Testing Semiconductor

Memories
Lab for Reliable Computing
Dept. Electrical Engineering
National Tsing Hua University
Cheng-Wen Wu
mbist1.10 Cheng-Wen Wu, NTHU
2
Outline
- Introduction
- RAM functional fault models and test
algorithms
- RAM fault-coverage analysis
- Cocktail-March for testing word-
oriented memories
- Testing multi-port RAMs
- Testing CAMs
- Testing flash memories
mbist1.10 Cheng-Wen Wu, NTHU
3
Introduction
- Memory testing is a more and more important issue
RAMs are key components for electronic systems
Memories represent about 30% of the semiconductor market
Embedded memories are dominating the chip yield
- Memory testing is more and more difficult
Growing density, capacity, and speed
Emerging new architectures and technologies
Embedded memories: access, diagnostics & repair, heterogeneity,
custom design, power & noise, scheduling, compression, etc.
- Cost drives the need for more efficient test methodologies
IFA, fault modeling and simulation, test algorithm development and
evaluation, diagnostics, DFT, BIST, BIRA, BISR, etc.
- Test automation is required
Failure analysis, fault simulation, ATG, and diagnostics
BIST/BIRA/BISR generation
mbist1.10 Cheng-Wen Wu, NTHU
4
Typical RAM Production Flow
Wafer
Full Probe Test
Marking
Final Test
Shipping
QA Sample Test Visual Inspection
Burn-In (BI)
Post-BI Test
Laser Repair Packaging
Pre-BI Test
mbist1.10 Cheng-Wen Wu, NTHU
5
Scope of RAM Testing
- Parametric Test: DC & AC
- Reliability Screening
Long-cycle testing
Burn-in: static & dynamic BI
- Functional Test
Device characterization
- Failure analysis
Fault modeling
- Simple but effective (accurate & realistic?)
Test algorithm generation
- Small number of test patterns (data backgrounds)
- High fault coverage
- Short test time
mbist1.10 Cheng-Wen Wu, NTHU
6
RAM Models
- Behavior Level
Verilog/VHDL
- Function Level
Verilog/VHDL/Block diagram
Normally not synthesizable
- Circuit Level
Spice/Schematic
- Layout Level
GDS-II/Geometry
Who should provide the model?
mbist1.10 Cheng-Wen Wu, NTHU
7
Memory Function Model Example
mbist1.10 Cheng-Wen Wu, NTHU
8
RAM Fault Models (Static)
- Address-Decoder Fault (AF)
No cell accessed by certain address
Multiple cells accessed by certain address
Certain cell not accessed by any address
Certain cell accessed by multiple addresses
- Stuck-At Fault (SAF)
Cell (line) SA0 or SA1
- Transition Fault (TF)
Cell fails to transit from 0 to 1 or 1 to 0
mbist1.10 Cheng-Wen Wu, NTHU
9
RAM Fault Models (Static)
- Bridging Fault (BF)
Short between cells
- AND type or OR type
- Stuck-Open Fault (SOF)
Cell not accessible due to broken line
- Neighborhood Pattern Sensitive Fault (NPSF)
Active (Dynamic) NPSF
Passive NPSF
Static NPSF
N
W BC E
S
mbist1.10 Cheng-Wen Wu, NTHU
10
RAM Fault Models (Static)
- Coupling Fault (CF)
State Coupling Fault (CFst)
- Coupled (victim) cell is forced to 0 or 1 if coupling
(aggressor) cell is in given state
Inversion Coupling Fault (CFin)
- Transition in coupling cell complements (inverts)
coupled cell
Idempotent Coupling Fault (CFid)
- Coupled cell is forced to 0 or 1 if coupling cell
transits from 0 to 1 or 1 to 0
mbist1.10 Cheng-Wen Wu, NTHU
11
RAM Fault Models (Dynamic)
- Recovery Fault (RF)
Sense Amplifier Recovery Fault (SARF)
- Sense amp saturation after reading/writing long run
of 0 or 1
Write Recovery Fault (WRF)
- Write followed by reading/writing at different location
resulting in reading/writing at same location
^ Write-after-write recovery fault
^ Read-after-write recovery fault
Results in functional faults---detected at high
speed (e.g., GALROW/GALCOL)
- Disturb Fault (DF)
- Victim cell forced to 0 or 1 if we read or write
aggressor cell (may be the same cell)
mbist1.10 Cheng-Wen Wu, NTHU
12
RAM Fault Models (Dynamic)
- Data Retention Fault (DRF)
DRAM
- Refresh Fault
^ Refresh-Line Stuck-At Fault
- Leakage Fault
^ Sleeping Sickness---loose data in less than
specified hold time (typically tens of ms)
SRAM
- Leakage Fault
^ Static Data Losses---defective pull-up
Checkerboard pattern triggers max leakage
BIST good for sync with refresh mechanism
mbist1.10 Cheng-Wen Wu, NTHU
13
Test Time Complexity (100MHz)
Size N 10N NlogN N
1.5
N
2
1M 0.01s 0.1s 0.2s 11s 3h
16M 0.16s 1.6s 3.9s 11m 33d
64M 0.66s 6.6s 17s 1.5h 1.43y
256M 2.62s 26s 1.23m 12h 23y
1G 10.5s 1.8m 5.3m 4d 366y
4G 42s 7m 22.4m 32d 57c
16G 2.8m 28m 1.6h 255d 915c
mbist1.10 Cheng-Wen Wu, NTHU
14
RAM Test Algorithm
- A test algorithm (or simply test) is a finite
sequence of test elements
A test element contains a number of memory
operations (access commands)
- Data pattern (background) specified for the Read
operation
- Address (sequence) specified for the Read and
Write operations
- A march test algorithm is a finite sequence of
march elements
A march element is specified by an address order
and a number of Read/Write operations
mbist1.10 Cheng-Wen Wu, NTHU
15
Classical Test Algorithms
- Zero-One Algorithm [Breuer & Friedman 1976]
Also known as MSCAN
For SAF
Solid background (pattern)
Complexity is 4N
)} 1 ( ); 1 ( ); 0 ( ); 0 ( { r w r w
mbist1.10 Cheng-Wen Wu, NTHU
16
Classical Test Algorithms
- Checkerboard Algorithm
Zero-one algorithm with checkerboard pattern
Complexity is 4N
For SAF and DRF
1 0 1
0 1 0
1 0 1
mbist1.10 Cheng-Wen Wu, NTHU
17
Classical Test Algorithms
- Galloping Pattern (GALPAT)
Complexity is 4N**2---only for characterization
All AFs,TFs, CFs, and SAFs are located
1. Write background 0;
2. For BC = 0 to N-1
{ Complement BC;
For OC = 0 to N-1, OC != BC;
{ Read BC; Read OC; }
Complement BC; }
3. Write background 1;
4. Repeat Step 2;
BC
mbist1.10 Cheng-Wen Wu, NTHU
18
Classical Test Algorithms
- Sliding (Galloping) Row/Column/Diagonal
Based on GALPAT, but instead of a bit, a complete
row, column, or diagonal is shifted
Complexity is 4N**1.5
1
1
1
1
1
mbist1.10 Cheng-Wen Wu, NTHU
19
Classical Test Algorithms
- Butterfly Algorithm
Complexity is 5NlogN
1. Write background 0;
2. For BC = 0 to N-1
{ Complement BC; dist = 1;
While dist <= mdist /* mdist < 0.5 col/row length */
{ Read cell @ dist north from BC;
Read cell @ dist east from BC;
Read cell @ dist south from BC;
Read cell @ dist west from BC;
Read BC; dist *= 2; }
Complement BC; }
3. Write background 1; repeat Step 2;
6
1
9 4 5,10 2 7
3
8
mbist1.10 Cheng-Wen Wu, NTHU
20
Classical Test Algorithms
- Moving Inversion (MOVI) Algorithm [De Jonge &
Smeulders 1976]
For functional and AC parametric test
- Functional (13N): for AF, SAF, TF, and most CF
- Parametric (12NlogN): for Read access time
^ 2 successive Reads @ 2 different addresses with
different data for all 2-address sequences
differing in 1 bit
^ Repeat T2~T5 for each address bit
^ GALPAT---all 2-address sequences
)} 0 , 0 , 1 ( ); 1 , 1 , 0 ( ); 0 , 0 , 1 ( ); 1 , 1 , 0 ( ); 0 ( { r w r r w r r w r r w r w l l
mbist1.10 Cheng-Wen Wu, NTHU
21
Classical Test Algorithms
- Surround Disturb Algorithm
Examine how the cells in a row are affected when
complementary data are written into adjacent cells
of neighboring rows
1. For each cell[p,q] /* row p and column q */
{ Write 0 in cell[p,q-1];
Write 0 in cell[p,q];
Write 0 in cell[p,q+1];
Write 1 in cell[p-1,q];
Read 0 from cell[p,q+1];
Write 1 in cell[p+1,q];
Read 0 from cell[p,q-1];
Read 0 from cell[p,q]; }
2. Repeat Step 1 with complementary data;
1
0 0 0
1
mbist1.10 Cheng-Wen Wu, NTHU
22
Classical Test Algorithms
- Zero-one and checkerboard algorithms do not
have sufficient coverage
- Other algorithms are too time-consuming for large
RAM
Test time is the key factor of test cost
Complexity ranges from N
2
to NlogN
- Need linear-time test algorithms with small
constants
March test algorithms
mbist1.10 Cheng-Wen Wu, NTHU
23
March Tests
- Zero-One (MSCAN)
- Modified Algorithmic Test Sequence (MATS) [Nair,
Thatte & Abraham 1979]
OR-type address decoder fault
AND-type address decoder fault
- MATS+ [Abadir & Reghbati 1983]
For both OR- & AND-type AFs and SAF
)} 1 ( ); 1 ( ); 0 ( ); 0 ( { r w r w
)} 1 ( ); 1 , 0 ( ); 0 ( { r w r w
)} 0 ( ); 0 , 1 ( ); 1 ( { r w r w
)} 0 , 1 ( ); 1 , 0 ( ); 0 ( { w r w r w l
mbist1.10 Cheng-Wen Wu, NTHU
24
March Tests
- Marching 1/0 [Breuer & Friedman 1976]
For AF, SAF, and TF
- MATS++ [Goor 1991]
Also for AF, SAF, and TF
Complete and irredundant
)} 1 , 1 , 0 ( ); 0 , 0 , 1 ( ); 1 (
); 0 , 0 , 1 ( ); 1 , 1 , 0 ( ); 0 ( {
r w r r w r w
r w r r w r w
l l
l l
)} 0 , 0 , 1 ( ); 1 , 0 ( ); 0 ( { r w r w r w l
mbist1.10 Cheng-Wen Wu, NTHU
25
March Tests
- March X
For AF, SAF, TF, & CFin
- March C [Marinescu 1982]
For AF, SAF, TF, & all CFs---redundant
- March C- [Goor 1991]
Also for AF, SAF, TF, & all CFs---irredundant
)} 0 ( ); 0 , 1 ( ); 1 , 0 ( ); 0 ( { r w r w r w l
)} 0 ( ); 0 , 1 ( ); 1 , 0 ( ); 0 (
); 0 , 1 ( ); 1 , 0 ( ); 0 ( {
r w r w r r
w r w r w

l l
)} 0 ( ); 0 , 1 ( ); 1 , 0 (
); 0 , 1 ( ); 1 , 0 ( ); 0 ( {
r w r w r
w r w r w

l l
mbist1.10 Cheng-Wen Wu, NTHU
26
March Tests
- Limitations
Sequential faults in address decoders
RF
NPSF
- O(9N-2) for 2-CF [Marinescu 1982]
- O(2NlogN+11N) for 3-CF [Cockburn 1994]
- Solutions
Address sequence variation
- Hopping
- Pseudorandom
mbist1.10 Cheng-Wen Wu, NTHU
27
Coverage of March Tests
MATS++ March X March Y March C-
SAF 1 1 1 1
TF 1 1 1 1
AF 1 1 1 1
SOF 1 .002 1 .002
CFin .75 1 1 1
CFid .375 .5 .5 1
CFst .5 .625 .625 1
Extended March C- (11N) has a 100% coverage of SOF
mbist1.10 Cheng-Wen Wu, NTHU
28
Testing Word-Oriented RAM
- Background bit is replaced by background word
MATS++:
- Conventional method is to use logm+1 different
backgrounds for m-bit words
m=8: 00000000, 01010101, 00110011, and
00001111
Apply the test algorithm logm+1=4 times, so
complexity is 4*6N/8=3N
)} , , ' ( ); ' , ( ); ( { ra wa ra wa ra wa l
mbist1.10 Cheng-Wen Wu, NTHU
29
Cocktail-March Algorithms
- Motivation:
Repeating the same algorithm for all logm+1
backgrounds has redundancy
Different algorithm targets different faults
- Approach:
Use multiple backgrounds in a single algorithm run
Merge and forge different algorithms and
backgrounds into a single algorithm
- Good for word-oriented memories
mbist1.10 Cheng-Wen Wu, NTHU
30
March-CW
- Algorithm:
March C- for solid background (0000)
Then a 5N March for each of other standard
backgrounds (0101, 0011):
- Result:
Complexity is (10+5logW)N, where W is word
length and N is word count
Test time is reduced by 39% if W=4, as compared
with extended March C-
Improvement increases as W increases
)} , , ' , ' , ( { ra wa ra wa wa
mbist1.10 Cheng-Wen Wu, NTHU
31
Comparison (Full Coverage)
mbist1.10 Cheng-Wen Wu, NTHU
32
Testing NPSF
- NPSF test approaches
Tiling
Multi-background march
- Easy BIST implementation
- 5-cell neighborhood
W
N
E
E
S
W
N
E
S
B
B
B
N, E, W, S
Deleted neighborhood cells: N, E, W, S
Base cell: B
B Neighborhood cells: and
mbist1.10 Cheng-Wen Wu, NTHU
33
NPSF Models
- Static NPSF (SNPSF)
BC forced to a certain state due to a certain deleted
neighborhood (DN) pattern
- Passive NPSF (PNPSF)
BC frozen due to a certain DN pattern
- Active NPSF (ANPSF)
BC content changes due to a change in DN pattern
- Change: a transition in one DN cell, with other DN cells & BC
containing a certain pattern
- Assumptions:
Single NPSF
Address scramble table is available
Memory is bit-oriented
- Word-oriented memory is tested as multiple bit-oriented ones
mbist1.10 Cheng-Wen Wu, NTHU
34
Test Strategy
- Multi-Background March
To generate all neighborhood patterns
X X X X 1 1 1
1
1 1
1
1
1 1
1
1 0 0 0 0
0
0
0
0
0
0
0
1
1 1
1
1
1
0 0
0 0
0 0
0 0 1 1
1 1
0 0
0 0
0 0
0 0
X
0 0
X
0 0
X 0 0 0
1 1
1 1
1 1
1 1 1
1 1
1
1
0
0
1
1
0
1 0
0
0
0
0
0
0
X
Solid BG
(FC < 30%)
Another BG
mbist1.10 Cheng-Wen Wu, NTHU
35
Testing PNPSF
- March 17N:
wa wb, ra, wa ra ( ); ( );
(ra) (rb, wa); (ra, wb);
wb wa, ra, wb
rb ( , wa);
(rb, ); ( , );
0 0 0 0
1 1
1 1
1 1
0 0
0 0
1 1
0 0
0 0
1 1
1 1
(w0); (r0); (w1, r1, w0); Alg1:
(w1); (w0, r0, w1); (r1); Alg2:
(w0); (w1); (r1); Alg3:
Alg4: (w1); (w0); (r0);
(w0); (r1); Alg5: (w1);
Alg6: (r0); (w0); (w1);
March Elements NWBES
mbist1.10 Cheng-Wen Wu, NTHU
36
Data Background Generation
- Data backgrounds
BG1: all zero
BG2: Ar[0], LSB of row address
BG3: Ar[1], second bit of row address
BG4: Ar[0]Ar[1]
00
01
10
11
00
01
10
11
00
01
10
11
00
01
10
11
0 0 0 0 1 1 1 1 1 1 1 1
00 01 10 11 00 01 10 11
BG.1 BG.2 BG.3 BG.4
00 01 10 11 00 01 10 11
0
0
0 0 0
0
0
0 0
0
0
0
1
0 0 0
0 0 1 1 1
0 0
0
0 0 0 0
1 1
1
1
1 0 1 0
0
0
0
0 1 1
0
0 0
0
1
0
0
1
1
1
mbist1.10 Cheng-Wen Wu, NTHU
37
Testing ANPSF
- March 12N:
(ra)
(wa); (wb, ra, wa); (ra, wb);
(rb, wa);
(rb, wa, wb);
0 0 0
0
0 0
0
0 0 0 0 0
0
0 0 0
1 1 1 1 1 1
1 1 1
1 1 1
1 1 1 1
0 0
0
1
0
1
1 1
0
1 1
1
0 0
0
1
( ,w1) ; ( ,w0) ; (w0) ;
( ,w1,w0) ;
( ,w0,w1) ;
(w0) ;
(w1) ; r1
(r0)
(r1)
(r0) r0
r0
r1
mbist1.10 Cheng-Wen Wu, NTHU
38
Time Complexity
- 12 N/BG X 8 BG = 96N
- Detects all NPSFs
00
01
10
11
00
01
10
11
0
0
0
0
00
01
10
11
00
01
10
11
00
01
10
11
00 01 10 11 00 01 10 11 00 01 10 11
00 01 10 11 00 01 10 11
00
01
10
11
00
01
10
11
00 01 10 11
00 01 10 11
00
00 01 10 11
01
10
11
0 0 0
0
0
0 0
0
0 0
0
0 0 1
0
1
0 1 0 1
0 0 0
1 1 1
0 0
1
1
0 0 0 0
0
0 0 1
1
1
0
0
1
1
0
1
0
1 0
1
1 0
0
1 0
BG.8
BG.4 BG.3 BG.2 BG.1
BG.5 BG.6 BG.7
1
1
1 0
0
0 1 0
1
0
1 1 0
1
0 1 1
0
1
1
1
1 1
1
1 0
0
0 0
0
1
0
0
1
0 1
1
1
1
0
1
1
0
1
0 1
0
1
0
1
1
1
1
0
0
0
0
0
0
0
1
0 0
0 1 1 0
1 1
mbist1.10 Cheng-Wen Wu, NTHU
39
Multi-Port Memories
- Popular architectures
k-port (k > 1)
n-read-1-write
FIFO
Storage
Address A
Data A
Control A
Port A
Address B
Data B
Control B
Port B
mbist1.10 Cheng-Wen Wu, NTHU
40
2-Port Topology
WL
Interport BL short
Interport WL short
WL
WL
WL
WL
WL
WL
WL
BL BL BL BL
BL BL BL BL
A
B
A
B
A
A
A
A
B
B
B
A
B
B
B
A
1
o |

2
3
mbist1.10 Cheng-Wen Wu, NTHU
41
Inter-Port Word-Line Short
* Functional test complexity: O(N
3
)
Faulty
Cell 1
Fault-Free
Cell 2
Address 3
Address 2
Cell 3
Address 1 Cell 1 Address 1
Port B
Port A
Cell 2 Address 2
mbist1.10 Cheng-Wen Wu, NTHU
42
Inter-Port Bit-Line Short
Port A
Fault-Free
Port B
Cell
Address Cell
Address o o
| |
Faulty
Address o
Address o
Cell o
Cell o
Address |
Address |
Cell |
Cell |
* Functional test complexity: O(N
2
)
mbist1.10 Cheng-Wen Wu, NTHU
43
Address Scrambling
B - I/O select
Logical Addr
Physical Addr
A B C D
7 6 5 4 3 2 1 0 8 7 6 5 4 3 2 1 0
row column
D - bit position in a word
A - word line select
C - bit line select
bit1 bit2 bit3 bit0
0 3
Address A
Data word A
mbist1.10 Cheng-Wen Wu, NTHU
44
Reading Neighboring Cells
- Read neighboring cells to detect inter-port faults:
r
N
, r
S
, r
E
, and r
W

0/1 1 1/0 0
1 1 1
0 0 0
E
N
W
S
B
0 0 0
0 1
1 1 1
0/1
mbist1.10 Cheng-Wen Wu, NTHU
45
TAGS-PS
Single-port
Test Algorithm
Multi-port
AF Test
Single-port
Test Algorithm
March Test
Section 1 Section 2 Section 3
Port 1
Port 2
Port m
Port 1
Port 2
Port m
(a)
(b)
Multi-port Inter-port
Test AF Test
Inter-port
MPF Test
mbist1.10 Cheng-Wen Wu, NTHU
46
Dual-Port RAM Test
- - - - -
- - - - - -
(b)
- - - - - - - -
(a)
- - -
1 r 0 w 1 r w
0
0 w
r
r 0
0 1 r w 1 r 0 w w r 0 1 0 0
0 r
0 1 r r 1 w 1
1
w w
0 r 1 w 1 r 0 w 1
r0
1
r
w0
r
N
1 r
S
w r 0
r
r
1
0
w
r
1 r0
S
r 1
N
r 0
S
r 0
N
r 1 w0 r1 w1
N
Section 1
r0
w
0
0
S
- - - - r
0
1
r
Section 2 Section 3
Section 1 Section 2
Section 3
- - - - - -
- - - -
w1 r0 w0 r1 w1 r0
Port 1
Port 2
Port 1
Port 2
mbist1.10 Cheng-Wen Wu, NTHU
47
Compacted Dual-Port RAM Test
(c)
-
-
(d)
1 r 0 w 1 r
N S
0
N
1 r w
0
r 1 r 0
w 1 r r
S
0 r
0 r 1 w 1
w
0 w 1
0
N
0
r
1
0 r
0
1 r
S
0
w0
w
w0
r
r1 w1 r0 r1 r0 r
r
Section 1
Section 2
- -
-
- -
S
r
- - - -
-
0
N
- -
Port 1
Port 2
Port 1
Port 2
* Time complexity: 10N
mbist1.10 Cheng-Wen Wu, NTHU
48
Four-Port RAM Test
- - - -
0 w 1 w 1 r 0 w 0 r
1 w 1 r 0 w 0 r
1 w 1 r 0 w 0 r
1 w 1 r 0 w 0 r
1 r
N
0 r
N
1 r
S
0 r
S
1 r
N
0 r
N
1 r
S
0 r
S
1 r
N
0 r
N
1 r
S
0 r
S
1 r
N
0 r
N
1 r
S
0 r
S
1 r
S
0 r
S
0 r
N
1 r
N
1 r
S
0 r
S
0 r
N
1 r
N
Port 1
Port 2
-
-
Port 3
Port 4
- - - - - - - - - - - -
- - - -
-
Inter-port Test
AF Test
- - - -
* Time complexity: 17N
mbist1.10 Cheng-Wen Wu, NTHU
49
Testing 6-Read-1-Write RAM
0 r
0 r 0 r
0 r
0 r
1 r
1 r
1 r
1 r
1 r
1 r
1 r
1 r
1 r
1 r
0 r
S
0 r
S
0 r
S
1 r
S
1 r
S
1 r
S
1 r
S
1 r
S
1 r
S
0 r
S
1 r
N
1 r
N
0 r
S
1 r
S
1 r
S
1 r
S
0 r
N
0 r
N
0 r
N
0 r
N
0 r
N
0 r
N
0 r
N
0 r
N
1 r
N
0 r
S
0 r
S
0 r
S
1 r
N
1 r
N
1 r
N
1 r
N
1 r
N
1 r
N
1 r
N
1 r
N
1 r
N
1 r
N
0 r
N
0 r
N
0 r
N
0 r
N
Port 1
- - - - - -
Port 3
- - -
Port 2
- - -
Port 4
- - -
Port 5
- - -
Port 6
- - -
Port 7
- - -
M0 M1 M2 M3 M4 M5
Test for ports of
M6
distance 1
Test for ports of
distance 2
distance 3
Test for ports of
0 w 0 w 0 w 0 w 1 w 1 w 1 w
0 r
0 r 0 r
* Time complexity: 13N
mbist1.10 Cheng-Wen Wu, NTHU
50
Flash Memory Testing
- Testing nonvolatile memories:
Masked ROM---exhaustive; pseudorandom
PROM (OTP) & EPROM---dummy row
EEPROM & flash memory---dummy row?
- Testing flash memory core is hard
Customized core and I/O
Isolation (accessibility)
Reliability issues: disturbances, over
program/erase, under program/erase, data
retention, cell endurance, etc.
Long program/erase time
mbist1.10 Cheng-Wen Wu, NTHU
51
Flash Memory Overview
- Flash memory can be programmed and erased
electrically
Has the advantages of EPROM and EEPROM
- A stacked gate transistor with both the control
gate (CG) and floating gate (FG):
G
D
S
P-Si
n
+
n
+
Source Drain
Control gate
Floating gate
mbist1.10 Cheng-Wen Wu, NTHU
52
Flash Memory Program & Erase
Program Erase
Program(1 to 0): channel hot-electron (CHE) injection or Fowler-
Nordheim (FN) electron tunneling
Erase (0 to 1): FN electron tunneling
By the entire chip or large blocks (flash erasure)
Different products have different program/erase mechanisms
mbist1.10 Cheng-Wen Wu, NTHU
53
Flash Memory Cell Types
- Stacked-gate Split-gate Select-gate





- Operations: Read, Program, Erase (Flash Erase)
As opposed to Read and Write in RAM
mbist1.10 Cheng-Wen Wu, NTHU
54
Programming Scheme Comparison
CHE Injection Channel FN Tunneling
High power (dual
external supplies)
Low power (single
external supply)
Low oxide field stress High oxide field stress
Faster program
operation (byte program
limited by power)
Slower program
operation (improved by
page program)


mbist1.10 Cheng-Wen Wu, NTHU
55
NOR-Array Structure
mbist1.10 Cheng-Wen Wu, NTHU
56
NAND-Array Structure
Select (drain)
Select (source)
WL 1
WL 2
WL 3
WL 4
WL 16
BL i
mbist1.10 Cheng-Wen Wu, NTHU
57
Disturbance Example (I)
WL
0
WL
1
WL
2
BL
0
BL
1
BL
2
BL
3
10V
0V
Program Disturbance
Drain-Disturb on "Programmed Cell"
Gate-Disturb on "Erased Cell"
6V 0V
0V
SL SL
0V
0V 6V
10V
0V
10V
Programming
NOR-Type Common Ground Standard (Stacked Gate)
mbist1.10 Cheng-Wen Wu, NTHU
58
Disturbance Example (II)
WL
0
WL
1
WL
2
BL
0
BL
1
BL
2
BL
3
5V
0V
Read Disturbance
Soft-Program on "Selected Cell"
1V
SL SL
mbist1.10 Cheng-Wen Wu, NTHU
59
Disturbance Example (III)
BL
1
Program Disturbance
Program '1'
WL
0
WL
1
WL
2
BL
0
SSL
GSL
2.8V
2.8V
18V
3.3V
3.3V
0V
10V
2.8V
BL
2
18V
0V
0V
10V
0V
3.3V
0V
0V
10V
0V
Program '0'
Gate-Disturb on "Erased Cell"
Gate-Disturb on "Programmed Cell"
mbist1.10 Cheng-Wen Wu, NTHU
60
Disturbance Example (IV)
BL
1
Read Disturbance
WL
0
WL
1
WL
2
BL
0
SSL
GSL
0V
5V
5V
5V
5V
BL
2
0V
5V
0.7V
5V
5V
V
th
=-3V V
th
=+2V
5V
5V
soft-program
BL
1
Erase Disturbance
WL
0
WL
1
WL
2
BL
0
SSL
GSL
0V
Floating
Floating
0V
BL
2
0V
0V
Floating
0V
0V
Floating
21V
21V
21V 21V
21V
21V 21V
21V
21V
mbist1.10 Cheng-Wen Wu, NTHU
61
Gate Program Disturb Fault (GPDF)
V(H)
V(H)
V(L)
V(L)
V(Gd)
Conditions:
1.Victim cell initial value is a logic 1
2.Aggressor 10 (program)
Victim 10 (program)
Control Gate
Floating Gate
Source Drain
Substrate
G
S D
B
mbist1.10 Cheng-Wen Wu, NTHU
62
Gate Erase Disturb Fault (GEDF)
V(H)
V(H)
V(L)
V(L)
V(Gd)
Conditions:
1.Victim cell initial value is a logic 0
2.Aggressor 10 (program)
Victim 01 (erase)
Control Gate
Floating Gate
Source Drain
Substrate
G
S D
B
mbist1.10 Cheng-Wen Wu, NTHU
63
Drain Program Disturb Fault (DPDF)
V(H)
V(H)
V(L)
V(Gd)
During programming, erased cells on
unselected rows on a bit-line that is being
programmed may have a fairly deep
depletion region formed under them
Electrons entering this depletion region can
be accelerated by the electric field and
injected over the oxide potential barrier to
adjacent floating gates
Conditions:
1.Victim cell initial value is a logic 1
2.Aggressor 10 (program) Victim 10 (program)
mbist1.10 Cheng-Wen Wu, NTHU
64
Drain Erase Disturb Fault (DEDF)
V(H)
V(H)
V(L)
V(L)
V(Gd)
Conditions:
1.Victim cell initial value is a logic 0
2.Aggressor 10 (program)
Victim 01 (erase)
Control Gate
Floating Gate
Source Drain
Substrate
G
S D
B
mbist1.10 Cheng-Wen Wu, NTHU
65
Read Disturb Fault (RDF)
Control Gate
Floating Gate
Source Drain
Substrate
G
S D
B
Conditions:
1. Occurs on the selected cell
2. Cell initial value is logic 1
Soft-Program
During the read operation,
hot carriers can be injected
from the channel into the FG
even if at low gate voltages
mbist1.10 Cheng-Wen Wu, NTHU
66
Over Erase Fault (OEF)
- Flash memory erase mechanism is not
self-limiting
- Threshold voltage can be low enough to
turn the cell into a depletion-mode
transistor
- Fault behavior:
An unselected cell in the same bit-line has
excessive source-drain leakage current
Reading that cell leads to incorrect value
(like DEDF)
Cannot be programmed correctly (like TF)
mbist1.10 Cheng-Wen Wu, NTHU
67
Basic RAM Faults for Flash Memory
- Address-Decoder Fault (AF)
- Stuck-At Fault (SAF)
- Transition Fault (TF)
- Stuck-Open Fault (SOF)
- Bridging Fault (BF)
Coupling faults need not be considered!
Replaced by disturb faults
mbist1.10 Cheng-Wen Wu, NTHU
68
Reliability Consideration
- Reliability characteristics of floating-gate
ICs depend on
Circuit density, circuit design, and process
integrity
Memory array type and cell structure
- Reliability stressing and testing must then
be oriented toward determining the relevant
failure rates for the particular array under
consideration
mbist1.10 Cheng-Wen Wu, NTHU
69
Data Retention Fault
- Retention time: the time from data storage to the
time at which a verifiable error is detected from
any cause
Intrinsic retention times exceed millions of years in
the operating temperature range
- Months at 300C
- 1 million years at 150 C
- 120 million years at 55 C
- Data Retention Fault (DRF)
Static leakage
Built-in data retention test circuit
mbist1.10 Cheng-Wen Wu, NTHU
70
Cell Endurance Fault
- Endurance: a measure of the ability to meet data-
sheet specifications as a function of accumulated
program/erase cycles
Endurance limit is a result of damage to the
dielectric around the floating gate caused by
electric stresses
In many flash devices, the end of endurance is
generally caused by hot electron trapping in the
charge transport oxide
- Cell Endurance Fault (CEF)
Threshold window shift due to increased
program/erase cycles
Built-in stress test circuit
mbist1.10 Cheng-Wen Wu, NTHU
71
Composite Failure Rate Determination
- 125C dynamic life stress
The 125C dynamic life stress is the standard MOS
memory continuous dynamic read in a burn-in
chamber
- Endurance test
The endurance test is the repeated data
complementing of floating-gate devices, possibly at
temperature extremes
- Extended data retention stress
This test is constituted by a high-temperature bake
with a charge polarity that is opposite to the
equilibrium state on the floating gate
mbist1.10 Cheng-Wen Wu, NTHU
72
Typical Test Modes (Characterization)
- Stress (row/column)
Reverse tunneling stress
Punch through stress
Tox stress
DC stress
- Mass program
- Weak erase
- Leak (thin-oxide, bit-line, etc.)
- Cell current; cell Vt
- Margin
- Etc.
mbist1.10 Cheng-Wen Wu, NTHU
73
Test Patterns
- A RAM test pattern definition includes both the
data pattern and the address pattern
The time to read a pattern is the same as the time
to write a pattern
- For flash memories, however, the address and
data pattern definitions must be segregated
It has long write times relative to the read times
- Typical data patterns:
Solid, checkerboard, random, etc.
- Typical address patterns:
Address increment/decrement, address
complement, column/diagonal galloping, etc.
mbist1.10 Cheng-Wen Wu, NTHU
74
Testing GPDF
1 Flash
2 Program the first column
3 Read all cells except the first column
4 Flash
5 Program any column except the first
6 Read the first column
*Assume reading and programming are done column-wise
Source: Saluja, et al., Int. Conf. VLSI Design, 2000
mbist1.10 Cheng-Wen Wu, NTHU
75
Testing GEDF
1 Flash
2 Program all cells
3 Read all cells except the last column
4 Program any column except the last
5 Read the last column
*Assume reading and programming are done column-wise
Source: Saluja, et al., Int. Conf. VLSI Design, 2000
mbist1.10 Cheng-Wen Wu, NTHU
76
Test Coverage: Previous Results
Fault DCP DCE DD EF GF
SAF 50% 50% 50% 100% 100%
TF 12.5% 50% 50% 87.5% 62.5%
AF 40% 0% 0% 44.5% 40%
SOF 0% 0% 0% 12.5% 6.2%
CFst 25% 25% 25% 50% 50%
GPDF 33.3% 0% 0% 100% 33.3%
GEDF 0% 100% 75% 100% 100%
DEDF 0% 75% 100% 100% 100%
DPDF 0% 0% 0% 0% 0%
Source: Saluja, et al., Int. Conf. VLSI Design, 2000
mbist1.10 Cheng-Wen Wu, NTHU
77
March-Based Flash Test: March-FT
- {(f); (r1,w0,r0); (r0); (f); l(r1,w0,r0); l(r0)}
This Flash memory is NOR type (Stacked gate).
Memory size (N) : 65536
Test length : 2(chip erase time) + 131072(word program time) + 393216(word read time)
Test time : 7.207173 sec
SAF : 100% (131072 / 131072) P.S.
TF : 100% (131072 / 131072) Flash Type = NOR
SOF : 100% (65536 / 65536) Gate Type = Stack
AF : 100% (4294901760 / 4294901760) Row Number = 256
CFst : 100% (17179607040 / 17179607040) Col Number = 256
GPD : 100% (16711680 / 16711680) Word Length = 1
GED : 100% (16711680 / 16711680) Chip erase time = 3 sec
DPD : 100% (16711680 / 16711680) Word program time = 9u sec
DED : 100% (16711680 / 16711680) Word read time = 70n sec
RD : 100% (65536 / 65536)
OE : 100% (65536 / 65536)
mbist1.10 Cheng-Wen Wu, NTHU
78
Test Length (Bit-Oriented)
- Notation:
F : Flash time
P : Program time
R : Read time
r : row number
c : column number
DCP 2(F) + 2r(P) + rc(R)
DCE (F) + (c+1)r(P) + rc(R)
DD (F) + (r+1)c(P) + rc(R)
EF 2(F) + (rc+2r+c-2)(P) +
(2rc+r+c-3)(R)
GF 2(F) + (rc+2r+c-1)(P) +
(2rc+c+r-2)(R)
FT 2(F) + 2rc(P) + 6rc(R)
mbist1.10 Cheng-Wen Wu, NTHU
79
Test Length (Word-Oriented)
- Word length = w:
2(F)+2rc(P)+6rc(R)+log(w)[2(F)+rc(P)+rc(R)]



Solid: 0000 (1111)
Standard: 0101 (1010), 0011 (1100)
- Ex: word length w = 4
6(F) + 4rc(P) + 8rc(R)
solid background
testing time
standard background
testing time
mbist1.10 Cheng-Wen Wu, NTHU
80
Test Algorithm Generation by Simulation (TAGS)
T(N) Test algorithms
2N
3N
4N
5N
6N
7N
8N
9N
10N
(f); (r1)
(f); (w0); (r0)
(f); (r1,w0); (r0)
(f); (r1,w0,r0); (r0)
(f); (r1,w0,r0); (r0,w0)
(f); (r0); (r1,w0,r0); (r0,w0)
(f); (r1,w0); (f); l(r1,w0,r0); (r0)
(f); (r1,w0); (r0); (f); l(r1,w0,r0); (r0)
(f); (r1,w0,r0); l(r0); (f); l(r1,w0,r0); (r0)
mbist1.10 Cheng-Wen Wu, NTHU
81
Embedded Memory Testing
- Memories are one of the most universal cores
In Alpha 21264, cache RAMs represent 2/3
transistors and 1/3 area; in StrongArm SA110, the
embedded RAMs occupy 90% area [Bhavsar, ITC-99]
In average SOC, memory cores will represent more
than 90% of the chip area by 2010 [ITRS 2000]
- Embedded memory testing is increasingly difficult
High bandwidth (speed and I/O data width)
Heterogeneity and plurality
Isolation (accessibility)
AC test, diagnostics, and repair
- BIST is considered the best solution
mbist1.10 Cheng-Wen Wu, NTHU
82
Embedded RAM Test Support
Test run Isolation only Isolation & BIST
Probe test Tester Tester/BIST
Pre-BI test Tester BIST
BI BI board BIST
Post-BI test Tester BIST
Final test Tester Tester/BIST
mbist1.10 Cheng-Wen Wu, NTHU
83
RAM BIST Approaches
- Methodology
Processor-based BIST
- Programmable
Hardwired BIST
- Fast
- Compact
- Interface
Serial (scan, 1149.1)
Parallel (embedded controller; hierarchical)
- Patterns (address sequence)
March
Pseudorandom
mbist1.10 Cheng-Wen Wu, NTHU
84
Typical RAM BIST Architecture
RAM
BIST Module
Controller
Comparator
Pattern
Generator
Go/No-Go
RAM Controller
mbist1.10 Cheng-Wen Wu, NTHU
85
Serial March (SMarch)
From March C-
Serial interface
One BIST for all
(cascaded)
One-bit read/write at a
time, but one pattern
per cycle
Slow
No diagnostics
} ) 0 0 ( ) 0 0 ( ; ) 0 0 ( ) 0 1 ( ; ) 1 1 ( ) 1 0 (
; ) 0 0 ( ) 0 1 ( ; ) 1 1 ( ) 1 0 ( ; ) 0 0 ( ) 0 x ( {
c c c c c c
c c c c c c
w r w r w r w r w r w r
w r w r w r w r w r w r

l l l
Source: Nadeau-Dostie et al., IEEE D&T, Apr. 1990
Memory Cell Array
X

D
e
c
o
d
e
r

Y Decoder
Transparent Serial Data-MUX
Addr
SI SO
D Q
c
c
mbist1.10 Cheng-Wen Wu, NTHU
86
Syntest MBIST
- Algorithms:
March C-
MOVI
March C++
Checkerboard
- Shared controller
for multiple RAMs
- Synthesizable RTL
code
FSM
Data Generator
Analyzer
ADR Control
CE
OE
WEB
A
D
Q
Pass
BistFail
Finish
Source: Syntest
mbist1.10 Cheng-Wen Wu, NTHU
87
NTHU/GUC EDO DRAM BIST
mbist1.10 Cheng-Wen Wu, NTHU
88
DRAM Page-Mode Read-Write Cycle
mbist1.10 Cheng-Wen Wu, NTHU
89
Programmable Memory BIST (PMBIST)
mbist1.10 Cheng-Wen Wu, NTHU
90
PMBIST Architecture
mbist1.10 Cheng-Wen Wu, NTHU
91
Controller and Sequencer
- Controller
Microprogram
Hardwired
Shared CPU core
IEEE 1149.1 TAP
- Sequencer (Pattern Generator)
Counter
LFSR
LUT
mbist1.10 Cheng-Wen Wu, NTHU
92
Controller
mbist1.10 Cheng-Wen Wu, NTHU
93
Sequencer
Combination Logic #0
Row Address Counter
Column Address
Counter
State
Combination Logic #1
eDRAM
control
signal
MCK MCK
Flags
eDRAM
BIST
Controller
D Q
D Q
D Q
D Q
D Q
D Q
D Q
D Q
Control Counter
Comparator
mbist1.10 Cheng-Wen Wu, NTHU
94
PMBIST Test Modes
1. Scan-Test Mode
2. RAM-BIST Mode
1.Functional faults
2.Timing faults (setup/hold times, rise/fall times,
etc.)
3.Data retention faults
3. RAM-Diagnosis Mode
4. RAM-BI Mode
mbist1.10 Cheng-Wen Wu, NTHU
95
PMBIST Controller Commands
Bit 4
Addressing order
Bit 3
Data type
Bit 2, Bit 1, Bit 0
Operations
1: l (increasing)
1: d = DB 000: EOT (End of test)
0: (decreasing)
0: d = ~DB 001: Rd (READ Cycle)
010: Wd (Early WRITE Cycle)
011: RdW~d (READ-WRITE) Cycle
EDO-PAGE-MODE
100: Wd (Early WRITE Cycle
101: RdW~d (READ-WRITE) Cycle
110: RdW~dR~d (READ Early WRITE Cycle)
111: Refresh
mbist1.10 Cheng-Wen Wu, NTHU
96
PMBIST Control Sequence
mbist1.10 Cheng-Wen Wu, NTHU
97
BIST Area Overhead
3%
0.3%
Overhead
Mem size
mbist1.10 Cheng-Wen Wu, NTHU
98
Processor-Based RAM BIST
Processor
mbist1.10 Cheng-Wen Wu, NTHU
99
On-Chip Processor-Based RAM BIST
- BIST program is stored in boot ROM during
design phase, and memory BIST is done by
executing BIST program
Address
bus
Embedded
memory
CPU core
BOOT
ROM
DATAI
bus
DATAO
bus
Control
bus
I/O port
mbist1.10 Cheng-Wen Wu, NTHU
100
Testing RAM Core by On-Chip CPU
- 6502 assembly program that performs March C-
test algorithm
.org 0HFF00
LDX #$$00
LDA #$$55

M0: STA 0000,X
INX
CPX #$$FF
BNE M0
LDX #$$00

M1: LDA 0000,X
CMP #$$55
BNE ERROR
LDA #$$AA
STA 0000,X
INX
CPX #$$FF
BNE M1
LDX #$$00
. . . . .
(W0)
(R0W1)
(R1W0)
(R0W1)
(R1W0)
(R0)
March C- algorithm
data background
write data background to memory
read from memory
write data background to memory
mbist1.10 Cheng-Wen Wu, NTHU
101
Test Speed Consideration
- Processor-BIST speed is lower than dedicated
BIST circuit
- Total clock cycles to implement MARCH C- is
O(114N)
IMM ABX IMP REL
LDA 2 4 - -
LDX 2 - - -
STA - 4 - -
INX - - 2 -
CPX 2 - - -
BNE - - - 2~4
CMP 2 - - -
Table 1. 6502 instruction cycles
mbist1.10 Cheng-Wen Wu, NTHU
102
NTHU Processor-Programmable BIST
BIST core
embedded
memory
I/O circuitry
BIST circuitry
embedded
CPU
1
0
1
0
0
1
0
1
mux_sel
mux_sel = 0 in normal mode
mux_sel = 1 in BIST mode
o
n
-
c
h
i
p

b
u
s
ADDR
DATAO
control
DATAI
A
DI
DO
control
ADDR_cpu
DATAO_cpu
clock_cpu
ctrl_cpu
DATAI_cpu
ADDR_bist
DATAO_bist
ctrl_bist
DATAI_sys
DATAI_bist
mbist1.10 Cheng-Wen Wu, NTHU
103
Advantages and Disadvantages
- Advantages
Reuse of on-chip CPU core
- Might need modification
Core March elements can be implemented in
hardware, allowing different March algorithms to be
executed via assembly programming
- Disadvantages
Some address space will be occupied by PPBIST
Area overhead
mbist1.10 Cheng-Wen Wu, NTHU
104
PPBIST Implementation
mbist1.10 Cheng-Wen Wu, NTHU
105
PPBIST Data Registers
Register Function
R
BG
Store background data
R
AL
Store lowest address
R
AH
Store highest address
R
ME
Store current March element
R
IR
Instruction register
R
FLAG
Status register
R
ED
Erroneous response of defective memory cell
R
EA
Address of defective memory cell


mbist1.10 Cheng-Wen Wu, NTHU
106
PPBIST Test Procedure
CPU write data back ground
CPU write start/stop address
CPU write MARCH element instruction
CPU write START instruction to wrapper
BIST core
(R0W1)
BIST core
(R1W0)
BIST core
(R0W1)
BIST core
(R1W0)
BIST core
(W0)
BIST core
(R0)
compare error?
complete?
write error flag
write faulty address
write faulty data
write complete flag
yes
yes
no
no
CPU take over
mbist1.10 Cheng-Wen Wu, NTHU
107
PPBIST Example Using 6502
- 6502 assembly program that performs March C-
test algorithm under the proposed BIST scheme
START: LDA #$$55
STA 0HFFE0
LDA #$$00
STA 0HFFE1
LDA #$$00
STA 0HFFE2
LDA #$$FF
STA 0HFFE3
LDA #$$0F
STA 0HFFE4
M0: LDA #$$00
STA 0HFFE5
JSR BIST

M1: LDA #$$01
. . . . . .
END: LDA #$$04
STA 0HFFE6
JMP FINISH

BIST: LDA #$$00
STA 0HFFE6
LOOP: LDA 0HFFE7
CMP #$$01
BEQ ERROR
CMP #$$FF
BNE LOOP
RTS

ERROR: LDA #$$03
STA 0HFFEA
JMP FINISH
mbist1.10 Cheng-Wen Wu, NTHU
108
PPBIST Example
- Addresses of the registers in the BIST experiment
Register Address
R
BG
FFE0
R
AL
FFE1 ~ FFE2
R
AH
FFE3 ~ FFE4
R
ME
FFE5



Register Address
R
IR
FFE6
R
FLAG
FFE7
R
ED
FFE8
R
EA
FFE9 ~ FFEA



- March elements and the corresponding R
ME

M
0
M
1
M
2
M
3
M
4
M
5

0
H
1
H
2
H
3
H
4
H
5
H



mbist1.10 Cheng-Wen Wu, NTHU
109
Experimental Results
- Total test time in terms of clock cycles
The sum of all the March elements' test time plus
30 clock cycles
10N clock cycles to perform March C-
- Test time of each March element :
M
0
M
1
M
2
M
3
M
4
M
5

1N 2N 2N 2N 2N 1N


mbist1.10 Cheng-Wen Wu, NTHU
110
Comparison of BIST Methodologies
BIST scheme Test time H/W overhead Routing overhead
Integrated BIST core Short Low High
On-chip processor Very long Zero Zero
Ours Short Very low zero


mbist1.10 Cheng-Wen Wu, NTHU
111
RAM BIST Compiler
- Use of RAM cores is increasing.
SRAM, DRAM, flash RAM
Multiple cores
- RAM BIST compiler is the trend.
- BRAINS (BIST for RAM in Seconds)
Proposed BIST Architecture
Memory Modeling
Command Sequence Generation
Configuration of the Proposed BIST
mbist1.10 Cheng-Wen Wu, NTHU
112
BRAINS Outputs
- Synthesizable BIST design
At-speed testing
Programmable March algorithms
Optional diagnosis support
- BISD
- Activation sequence
- Test bench
- Synthesis script
mbist1.10 Cheng-Wen Wu, NTHU
113
BIST Synthesis Flow
GUI
RAM/BIST
Description
Parser

Compile
Engine
Memory
Library
BIST
Template
Synthesis
Cell
Library
RTL
Netlist
mbist1.10 Cheng-Wen Wu, NTHU
114
NTHU/GUC PMBIST Architecture
Controller
Memory
MBS
T
e
s
t

C
o
l
l
a
r
Sequencer
Comparator
MBC
MBR
MSI
MSO
MRD
MBO
MCK
Normal
Access
Controls
Address
D
Q
Programmable Memory BIST
TPG
mbist1.10 Cheng-Wen Wu, NTHU
115
Controller Sequencer
Test
Pattern
Generator
Test Command/Information Storage
Module
Serial
data in
Serial
data out
Error Error
To
Memory
Address
BIST
control
signals
Memory
Command
Command
Hand-
shaking
PMBIST with Scan
Source: Cheng, et al., DFT00
mbist1.10 Cheng-Wen Wu, NTHU
116
Sequencer
Control
Module
Address Generator
Sequence Generator
Command Generator
Error Handling Module
address
command
error
info.
go
error
signature
mbist1.10 Cheng-Wen Wu, NTHU
117
State Diagram of Control Module
BIST
apply
BIST
done
BIST
idle
BIST
apply
BIST
done
BIST
idle
BIST
active
For DRAM For SRAM
mbist1.10 Cheng-Wen Wu, NTHU
118
DRAM Page-Mode Operation
mbist1.10 Cheng-Wen Wu, NTHU
119
Memory Specification Techniques
Memory Specifications
I/O Specification
Command Specification
Task Specification
Delay Constraint Specification
AC Parameter Specification
Support customized memories.
mbist1.10 Cheng-Wen Wu, NTHU
120
I/O Specification
Four parameters
IO_type
IO_width
IO_latency
IO_packet_length
IO_type: input, output, or inout
IO_width: port width (#bits), can be a constant
or specified by user
mbist1.10 Cheng-Wen Wu, NTHU
121
I/O Specification
- IO_latency: port latency
mbist1.10 Cheng-Wen Wu, NTHU
122
I/O Modeling
- IO_packet_length: #bits packed within a clock
cycle for the port

mbist1.10 Cheng-Wen Wu, NTHU
123
Command Specification
- Specifies the memorys instructions
mbist1.10 Cheng-Wen Wu, NTHU
124
Task Specification
- Specifies a complete memory operation
- A task can be a single command or a sequence
of commands.
mbist1.10 Cheng-Wen Wu, NTHU
125
Delay Constraint Specification
- Specifies the minimal time interval between any
two tasks
mbist1.10 Cheng-Wen Wu, NTHU
126
AC Parameter Specification
- Specifies input and output delays
- Specified parameters will be inserted into the
synthesis script.
mbist1.10 Cheng-Wen Wu, NTHU
127
Memory Specification Example
- For ZBT SRAM:
Method A:
- @latency D = 1;
- @task write = {write};
Method B:
- @latency D = 0;
- @task write = {pre_write, post_write};
- The BIST circuit from method A is faster than
the one from method B, but it has higher area
overhead
mbist1.10 Cheng-Wen Wu, NTHU
128
Sequence Generation
- For each March element, the compiler generates
the command sequence according to the read
task, write task, and minimum delay between the
two tasks
- For example:
task read = {A}
task write = {B, C}
minimum delay between read and write = 10ns
clock period = 10 ns
Then the (rw) element becomes {A, nop, B, C}
- One can also optimize the command sequence
mbist1.10 Cheng-Wen Wu, NTHU
129
Fast Access Mode
mbist1.10 Cheng-Wen Wu, NTHU
130
Diagnosis Support
- The BIST circuit scans out the error information
(element, address, signature, and polarity) during
the diagnosis mode.
- Assume address 20h stuck-at 64h:
mbist1.10 Cheng-Wen Wu, NTHU
131
Multiple RAM Cores
- Controller and sequencer can be shared.
controller
Test pattern
generator
Test pattern
generator
sequencer
Ram Core A
Ram Core B
Ram Core C
Test pattern
generator
sequencer
mbist1.10 Cheng-Wen Wu, NTHU
132
Experimental Results
- The Built-In Memory List
DRAM
- EDO DRAM
- SDRAM
- DDR SDRAM
SRAM
- Single-Port Synchronous SRAM
- Single-Port Asynchronous SRAM
- Two-Port Synchronous Register File
- Dual-Port Synchronous SRAM
- Micron ZBT SRAM
- BRAINS can support new memory architecture
easily
mbist1.10 Cheng-Wen Wu, NTHU
133
Experimental Results
mbist1.10 Cheng-Wen Wu, NTHU
134
Experimental Results
- Four single-port SRAM BIST circuits share the same
controller and sequencer.
- Size of the SRAM core: 8K x 16
Original
BIST area for single-port
SRAM: 1438 (gates)
Total area = 1438 * 4 =
5752 (gates)

Shared
gate count: 3350

mbist1.10 Cheng-Wen Wu, NTHU
135
Experimental Results
- 8K x 16 single-port synchronous SRAM (0.25um)
- Area:
Die size: 1780.74 x 755.07 um
2

BIST area: 80.1 x 583.48 um
2

Area overhead : 3.4%

mbist1.10 Cheng-Wen Wu, NTHU
136
Experimental Results
- 2K x 32 two-port register file (0.25um)
Die size: 1130.74 x 936.34 um
2

BIST area: 77.88 x 620 um
2

Area overhead: 4.5%
mbist1.10 Cheng-Wen Wu, NTHU
137
Why RAM Diagnostics?
- Memory testing is more and more important
Memories are key components
- Represent about 30% of the semiconductor market
- Dominate the chip area/yield
- Memory testing is more and more difficult
Growing density, capacity, and speed
Emerging new architectures & technologies
Growing need for embedded memories
- Why diagnostics?
Yield improvement
- Repair and/or design/process debugging
mbist1.10 Cheng-Wen Wu, NTHU
138
Fault Model Subtypes
mbist1.10 Cheng-Wen Wu, NTHU
139
NTHU-FTC BIST Architecture
mbist1.10 Cheng-Wen Wu, NTHU
140
Test Mode
- In Test Mode it runs a fixed algorithm for
production test and repair
Only a few pins need to be controlled, and BGO
reports the result (Go/No-Go)
mbist1.10 Cheng-Wen Wu, NTHU
141
Fault Analysis Mode
- In Fault Analysis Mode, we can apply a longer
March algorithm for diagnosis
FSI captures the error information of the faulty
cells
EOP format:
mbist1.10 Cheng-Wen Wu, NTHU
142
Error Catch and Analysis
- Locate the faulty cells
- Identify the fault types
mbist1.10 Cheng-Wen Wu, NTHU
143
How to Identify Fault Type?
RAM Circuit/Layout Tester/BIST Output
mbist1.10 Cheng-Wen Wu, NTHU
144
March Dictionary
March 11N
E0 E1 E2 E3 E4 E5 E6 E7 E8 E9 E10
mbist1.10 Cheng-Wen Wu, NTHU
145
March Signature and Error Map
March Signature (Syndrome)
Error Map
mbist1.10 Cheng-Wen Wu, NTHU
146
MECA System
mbist1.10 Cheng-Wen Wu, NTHU
147
Error Analyzer
Data log parser
Tester/BIST data log
Fault analysis
Error maps
Fault maps
March
Dictionary
mbist1.10 Cheng-Wen Wu, NTHU
148
Fault Analysis
- Derive analysis equations from the fault dictionary
- Convert error maps to fault maps by the
equations
mbist1.10 Cheng-Wen Wu, NTHU
149
Test Algorithm Generation
Start from a base test: generated by TAGS or
user-specified
Generation options reduced to read-insertions
mbist1.10 Cheng-Wen Wu, NTHU
150
Diagnostic Resolution
- Diagnostic resolution
faults detectable of #
faults hable distinguis of #
resolution Diagnostic =
mbist1.10 Cheng-Wen Wu, NTHU
151
Experimental Results
- Proposed diagnosis framework has been
applied to commercial embedded SRAMs
- Results for a 16Kx8 embedded SRAM
(FS80A020) are shown
- Tester log from Credence SC212 is examined
- Address remapping (logical to physical) is
applied
mbist1.10 Cheng-Wen Wu, NTHU
152
The Total Error Bitmap
mbist1.10 Cheng-Wen Wu, NTHU
153
Fault Bitmaps
Idempotent Coupling Fault Stuck-at 0
mbist1.10 Cheng-Wen Wu, NTHU
154
Redundancy and Repair
- Problem: We keep shrinking the feature size and
increasing the chip density and size. How do we
maintain the yield?
- Solutions:
Fabrication
- Material, process, equipment, etc.
Design
- Device, circuit, etc.
Redundancy and repair
- On-line
^ EDAC (extended Hamming code; product code)
- Off-line
^ Spare rows and/or columns
mbist1.10 Cheng-Wen Wu, NTHU
155
From BIST to BISR
BIST BISD BIRA BISR
BIST: built-in self-test
BIECA: built-in error catch & analysis
-BISD: built-in self diagnosis
-BIRA: built-in redundancy analysis
BISR: built-in self-repair
mbist1.10 Cheng-Wen Wu, NTHU
156
RAM Built-In Self-Repair (BISR)
RAM
BIST
Redundancy
Analyzer
Reconfiguration Mechanism
S
p
a
r
e

E
l
e
m
e
n
t
s

mbist1.10 Cheng-Wen Wu, NTHU
157
RAM Redundancy
- 1-D: spare rows (or columns) only
SRAM
Algorithm: Must-Repair
- 2-D: spare rows and columns
Local and/or global spares
NP-complete problem
Conventional algorithm:
- Must-Repair phase
- Final-Repair phase
^ Repair-Most (greedy) [Tarr et al., 1984]
^ Fault-Driven (exhaustive, slow) [Day, 1985]
^ Fault-Line Covering (b&b) [Huang et al., 1990]
mbist1.10 Cheng-Wen Wu, NTHU
158
Redundancy Architectures
mbist1.10 Cheng-Wen Wu, NTHU
159
An SRAM with BISR
[Kim et al., ITC 98]
mbist1.10 Cheng-Wen Wu, NTHU
160
A DRAM Redundancy Example
4 local spare rows per block
2x4=8 global spare columns
mbist1.10 Cheng-Wen Wu, NTHU
161
Definitions
- Faulty line: row or column with at least one faulty
cell.
- A faulty line is covered if all faulty cells in the line
are repaired by spare rows and/or columns.
- A faulty cell not sharing any row or column with
any other faulty cell is an orthogonal faulty cell.
- r: number of (available) spare rows
- c: number of (available) spare columns
- F: number of faulty cells in a block
- F:number of orthogonal faulty cells in a block
mbist1.10 Cheng-Wen Wu, NTHU
162
Example Block with Faulty Cells
mbist1.10 Cheng-Wen Wu, NTHU
163
Repair-Most (RM)
Run BIST and construct
bitmap.
Construct row and column
error counters.
Run Must-Repair algorithm.
Run greedy Final-Repair
algorithm.
mbist1.10 Cheng-Wen Wu, NTHU
164
Worst-Case Bitmap (After Must-
Repair)
r=2; c=4
Max F=2rc.
Max F=r+c.
Bitmap size: (rc+c)(cr+r).
mbist1.10 Cheng-Wen Wu, NTHU
165
Local Repair-Most (LRM)
- RM is not good enough for embedded RAM.
Large storage requirement: bitmap and counters
Slow
- LRM improves the performance.
Repair-Most based
Improved heuristics
Early termination rules
Concurrent BIST and BIRA
No separate Must-Repair phase
- LRM reduces the storage required.
Smaller local bitmap
- From (rc+c)x(cr+r) to mxn
mbist1.10 Cheng-Wen Wu, NTHU
166
LRM Algorithm
- Activated by BIST whenever a faulty cell is
detected.
- Fault Collection (FC)
Collects faulty-cell addresses.
Constructs local bitmap.
Counts row and column errors.
- Spare Allocation (SA)
Allocate spare rows or columns when bitmap is full.
Allocate spare rows or columns at end.
mbist1.10 Cheng-Wen Wu, NTHU
167
LRM: FC and SA
(1,0), (1,6), (2,4), (3,4), (5,1), (5,2)
mbist1.10 Cheng-Wen Wu, NTHU
168
LRM Example
(5,2) (5,4),(5,6),(5,7) (7,3)
mbist1.10 Cheng-Wen Wu, NTHU
169
Local Optimization (LO)
- LMR has drawbacks:
Selecting line with largest fault count may be slow.
Multiple lines may need to be selected for repair.
Area overhead is still high.
Repair rate depends on bitmap size.
- LO has a better repair rate based on same
hardware overhead, i.e., a higher repair efficiency.
Fault Collection (FC)
- Records faulty cells in bitmap until it is full.
Spare Allocation (SA)
- Exhaustive search performed for repairing all faults.
Bitmap cleared; process repeated until done.
mbist1.10 Cheng-Wen Wu, NTHU
170
LO: Column*/Row Selection for SA
A 1 means that the corresponding
col is selected for repair, unless empty.
* Assume column selection has a lower cost than row selection.
Col selection vector
1. Col 5 selected for repair.
Row selection vector
2. Row 5 is selected for repair.
mbist1.10 Cheng-Wen Wu, NTHU
171
LO Example
mbist1.10 Cheng-Wen Wu, NTHU
172
Essential Spare Pivoting (ESP)
- Maintain high repair rate without using a bitmap.
Small area overhead.
- Fault Collection (FC)
Collect and store faulty-cell address using row-
pivot and column-pivot registers.
- If there is a match for row (col) pivot, the pivot is an
essential pivot.
- If there is no match, store the row/col addresses in
the pivot registers.
If F > r+c, the RAM is unrepairable.
- Spare Allocation (SA)
Use row and column pivots for spare allocation.
- Spare rows (cols) for essential row (col) pivots.
SA for orthogonal faults.
mbist1.10 Cheng-Wen Wu, NTHU
173
ESP Example
mbist1.10 Cheng-Wen Wu, NTHU
174
Cell Fault Size Distribution
Mixed Poisson-exponential distribution.
mbist1.10 Cheng-Wen Wu, NTHU
175
Repair Rate Comparison
1,552 RAM blocks.
1,024x64 bits per block.
r from 6 to 10.
c from 2 to 6.
LRM bitmap: rxc.
LO bitmap: 8x4.
mbist1.10 Cheng-Wen Wu, NTHU
176
Normalized Repair Rate
mbist1.10 Cheng-Wen Wu, NTHU
177
Repair Rate (r=10)
mbist1.10 Cheng-Wen Wu, NTHU
178
Normalized Repair Rate (r=6)
mbist1.10 Cheng-Wen Wu, NTHU
179
Area Overhead
Overhead is about 5-12% for 16Mb DRAM, r=8, and c=4.
mbist1.10 Cheng-Wen Wu, NTHU
180
Computation Time (Simulated)