Title: eVIKINGS II Meeting
1Design and Testof Digital Systems
Raimund Ubar Tallinn Technical University raiub_at_pl
d.ttu.ee
2Roadmaps
- International Technology roadmap for
semiconductors - MEDEA Industry-driven pan-European programme for
advanced co-operative RD in microelectronics - A dependability roadmap for the information
society in Europe
3System Design at a Crossroad
- Productivity Crisis
- 21 Productivity Increase / Year vs.
- 58 Complexity Incease / Year
4Electronic Design Automation (EDA)
- System on a Chip development Yesterdays chip
is todays functional block!
New design methodologies are needed
5Technology trends
6Impact of new design technologies on cost
Tall thin engineer
Large block reuse
Intelligent test bench
7Landscape of design technology
8EDA challenges
- Designers need to increase the level of
abstraction, because the amount of information in
the designs exceeds their ability to track the
information? increased use of synthesis
technologies - Necessity to reuse already designed, tested and
synthesised functional blocks ? Intellectual
Property (IP) - Validation High-Performance simulation tools for
the entire design are needed (including early
high-level timing validation) - Improvement of the cooperation of different
EDA-tools? Standards have to be defined
9FW 6 - EuroSoC Key Nodes
- Network on chip (Hannu Tenhunen, Sweden)
- Multiphysics (Marta Rencz, Hungary)
- Mixed signals (JoseLuis Huertas, Spain)
- Low power (Christian Piguet, Switzerland)
- Integrated intelligence (Patrick Dewilde, The
Netherlands) - Test and debug (Ch. Landrault, France)
- Formal mehtods (Dominique Borrione, France)
10FV 6 - EuroSoC Key Nodes
- Modeling (Alain Vachoux, Switzerland)
- Reconfigurable SoC (W. Rosenstiel, Germany T.
Arslan, UK) - Design methods and IP reuse (J. Vounckx, Belgium
J.C. Lopez, Spain) - SoC programming model (E. Villar, Spain)
- SoC application (N. Wehn, U. Germany L.Lavagno,
Italy) - SoC education (J. Madsen, Denmark)
11EuroSoC Thematic Areas in Test
- Research topics
- Digital, Analog, Mixed-Signal, and RF
- Failure Analysis, Defect and Fault Modeling
- ATPG, BIST, DFT
- Test Synthesis , Test Resource Partitioning and
Embedded Test - Defect/Fault Tolerance and Reliability
- Debug and Diagnosis
- Research communities
- European Group of the IEEE Test Technology
Technical Council - International Test Conference (ITC), VLSI Test
Symposium (VTS) - European Test Workshop (ETW), DATE C, Online Test
Symposium (IOLTS)
12Paradigm Shift in Test
- Sky-rocketing complexities with decreasing
pin/gate ratio, higher frequencies and decreasing
product life cycles - New defects due to technology evolution
- Hardware/software designs, test of programmable
systems - IP(core)-based designs
- Increasing use of analog, mixed-signal and RF IPs
- Increase of temporary faults (transient and
timing faults)
13Current Status and Barriers to Overcome
- Current status
- Testing cost may represent up to 50 of the
manufacturing cost of SoC - Some of the barriers
- Development of nanometer fault models
- ATPG, BIST for real defects
- Reduction of test execution time
- Power consumption during test
- Optimized Test Resource Partitioning strategies
- Obsolescence of Iddq testing
- Analog BIST
- Tolerance of temporary faults
- DfD methodology
14Testing and Quality
How much to test?
The problem of testing can only be contained not
solved T.Williams
15Complexity vs. Quality
- Problems
- Traditional low-level test generation and fault
simulation methods have lost their importance
because of the complexity reasons - Traditional Stuck-at Fault (SAF) model does not
quarantee the quality for deep-submicron
technologies
16Complexity vs. Quality
- New solutions
- The complexity is handled by raising the
abstraction levels from gate to
register-transfer, instruction set architecture
(ISA) and behavioral levels - But this moves us even more away from the real
life of defects (!) - To handle adequately defects in deep-submicron
technologies, new fault models and
defect-oriented test generation methods should be
used - But, this is increasing even more the complexity
(!) - To get out from the deadlock, the both trends
should be merged into hierarchical approaches
17Research in TTU (ATI)
- Modeling of Digital Systems at different levels
to cope with the complexity - Test generation and fault simulation
- Hierarchical approaches to test generation and
fault simulation - Intensive international cooperation
- Joint design and test flow methods and tools
(integr. research at ATI) - BIST (cooper. with U Stuttgart, Uni Linköping,
Artec Design) - Defect-based approach (cooper. With TU Warsaw,
Slovak Academy) - Collaborative design via Internet (Fraunhofer
Inst. In Germany) - New teaching methods, tools, research scenarios
(TU Ilmenau) - Subcontractors in red
18DECIDER Hierarchical ATPG
Modules or subcircuits are represented as
word-level Decision Diagrams
19Joint VHDL Design and Test Flow
Cooperation with LIU (subcontractor) and in TTU
- High-level synthesis
- CAMAD developed at Linköping University
- Test pattern generation
- DECIDER developed at Tallinn Technical University
- Interfaces to
- behavioral and RT-level VHDL and
- EDIF netlist formats
20BIST (Optimization and Hybrid Solutions)
Cooperation with LIU (subcontractor)
- Combining
- on-line generated pseudo-random patterns
- with pre-generated and stored test patterns
- Problems
- To find the best characteristics for test
generator (PRPG) - To find the best level of mixing pseudo-random
test and stored test as the tradeoff between
memory cost and testing time
SoC
ROM
. . .
Core
Test Generator
.
CORE UNDER TEST
Controller
Response Analyzer
21International Virtual Lab Tool integration
Cooperation with EAS/IIS, Dresden EV-subcontractor
22Tools for test generation and fault simulation
Methods Single fault Parallel Deductive
Fault models Stuck-at-faults Stuck-opens Delay
faults
Methods Deterministic Random Genetic
Levels Gate Macro
Test Generation
Fault Simulation
Fault Location
Design
Test
BIST Simulation
Methods BILBO CSTP Store/Generate
Fault Table
Fault Diagnosis
Test Optimization
23Applets for Learning RT Level Test
Cooperation with TU Ilmenau, Germany EV
subcontractor
- For learning problems of RT-level design and
test - - Design
- - Simulation
- - Fault simulation
- - Test generation
- - DFT
- - BIST
24Conclusions
- Long term goal of the NoE EuroSoC in Framework 6
2003
2005
2007
2010
Get funding from second FP6 Call and other
public funding
Start
Convince industry to partially fund the network
Network Fully funded by Industry