Jin Hu


August 08, 2005 - August 14, 2005





Monday August 08, 2005

This morning I practiced my presentation in the conference room (Sudarshan was nice enough to open up the door me, thank you!). After a break for lunch, I met with Professor Eli and gave her a rundown of what she could expect. She gave me a lot of comments, especially toward the beginning of my presentation. We also talked about the experiments that I will run after the presentation. After the meeting I talked to Love and he mentioned that I will start working after I present on Wednesday.

Tuesday August 09, 2005

Today I practiced some more on my presentation. Hopefully, everything will go smoothly and that I won't have any major mistakes. I also uploaded the data for the dual multipliers - if placing them far apart vs. close together makes a difference. I found out that with the top and bottom ones, there was hardly any change. With the left and right ones, I found that the difference is not significant, but there was more variation than the top and bottom ones.

Wednesday August 10, 2005

This morning, I did a dry-run of my presentation in front of Richa right before we presented for real. Surprisingly, the presentations took an hour and a half instead of the scheduled hour or so - I'm not sure if it was a good thing or not. Overall, I hope that the presentation went well and that there wasn't too much that I screwed up. Professor Eli also mentioned that I should be writing the first half of my report and incorporating the comments that she made during my presentation. After a relaxing lunch with Professor Eli, Love, Sudarshan, and Richa, I went and talked to Love about the remaining parts of my project. We talked for about 90 minutes in detail on what exactly I should do until the end of the quarter.

Thursday August 11, 2005

The first day after the presentation! Based on what Love and I talked about yesterday, I started running some experiments to look at the performance, frames, and bits for dataflow graphs (dfgs). I started some pre-lim tests but currently do not have any strong conclusions. I have seen that when I do full mapping under exact conditions, the tests fail every time. However, with leverage, the trial completes. I have not compiled the data for the traits to see any trends. I also met with Professor Eli today. She mentioned that I should start writing the first half of my final report to hand in for next Thursday.

Friday August 12, 2005

I continued running my experiments for the dfgs. At this point, I figure I can let all the data run before I start looking at the trends. I started also doing some partial mapping and hopefully for that, we will see some variations. Love and I looked at some of the fully mapped dfgs and Love was disappointed that there wasn't more variation with the timing. Perhaps there will be more variation in the other experiments. On a side note, Love and I was kindly reminded today by André to keep our voices down or to go somewhere else and discuss; I didn't realize how long Love and I had been talking until André popped his head in.

Weekend August 13-14, 2005

I ran some scripts over the weekend. What I've noticed is that these experiments take up a lot of hard-drive space - I'm worried that I'm going to run out. I hope that new computer will be on the network soon.

© 2005 Jin Hu