Facing up big challenges

 By Eniday Staff

What do our car satnavs, history’s most famous lunar mission, a young woman and Eni’s HPC4 supercomputer have in common? Let’s find out!

Female ingenuity taking us to the Moon

The name of the project was Apollo. Among the many challenges involved was one in particular, guiding the lunar module in the extremely delicate lunar landing phase. And the lives of a three-man crew were on the line.
Nowadays, we’d assume this was a job for a computer. More specifically, we’d get software to provide instructions, the same thing that happens every time autopilot is engaged on an airliner. Things were different in 1969: it wasn’t a given by any stretch of the imagination. We all know what happened now, but what went into making it happen is probably less well known. On one hand is the courage of three astronauts prepared to face the unknown and on the other is the technology and ingenuity of man. Or woman, in point of fact: Margaret Hamilton, plus her team of programmers.
Margaret wasn’t yet 30 in 1969, but already had a small daughter who followed her everywhere, including inside the MIT Instrumentation Laboratory. On the smaller side, she had big glasses and a charming smile. At first glance you might have said she was 18. She oversaw the software design for the on-board Apollo Guidance Computer for the Apollo 11 Mission command and lunar modules. This computer had to be programmed for an absolutely critical and complex mission, but this was at the very beginning of the discipline now known as software engineering. Assembly language, more machine than human, was the only game in town. The solution was to be found in Margaret’s insight and in the skill of her development team. She foresaw the possibility of overload and conflict in terms of information during the lunar landing. This made it necessary to organise tasks by priority, excluding unnecessary ones if applicable. Her team wrote the program that allowed humanity to make the historic flight of 20 July, 1969.

Margaret Hamilton in action (NASA)

Apollo Guidance Computer

Hamilton and her team had to work within the Apollo Guidance Computer’s memory – a grand total of 152k – and clock speed of 0.043 to 2 MHz. They developed software for the machine that allowed her to coordinate the flow of data from the gyroscope navigation system, telescope and two radars and give the astronauts control over the engines and all on-board operations. The astronauts weren’t specialist IT technicians but could make use of a simplified interface involving a numeric keypad and indicator lights. This allowed them to run programs or perform operations using combinations of two digits.
It officially marked the start of a revolution in the relationship between man and machine, with software engineering – Margaret’s child – starting to take its first steps, like her actual daughter did perhaps. Not even 50 years have passed since the night when mankind succeeded in surpassing its imagination. And everything’s changed when it comes to IT. Computers and programming languages, were once the preserve of government and the military: today they’re in our homes and part of our everyday lives. That meant basic games to begin with, and rudimentary graphics. Something approaching a realistic level of graphical detail was to come in later, more complex games. At the same time, those of a geekier bent started getting worked up about code, from C++ to Java. Programming courses were added firstly in humanities departments and then in secondary school. Today it’s the turn of primary schools, in defined ‘dedicated contexts’. There’s something fun in just sharing your activity when it comes to programming, but it also offers an opportunity to develop skills and tools for solving complex problems, creating project and getting your ideas across. It represents a whole world of possibilities and, thanks to the wealth of knowledge offered by the web and the extensive presence of computers, it’s truly accessible to all. Since Hamilton’s time, enormous progress has been made. What was once used to reach the Moon is now used for completely different purposes, like oil exploration and production from fields.

The user interface unit (display and keyboard) of the Apollo Guidance Computer DSKY (NASA)

The HPC4 supercomputer and internal expertise: a winning combination

Eni officially presented the HPC4 supercomputer just a few days ago. Its 18.6 petaflops (peak) and graphics processing units (GPU) are now in action at the Green Data Center, processing along siide conventional CPUs.
This super-powerful computer is also called a cluster, and means expertise too, as far as Eni is concerned. You can’t expect to buy and use such a big and complex computer right out of the box. For one thing, you need to know how to use the same components as everyday computers in order to benefit from the ever decreasing power consumption that means we can go longer without plugging our laptops in. So you need to know what’s out there commercially. And after you’ve found the components you need, you need expertise in designing the cluster architecture. We have that too. The additional distinctive element was and is Eni’s capacity to develop software. It once took us to the Moon, and now it helps us make amazing discoveries: last but by no means least, Zohr.
Eni chose to develop proprietary software in house, and as early as the late 1990s had started to hire the first researchers, physicists, mathematicians, engineers and computer scientists to permit autonomous software development. This was supported by researchers and developers at universities, supercomputing centres and research institutions, all Italian. It takes knowledge of application areas and the ability to make it a part of computer-executable programs to develop tools for processing data from the field and provide Eni technicians with results that let them make new discoveries and work fields that have already been discovered as intelligently as possible.

Farsighted tools

We’re counting lines of code in millions now, code that gives us images with detail down to just a few metres: we see what’s really hiding in the heart of the rocks and depths of the ocean.
But it’s more than that. Simulating oil systems amount to time travel. You look back millions of years and retrace the development of sedimentary basins, establishing whether the conditions arose for the generation and accumulation of hydrocarbons in the most distant geological eras.
And as soon as we’ve made a new discovery today, it’s already time to look to the future instead, to predict it. That is, modelling the field and its fluids so that we can tell in advance how it will react to new wells and to production, in order to proceed safely and efficiently in all circumstances.
Visualising what we can’t touch, modelling complex systems, predicting future responses: all these abilities are rooted in Margaret Hamilton’s pioneering work. Her lunar navigation system is the precursor to all the software that outdoes even the latest video games as far as the men and women at Eni are concerned. If possible, it makes this work even more fascinating: fun even!
It takes some brute computing force to do this – courtesy of HPC4, a supercomputer with 1600 nodes – as well as a wealth of expertise, new ideas, programs and programming languages, teamwork and engineering. You might not expect this undoubtedly stimulating area to have a place in an oil and gas company. It just goes to show that while the worlds oil and IT may seem poles apart, you always find genuine innovation in the most unexpected places.

Apollo Moon landing 11, 20 July 1969

READ MORE: All about HPC4 supercomputer by Eniday Staff


about the author
Eniday Staff