I just spent the last week at IPDPS in Boston. It was a good time. I got to meet a few new people, and connect with a lot of friends who are now living in the Boston area. I also presented our work on Rust for the GPU at HIPS. In the course of watching a lot of presentations, I came up with a few tips. I admit I did not follow all of these in my own presentation, but hopefully all of us can learn from these.
I know where I am.🔗
I was amazed at how many people felt the need to put "IPDPS 2013" on every single slide in their presentation, as if I might forget what conference I'm attending. I suspect this is the result of some Beamer template being helpful, but really it's unnecessary. Some helpful presenters even thought it'd be useful to remind me on every slide that I'm in Boston. Even if I was somehow unaware of my location, the fact the we are in Boston likely does nothing to help me understand the information you are trying to convey to me. In the same way, if I wanted to know the date, I'd look at the cell phone in my pocket, rather than at your slides.
That said, often slides are posted on web sites afterwards for archival purposes, and in this case the conference name and date actually are somewhat useful. Putting this information on the title slide is sufficient. It doesn't need to go on every slide.
I have probably forgotten what your talk is about.🔗
One piece of information that I realized would actually be useful to have on every slide is the title of the talk. I was surprised at how many times we were on about slide 3, chest deep in technical details and I realized I had completely forgot what the point of it all was. The fact is, most of us in the audience are not giving you our complete attention. Make it easy for us to try to tune back in by keeping some kind of a reminder of the big picture on every slide.
If I wanted to read, I'd read your paper.🔗
Put as little text as possible on your slides. If you can, come up with a clear graphic or tasteful animation that makes your main point clear. The sooner I can understand your main point, the easier it will be for me to put the rest of your slides in context and avoid the existential crisis mentioned in the previous point.
I don't want to think about your graphs either.🔗
Make your graphs easy to understand. Often they will have some of the smallest text in your presentation, which means I won't be able to read it. That's fine, since graphs are about presenting information visually. Generally, the only thing I'm looking for in your graphs is how much better your system is than the ones you compare against. Make this easy for me. Clearly indicate whether up is better or down is better. Even better, be consistent and set up all your graphs so that up is always better or down is always better. Make it really clear which bars are lines are your system. Make them a brighter color, bold your system's name in the legend, or put an asterisk next to it. This is particularly important if you are comparing a bunch of systems whose names are all acronyms.
You don't need to include every graph from your paper. Focus on the key ideas. What are the major results that will make me want to read the rest of your paper? I can get all the nitty gritty details once I decide to read your paper.
For each graph you do include, spend some time telling me what I am supposed to see in it. Remember the saying, a picture is worth a thousand words. You can't adequately explain your picture with a dozen words. You need to say more than "here is our memory usage." Why is this result significant? Does it show your system is far faster than anything else? Does it show you have excellent scaling characteristics? Does the chart show you are only marginally worse than other systems, but yours provides other important benefits to make it worth the sacrifice? You should be trying to convince me to adopt your ideas in my own work. How does your graph help convince me of that?