Kevin Luong joined Berkeley SETI Research Center as an intern in summer 2015. Kevin worked with David Anderson to revise the NTPCKR system to run on cloud computing servers. There are still a few kinks to sort out, but if it works, this should let us run our data base through NTPCKR in a few weeks instead of more than a year. Kevin transferred to UCLA in the fall, but is continuing to work with SETI to explore cloud computing for other projects.

We'll be launching an expanded undergraduate internship program soon. Details will be announced later this year.

Watch the video at https://youtu.be/gR91rObbfqs

Awesome. I was wondering why ntpckr wasn't ported to cloud computing but this answers that. I like the idea that we will be analyzing data once again.

Good job. Keep those undergrads coming, Young bright minds are a real plus IMO.

Bob

---

Sometimes I wonder, what happened to all the people I gave directions to?
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
SETI@home classic workunits 4,321
SETI@home classic CPU time 22,169 hours

Great video. Thanks for helping out Kevin! And thanks for posting this Steve. Keep 'em coming!

Thanks for having joined SETI@home, Kevin.

Even if I did not catch a lot from the video : it's not your fault, it's mine because of my poor English (when reading, it's almost OK but when listening, it becomes an other game !).

We need young people as you to continue the job started décades ago and finally get success.

Cheers.
Tiers

Watch the video at https://youtu.be/gR91rObbfqs

Always interesting to see what is in someone's youtube search history. :)

---

Taking the Database as V7 and needing 1-2 years without cloud servers for NTPCKR I understand. (Clear cut starting point)

What I don't quite grasp from the video (and it may depends on the size of the Tiny Sky that the estimates are based off) is why then the deltas then need to hopefully be done in the cloud every 1-2 weeks.

keeping the out-sorced costs down (if your paying for the server time I understand)

but I thought Telescope time had to be scheduled months in advance so last quarters NTPCKR re-observation table wasn't so important until the scheduled time was approaching?

the deltas then need to hopefully be done in the cloud every 1-2 weeks.

keeping the out-sorced costs down (if your paying for the server time I understand)

I took it to mean: 'Instead of (potentially) 10 years of Nitpicker, let's try for "near-real time" as a goal', with 1-2 weeks being a 'close enough' goal.
And the Amazon EC2 then became an option after trying it out with Tiny Sky. (As a test/proof-of-concept?)

I thought Telescope time had to be scheduled months in advance so last quarters NTPCKR re-observation table wasn't so important until the scheduled time was approaching?

Isn't part of the scheduling having the telescope team/owner/whatever review the Request For Telescope Time? And the Nitpicker analysis would have to be a part of the request, along the lines of "Where do we want the telescope aimed?"

---

the deltas then need to hopefully be done in the cloud every 1-2 weeks.

keeping the out-sorced costs down (if your paying for the server time I understand)

I took it to mean: 'Instead of (potentially) 10 years of Nitpicker, let's try for "near-real time" as a goal', with 1-2 weeks being a 'close enough' goal.
And the Amazon EC2 then became an option after trying it out with Tiny Sky. (As a test/proof-of-concept?)

I thought Telescope time had to be scheduled months in advance so last quarters NTPCKR re-observation table wasn't so important until the scheduled time was approaching?

Isn't part of the scheduling having the telescope team/owner/whatever review the Request For Telescope Time? And the Nitpicker analysis would have to be a part of the request, along the lines of "Where do we want the telescope aimed?"

As far as I know, SETI@home has never had enough funds to request any telescope positioning ever, instead just having to piggyback and take their data from whatever other observations were taking place and where they were pointed at.

I believe that has always been true.

---

"Time is simply the mechanism that keeps everything from happening all at once."

Maybe I was thinking more of the Hubble?
You do have a point, that SETI tends to piggyback on wherever Arecibo was aimed, for general purposes; but if 'we' wanted to look closer at a specific point, then Nitpicker most likely needs to have been run at least 3 weeks ahead of time (I think)- 2 weeks for the EC2 run, and one week to look over the result and planning for when to piggyback to make sure the telescope is pointed in the right area. And, if there's no announced plans for that area of the sky, putting in a request somewhere to see if they "wouldn't mind".

---

One thing to remember when looking at the data from Nitpicker is that the newest tapes being run today a couple date from October 2015, but most date back to 2011, so there is very little time criticallity in the SETI world. And of course "grabbing" un-booked time on a telescope still costs money - about $1500 per hour, and I would guess that you can't book a one hour slot due to set-up and configuration time.

---

Bob Smith
Member of Seti PIPPS (Pluto is a Planet Protest Society)
Somewhere in the (un)known Universe?

@Kevin Luong

Hey just want to say a Massive Nanu Nanu Welcome from all of us here at SETI@Home



Looking forward to some of your proposals you mentioned via the Video, anyway Welcome Pal.

---

How can we introduce an Alien Race to the people of Earth, Without the power-hungry Governments of this Planet fighting for control & Technology? all because people Fear what they do NOT Understand!

Thanks everyone

You raised points that I thought didn't quite mesh with the video.

Guess my "wouldn't it be nice for Science filter" experienced a temporary fault. It's all good now.

feature request:

When you get a ntpckr going, can you also get it to toss out a few stats for general consumption? Thinking along the lines of average number of observing seconds per pixel. Average number of observations per pixel. Also the maximum observed pixel. The minimum observed pixel with >0 observations. The number of pixels with 0 observations. As we get data from different sources perhaps a sensitivity of search number(s) as well to go along with these.

Just some numbers so we have some idea of how much actual searching has been done and how much still needs to be done until we have some level of confidence that there isn't another radio capable civilization broadcasting a homing beacon within X parsecs of earth.

---