As a modern living being it is quite often to be distracted by too many things. The one downside of this is that I procrastinate a lot. The other is amnesia, as I tend to forget things I did and ideas I had.
For this reason, I have always wanted to start a blog, documenting everything that have crossed my mind. I also aim to do this frequently, which could be useful to organize what I have done already and to plan for the next step, and allow me to focus on work, avoiding the distractions. Plus, I could perhaps gain a little audience of my blog in the process.
Jekyll
To start this blog I spent some time learning about Jekyll, and how
to publish a Jekyll blog on GitHub. Initially I started with the
default Cayman theme from GitHub but soon found out I couldn’t write blog
posts with it, so discovered a Cayman blog theme from lorepirri,
which enables blog posts on GitHub pages. I tweaked a little bit of this
theme to suit my needs better. Moreover, I enabled MathJax by
inserting a <script>
into the default.html
file (as shown below).
Math constructs in LaTeX, for example, $$E = mc^2$$
, can be rendered
directly into a nicely formated version, . This will come in handy
as we introduce formulae and equations in subsequent posts.
In the process of setting up this blog, I learned how to add a comment section to posts using Disqus, and also display a comment count on the home page for individual posts. This process is surprisingly painless, simply requiring me to register on Disqus, and adding an embedded comment section is as simple as pasting this snippet (provided by Disqus) to the template HTML file:
then I added comment counts to individual posts with:
and that’s it! So you can expect fast responses from me by commenting :laughing:.
Things to write about
I plan to write a few posts about things I did in my PhD, which focuses on numerical programs, and how we automatically optimize them to run as accurate, fast and resource-efficient as possible on FPGAs. I expect this to be quite a dense introduction to the underlying techniques, so this will be a 3-part series at least, in which I will respectively go through the rationale for doing so, how programs can be analyzed for accuracy, latency and area when synthesized into circuits, and finally how we can use the information to guide the optimization process.
I also envision new stuffs to be posted here as work progresses, which I do not have specifics to share for the moment, except the keywords FPGA and deep learning. I have a lot to learn and a lot of work to be done, I certainly have high expectations for future updates!
Although this blog will be mostly work related, please anticipate for other non-technical things (or technical things that are highly uncorrelated with work), as I master the art of procrastination. I guess that’s it for now, and I can’t wait to update on the things promised to I write about in this post in the following weeks to come.