multi-cores: architecture/vlsi perspective
DESCRIPTION
Multi-Cores: Architecture/VLSI Perspective. The Hardware-Software Relationship: Date or Dump?. Embedded Applications -- Spencer. From discreet cochlear implants to high-end biomedical imaging! Multi-cores speed up performance by 50x! Creating new application domains!. - PowerPoint PPT PresentationTRANSCRIPT
![Page 1: Multi-Cores: Architecture/VLSI Perspective](https://reader036.vdocuments.mx/reader036/viewer/2022062500/56815448550346895dc25b23/html5/thumbnails/1.jpg)
Oct 31st 2007 University of Utah 1
Multi-Cores: Architecture/VLSI Perspective
The Hardware-Software Relationship:
Date or Dump?
![Page 2: Multi-Cores: Architecture/VLSI Perspective](https://reader036.vdocuments.mx/reader036/viewer/2022062500/56815448550346895dc25b23/html5/thumbnails/2.jpg)
University of Utah
Embedded Applications --Spencer
From discreet cochlear implants to high-end biomedical imaging!
Multi-cores speed up performance by 50x!
Creating new application domains!
![Page 3: Multi-Cores: Architecture/VLSI Perspective](https://reader036.vdocuments.mx/reader036/viewer/2022062500/56815448550346895dc25b23/html5/thumbnails/3.jpg)
University of Utah
How to use “multiple” cores?
Oct 31st 2007
3
Parallel programmin
g
Synchronization
Deadlock
Livelock
Memory management
![Page 4: Multi-Cores: Architecture/VLSI Perspective](https://reader036.vdocuments.mx/reader036/viewer/2022062500/56815448550346895dc25b23/html5/thumbnails/4.jpg)
University of Utah
Oct 31st 2007
4
How to use “multiple” cores?
Program = Communicat
ion + Computatio
n
Global restructuring and parallelization
![Page 5: Multi-Cores: Architecture/VLSI Perspective](https://reader036.vdocuments.mx/reader036/viewer/2022062500/56815448550346895dc25b23/html5/thumbnails/5.jpg)
University of Utah
Oct 31st 2007
5
Structured “Communication”
Lang: StreamIt, MPI
Compilers: RAW, CoGenE
Architecture: TRIPS, HWRT
Key: Help other levels and leverage communication
![Page 6: Multi-Cores: Architecture/VLSI Perspective](https://reader036.vdocuments.mx/reader036/viewer/2022062500/56815448550346895dc25b23/html5/thumbnails/6.jpg)
University of Utah
Another Constraint?
Oct 31st 2007
6
Parallel programmin
g
Synchronization
Deadlock
Livelock
Memory management
Hey..
Surprise!!!
Communication
Scheduling
![Page 7: Multi-Cores: Architecture/VLSI Perspective](https://reader036.vdocuments.mx/reader036/viewer/2022062500/56815448550346895dc25b23/html5/thumbnails/7.jpg)
University of Utah
Another Constraint?
Oct 31st 2007
7
Oh God!!!
Communication
Scheduling
![Page 8: Multi-Cores: Architecture/VLSI Perspective](https://reader036.vdocuments.mx/reader036/viewer/2022062500/56815448550346895dc25b23/html5/thumbnails/8.jpg)
University of Utah
Focus of Architecture Research
Reduce the load of programmers Hardware transactional memory
Aggressive pre-fetching
Dynamic reconfiguration at every possible level
Keep the architectural innovations transparent
to compilers or programmers Learn from the mistakes of ITANIUM !
Remember the success of OOO execution
Oct 31st 2007
8
![Page 9: Multi-Cores: Architecture/VLSI Perspective](https://reader036.vdocuments.mx/reader036/viewer/2022062500/56815448550346895dc25b23/html5/thumbnails/9.jpg)
University of Utah9
Reliability Issues --Niti
Shrinking transistor sizes & lower voltages Increased transient faults, process variations – leakage
power and frequency variations, hard errors, interconnect noise
Many-core – “Many cores” may not work reliably Some cores will end up providing redundancy
Heterogeneous cores may be able to help Simple in-order cores can provide redundancy at low cost
The compute power gain of many-core can get offset by reliability requirements of the system
![Page 10: Multi-Cores: Architecture/VLSI Perspective](https://reader036.vdocuments.mx/reader036/viewer/2022062500/56815448550346895dc25b23/html5/thumbnails/10.jpg)
University of Utah
Oct 31st 2007
10
On-Chip Sensor Networks --Nathaniel, Amlan
Analog sensors everywhere! Need to monitor power, voltage droop,
variation, critical paths, delays, slew
rates, etc.
Control system to react to changes.
In multi-core, sensor network will only
grow.
Xeon and Itanium processors JSSC Jan 06 & 07
On-chip sensors