s) is the unique solution to the following equation: prα(s) = αs + (1 − α)...
https://ocw.mit.edu/courses/18-409-topics-in-theoretical-computer-science-an-algorithmists-toolkit-fall-2009/5823d396bfe2ae2c5bea9c431601276e_MIT18_409F09_scribe8.pdf
O( α)). 3.2 Properties We now prove three properties about the stationary distribution prα. Proposition 2 (Uniqueness) prα(s) is unique. Proof We must show that Equation (1) has a unique solution. Rewrite the equation as (I − (1 − α)W )prα(s) = αs. The matrix I − (1 − α)W is strictly diagonally dominant2 because t...
https://ocw.mit.edu/courses/18-409-topics-in-theoretical-computer-science-an-algorithmists-toolkit-fall-2009/5823d396bfe2ae2c5bea9c431601276e_MIT18_409F09_scribe8.pdf
α)W (cprα(v) + dprα(w)) = αcv + (1 − α)W cprα(v) + αdw + (1 − α)W dprα(w) = cprα(v) + dprα(w) (cid:3) = x . By Proposition 2, the equation has a unique solution, so that x = x(cid:3) and the result follows. Proposition 4 (Commutativity with W ) prα(W s) = W prα(s). Proof By definition, the vector x ≡ prα(W s) sat...
https://ocw.mit.edu/courses/18-409-topics-in-theoretical-computer-science-an-algorithmists-toolkit-fall-2009/5823d396bfe2ae2c5bea9c431601276e_MIT18_409F09_scribe8.pdf
Starting with initial values p = 0 and r = s, in each iteration, we pick a vertex u, and update the two vectors p and r to the new vectors p(cid:3) and r(cid:3) defined as follows: p (cid:3) = p + αr(u)χu , r (cid:3) = r − r(u)χu + (1 − α)r(u)W χu . The vector χu is the characteristic vector of u, i.e., the vector wit...
https://ocw.mit.edu/courses/18-409-topics-in-theoretical-computer-science-an-algorithmists-toolkit-fall-2009/5823d396bfe2ae2c5bea9c431601276e_MIT18_409F09_scribe8.pdf
Using the triangle inequality, ||r (cid:3)||1 = ||r − r(u)χu + (1 − α)r(u)W χu||1 ≤ ||r − r(u)χu||1 + (1 − α)r(u)||W χu||1 . However, ||W χu||1 ≤ 1. Indeed, the ith element of W χu is 1 when i =(cid:8) u and 1 when i = u. Therefore, 2d(u) 2 ||r (cid:3)||1 ≤ ||r||1 − r(u) + (1 − α)r(u) = ||r||1 − αr(u) , as desired...
https://ocw.mit.edu/courses/18-409-topics-in-theoretical-computer-science-an-algorithmists-toolkit-fall-2009/5823d396bfe2ae2c5bea9c431601276e_MIT18_409F09_scribe8.pdf
9). We deduce that the algorithm must terminate in at most O (cid:4)α 1 Next, by definition, the process terminates when there are no more vertices u such that r(u) ≥ (cid:9)d(u). iterations. Therefore, condition (1) is automatically satisfied. Moreover, if we let T denote the number of iterations that the algorithm ...
https://ocw.mit.edu/courses/18-409-topics-in-theoretical-computer-science-an-algorithmists-toolkit-fall-2009/5823d396bfe2ae2c5bea9c431601276e_MIT18_409F09_scribe8.pdf
C with conductance φ2, we will find with probability at least 0.5 a set with volume at least vol(C)/2. However, this holds only if we use “appropriate” parameters α and (cid:9), which we do not know! The fix is to binary search over the (C) . α 8-5 possibilities, incurring an additional cost that is only a logarit...
https://ocw.mit.edu/courses/18-409-topics-in-theoretical-computer-science-an-algorithmists-toolkit-fall-2009/5823d396bfe2ae2c5bea9c431601276e_MIT18_409F09_scribe8.pdf
fication Sparsification is a technique used in dynamic graph algorithms to reduce the dependence of an algorithm’s time on the number of edges in a graph. We briefly motivate this technique now, and will discuss it next time. Suppose that we have a graph G = (V, E) with m = Θ(n2) edges. We would like to solve some cut...
https://ocw.mit.edu/courses/18-409-topics-in-theoretical-computer-science-an-algorithmists-toolkit-fall-2009/5823d396bfe2ae2c5bea9c431601276e_MIT18_409F09_scribe8.pdf