TY  - THES
U1  - Master Thesis
A1  - Jung, Richard
T1  - Radiation Qualification of the Cologne Chip GateMate A1 FPGA
N2  - In this thesis, the radiation sensitivity of the novel Cologne Chip GateMate A1 field-programmable gate array (FPGA) is evaluated. An initial introduction of radiation mechanisms and their effects on electronics is given, followed by a brief overview of radiation test standards. The common elements present in FPGAs are discussed, which is followed by details of the GateMate FPGA device and a description of the software design flow. Afterwards, the development of a purpose-built printed circuit board (PCB) for radiation tests with the GateMate FPGA is detailed.
Four components of the GateMate have been tested during three radiation campaigns, as well as a benchmark circuit to compare the radiation performance of the GateMate with other FPGAs tested at the European Organization for Nuclear Research (CERN). The test architecture consists of the device under test (DUT) FPGA and a TESTER FPGA whose task is to provide inputs to the DUT and record its response. The DUT and TESTER designs developed for all tests are discussed in detail. Finally, the results obtained during the irradiation campaigns are presented, showing that the GateMate FPGA performs similarly to other FPGAs using the same process technology. Only the benchmark test was not finalized, as implementation problems prevented its completion in the given time frame. The thesis concludes with a comprehensive summary and outlook.
N2  - In dieser Arbeit wird die Strahlungsempfindlichkeit des neuartigen Cologne Chip GateMate A1 FPGA untersucht. Zunächst wurde eine Einführung in Strahlungseffekte und ihre Auswirkungen auf elektronische Komponenten gegeben, gefolgt von einem kurzen Überblick auf aktuelle Strahlungsteststandards. Die üblichen Elemente in FPGAs werden diskutiert, gefolgt von Details über GateMate spezifische Elementen sowie eines Überblicks über den Software-Design-Flow für GateMate FPGA Anwendungen. Im Anschluss wird die Entwicklung eines PCBs für Bestrahlungstests des GateMates detailliert.
Vier Komponenten des GateMate wurden während drei Strahlungskampagnen getestet, sowie eine Benchmark-Schaltung, um die Strahlungsempfindlichkeit des GateMate mit anderen am CERN getesteten FPGAs zu vergleichen. Die Testarchitektur besteht aus dem DUT FPGA und einem TESTER FPGA, dessen Aufgabe es ist, Eingaben an das DUT zu liefern und dessen Reaktion aufzuzeichnen. Die für alle Tests entwickelten DUT- und TESTER-Designs werden im Detail diskutiert. Schließlich werden die während der Bestrahlungskampagnen erzielten Ergebnisse vorgestellt, die zeigen, dass der GateMate FPGA ähnliche wie andere FPGAs mit vergleichbarer Prozesstechnologie liefert. Lediglich der Benchmark-Test wurde nicht finalisiert, da Probleme bei der Implementierung die Fertigstellung im vorgegebenen Zeitrahmen verhinderten. Die Arbeit schließt mit einer umfassenden Zusammenfassung und einem Ausblick ab.
KW  - FPGA
KW  - Radiation qualification
KW  - CERN
KW  - LHC
KW  - GateMate
Y2  - 2023
UN  - https://nbn-resolving.org/urn:nbn:de:hbz:dm13-33643
U6  - https://doi.org/10.26205/opus-3364
DO  - https://doi.org/10.26205/opus-3364
SP  - 118
S1  - 118
ER  - 
TY  - THES
U1  - Master Thesis
A1  - Shi, Yanchen
T1  - Power Simulation of a MIPS microAptiv UP Core implemented as a virtual ASIC prototype in a 65nm CMOS technology
N2  - This thesis presents a power simulation of a MIPS MicroAptiv UP Core implemented as
a virtual ASIC prototype using Taiwan Semiconductor Manufacturing Company(TSMC)
65 nm CMOS technology. Based on the MIPS instruction set program data is generated
and introduced in the simulation by means of initialization files. Before the simulation,
technology specific SRAM modules are integrated into theMIPS core. Two different programs
are used for power characterization. The first program performs frequent memory
accesses by means of load/store word instructions, while the second program is a loop
which operates on registers only and mainly increments addresses. The simulation is
based on a virtual prototype which is generated by synthesis and place & route including
post-layout parasitic extractions. The stimuli for the power extraction is generated
via gate-level simulation and forwarded to the power calculation engine. The effect of X-propagation
on gate-level simulations is avoided by modifying the address-related statements
in the execution data path module, which use another form of 2 to 1 multiplexer,
setting the output to zero for all input signals even with an initial value of ’x’ without
changing the functionality. Finally, the consumed power is provided by reports generated
by the power simulation engine. The memory-centric program consumes 35.39mW
of internal power using instructions, which is 0.73mW less than the internal power of the
register-centric program, and the overall average power is also lower by almost 0.7mW.
Y2  - 2022
UN  - https://nbn-resolving.org/urn:nbn:de:hbz:dm13-32171
U6  - https://doi.org/10.26205/opus-3217
DO  - https://doi.org/10.26205/opus-3217
SP  - 93
S1  - 93
ER  -