Posted on by Muhammad Omer Posted in Blogs 2020 — No Comments ↓

The Beginning of the Journey to SoHPC 2020

It was a normal office working day. I was working on my research project and suddenly I heard a “ding” bell. It was an email from my project research supervisor which mentioned Summer of HPC 2020 (SoHPC 2020). This was the first time I ever heard about this placement. In the first instance, it looked interesting so I spared sometime to read more about it on the official website. After reading online and discussing it with my supervisor, I felt very excited and was convinced to fill an application.

Now, came the first challenging bit, write an effective and interesting application to compete with participants from all over the world. I don’t know that’s some billion students. I was confident and optimistic but at the same time was a bit nervous when I imagined the number of applicants will be applying. I really wanted this opportunity because it completely was aligned with my research project and offered skills and experience which was required to step-up my research.  I just could not let the opportunity slip out of my hands.

Who Am I

Oh sorry! I forgot to introduce myself. My name is Muhammad Omer and I am 27 years old. I am based in Manchester, UK. Currently, I am a second year postgraduate research student at the University of Manchester.

My Research

My research is in use of 3D scanning, Machine Learning, High Performance Computing and Virtual Reality to automate inspection of Civil Infrastructure. The work is novel and of timely importance. I have successfully published a journal article titled “Use of Gaming Technology to Bring Bridge Inspection to the Office” in the Journal of Structure and Infrastructure Engineering (first quartile). Click the link below to for an interesting read.

https://www-tandfonline-com.manchester.idm.oclc.org/doi/full/10.1080/15732479.2019.1615962

When people ask me to describe my research, I use this one line statement, “Instead of Office Going to the Bridge, Bridges are coming to the Office”. Sounds crazy right! Since when bridges learned to walk and fit into a room. I hope the picture line below will make things clearer.

My Research

The Tragedy

Coming back to the story! I invested a lot of energy to prepare my application and then finally submitted it a few hours before the deadline. While I was dreaming to work in different super-computing facilities across Europe, all of a sudden there was breaking news everywhere, a state of panic, “these many people are infected, these many deaths and so on”. Our planet just turned from the most colourful place to the scariest place in the solar system. Corona had over ruled our planet! The only way to protect others and oneself was to “STAY HOME!”

I really felt demotivated imaging the program to be postponed/cancelled Thanks and many applause to the administration for bringing in the COVID – 19 version of SoHPC.

A Worst Nightmare

Still one the biggest challenge remained. That email which says “you are selected”. It was a normal home office working day. Suddenly I heard a “ding” bell. It was an email from SoHPC 2020 titled

“[PRACE events] Your PRACE Summer of HPC 2020 application is rejected”

I was really upset and wondered where I could have improved my application to make it more effective. I still remember that day, I didn’t work on anything, and my mood was completely off. Obviously, that didn’t changed anything, so I just convinced myself that maybe next year I will try harder. This years its anyways online, never mind!

The Miracle

I slept late that night. I woke up in the afternoon and checked my email. There was an email titled

“[PRACE events] Your invitation to PRACE Summer of HPC 2020”.

For the first few seconds, I was numb and didn’t knew how to react. Slowly, I started bouncing off my chair in excitement and then immediately contacted my supervisor to share the great news. I couldn’t believe what happened a day before but just a worst nightmare. I was elected for Project 2001: High Performance Machine Learning”.  

SoHPC 2020 kicks-off

Honestly, I cannot compare the excitement of face to face training week in Vienna with COVID -19 version of SoHPC 2020. Now, the whole program is online. But I must acknowledge the efforts of the organisers. The training week turned out to be more than my expectations. We had great lectures from the experts. Hand-On exercises to get feel of super-computing. Yoga sessions for physical fitness and breakout sessions to meet fellow colleagues. So far it has been a great experience, I really look forward to what’s coming next.

I hope and pray everybody remains safe and things go back to normal asap. The first thing I would like to do as things reopen is go for hiking and do my superman jump again!

A trip to Patterdale. Superman jump!

Posted on by Jerónimo Sánchez García Posted in Blogs 2020 — No Comments ↓

During the last 5 years, a big chunk of computations done previously on the CPU are being moved into GPUs because graphic cards are really good at working with enormous amount of data, so, why is x86 trying to improve itself by adding instructions sets like AVX-512 when that kind of computations are done now on GPUs? Isn’t there other kind of CPU architecture more suited for general purpose server CPUs? This series of blogs will give some hints about that, while explaining some interesting projects related to ARM, but…

First things first

I am Jerónimo Sánchez, a 20-years-old and a humble Computer Science student at University of Almería (UAL), Spain. I am about to start my 4th and last year studying CS here.

Me, acting weird, circa 2019

I am specializing on IT systems (there is not many varieties where I can choose from), although my main interests are in HPC and in working close to the hardware.

Getting personal, I am into soft sports, such as swimming, cycling, … because of a knee injury. I also really enjoy spending time with my friends playing board and video games. I especially like playing Dungeons and Dragons, but our sessions are quite scarce as this game is really a time consumer.

As a fellow Computer Scientist, I am also into solving problems. That means that I spend part of my day thinking and coding solutions to said problems. I mainly use a Raspberry Pi to automate some of my room tasks, but, since I am sort of power outlets, I do this problem-solving exercise as a puzzle.

Regarding Summer of HPC, I have found it to be the perfect way to spend this summer, as, since the start of the pandemic, my not-now-current summer internship was ended, so the timing was excellent.

Here, in Summer of HPC, I will be working on porting and benchmarking a fully ARM based cluster (project #2006). As part of this enterprise, I will be working on two sub projects:

  • The first one will give some insights about the implementation of the file-system drivers of the cluster, based on benchmarks powered by IOR tool.
  • The next and last subproject will try to “cheat” the MPI library so it can work correctly with SMT-X – Simultaneous Multithreading – levels where X > 1. In doing so, project #2006 can yield some valuable and helpful insights about ARM on the server space.
Behold one of the cabinets of the ARM cluster

This cabinet is part of the Fulhame cluster, owned by EPCC (Edinburgh Parallel Computer Centre). Bellow its components:

This HPE Apollo 70-based system consists of 64 compute nodes , each with two 32-core Cavium ThunderX2 processors (ie 4096 cores in total),  128GB  of memory composed of 16 DDR4 DIMMs, and Mellanox InfiniBand interconnects.

Retrieved from the Fulhame website

So, why there is an ARMageddon?

Good question dear reader. ARM is a “novel” computer architecture, mostly known by the general public for powering their smartphones. Until now, ARM was dismissed as a high performance architecture mainly for the fact of being mostly designed for devices with power constrains (smartphones, embedded devices, …), but now it powers the TOP 1 supercomputer in the world: Fugaku.

To sum up:

As it is said, a picture is worth a thousand words

ARMageddon is an epic battle for the dominance of the server space with three main participants: ARM, x86 and PowerPC; being at the same time an architectural design battle: RISC vs CISC, but that is for another post.

For what is of our concern, given that all the computations are being moved into GPUs and server/supercomputer CPUs are getting a controller role in the data pipeline, processors can become low powered chips as they are only sending and controlling data, and the remaining calculations that are still done on CPUs don’t need to the quickest, just good enough.

For the reasons stated above, ARM is a great candidate for CPUs on the server space, also in the consumer space, and ARM offers some other advantages:

  • Custom chips can be designed. You are not tied to a Xeon CPU with features you do not need.
  • You are not also tied to a supplier; your custom design can be manufactured by different companies.
  • Intel Tax is skipped, although consumer Nvidia GPUs cannot be used on servers, so where is always a tax.
  • Performance / Price is better.
  • More energy efficient, so it can cut the electric bill.
  • Related to previous. Chips generate less heat, so infrastructure to dissipate that heat is cheaper and easier to maintain.

For now, thanks for reading, and it is a pleasure to be part of the ARMageddon. Until next post dear reader.

Posted on by Francesca Schiavello Posted in Blogs 2020, PRACE Summer of HPC 2020 — No Comments ↓

Well that’s an odd question… you just file up in a queue obviously, or the shortest queue if there are multiple ones, and get in the back of such line. Well yes, that is one way to do it, and this way is what we are mostly used to in our daily lives, but as we will see in this post, that’s not always the best way to do it.

Queueing theory, a mathematical area developed to solely study waiting lines or queues, tells us that there are several ways to get in a queue, each one with its own benefits and costs. The one method that we are most used to, is called FIFO, short for first-in first-out, where the users are serviced on a first come first serve basis. Now, whilst this seems fair, this isn’t always the most convenient or efficient way to form a queue. Shortest Job First (SJF), builds a queue based on a user’s servicing time such that users with smaller jobs get serviced first.  A job can refer to a multitude of things, think for example if you are in the grocery store, and you only want to buy a carton of milk, but the person in front of you has a whole shopping cart full, in the SJF scenario you could skip the person in front of you, and the cashier would service your job (i.e. make you pay for your single carton of milk) first. In SJF the average waiting time per person is decreased with respect to FIFO, but this scheduling of queue’s presents us with another problem called starvation, which means that the poor person with a full shopping cart of items might have to wait an undetermined amount of time before he gets to check-out. Another type of queueing type is based on the priority of users. Think of airports, where it doesn’t matter when a business class ticket owner arrives or how long it takes to service her, she gets priority and skips the line. These are just a few of the examples of scheduling algorithms used in queueing theory, this field can be studied much more in depth looking at the mathematics behind the arrival times of users, the average waiting times and servicing times.

Ok, great…. but what does all of this have to do with high performance computing (HPC)? I’m glad you asked, this in fact is a very big part of HPC and how clusters work. At Hartree Centre, different users, think master students, PhD students, staff, etc) can submit their programs to run on the cluster. As we saw in the previous post, these users have to submit an estimate of their running time such that the cluster can schedule the jobs in an appropriate manner, based on the scheduling algorithm chosen.

Hartree Centre uses the SLURM workload manager to schedule the cluster’s jobs. This is a flexible tool that allows the user, in this case the Hartree cluster, to decide how to best manage their scheduling algorithm. SLURM in fact lets the user decide how much weight to give to factors like job size, priority status, age of the job (which avoids the indefinite starvation of a job, i.e. at a certain point the shopping-cart full person gets priority) and more. All these factors can be modified at the discretion of the user, such that queueing can give more importance to any and all of these different aspects. The default scheduling algorithm for SLURM remains FIFO.

Apart from giving the SLURM user the power to give priority based on their needs, this is also important because it allows the cluster to be run efficiently, to consume less resources or energy if properly configured. To give a simple example, the figure below illustrates 4 jobs submitted to the cluster, in it we can see the difference between two scheduling algorithms on the left and right.

On the left we see the jobs compiled with SJF scheduler, and on the right we have the jobs compiled with SJF using backfilling. Note with the latter algorithm, the submitted jobs will finish running faster.

In the figure we can see the difference between what is commonly referred to as backfilling. In it, the 4th job is the longest job in the queue, but given that it needs fewer resources from the cluster, and that it doesn’t delay the 2nd or 3rd jobs, it can be back-filled.

This is some of the background theory I learnt the past week about the scheduling algorithms used in clusters and through SLURM. As mentioned in the previous post, these are effectively rendered inefficient if users don’t submit correct estimates of running times.

Stay tuned for the next blog post, where we will be exploring different regression methods and machine learning algorithms to predict more accurate running times to feed to the SLURM scheduler.

Posted on by Shiva Dinesh Chamarthy Posted in Blogs 2020, Participants 2020 — No Comments ↓

About Me:

My name is Shiva Dinesh. I come from India, but I am right now in the beautiful Franconian city of Erlangen. I am currently pursuing my Masters in Computational Engineering at the Friedrich-Alexander-Universität Erlangen-Nürnberg in Germany. In my free time, I love playing strategy games.

How I met HPC and SoHPC:

I got introduced to the High-Performance Computing in a course in my first semester of my Master’s. I was intrigued by the concept of HPC and realized its need and importance. I then took another course on HPC in which they recommended me to go to ISC Frankfurt – an HPC event. This was where I came to know about PRACE, and they made me aware of SoHPC. I undoubtedly wanted to take part in SoHPC, so I decided to apply for it in the next application period. Voila! I am here writing a blog on their website.

Me awkwardly posing at the ISC Frankfurt 2019 at the release of the 53rd edition of Top 500 List. This was the first time all top 500 HPCs were faster than one Petaflop.

The Project:

The idea of simulating flow around the submarines has captured my interest, so I have chosen the project #2021:Submarine Computational Fluid Dynamics. Considering my background in mechanical engineering, I believed that this would be the ideal topic where I could integrate fluid mechanics and HPC.

Training Week in Vienna:

Due to the pandemic, most of the summer schools got canceled but, luckily for me, not SoHPC. I am glad that the organizing team has managed to set up an online version of the summer school and have done a commendable job till now. The training week was well scheduled, which introduced us to the basics of parallel programming and gave us hands-on experience on the Vienna Scientific Cluster (VSC). The exercise and yoga breaks in between the lectures were well thought out, as it released the stress of sitting in front of a screen for long hours. It also gave me an opportunity to interact with like-minded people from the community, and I had fun interacting with them.

First Weeks at Luxembourg:

I am still daydreaming that I am in Luxembourg, and we will let it be that way for the moment. I got started to work on my project after the training week mentored by researchers from the University of Luxembourg. I began getting acquainted with the CFD concepts and visualization tools required to simulate the flow around the submarines, alongside my project partner Matt. So, for the next six weeks, I would delve into the realms of HPC and CFD.

If you are still here, Thank you for reading, and I will keep you updated about my project here.

“There is only one way to learn. It’s through action. Everything you need to know you have learned through your journey.”

Paulo Coelho The Alchemist

Hoping to have a great journey here and learn a lot of new things. Have a nice day, and stay safe! See you soon.

Posted on by Neli Sedej Posted in Blogs 2020, Participants 2020 — No Comments ↓

Welcome to my Summer of HPC blog! In this post I will tell a little about myself and about the very exciting start of the SoHPC program.

About me

My name is Neli and I am finishing my Master’s degree in chemistry at University of Ljubljana, Slovenia. Before enrolling into university I had to decide: biochemistry or computer science? I chose the former, but later I found the way to combine both. I got acquainted with supercomputing during my Bachelor’s thesis. This convinced me that big computers are much more thrilling for me than a normal biochemistry laboratory. I continued in this field in my Master’s thesis and I intend to stick with it for PhD. Computational biochemistry is a more and more important dicsipline of biochemistry and can be used to tackle many problems inaccesible to the experimental methods. For now, I have considered myself more of a “user” of HPC, but I desire to learn more about how it works and how to write my own, efficient software. When I heard about the SoHPC it seemed like a perfect program for me! Naturally, I chose a project that is connected to chemistry: Neural networks in quantum chemistry at CCAS in Bratislava, Slovakia. I am glad that I was chosen for a project about neural networks, since they are a very powerful method, becoming more and more popular in chemistry. I have a great interest in Neural networks, but I have not had the chance to program my own. And here is my chance!

Besides science, I also love nature, hiking and climbing. When I can not be outside, I spend my time with learning languages and playing guitar.

When I take a break from chemistry, mountains are where you will most likely find me. (Jezerska Kočna, Slovenia)

Starting off

I will not lie, I was very excited about spending a week in Vienna for the training week and another seven weeks in Bratislava. Due to COVID-19 the whole program is organized online. I have to say I was positively surprised about how well the online training week turned out! We could learn a lot about supercomputers and parallel programming from the comfort of our homes and we still got to meet a lot of new people. Another great thing is that we get to work on project in teams. I might have to find another reason to learn Slovakian, but nevertheless, I am sure that this is going to be an exciting summer!

Posted on by Benedict Braunsfeld Posted in Blogs 2020, Participants 2020, PRACE Summer of HPC 2020 — No Comments ↓

Hey Folks, it is been a while since I started my adventure in quantum computing at the Irish Center for High-End Computing and so far it is mind-blowing. Together with my colleague Sara, we started implementing an algorithm to compare strings using a quantum computing approach. So how does a quantum computer even work?

The basics

A classical computer performs operations using classical bits, which can be either 0 or 1. With many bits, we can create binary code that can represent texts, computer operations or any other kind of data and can be processed by a classical computer.

A quantum computer uses quantum bits or short qubits. There are a number of physical objects that can be used as qubits, for example, a photon or electron. The object needs to be small enough for quantum physics to apply and it needs a spin that can be influenced with a pulse from the outside. A qubit can be both in state 𝛼|0> and in state 𝛽|1> at the same time before measurement. The coefficients (𝛼, 𝛽) represent the state’s amplitude and the absolute values of them squared are the probability of measuring those states.

Blochsphere’s representing a single qubit. From the left to the right the qubits are in state |0>, |1> and both states at the same time.

So far so good but how does this enable us to perform interesting computations? For this, we have to take a look at multiple qubits interacting with each other. For example, if two qubits interact with each other we can create an entangled quantum state and put it in superposition, namely the state of 𝛼|00> + 𝛽|01> + γ|10> + 𝛿|11>.

The main difference between classical bits and quantum bits is the amount of information saved. While we could create the states 00, 01, 10 and 11 with two classical bits we still only have two bits of information, because which of the four numbers we have is determined by the value of the first and the second bit. In quantum computing, the state of the two qubits is determined via the four coefficients. You could say that two qubits hold the information worth of four classical bits. Saying this in a more general way we have classical information of two to the power of the number of qubits stored in our quantum states. Sadly we can not just measure a superposition, all we can measure are our qubits in their basis states |0> and |1>. So we have to manipulate our superposition in a way that the final result can be represented with a unique state expressed through the basis states of our qubits to be able to measure.

The actual project

In our project, we encode non-binary strings in quantum states and compare them with each other to determine the degree of similarity. This kind of application is widely used in bioinformatics to match gen or protein patterns. More about the actual implementation and the string comparison method will soon be posted on Sara’s blog. So stay tuned and stay safe!

Posted on by Rafał Felczyński Posted in PRACE Summer of HPC 2020 — No Comments ↓

Hi, I’m Rafał and this year I wanted to make something interesting during my 3 months long summer break and… SoHPC came with help!

I come from Poland and I’ve studied Biomedical Engineering and Computer Science simultaneously at the Wroclaw University of Science and Technology. Just before the end of 2019 I finished writing my biomedical engineering thesis, so since the beginning of 2020 I’ve had much more time to do computer stuff.

Don’t let the dirty water discourage you, Wrocław is really beautiful.

I love optimization – making programs faster and more reliable always brings a smile to my face. It is very important especially regarding real time applications running on microcontrollers, where every CPU cycle matters. My programming journey’s begun from Java but C and C++ really stole my heart.

Optimization is just making things better and life easier.
Image taken from here

About my project

In the January I saw SoHPC ad on my University facebook page. I checked the site and became interested in the programme, especially since I was looking for some kind of internship. I applied, got in and will be taking part in the project 2022 – Novel HPC Parallel Programming Models for Computing. I had never used supercomputers before and I thought this programme would be a great way to start, since I won’t be left alone and have some experienced mentor watching over me.

The project will focus on profiling and benchmarking some parallel programming models. I’m going to write some algorithms in C/C++ and compare the results obtained from running these models. Maybe I will find some bottlenecks or applications for which some models work better than the other ones? Who knows…

What I know for sure is that for me this summer will be a lesson on multitasking because since SoHPC is online this year, I’ve decided to take part in another internship programme organised by my University in my city. So at the end of August I’m going to find out if a human can be such a multitasker as a computer is but I’m very optimistic and I think it’s going to be fun!

Posted on by Andres Vicente Arevalo Posted in Blogs 2020, PRACE Summer of HPC 2020, Projects 2020 — No Comments ↓

How I end up here

Hi, I’m Andres Vicente but everyone calls me Andreu, I’m an Astrophysicist (it’s official since yesterday in fact) based on the Canary Islands. I heard about the Summer of High performance computing program (SoHPC for short) in the course of programming techniques at my university and I couldn’t resist joining. I can tell that it was an awesome decision. Before I can talk to you about my project and how I planned to use Deep Neural Networks (DNN) in the field of galaxies let’s start from the beginning:

Let’s start by warming up our neurons: Training week

SoHPC starts with a training week where we were instructed with the basics of HPC and where our brains start to get used to thinking about performance and parallelization. The four days of the training fly by even quicker than the executions of our programs in the Supercomputers and then we are on our own, facing the project we dreamed about. But we are not worried, we don’t feel alone, our mentors and peers will make the journey easier.

In my case The project was ” Object Detection Using Deep Neural Networks – AI from HPC to the Edge ” and it was the perfect opportunity to unify both of my passions, the Astrophysics and the High-Performance Computing.

HPC from home to the Universe

The objective of the project is to use DNN to detect objects in images and we are lucky because, in the Astrophysics world, a very big portion of the data obtained from the Universe is in form of pictures. In the past, most of the classification of the galaxy morphologies was done by simple human inspection. Nowadays, we have better tools to classify the galaxies but almost all of them need to be applied using an analytical model and fit it to every one of the observations made, which is obviously time-consuming and computationally expensive. DNN and his object detection capabilities open a new world of possibilities in Astronomy and Astrophysics because we will not only be able to detect morphologies of galaxies (which is a rather easy task), but we also could go beyond that and infer physical properties of the galaxy just by looking at the raw image!!

In our case, we will try to detect the orientation of galaxies (which is closely related to the angular momentum vector if you are wondering). This is not an easy task since we don’t have “training data” to feed our network because we don’t know the ground truth of this magnitude in the observed galaxies but… Here comes the HPC again to rescue us again.

We can mock the observed data with high-resolution simulations of galaxies were we know all the parameters. These simulations are done for experts in the field in the most powerful supercomputers in the world as for example the Illustris simulation done at PRACE supercomputers: https://prace-ri.eu/universe-simulation-illustris-is-an-ongoing-success-story/

From those simulations, we can render images of galaxies in any orientation and point the angular momentum as a vector as we can see here:

Figure 1: Rendered Images from the NIHAO simulation with the angular momentum superimposed as a green arrow. These images have been created to obtain our dataset to train the network.

These physical properties of the galaxies are relevant because they tell us the story of the galaxy evolution and how it has been formed and evolved. This helps us to understand our own galaxy and somehow why we the Universe is as beautiful as it looks.

With the Idea and the training the data, it’s time to get our hands dirty!!

We plan to use a convolutional deep neural network to do the job. The full architecture is not clear yet but the general scheme (simplifying a bit) will look something like figure 2.

Figure 2: Illustration of the DNN for computing the angular momentum of a galaxy.

I’m very excited to see how the project evolves and I hope you will join my journey and share some of my emotions to unveil the mysteries of the Universe through HPC!!

Keep an eye on the SoHPC blogs if you want to stay posted.

Posted on by Anssi Tapani Manninen Posted in Blogs 2020, Participants 2020, PRACE Summer of HPC 2020, Projects 2020 — 1 Comment ↓

Prelude:

Greetings dear blog intruder, this blog belongs to Anssi from the northern forests of Finland. My home office for this summer will be in secluded Kuopio. At this point, I would like to rectify common fallacy; there are no polar bears in here.

The PRACE Summer of HPC (HPC actually stands for high-performance computing) consists of annually organized projects in different supercomputing centres around Europe. As I am always watchful about embracing new information on the mathematics-related things, it was an easy decision to apply for the SoHPC. 

Figure 0: Transition between states of mine (Quantum incoherence was faint)

Enigmatic problems emerging from all kinds of disciplines have always been intriguing for me. Such as statistical inverse problems which I’ve been studying in the computational physics department at Eastern-Finland university. Albeit this kind of problems often crack one’s head and diminish the remaining sanity, they also offer the dire moments of euphoria. Similar challenges arise when designing well-functioning HPC algorithms and protocols. 

I consider myself as a sloppy coder; nonetheless, it would be still nice to manifest computed results at times. From SoHPC, I expect to learn some basic concepts of HPC and to extend my programming toolkit for the future. Furthermore, there is also a hazard to be comprehending new neat stuff from physics/mathematics. 

The project under code name 2011 and the training week:

My projects organizer is the Jülich Research Centre in Germany for which I’ll be working remotely for the next seven weeks. The projects essence is the lattice discretization of Quantum Chromodynamics (LQCD), which allows one to compute the behaviour of gluons and quarks by machinery (see https://summerofhpc.prace-ri.eu/high-performance-quantum-fields/). As the lattice tends to have an enormous amount of estimated parameters (e.g. colours and spins), the large scale parallel computing is harnessed to drastically reduce the computational time requirements.

So far, the SoHPC offered the training week for the participants in which some essential HPC concepts were introduced. Namely, how to use OpenMP, MPI and Cuda and how HPC architectures are often constructed. Overall, the training week was an enjoyable experience and had some refreshing daily embedded activities such as Yoga with professional (thanks to Cornelia, the yogi).   

Figure 1: Not all heroes wear blades, but this one does. Kept me cool during the training week.

Since the program is done remotely, I am not currently capable of providing any local views from Jülich. Instead, you can check out the Jülich blogs from the previous years. The plan is to publish sceneries from the vicinity of mine as quick as I get to outside.

To finalize the post, I would like to provide a few futile facts (someone could call these as opinions).

The best GPU: Nvidia GeForce GTX 770 GPU – Has served me well during the years, with a minimum amount of protests. 

The best juggling pattern: Mills mess (4 balls) – Fairly easy and splashy.

The best basis: Bell basis – Entanglement for everyone.

The best (prog) music composer: Steven Wilson – Dude just knows how to do amass the instruments.

The best type of graph: Markov random fields – I just like the approach.

The best paradox: Banach-Tarski paradox – Virtually, solves the worlds food shortage problem.

The best type of toothbrush head: One with medium-strong brushes – Works gently. 

The best type of book: Hardcover – Allows one to tap audible patterns.

Within the upcoming weeks, you can find about my progress via this blog. Also, check out Aitor’s blog, who is also working with the high-efficiency quantum fields project.

To keep things interactive, there’s brief juggling video below

Video 01: Providing (below) mediocre juggling, because why not?

Posted on by Aitor López Sánchez Posted in Blogs 2020, PRACE Summer of HPC 2020 — No Comments ↓

Hi everyone,

we (my partner Anssi Manninen and me) have started our development late our project High Performance Quantum Fields but finally, we’ve been able to start working. And I’m going to tell you everything that happened.

The first week after completing the online training week of PRACE Summer of HPC, it was time to meet with the mentor. Unfortunately, our mentor Stefan Krieg was sick, and he only recommended to us to read some interesting papers about simulations of classical or quantum field theories that we were going to work with. In these papers, three of them about simulations of Lattice Quantum Chromodynamics and another two about graphene and its famous honeycomb structure, we had to choose to work on either the lattice QCD or work with graphene. On Friday, when Stefan was better, I chose graphene. So we decided to meet Monday to speak about the main propose of the project.

Honeycomb structure of graphene

This Monday was our first meeting, and then we are going to meet every week. In this one, Stefan gave me all the necessary material to understand the tasks that we were going to do, he gave me access to the Jülich Supercomputing Centre (JSC) to use the supercomputer and put me in touch with his team to work together. He and his team are working to improve the performance of a Quantum Monte Carlo algorithm applied to Carbon Nanotubes.

Our work throughout SoHPC will be to improve this algorithm in CPU (with openMP) and then in GPU (with openACC). And for this, we will work step by step to achieve it. Thus, our first step will be learning about openMP, openACC and some common parallelization techniques, and also understanding the algorithm well along with the parts of the code that can be optimized.

If you want to know what results we are getting and what interesting things we are learning, don’t forget to visit my blog weekly. See you soon!

Posted on by Batuhan Hangün Posted in Blogs 2019 — No Comments ↓

This is Batuhan here. I am writing this blog post from Istanbul, where I live, work and study. I am a graduate student at Marmara University, Istanbul and I am pursuing a MSc degree at Electrical and Electronics Engineering Department. My main areas of interest are image processing and computer vision. After I realized that I could improve time performance of image processing/computer vision algorithms by using parallel computing and high performance computing I started to learn basics of parallel computing and high performance computing. I took a course about introduction to parallel computing and I started to like this new world.

My Application Process to SoHPC

While I was surfing on web and looking for some basic information to read about HPC, I ended up website of Turkish National Center for High Performance Computing (UHeM) and this was where I first heard about the PRACE and SoHPC. I realized that SoHPC would be the best technical event which I could learn the basics of HPC. One week of training at Vienna and then spending months at project site country seemed like an offer that I cannot refuse and I decided to apply. There were lots of amazing projects those one can learn many new useful information. I chose the top three projects which are more related to my studies and completed my application process. I was waiting the results impatiently. And then… COVID-19 happened….

Effect of COVID-19 to SoHPC

I was at Hangzhou, China in 2018 on an academical visiting process. City was good, food was good, people at Zhejiang University were helpful and hardworking and at that day I would never think about that there would be a pandemy appeared from this beautiful country can affect the whole world. Below you can see my photo which was taken at the central campus of Zhejiang University. That was very hot day so I was sweating like river.

From the central campus of Zhejiang University

After the worrying results of COVID-19 were started to emerge, countries of the world taken some protective measures immediately. Our beloved SoHPC was also affected from the current situation and I had an informative e-mail from the Dr.Kos which tells that SoHPC is going to be a remote summer school and asks that if I’d like to attend that kind of summer school, I informed Dr.Kos that I would gladly attend any kind of SoHPC.

A wheel of fortune that I spin

The day I waited for some time was at the door. An e-mail about the result of my application to SoHPC. That e-mail has had some bad news and good news at the same time. Bad news is, that I wasn’t chosen for the projects I wanted to take part in, and good news is, there was a project that I could be able to chosen as an attendant. Since the project that I could attend was very different from the topics I study, I spinned the wheel of fortune and I tried to be selected to a more relevant project at the next turn of selection. And I have become a victim of my own greed. I wasn’t chosen at next turn but the story hasn’t ended there.

Some Bittersweet News

In my darkest hour I got an email which brings smile to my face. It was saying that the one week training is going to be remotely available to all applicants. I completed all required actions and I started to wait for the training week.

Training Week

The training week was very useful for me to understand some basic concepts and tools those are used in HPC. All instructors were aware of that there were attendees who were not familiar to HPC so they were very explanatory at the lectures they gave. Topics were about foundations of super computing, how to use super computers, using super computers to make them super compute by using MPI, OpenMP and CUDA. A great bonus was that we were able to use a super computer at Austria which is named VSC-3 (Vienna Scientific Cluster 3). This supercomputer is amongst the top 500 supercomputers around the world and also known for its innovative oil based cooling system. You can check here and here to know more about the VSC-3.

Beside the technical knowledge, this training week helped me to make a new friend. I’ve met Pavel at training week. He’s a Phd student at Canary Islands. Both of us haven’t been selected for SoHPC so we were paired to be classmates during the training week. We talked about studies we were doing and tried to see how could we collaborate. Below you can see the photo of me and Pavel.

Me and Pavel from the training week

A New Hope

After some days, I got an e-mail which informs me about that there was an attendee who stepped out of a project and if I’d like to jump in to this project. Even though the topic (plasma physics) wasn’t relevant to my studies I took that e-mail as a sign and I accepted the offer without thinking.

“And here I am… In a project about plasma physics, to feel the amusement of the remote SoHPC. “

-Batuhan Hangün, The attendant of SoHPC

While I was writing my final words on this self introductory blog post that I had such fun at writing process, maybe this journey of mine be an inspiration for somebody to be hopeful in any situation. I suggest you to stay tuned for my upcoming blog post(s) about the project that I work in. I can’t wait to share my experience with you. Stay safe out there.

Posted on by Jake Love Posted in Blogs 2019 — No Comments ↓

Hello there, I am Jake and as you can see am one of the participants in the PRACE summer of HPC program. I am 22 years old and have just graduated from my Master’s in theoretical physics at the University of York. I have been interested in scientific computing for a long time and have just accepted a PhD position at the Johannes Gutenberg University of Mainz to pursue this interest further. My future work will be based upon using spintronics as a basis for unconventional computing devices.

In my spare time I enjoy programming and classical music. I am also a big fan of keyboards…

But which one is the right choice for SoHPC?

Project 24

This project is focused on generating a polygonal mesh that defines an iso-surface of a 3D scalar field. What is an iso-surface? Consider a two dimensional scalar field, such as a relief map. Relief maps often contain contours, or iso-lines, which define a continuous set of points of equal values. If this idea is extended to 3D space, the set of points with equal values become a continuous surface known as an iso-surface.

On method of creating such iso-surfaces is known as the ‘marching tetrahedra algorithm’. The end goal of the project is to develop a program that uses this algorithm to render ISO-surfaces of arbitrary 3D scalar fields using CUDA technology.

A 3D space divided into a tessellating structure of tetrahedra, the building blocks of the marching tetrahedra algorithm

Posted on by Theresa Vock Posted in Blogs 2020 — No Comments ↓
Me at my workstation

…there is a very simple solution to this this problem: Take one of the most powerful computers in the world and let it do all calculations for you. In the twinkling of an eye, it can perform more arithmetic operations than one human being could do in their whole lifetime – a good reason to familiarize yourself with this topic, isn’t it?

From mathematics to computational sciences

My name is Theresa Vock and during my studies of Technical Mathematics at Vienna University of Technology I always appreciated the beauty of elegant, efficient and sometimes unconventional solutions to different problems. Getting in touch with programming in order to solve mathematical tasks, I discovered my interest in high performance computing and parallel programming. As a consequence, I am excited to enlarge my field of expertise during this summer of HPC.

Matrix exponentiation, France and me

I was very happy when I discovered a project based on linear algebra and even happier when I found out that it was taking place in France because I absolutely fell for that country the first time I visited it. So I applied immediately and I was very excited when I got accepted. For the next weeks, I am going to work on different algorithms to calculate the exponential of a complex matrix remotely from Vienna. The final goal is an implementation that works efficiently for a GPU architecture using the MAGMA library. I cannot wait to work on the Jean-Zay machine, one of the fastest clusters in Europe. Of course I am going to keep you up to date on the progress we are making!

For now, I am trying to get at least a little bit of French flair to my flat during the summer by having some croissant and Pain au Chocolat in the morning and some baguette and Camembert for dinner.

Posted on by Cathal Maguire Posted in Blogs 2020, Participants 2020 — 37 Comments ↓

Hi, my name is Cathal Maguire. I am a Physics & Astrophysics graduate from Dublin, Ireland. I obtained a First Class Honours Degree in Physics & Astrophysics from Trinity College Dublin in 2020. I am also a big soccer and gaelic football fan, however I’ll realistically watch any competitive sport.

As part of SoHPC 2020, I will be working on Project #2003 – ‘Visualization of Supernova Explosions in a Magnetised Inhomogeneous Ambient Environment’. This project will be completed under the supervision of Project Mentor Salvatore Orlando, in collaboration with my fellow Trinity classmate and friend, Seán McEntee.

Project #2003

Project #2003 was due to take place in the CINECA Computing Centre in Bologna, Italy. However, due to the Global COVID-19 Crisis, the Summer of HPC 2020 is being carried out remotely. Although I would have enjoyed the summer sunshine of Northern Italy, the wind and rain of West Dublin will have to suffice.

I have previous experience running MHD (Magnetohydrodynamic) computer simulations, such as that outlined in the project description. As part of my final year, I undertook a Research Project investigating the circumstellar environment of the red supergiant star Alpha Orionis (aka Betelgeuse), which was ultimately accompanied by a thesis titled: ‘What can radio emission tell us about the stellar winds of Alpha Orionis?’.

Posted on by Ömer Faruk Karadaş Posted in Blogs 2020 — No Comments ↓
  • Fond of nature and sea!
  • Top view of the Institute I studied

Introducing myself

Hello everyone and greetings from Turkey! I am an electronics and communications engineering student at Izmir Institute of Technology (IZTECH), but I am in love with artificial intelligence and practices as well as the field I chose to study, and I love shaping my life with this passion. My interest in artificial intelligence, which has been coming from childhood, has become deeper as I learn new things in this field day by day. When I was learning machine learning, artificial neural networks, generative adversarial networks, I found myself in this field. With multi-disciplinary studies, I saw that artificial intelligence can be applied in every field. Freediving and mountain biking are among my professional hobbies. Also, I am interested in camping, being in touch with nature and the sea, meeting new people, and discovering new things. I have always dreamed of exploring the bottom of the sea.

Training week

As someone who has never been abroad before, the Summer of HPC (SoHPC) program was very scary for me at first. I would have to contact tens of people whose native language was different from me, but seeing the warmth of the coordinators in the first minutes of the training week helped me overcome this fear, and thanks to the Social events, all that political atmosphere left the place to a friendly atmosphere. Due to the shifting of timings around the world caused by the pandemic, the SoHPC training week coincided with the final exam period, but since they recorded each lesson, I was able to easily watch all the lessons I missed. The presence of an event like Summer of HPC has made these days much easier and more bearable for us today when the pandemic has become the reality of life.

Meeting with Teammates & Mentors

Meeting my group mates and our mentor was the most exciting part of the first week’s process. It is very important for me to know the people I work with, to communicate with them. I first met my teammates and quickly got used to each other. In the following hours, I met our mentors and got a short summary of my project. In the following days, I had more extensive meetings with our mentors and learned the details of the project. I believe that conducting a professional project sincerely in such a large organization will contribute to me a lot in my academic and business life. As far as I can imagine, we will be able to carry out an instructive project as I imagined with my mentor Anton and my teammate Francesca.

Posted on by Pablo Antonio Martínez Sánchez Posted in Blogs 2020 — No Comments ↓

Hard times that we have to live

I don’t think any of us are going to disagree, but we must show it to the world. We have to demonstrate, each of us (from home), that science and HPC are not going to stop due to coronavirus. This Summer of HPC edition will always be remembered because of coronavirus, but we must do an effort to make it also remembered because of all the great work and wonderful findings in HPC.

About myself

I’m Pablo and I should be enjoying Paris’s heat by now. However, I’m suffering (enjoying) the Murcia’s (Spain) one. Well…Murcia isn’t a bad city at all, is it?

Murcia’s cathedral timelapse I made some months before the pandemic started

As you can see, I love photography, but also computer science and HPC. I’m 22 and I graduated from computer science last year and I have recently finished my MSc in computer science, more precisely, in high performance computing (HPC), both at the University of Murcia. The project I will be working on is “Matrix exponentiation on GPU for the deMon2k code“.

What I have been working on

For the past two years, I have been working very intensively in the HPC world. My first work was focused on optimizing Ant Colony Optimization algorithm applied to the TSP problem (see image below). I started with a highly efficient, parallelized and vectorized implementation and, as part of my bachelor thesis, I improved the performance, obtaining a ~300x speedup on Xeon Phi KNL. Before me, running the code could take 12 days, and now it would take 1 hour. Not bad, right? Anyhow, since my bachelor thesis I have been working on this code a lot, so the comparison now would look extremely ridiculous. I also have written a paper about it which I expect to be published soon!

TSP Problem solved for 48 cities
Image taken from this webpage

This year I have been working in neural networks. I have been studying the performance of a Caffe (a machine learning framework) implementation which uses PHAST library (which is somehow similar to OpenCL). This Caffe version is coded once, but we can target both CPU and NVIDIA GPUs without touching the code at all (just changing the compilation flags!). However, the performance was poor, so my job was to understand why and fix it. The short story is that the problem was some Caffe layers which were working so bad with PHAST, because the implementation was very poor.

Image taken from the PHAST library webpage

I’m pretty excited about the project I’m going to work on. I always wanted to know how a supercomputer was managed and how is it to have access to one of them. I am excited to start working and to evaluate the performance of the GPUs in the Jean Zay machine. I hope to obtain great performance and exploit this huge computations as much as I can!

In case you are interested in my work, you can also check my Github! (https://github.com/Dr-Noob). Many interesting projects are waiting for you…

My github image profile (own content)

Posted on by Matthew Asker Posted in Blogs 2020, PRACE Summer of HPC 2020 — 40 Comments ↓

Hi there! My name’s Matt and I’m getting involved in the work-from-home Summer of High Performance Computing (SoHPC) internship programme this year. I live near Cambridge, England and I’m going into my final year, studying Physics at The University of Manchester. During the SoHPC, I’ll be working on project 21 – Submarine Computational Fluid Dynamics.

A photo of me in Barcelona, looking sufficiently smug.

Computational physics problems to me are like LEDs to a moth. This might lead you to assume that I can’t wait to get stuck into the internship… and you’d be right! At the end of my first Fluid Mechanics course I learnt how to calculate the lift on an aerofoil (plane wing) and it truly felt like magic. Since then, I’ve been chasing the same feeling in all of my studies. Excited is an understatement. Now I just can’t wait to start on the submarines!

Educational video on how submarines work. My only criticism is that they aren’t all yellow.

Outside of my academic interests, I’m an avid rock climber (every wall is just another problem to solve!) and I love to run around the countryside. I’ll give pretty much anything that involves going out in nature and being a bit outdoorsy a go! I’ve also started learning guitar recently so I can at least feel productive over lockdown. I like to think I’m doing okay, but anyone in earshot may disagree.

Canyoning in the beautiful Peneda-Gerês National Park, Portugal.

Having completed the online training week, I’ve been blown away by how well each session’s been organised – especially considering current events. Parallel programming’s brand new to me, so I’d be lying if I said I wasn’t a tad nervous about jumping straight into programming on the Vienna Scientific Cluster (VSC). We were introduced to a few different types of parallel programming, and while OpenMP is a standout favourite currently, it is still all to play for in the upcoming weeks! It’s also provided a fantastic opportunity to meet the like-minded individuals that I’ll be working alongside, and subdue some of the imposter syndrome that I think was felt unanimously before we began.

Everyone’s lovely faces from the SoHPC training week.

Now… if you’re wondering why this year’s SoHPC is a work-from-home project, there truly is no hope for you. Nevertheless, I am delighted that the guys at PRACE were able to keep the programme running – even if it means missing out on a nice trip abroad. To make up for it, I’m thinking of bringing a bit of Luxembourg to me. If you have any suggestions let me know!

Thanks for reading! Any questions, please let me know in the comments below.

Posted on by Josip Bobinac Posted in Blogs 2020 — No Comments ↓

Hi everybody! My name is Josip Bobinac and I am 24 years old. Just before the start of the Summer of HPC, I finished my second semester of Master’s program in Computational Science and Engineering at Vienna University of Technology.

When I am outside of the scientific computing domain, I enjoy sports like volleyball, running, swimming and have recently taken up yoga and playing Ukulele. As a big fan of healthy eating, you can find me feeding my pet Liza every morning – she is a Kefir-producing mushroom. If you are interested in the wonderful milk bacteria, I can give you an introduction to the art of kefir. Additionally, I enjoy meeting people of different cultures and nationalities, and of course – travelling.

Motivation

My frequent encounters with texts and videos from highly motivated and inspired people often leave me craving for clarity about what I want to do professionally. As the Master’s program sparked my interest for efficient simulations, I decided to make an additional step in that direction and immerse myself in the world of High Performance Computing. I believed and still believe I will gain some of that clarity through additional commited work. Hence, when one of my colleagues told me about Summer of HPC earlier this year, I applied immediately.

It is now early July and the months following that decision were quite unusual. Although not exactly as imagined, I have a great pleasure to be working on the project “Got your ducks in a row? GPU performance will show” in Jülich Supercomputing Center in Germany. However, I will be working on it from my humble “home-office” in Vienna that you can see in the picture below.

My workspace in Vienna – a lot of green tea went through that white cup

This is also the place where I participated in the training week organized by Vienna Scientific Cluster. Although the circumstances were not ideal, meeting my SoHPC peers through video conference and going over the introductions to blogging, MPI, OpenMP and CUDA programming made the whole experience worthy already. Below you can see a picture from a stroll around the city that helped me process the new information right after the training week.

A view of Maria-Theresien-Platz during my Vienna Ringstraße stroll in the center of Vienna

Stay tuned for my further posts about the project development!

Posted on by Jesús Molina Rodríguez de Vera Posted in Blogs 2020, Participants 2020, PRACE Summer of HPC 2020 — No Comments ↓

My name is Jesús Molina Rodríguez de Vera. I am from Murcia, which is located in the southeast of Spain.

Region of Murcia. Image is taken from Wikimedia Commons.

I have just finished my double degree in Mathematics and Computer Science at the University of Murcia.

When the time came to choose what degree to study, I was pretty clear about it. On the one hand, I have always been attracted to problem solving using the latest technologies. On the other hand, I believe that a deep knowledge of mathematics is fundamental for acquiring not only a range of powerful tools for problem solving, but also for facing these problems from new perspectives.

Although I have studied the specialization that is most related to Artificial Intelligence, I have always been interested in HPC and computer architecture too. Actually, I’ve been an intern in my faculty’s computer architecture department for some time.

I applied for the “Summer of HPC” to learn first-hand what it’s like to work in an HPC centre, and to be in touch with sector specialists from whom I can learn. In particular, I decided to participate in the project “2017 – Benchmarking and performance analysis of HPC applications on modern architectures using automating frameworks” because, from my point of view, it is needed to know how a system behaves in order to use it efficiently and responsively. In this project, I will learn how to perform benchmarking and regression tests that will give a picture of how the system works.

While it is a shame that we cannot go to the HPC centres this year because of the pandemic, I am very grateful for the opportunity of working at a leading supercomputing centre in Europe during this summer, even if it is remote. I think it will be an unforgettable experience in which we will learn not only the technical basics of working with supercomputers, but also be better prepared to work in a global world where remote cooperation is becoming more and more necessary.

Indeed, the HPC experience started last week with the “Summer of HPC” training week. Despite the fact that it was held remotely, the organizers have done an exceptional job in making everything work smoothly. It has been a great opportunity to learn some basic concepts of MPI, OpenMP and CUDA; and we even had a virtual tour to the VSC supercomputer infrastructure.

I’ll be updating the status of the project on this blog, so keep an eye on it if you’re interested in learning more about this exciting world.

Posted on by Elman Hamdi Posted in Blogs 2020 — No Comments ↓
My new friend HPC

Hi, Hoi , Hola, Merhaba … , I don’t know how many different languages I should say ‘Hello’, because now I have a lot of friends around Europe, and it’s time to meet with me. It is Elman from Turkey. I am an undergraduate student in Izmir, studying Computer Engineering and I have finished my 3rd grade in university just 8 days ago and on the same day, this travel started.

Now I am an intern at SURFsara on ‘Benchmarking and performance analysis of HPC applications on modern architectures using automating frameworks’ in Netherland with my new friend Jesus. Indeed, PRACE, SURFsara and my colleagues are very friendly and thanks to them I am learning a lot of new things.


I know this information never excited you, so I want to talk about my hobbies. I love to code something absolutely, but I love making videos and animations, drawing pictures but actually drawing new project it is can be a design of an application or a logo, interesting on electricity, for example, I have an electricity longboard which is I bought all its parts and merge them and now I have an ‘avelboard’ this is a name I put on it. Hopefully, one day it will be ‘avelcar’. If you are wondering about me please visit avelman.com.


Now, thanks to PRACE and its partners I am a participant in Summer Of HPC, and this summer I am going to learn a lot of thing about HPC. I can say now I spend two weeks for it and It will be the most precious summer of my life although working remove.

You can find me on my LinkedIn. Please feel free to send my a message.

See you in my other articles!!!

Facebook
Twitter
Follow by Email