Photo:

Dave Bond

Loving the live chat - Great questions

Favourite Thing: As I am involved in Scientific Computing I enjoy the small projects that are required to enhance an overall system. One of the latest ones was writing an application to watch real time performance statistics from one of Diamonds large storage arrays.

My CV

Education:

University of Portsmouth

Qualifications:

BSc(Hons)

Work History:

Hitachi Data Systems then Diamond Light Source

Current Job:

Scientific Computing Systems Administrator (its more interesting than it sounds)

Employer:

Diamond Light Source

Me and my work

Computing geek, involved with developing and maintaining Scientific Computing at Diamond

We are a small team of eight people at Diamond maintaining and developing all the computing systems, involved in data collection, analysis, data storage and running of the synchrotron (a type of particle accelerator).  Though being a small team I get to do a bit of everything, I specialise in the data storage systems.  A typical system my be between one to two petabytes.

  • If the average MP3 encoding for mobile is around 1MB per minute, and the average song lasts about four minutes, then a petabyte of songs would last over 2,000 years playing continuously.
  • If the average smartphone camera photo is 3MB in size and the average printed photo is 8.5 inches wide, then the assembled petabyte of photos placed side by side would be over 48,000 miles long – almost long enough to wrap around the equator twice.
  • One petabyte is enough to store the DNA of the entire population of the US – and then clone them, twice.

We also require this storage to record data at about 30 gigabytes per second for our fastest file system.

  • So if an average DVD is 4.7GB we are storing approximately 6.4 DVD films per second.  We have three and soon to be four of these file systems operational at Diamond.

Other than that there are about two thousand CPU cores in our clusters and a few hundred servers that we operate.

 

It might sound busy but a large part of our work is ensuring the system can look after itself allowing only eight people to run all of this.

 

 

 

My Typical Day

Ensuring the scientific computing systems are avaliable allowing Science to happen at Diamond

Start … Coffee …. But I am fortunate that most of what I do is innovative, so I do not need to do much repetitive work.  I let computers do that as that is one thing they are very good at.  I do not have a typical day in many ways as every days problems are different.

One day I might be on a beamline (where the experiments take place) repairing at a detector server.  This is the device involved with looking at the sample much like you can take a picture with a digital camera but this can see X-Rays and can take a picture a hundred times a second.

Other days I may be involved with buying and testing the storage systems that will ensure Diamond will be able to store the experimental data in three years time.  Currently this is 30 gigabytes a second so 6.4 DVD’s a second but in three years this could double.

Or even writing software to gather statistics, automate repetitive tasks or software based tools.

 

 

What I'd do with the money

Create a demo of how Scientific Computing at Diamond works

Diamond Light Source has a number of public open days where public and students can come into Diamond and see what we do.  The is one of the few points in the year that you can stand next to the accelerator.  This is normally closed off because of the radiation risk.    On these open days different teams at Diamond show what they do with a demo.  My last demo was creating a small example of how the cluster computing works at Diamond using Raspberry Pi’s.  I have 6 of them doing a demo of how splitting up one task over 6 computers works and the expected decrease in time you see that this task takes.  This was nice as if you were interested in this there is no reason why you could not reproduce this at home for little money, and you would learn a lot.

We have a open day coming up in the summer where we are teaming up with the Rutherford Appleton Lab who we share the same site and for this and the other open days I would like to create a better demo of how scientific computing happens at Diamond.

 

From what happens when you collect an image to how it is processed on a high performance cluster computer and stored on one of the high performance file systems.  Most of what makes my job interesting you cannot see as it is hidden inside a computer.  What I want to demo is all these steps and tasks using things that everyone has access to.  So Raspberry Pi’s and Lego, there would be instructions so that anyone could create this demo or part of it as a project.

My Interview

How would you describe yourself in 3 words?

food loving geek

Who is your favourite singer or band?

Avicii

What's your favourite food?

Curry

What is the most fun thing you've done?

Driven a steam roller

What did you want to be after you left school?

Mechanical Engineer

Were you ever in trouble at school?

Not often I was good at not getting caught

What was your favourite subject at school?

DT

What's the best thing you've done as a scientist?

Completly unrelated to IT but held one of the oldest things on earth at half a billion years old

What or who inspired you to become a scientist?

5 years working in a real job I liked the University work style more, so came to Diamond

If you weren't a scientist, what would you be?

unemployed… or working on less interesting things for another company

If you had 3 wishes for yourself what would they be? - be honest!

three more wishes, never having to worry about bills, illness etc and to live life to the full

Tell us a joke.

Why do programmers always mix up Halloween and Christmas? …… Because Oct 31 == Dec 25!

Other stuff

Work photos:

myimage1

This is the view from just outside my office looking over the experimental hall

myimage2

This is the view inside the storage ring, the part of the accelerator that holds the electron beam and creates the synchrotron radiation (light) for experiments.  The part of the spectrum produced is from infra red to xrays including visible light.

myimage3

This is one of the experiment hutches where the sample is viewed using the detector.  This is the square item in the blue cradle.

myimage4

Though not as interesting, just showing even in a place like this a messy office is allowed.

myimage5 myimage6 myimage7

These are views of the main data centre.  Where most of our computing is housed.

This is an aerial view if Diamond.  If you drive down the a34 in Oxfordshire you will be able to see it