Agenda for our January meeting, this Monday, January 22nd. See you there!

[Ray Simar]

Hi all,

If I haven’t already told you, Happy New Year!

First, here’s our Zoom link for our meeting (8 AM US Pacific time): https://riceuniversity.zoom.us/j/94249046147?pwd=YzBHTzdHWWVFY09YNHdvTjREME84Zz09 

Next, a big thank you to everyone’s efforts on our most recent paper.  We all owe Dave a big thanks for shepherding this paper along.

Additionally, I am happy to welcome a new member to our team, Lizy K John who is a prof at UT.  She won’t make our January meeting but is taking a look at how she might juggle her busy schedule to join us in the future.

It’s time to begin to plan our next big steps on development of our next benchmarks, namely Embench IOT 2.0 and Embench DSP 1.0.

Regarding Embench IOT 2.0 we are building on the success of 1.0 and have a good running start.  Here are a few actions we already have identified:

• Asked Jeremy to come up with a list of the four or so benchmarks we might delete.  
• Nbody and Cubic are on the list.
• Per Jennifer, the ud benchmark has a * 2.0 in it that should be changed to a * 2 to avoid an unnecessary floating-point operation.
• Jeremy has a good bitmanip operation he’d like to add to the suite.
• We could use help cleaning up the flows to support the change in benchmarks.  Jeremy created these, but he could use some help!
• Jennifer also identified (around December 18th) a next level of issues we need to take a look at to improve 2.0.

I have the ball on Embench DSP 1.0.  We have the example of the FIR benchmark and want to extend that suite to include some other DSP benchmarks like IIR, FFT, and adaptive filtering.  To get to 2.0, we don’t need a lot, but we do need more than one!

Another action I have is to develop an embedded ML benchmark.  I have a new member of my student team and we are looking into having them take a look at this benchmark this semester.

For Monday, let’s talk about these above areas and propose some next steps.  This will be an exciting chapter for the next phase of Embench!

All the best,

Ray

[David Harris]

Adding to the list:

wikisort also has avoidable floating-point involving * 1.0 

I think nbody might be salvaged if the x coordinates were updated based on the velocities after each iteration.  Then bodies_energy would not be invariant.

David

--
You received this message because you are subscribed to the Google Groups "Embench" group.
To unsubscribe from this group and stop receiving emails from it, send an email to embench+u...@lists.librecores.org.
To view this discussion on the web visit https://groups.google.com/a/lists.librecores.org/d/msgid/embench/0478E242-3961-4DA2-B4F8-DC7B75A28CC4%40rice.edu.

[Ray Simar]

Excellent!

All the best,

Ray

To view this discussion on the web visit https://groups.google.com/a/lists.librecores.org/d/msgid/embench/99812DF3-CB9B-450E-9B1F-5CA2284898EC%40g.hmc.edu.

[Paolo Savini]

Hi Ray, all,

I have a "service notice" about the Embench chat on Matrix. I was thinking to ask you to add an entry in the agenda for this but it's probably better to just say it here as not everyone is always in the call.

In short: there's an issue with the Matrix chat due to some changes in the Fossi foundation subscription rules.
We're advising not to use the chat for now. We'll be able to give you updates in a week or two.

Thanks for the patience.

Talk to you later,

Paolo

[Ray Simar]

Hi Paolo,

Thanks for the heads up on the Matrix issue.  We look forward to hearing an update when it’s working.

All the best,

Ray

[Ray Simar]

Hi all,

I wanted to let you all know that we will move our scheduled meeting from this Monday, 19 February, to the following Monday, 26 February.

We are moving this due to the US Presidents’ Day holiday.  I had overlooked this and Dave was kind enough to let me know.

Have a great weekend!

All the best,

Ray