![Page 1: T03160010220104036 multipleproc week11-1-pert 21](https://reader036.vdocument.in/reader036/viewer/2022062420/55c3a540bb61eb3d338b482d/html5/thumbnails/1.jpg)
![Page 2: T03160010220104036 multipleproc week11-1-pert 21](https://reader036.vdocument.in/reader036/viewer/2022062420/55c3a540bb61eb3d338b482d/html5/thumbnails/2.jpg)
Multi-Processing SystemWeek 11 - 1
Course : T0316 - Operating System Year : February 2011
![Page 3: T03160010220104036 multipleproc week11-1-pert 21](https://reader036.vdocument.in/reader036/viewer/2022062420/55c3a540bb61eb3d338b482d/html5/thumbnails/3.jpg)
Multiple Processor System
Bina Nusantara University 3
(a) A shared-memory multiprocessor. (b) A message-passing multicomputer. (c) A wide area distributed system.
![Page 4: T03160010220104036 multipleproc week11-1-pert 21](https://reader036.vdocument.in/reader036/viewer/2022062420/55c3a540bb61eb3d338b482d/html5/thumbnails/4.jpg)
UMA Multiprocessors with Bus-Based Architectures
Bina Nusantara University 4
Three bus-based multiprocessors. (a) Without caching. (b) With caching. (c) With caching and private memories.
![Page 5: T03160010220104036 multipleproc week11-1-pert 21](https://reader036.vdocument.in/reader036/viewer/2022062420/55c3a540bb61eb3d338b482d/html5/thumbnails/5.jpg)
UMA vs. NUMA
Bina Nusantara University 5
• Unified Memory Access (UMA)– also known as Symmetric MultiProcessors (SMP)
• Non-Unified Memory Access (NUMA)• SAME
– single memory space– pitfall: Compared between shared memory and
distributed memory.
(Irrelevant Comparison)• DIFFERENT
– access time– # of processors– bus vs. network implementation
![Page 6: T03160010220104036 multipleproc week11-1-pert 21](https://reader036.vdocument.in/reader036/viewer/2022062420/55c3a540bb61eb3d338b482d/html5/thumbnails/6.jpg)
NUMA Processors(1)
Bina Nusantara University 6
Characteristics of NUMA machines:1.There is a single address space visible to
all CPUs.2.Access to remote memory is via LOAD and
STORE instructions.3.Access to remote memory is slower than
access to local memory.
![Page 7: T03160010220104036 multipleproc week11-1-pert 21](https://reader036.vdocument.in/reader036/viewer/2022062420/55c3a540bb61eb3d338b482d/html5/thumbnails/7.jpg)
NUMA Processors(2)
Bina Nusantara University 7
(a) A 256-node directory-based multiprocessor.
![Page 8: T03160010220104036 multipleproc week11-1-pert 21](https://reader036.vdocument.in/reader036/viewer/2022062420/55c3a540bb61eb3d338b482d/html5/thumbnails/8.jpg)
Each CPU Has Its Own Operating System
Bina Nusantara University 8
Partitioning multiprocessor memory among four CPUs, but sharing a single copy of the operating system code. The boxes marked Data are the operating
• system’s private data for each CPU.
![Page 9: T03160010220104036 multipleproc week11-1-pert 21](https://reader036.vdocument.in/reader036/viewer/2022062420/55c3a540bb61eb3d338b482d/html5/thumbnails/9.jpg)
Master-slave Multiprocessor Model
Bina Nusantara University 9
![Page 10: T03160010220104036 multipleproc week11-1-pert 21](https://reader036.vdocument.in/reader036/viewer/2022062420/55c3a540bb61eb3d338b482d/html5/thumbnails/10.jpg)
Multi Processor Synchronization(1)
Bina Nusantara University 10
• Figure 8-10. The TSL instruction can fail if the bus cannot be locked. These four steps show a sequence of events where the failure is demonstrated.
![Page 11: T03160010220104036 multipleproc week11-1-pert 21](https://reader036.vdocument.in/reader036/viewer/2022062420/55c3a540bb61eb3d338b482d/html5/thumbnails/11.jpg)
Multi Processor Synchronization(2)
Bina Nusantara University 11
• Figure 8-11. Use of multiple locks to avoid cache thrashing.
![Page 12: T03160010220104036 multipleproc week11-1-pert 21](https://reader036.vdocument.in/reader036/viewer/2022062420/55c3a540bb61eb3d338b482d/html5/thumbnails/12.jpg)
Time Sharing
Bina Nusantara University 12
• Figure 8-12. Using a single data structure for scheduling a multiprocessor.
![Page 13: T03160010220104036 multipleproc week11-1-pert 21](https://reader036.vdocument.in/reader036/viewer/2022062420/55c3a540bb61eb3d338b482d/html5/thumbnails/13.jpg)
Space Sharing
Bina Nusantara University 13
• Figure 8-13. A set of 32 CPUs split into four partitions, with two CPUs available.
![Page 14: T03160010220104036 multipleproc week11-1-pert 21](https://reader036.vdocument.in/reader036/viewer/2022062420/55c3a540bb61eb3d338b482d/html5/thumbnails/14.jpg)
Gang Scheduling (1)
Bina Nusantara University 14
• Figure 8-14. Communication between two threads belonging to thread A that are running out of phase.
![Page 15: T03160010220104036 multipleproc week11-1-pert 21](https://reader036.vdocument.in/reader036/viewer/2022062420/55c3a540bb61eb3d338b482d/html5/thumbnails/15.jpg)
Gang Scheduling (2)
Bina Nusantara University 15
The three parts of gang scheduling:1.Groups of related threads are scheduled
as a unit, a gang.2.All members of a gang run simultaneously,
on different timeshared CPUs.3.All gang members start and end their time
slices together.
![Page 16: T03160010220104036 multipleproc week11-1-pert 21](https://reader036.vdocument.in/reader036/viewer/2022062420/55c3a540bb61eb3d338b482d/html5/thumbnails/16.jpg)
Gang Scheduling (3)
Bina Nusantara University 16
![Page 17: T03160010220104036 multipleproc week11-1-pert 21](https://reader036.vdocument.in/reader036/viewer/2022062420/55c3a540bb61eb3d338b482d/html5/thumbnails/17.jpg)
The End
Bina Nusantara University 17