Posted in

Three Mid-Year’s Resolutions: Learn Some HPC, AI, and Cloud

by James Reinders•

With 2018 halfway over, I call on everyone to forget your organization for a moment and make a conscience plan to invest in yourself. 1-2-3. Think of it as “three steps to a better me.”  It’s fun too!

Here is an assignment for us now (should you decide to accept it): do one thing in each of three categories: HPC, AI, Cloud.

HPC

Our assignment for HPC is this: Complete at least one of the following three choices.  Extra credit: splurge on yourself and do all three.

  1. Ask vendors for recommendations on how to get a day of technical training that sounds interesting to you and your team.  Of course, you can do a Google search and find many online courses and videos on your own.  Regardless of how you find the content “ give it your full attention and learn!
  2. Attend an HPC conference, and find time to soak it in. The International Supercomputing Conference in Germany (just completed this year), and Supercomputing in the USA (coming up in November!), are the two largest to consider. The key is to give yourself time to engage!
  3. Attend a User Group for a software package (CFD, CAD, etc.) you use or think you should learn about. This is the hidden gem in these recommendations! User groups are the least commercialized of these three options, and the most overlooked. I’ve found plenty of useful feedback and advice at User Group meetings.

Here is some help – a ready-to-click option that blends A and B is this: Intel’s HPC Developer Conference in 2018 had many wonderful talks, many of which are online to view (no travel required) and highly relevant long-term:

AI / Machine Learning (ML)

Our assignment for AI/ML is this: Complete at least one of the following three choices.  Extra credit: splurge on yourself and do all three.

Here are my big, small and an in-between option that for many may be just right.  Because you are investing in yourself, you can’t go wrong picking any of these as education for yourself!

  1. Dip your toe in gently with a Kaggle competition turned into a cut your teeth step-by-step exercise. Check out Titanic: Machine Learning from Disaster .  If you want to learn a little about ML, and you can’t finish this, then you really aren’t taking time to invest in yourself.
  2. Bite-sized pieces, three courses step-by-step, courtesy of Intel’s Nirvana team.
    1. Check out COURSE 1: MACHINE LEARNING 101 .  Do you at your own pace ¦ 12 lessons ( weeks ).  You will learn about Jupyter, Python, and much more for ML by the end.
    2. Check out COURSE 2: DEEP LEARNING 101 .  Do you at your own pace ¦ 12 lessons ( weeks ).   You will learn about neural nets, training methods, MNIST, CNN, RNN and much more.
    3. Check out Course 3: TensorFlow 101 .  Do you at your own pace ¦ Eight lessons ( weeks ).  You will learn how to use one of the most popular ML frameworks (created and used by Google).
  3. Jump in deep with Stanford Professor Andrew Ng.  Check out Machine Learning on Coursea .  This is a full college course style class, online for a do at your own pace.   I know from experience; this is a wonderful course which requires more than a little discipline to finish if you have a full-time job to deal with as well. I highly recommend it for those with the time and discipline.

Cloud

Our assignment for Cloud is this: get your own cloud account, and use it on your own. Using a remote machine at work does not count for this exercise.

An excellent choice here is Amazon AWS, but feel free to find other cloud services (many are eager to catch your attention).  You get credit for helping yourself if you sign up for a cloud account, write some code, and run it in the cloud. Finding tutorials to follow is not cheating.

Extra credit: Program an FPGA to do something in the cloud. I wrote a step-by-step guide to FPGA Programming with the OpenCL Platform with all programming done in the cloud. This is not for the faint at heart, but hopefully my step-by-step makes it an approachable investment if you want to learn about both using the cloud, and about FPGAs.

Summary: Invest in Yourself!

It is all too easy to be a bystander while advances in HPC, AI and the Cloud all seem too far away. For mid-2018, I offer you my 1-2-3 suggestions for a few steps to make them closer!  And most of my suggestions cost only your time, the notable exceptions being the encouragement to travel to a conference or user group and interact with others.

I like to take stock in what I am learning, every year.  July is a good time to make sure we are not putting off trying things out first hand, learning directly (hands on). It’s not that scary; I hope you take a day off and do a few of these challenges.  Post your results!

Good luck!

I like fast computers and the software tools to make them speedy. My experience in High Performance Computing (HPC) and Parallel Computing spans four decades, and includes 27 years at Intel Corporation (retired June 2016). I have authored of eight books in the HPC field (and I'm working on another with some amazing co-authors!), numerous papers and blogs. I am currently working on a couple more book projects, teaching/writing/blogging, and working for Pattern Computer - a startup hoping to use machine learning style approaches to  make the world a better place.

We are all on a journey that incudes lifelong learning. I continue to accept opportunities to work with others from whom I can learn. In this vein, I do accept compensation for my time interviewing and/or writing blogs from time-to-time. Anytime that I attach my name to an article or blog, I stand behind them as my opinion (not an opinion that was purchased). If you disagree with them, you are disagreeing with something I believe - and I am interested in feedback.

 

Privacy Overview

This website uses cookies so that we can provide you with the best user experience possible. Cookie information is stored in your browser and performs functions such as recognising you when you return to our website and helping our team to understand which sections of the website you find most interesting and useful.