is conservative. PCM subsystems would more likely experience a mix of compute and memory-intensive workloads.
Expected lifetimes would be higher had we considered, for
example, single-threaded SPEC integer workloads. However,
such workloads are less relevant for a study of memory subsystems. Moreover, within memory-intensive workloads, we
would expect to see a mix of read and write intensive applications, which may further increase lifetimes.
Scalability is projected to improve PCM endurance
from the present 1E+08 to 1E+ 12 writes per bit at 32nm
with known manufacturable solutions. 17 This higher
endurance increases lifetime by four orders of magnitude in our models. ITRS anticipates 1E+ 15 PCM writes
at 22nm although manufacturable solutions are currently
The proposed memory architecture lays the foundation
for exploiting PCM scalability and nonvolatility in main
memory. Scalability implies lower main memory energy
and greater write endurance. Furthermore, nonvolatile
main memories will fundamentally change the landscape
of computing. Software cognizant of this newly provided
persistence can provide qualitatively new capabilities. For
example, system boot/hibernate will be perceived as instantaneous; application checkpointing will be inexpensive7; file
systems will provide stronger safety guarantees. 6 Thus, we
take a step toward a new memory hierarchy with deep implications across the hardware–software interface.
1. Aslot, V., eigenmann, R. Quantitative
performance analysis of the sPeC
oMPM2001 benchmarks. Sci. Program.
11, 2 (2003).
2. bailey, D. et al. nAs parallel benchmarks.
In Technical Report RNR-94-007, NASA
Ames Research Center, March 1994.
3. bedeschi, F. et al. A multi-level-cell
memory. In International Solid-State
Circuits Conference, 2008.
4. Chen, y. et al. ultra-thin phase-change
bridge memory device using Gesb. In
International Electron Devices Meeting, 2006.
5. Choi, y. under the hood: DRAM architectures: 8F2 vs. 6F2. EE Times, February 2008.
6. Condit, j. et al. better I/o through
byte-addressable, persistent memory.
In Symposium on Operating System
Principles, oct 2009.
7. Dong, X. et al. leveraging 3D PCRAM
technologies to reduce checkpoint
overhead in future exascale systems.
In Conference on Supercomputing,
8. horii, h. et al. A novel cell technology
using n-doped Gesb Te films for phase
change RAM. In Symposium on VLSI
9. lai, s. Current status of the phase change
memory and its future. In International
Electron Devices Meeting, 2003.
10. lee, b., Ipek, e., Mutlu, o., burger, D.
Architecting phase change memory
© 2010 ACM 0001-0782/10/0700 $10.00
Benjamin C. Lee ( firstname.lastname@example.org),
Engin Ipek ( email@example.com),
university of Rochester.
as a scalable DRAM alternative. In
International Symposium on Computer
Architecture, june 2009.
11. lee, K.-j. et al. A 90 nm 1. 8 V 512 Mb
diode-switch PRAM with 266 Mb/s read
throughput. J. Solid State Circuit. 43, 1
12. Micron. 512 Mb DDR2 sDRAM component data sheet: MT47h128M4b6-25.
In www.micron.com, Mar 2006.
13. nirschl, T. et al. Write strategies for 2
and 4-bit multi-level phase-change
memory. In International Electron
Devices Meeting, 2008.
14. Pirovano, A. et al. scaling analysis of
phase-change memory technology. In
International Electron Devices Meeting, 2003.
15. Raoux, s. et al. Phase-change random
access memory: A scalable technology.
IBM J. Res. Dev. 52, 4/5 (jul/sept 2008).
16. Renau, j. et al. sesC simulator. In
17. semiconductor Industry Association. Process integration, devices &
structures. International Technology
Roadmap for Semiconductors, 2007.
18. sinha, M. et al. high-performance and
low-voltage sense-amplifier techniques
for sub- 90 nm sram. In International
Systems-on-Chip Conference, 2003.
19. Woo, s. et al. The sPlAsh- 2 programs:
Characterization and methodological considerations. In International Symposium
on Computer Architecture, june 1995.
Onur Mutlu ( firstname.lastname@example.org), Carnegie
Doug Burger ( email@example.com),
You’ve come a long way.
Share what you’ve learned.
ACM has partnered with MentorNet, the award-winning nonprofit e-mentoring network in engineering,
science and mathematics. MentorNet’s award-winning One-on-One Mentoring Programs pair ACM
student members with mentors from industry, government, higher education, and other sectors.
• Communicate by email about career goals, course work, and many other topics.
• Spend just 20 minutes a week - and make a huge difference in a student’s life.
• Take part in a lively online community of professionals and students all over the world.
Make a difference to a student in your field.
Sign up today at: www.mentornet.net
Find out more at: www.acm.org/mentornet
MentorNet’s sponsors include 3M Foundation, ACM, Alcoa Foundation, Agilent Technologies, Amylin Pharmaceuticals, Bechtel Group Foundation, Cisco
Systems, Hewlett-Packard Company, IBM Corporation, Intel Foundation, Lockheed Martin Space Systems, National Science Foundation, Naval Research
Laboratory, NVIDIA, Sandia National Laboratories, Schlumberger, S.D. Bechtel, Jr. Foundation, Texas Instruments, and The Henry Luce Foundation.