
Adedotun Adeyemo is originally from Nigeria. He joined
Oxford Brookes as a research student in January 2014 and his thesis title is ‘Design
and Analysis of Reliable Memristor-Based Architectures’.
How did you hear about Oxford Brookes University?
I got to know of Oxford Brookes through my MSc dissertation
supervisor at the University of Bristol. He was involved in some research with
my current supervisor here at Oxford Brookes.
What attracted you to Oxford Brookes University to conduct your research?
The nature of the research being carried out by the Advanced
Reliable Computer System (ARCoS) group attracted me to Oxford Brookes. ARCoS
presented me with the opportunity to work on an emerging technology that
involves a mix of electronics and computer science based research.
What were you doing before?
After completing a bachelor’s degree in Computer Science (2010)
in Nigeria, I worked for two years in an international bank before being
awarded a postgraduate (MSc) scholarship by the Nigerian government. I moved to
England in 2012 and completed an MSc in Advanced Microelectronic Systems
Engineering at the University of Bristol in 2013.
How easy did you find it to settle into the research environment?
I had a fairly good research experience during the MSc
programme at the University of Bristol. This prior experience made it easy for
me to adapt quickly into full research at Oxford Brookes despite the default
challenges that comes with research. Adequate support from the faculty,
especially my supervisory team went a long way in making my research journey a
rewarding one thus far.
Tell us about your research.
My research project focuses on
design and analysis of architectures made from an emerging technological device
called memristor. The memristor, a portmanteau of MEMory and ResISTOR (because
it behaves like a linear resistor with memory) is a newly discovered two
terminal circuit element (like the existing trio of resistor, capacitor and
inductor).
As transistor draws near to the
end of it scaling limit, there is the dire need for the semiconductor industry
to explore new technologies to ensure continuous evolution of computing
devices according to International Technology Roadmap for Semiconductors (ITRS).
Memristor – a novel electronic device is among the emerging technologies tipped
to usher in a new generation of computer architectures because of its simple
structure, non-volatility and smaller feature size than the transistor. Since
HP Lab fabricated the first physical memristor in 2008, there has been an
exponential rise in memristor-based research. Since HP Lab’s discovery, several
applications of this device have been proposed and demonstrated such as switch
to replace Complementary Metal-Oxide Semiconductor (CMOS) transistors in
digital logic design, analog circuit design, synapse in neural networks
implementation and storage cell in computing memory (primary and secondary).
Among these applications, the memristor’s use as a storage cell in memory
design is the most widely researched area especially non-volatile memory
(Resistive Memory) capable of eliminating the slow boot up process in our PCs.
Memristor-based memories
otherwise known as Resistive Random Access Memory (ReRAM) are fast gaining
popularity because of their unprecedented device density that can be achieved
via crossbar structures, high retention rate and their simplicity. Resistive
RAMs are considered a promising candidate to replace the current Flash memories
and Dynamic RAMs (DRAM) in future computers and mobile devices. High density
and low power memories can be designed with memristor as the main storage cell,
the memristor cell has an estimated area of 4F2 (F is the minimum
feature size). Memories are present in all electronic and computing systems;
their usage and demand for a more robust type keep increasing. As expected with
any technology at its infancy, memristor-based architectures are not foolproof.
Memristor-based architectures can successfully replace existing technologies if
their reliability problems are effectively resolved. Recent research made it widely
evident that reliability aspects of resistive memories radically differ from
those of their traditional counterparts. Although resistive memories are more
resilient to transient error but their yield and device fault rate will be
higher than conventional memories. Memristor-based memories are however
susceptible to other problems such as resistance drift, which affects data
integrity and current leakages (sneak-path) that leads to excessive power
consumption in the system. Sneak-path is a major problem with memristive-based
architecture. These leakages are a direct consequence of the bidirectional
properties of the memristor (allows current to flow in both direction) and also
aided by the crossbar architecture.
My research project aspires to be
a key enabler for the effective deployment of emerging memristor-based
resistive memory technologies in all segments of computing systems by
innovations towards improving their weakest aspect: reliability. Resistive
memories bear the hope of extending the validity of Moore’s law for several
decades, in terms of both memory devices density and performance: this is a
unique opportunity for the university to stay on the leading edge of computing
systems developments for many years. The application of these project results
will span most spectrum of the electronics domain. All major industrial players
of the semiconductor marketplace have a big interest in next generation
memories that are anticipated to resolve the scaling problems of current scaled
CMOS technologies. Such emerging memories are planned to be used both as
stand-alone products, or integrated in complex chips, in a diverse range of
markets from high performance computing (servers and data centres), to mobile
devices, to critical electronics for transportation (automotive, railway) as
well as avionics (satellites and space-crafts).
What do you enjoy about being a research student?
I love the freedom to work on what I am passionate about. I
am excited about the opportunity to be pioneer of novel and game-changing
concepts. On the flip side, research can be a daunting task, especially when
the project is based on an area that has not been previously explored in depth.
Perseverance, diligence and a clear goal keep researchers on course.
What are your future plans?
I find the research and academic experience so far to be an
exciting and rewarding experience. I plan to delve into full research after my
PhD so as to be consistent contributor to the knowledge domain.