Skip to content

THE COURSE / M10 MEMORY / M10.4

Sense, test and repair

After this lesson you can:

  • Say how wafer test finds a die's failing bits and maps them.
  • Cover a die's faults with spare rows and columns, using the must-repair rule.
  • Explain how the laser swaps the spares in, and what burn-in is for.

Loading the lesson…

Sources

  1. P. Öhler, S. Hellebrand and H.-J. Wunderlich, "An Integrated Built-in Test and Repair Approach for Memories with 2D Redundancy," 12th IEEE European Test Symposium, May 2007. The most common memory repair is 2D redundancy, spare rows and spare columns; the memory is tested and its faults collected in a failure bitmap; repair analysis then allocates spares to cover every fault at minimum cost, or finds the memory not repairable; Kuo and Fuchs showed optimal 2D redundancy allocation is NP-complete; must repair: more faults in a row than there are spare columns means that row must take a spare row, and likewise for columns.
  2. ESI, 9850TPIR+ (product page, archived 2019). A dual-beam, tailored-pulse laser link-processing system for DRAM, flash and embedded memory: fuse processing through Cu and Al interconnect layers; a "chair-shaped" pulse whose fast high peak ruptures the fuse's overlying dielectric, then lower energy cleanly removes the remaining fuse; ESI Pyro 2 W laser, 1.0 µm, 1.5 µm minimum spot, 0.15 µm accuracy, 150 kHz, up to 400 mm/s; 200 and 300 mm wafers.
  3. Micron Technology (K. W. Marr), "Method, circuit and system for determining burn-in reliability from wafer level burn-in," US 6,943,575 B2 (filed 29 July 2002, granted 13 September 2005). The reliability curve's three regions: infant mortality, random failures and wearout; infant-mortality failures, though a small fraction of parts, are the largest contributor to early-life failure rates; burn-in stresses ICs with high temperature and raised supply voltage; conventional burn-in tests packaged parts.
  4. J. Liu, B. Jaiyen, R. Veras and O. Mutlu, "RAIDR: Retention-Aware Intelligent DRAM Refresh," ISCA 2012. A DRAM cell is a capacitor and an access transistor, connected to a bitline and controlled by a wordline; cells sharing a wordline form a row; each bank has a row of sense amplifiers, one per bitline; opening a row from bitlines precharged to VDD/2 shares each cell's charge with its bitline, and the sense amplifier drives the bitline fully to VDD or 0 V, recharging the cells; a row is refreshed by opening it; the refresh interval has stayed 64 ms for several generations; tREFI 7.8 µs below 85 °C, 3.9 µs from 85 to 95 °C; retention log-normal, set by the leakiest cell: in 32 GB, over 10¹¹ cells, about 30 can't wait twice as long and about 1,000 four times as long.