200 likes | 289 Views
Emerging Technologies: A CompSci Perspective. Tim Sherwood. UC SANTA BARBARA. Software Beware. The End of an Era. $381B / year. The Beginning of a New Era. 3D Stacks of Dies. 80 Cores. Integrated MEMS. The Role of Architecture. Demands. SW. HW. Constraints.
E N D
Emerging Technologies:A CompSci Perspective Tim Sherwood UC SANTA BARBARA
The End of an Era $381B / year
The Beginning of a New Era 3D Stacks of Dies 80 Cores Integrated MEMS
The Role of Architecture Demands SW HW Constraints (Battery Life, Performance, Programmability ) Applications Runtime System Architecture Emerging Technology Circuit Device Package (Noise, Thermal, Yield)
A Simple Performance “Ecosystem” package temp total power dynamic power leakage V communication utilized area freq parallelism feedback OS or runtime No multicore, no spatial variance, no temporal variance, no metrics of cost or error or yield app chip performance
3D Integration 3D Stacks of Dies 80 Cores Integrated MEMS
3d technology 5x5μm • Science Fiction? • Intel, IBM, Ziptronix invest heavily in 3d integration research • Many demonstrated 3d prototype systems activelayer ThroughSiliconVias (TSV) CMP Reduced Si Layer Standard Si Substrate
Work at UCSB • Shashidhar Mysore, Banit Agrawal, Sheng-Chih Lin, Navin Srivastava, Kaustav Banerjee and Timothy Sherwood. Introspective 3D Chips , Proceedings of the Twelfth International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS), October 2006. San Jose, CA • Gian Luca Loi, Banit Agrawal, Navin Srivastava, Sheng-Chih Lin, Timothy Sherwood, Kaustav Banerjee. A Thermally-Aware Performance Analysis of Vertically Integrated (3-D) Processor-Memory Hierarchy, Proceedings of the 43nd Design Automation Conference (DAC), June 2006. San Francisco, CA
Basic Savings in 3D Area: 4 Dist: √8 ≈ 2.8 Area: 2 Dist: √4 ≈ 2 + 1L Area: 1 Dist: √2 ≈ 1.4 + 3L BW: √8 ≈ 2.8 BW: 4√2 ≈ 5.6 BW: 2√4 ≈ 4 On-chip Latency improvedBandwidth could improve even more UCSB First to Successfully Model Thermal/Performance
Addressing more than Performance • The hardware/software boundary is uniquely situated • Ultimately, Everything is an instruction • Used by Intel, AMD, Freescale, to guide their development • Could Provide Unprecedented Visibility • Not just data capture, we need the ability to put togethera cohesive picture of system interactions and correlate between them in a sound and non-intrusive manner
Cutting Through Abstraction Complex interactions across levels of abstraction make debugging, optimizing, securing, and analysis in general difficult
What programmers want L1_BPU Decode L2_BPU Bus Control Trace CacheTop MOB ITLB Trace Cache Bottom DTLB L1CacheTop L2 Cache UROM FP Exec 2 L1 CacheBottom 320 3 2 FP Reg To Integrated Monitoring Hardware Int Exec MemCtl Alloc Retire 790 Int Reg Rename InstrQ1 Sched InstrQ2 Less buggy systems ($54 Billion / Year ) 32 bit Memory Address 32 bit Memory Value 10 bit Opcodes 2, 5 bit Register Names 2, 32 bit Register Values 10 bits of “status” 4x 4x 4x 4x 4x 4x 3x 3x 3x 3x 3x 3x 1892 bits per cycle = 1 terrabyte/sec @ 4Ghz
Why programmers cant have it L1_BPU Decode L2_BPU Bus Control Trace CacheTop MOB ITLB Trace Cache Bottom DTLB L1CacheTop L2 Cache UROM FP Exec 2 L1 CacheBottom 320 3 2 FP Reg To Integrated Monitoring Hardware Int Exec MemCtl Alloc Retire 790 Int Reg Rename InstrQ1 Sched InstrQ2 • Interconnect is not free • Huge cross chip busses • OptBuf 285um • 20,000 buffers • Analysis is not free • Significant processing required • Extra cost of added heat • $15 budget for cooling • Used by developers
3D Introspection 5x5μm activelayer Observer PrimaryProcessor CMP Reduced Si Layer Standard Si Substrate
Thermal Impact P4 – Base Case w/ 3D Layer w/ 4x Processing w/ 8x Processing Processing Layer Analysis Layer
Conclusions • Emerging Technologies will play a significant new role • Risk is hard to avert right now • UCSB Computer Science and Engineering • Collaborate across disciplines to consider the entire SW/HW • Research that is driving industry • UCSB Technology in use in most Microprocessors and Networks • Always looking for more collaboration with industry partners
http://www.cs.ucsb.edu/~arch/ NSF CNS 0524771, NSF CCF 0702798, NSF CCF 0448654