2. Outline Motivation and background Read Write
Read “seek” time
Good sequential read
Write
Write log block (buffer)‏
“1” !=”0”?
Conclusion and future work

3. Motivation Era for Flash? We only get BLACK BOX.
Nice properties: Fast random access; Shock / Temperature resistance ; Smaller size and weight.
Energy saving.
Reducing price + increasing volume.
We only get BLACK BOX.
Industry company may not tell?
No systematic performance study literature?

4. Flash Memory Organization Read and Write Each page: 2KB,4KB
Each block: 64,128 pages
Read and Write
Read in pages ~50us
Write in pages ~800us
Any write needs block erase ~1500us
Block 0
Block 1
Page 0
Block n-1
Page 1
Page 2
......
Page 3
......
Page m-1
Chip organization

5. Outline Motivation and background Read Write
Read “seek” time.
Good sequential read.
Write
Write buffer size.
“1” !=”0”?
Conclusion and future work

6. Read Assumptions and basic knowledge Experiments Setup
Uniform random read time?
Good sequential read performance?
Experiments Setup
Fedora Core 7.0,1GHZ CPU, 1G Memory
Flash memory
(I) Kingston DataTraveler 1G
(II) PNY Attaché 2G
(III) Unknown 1G

7. Random Read-- “seek” time

8. Sequential Read –good!

9. Sequential Read –Scale up

10. Read –what we can do? Technology aware FS
Block group VS Cylinder group
To Group files.
Random read is good but not perfect.
To decrement random accesses.

11. Outline Motivation and background Read Write Conclusion
Read “seek” time
Good sequential read
Write
Write log block (buffer)‏
“1” !=”0”?
Conclusion

12. Write Assumptions and basic knowledge Bad random write performance
Needs block erase (1 page--> block erase)‏
Good sequential write performance
Limited block erase times (64 pages--> block erase)
Reason : Log (buffer) and Merge

13. Write -- merge 1.Write valid valid valid 2.merge 2.merge valid valid
Log Block Pool
Data Block
Free Data Block
Log Block
3.erase
3.erase

14. Random Write -- bad!

15. Continuous write – great relief from erase

16. Write -- log block What is it? Why is it used?
Flash block as write buffer
Correspond to one flash block at one time
Why is it used?
Hard disk : clustering writes; save redundant
Flash disk: reduce erase times
Interesting: Log block size and usage

17. Log block -- size Motivation Size: 64 pages 128 pages
Trade off: Large merge time VS frequent merge
Size: 64 pages pages

18. Determine size of Log block pool
Repeat writing one page into set of continuous flash blocks sequentially. Check the time cost.

19. Log block pool -- Use Strategy
Log block pool use FIFO to reclaim used blocks
Repeatedly writing less than 16 pages into one flash block does not trigger data merge.

20. “0” != “1”

21. How about 50%?

22. Write -- what we can do? New file system for flash
Modified LFS Log block !=write buffer
Special policy for frequently updated data, e.g inodes
Anticipatory scheduling
More flexible. Directly execute any write request in one data block associated with log block.
Flip “1” to “0”, “0” to “1” may save time (attributed to Remzi)‏

23. Conclusion
Comprehensive study of the read and write performance of flash memory
Design a relatively systematic method to study the flash memory as a black box
Find some interesting and potentially useful properties, e.g. “1”!=”0” ; “seek” time
Apply similar performance study strategy to SSD and check whether the properties still hold

24. Q&A

25. Random Read-- “seek” time

26. Random Read-- “seek” time