Question 3. (45 marks)
The Isaac Research Centre has sent the fastest spaceship named the Explorer X to
the Mars for measuring a comprehensive set of the environmental data including the
atmospheric pressure, humidity, temperature, winds and also ultra-violet radiation levels, etc.
Copyright By PowCoder代写 加微信 powcoder
sampled simultaneously at 1,500 spots over the surface of the Mars with flying drones
installed with specific sensors. All data communication between the Mars and Earth has to go
through satellites with an overall transmission delay of 3 to 5 minutes due to the huge
distance involved. At each sample spot, each flying drone installed with sensors can measure
the data per minute, and then send the measured data through satellite communication back to
a high-performance computer system for further analyses using machine learning algorithms.
A team of computer engineers and researchers is trying to design the targeted computer
system, and has considered the possibility of building:
i) a scoreboard computer with a very low-latency data interconnect linking up a cluster of
ii) a Tomasulo computer integrated with a very low-latency common data bus to link up a
cluster of GPs as the reservation stations;
ili) a public/private cloud computing platform integrated with a cluster of servers installed
with GPUs;
iv) a hybrid computer system combining any 2 of the above options from i) to ili), e.g. a
hybrid model combining the scoreboard computer in i) with the Tomasulo computer in il)
to run machine learning algorithms for real-time data analyses on all the collected raw data.
Based on the above possible solutions considered by the team or other new design(s) that you
may think of, clearly describe the most appropriate high-performance computer system design
for real-time data analyses in this Mars exploration project, and give a critical evaluation
based on various aspects including the cost-effectiveness, extendibility, reliability, security
and scalability, etc. of the resulting high-performance computer system.
程序代写 CS代考 加微信: powcoder QQ: 1823890830 Email: powcoder@163.com