We had an outage on the short side. 22se19ab is splitting at the moment.

We were just talking about MeerKat with this pops into my feed, so I thought I'd share.
https://www.popsci.com/are-we-alone-in-the-universe/

Makes it sound like they will send the MeerKat data directly to one of their machines to analyze. Collectively we are the best supercomputer, so I'm puzzled they haven't sent us this data somehow. Anyone have anymore insight? maybe I'll dig around and see what other articles/info I can find.

Collectively we are the best supercomputer, so I'm puzzled they haven't sent us this data somehow. Anyone have anymore insight?

That up-to-two-month latency may have something to do with it. :^)

Edit: Also all of BOINC combined is in the Top Ten of supercomputers, but all of SETI@Home is barely in the Top 1000 (minimum of a PF required these days.) SETI@Home has a total RAC of just over 200M which is only 1PF, and they indicate will need a hundred times that. Incidentally? Coincidentally? that is also what BL needs.

---

Methinks our limited programing power plays into it.

---

https://www.skatelescope.org/software-and-computing/

Processing the vast quantities of data produced by the SKA will require two (one in South Africa, one in Australia) very high-performance central supercomputers capable of in excess of 100 petaflops (one hundred thousand million million floating point operations per second of raw processing power), which is equivalent to the fastest supercomputer on Earth at the present time (Source: Top500; November 2018).

---

SETI@home classic workunits 103,576
SETI@home classic CPU time 655,753 hours

https://www.skatelescope.org/software-and-computing/

Processing the vast quantities of data produced by the SKA will require two (one in South Africa, one in Australia) very high-performance central supercomputers capable of in excess of 100 petaflops (one hundred thousand million million floating point operations per second of raw processing power), which is equivalent to the fastest supercomputer on Earth at the present time (Source: Top500; November 2018).

If that was the case it would be cool because it would be done in a matter of seconds, on the other hand though I think it would melt the Internet connection and storage capacity he he

---

Collectively we are the best supercomputer, so I'm puzzled they haven't sent us this data somehow. Anyone have anymore insight?

That up-to-two-month latency may have something to do with it. :^)

Edit: Also all of BOINC combined is in the Top Ten of supercomputers, but all of SETI@Home is barely in the Top 1000 (minimum of a PF required these days.) SETI@Home has a total RAC of just over 200M which is only 1PF, and they indicate will need a hundred times that. Incidentally? Coincidentally? that is also what BL needs.

Can you use RAC as a measure?

Seti RAC deprecates with increases in computing power, while Einstein with its fixed credits gives an RAC over 20X greater than Seti.

Can you use RAC as a measure?

Seti RAC deprecates with increases in computing power, while Einstein with its fixed credits gives an RAC over 20X greater than Seti.

Unfortunately we don't have any other metric... I'm still using the Cobblestone definition where 1GF over a day gives a RAC of 200, so one credit is 432 billion floating-point operations.

Also need to consider that over half our computation power disappears due to quorum.

Edit: When I moved my farm over to Einstein I was getting about 2-4x the RAC depending on the workload here and there, nowhere close to 20x.

---

Can you use RAC as a measure?

Not for Seti as it uses Credit New, which does not award Credit in keeping with the original definition of a Cobblestone.

---

Grant
Darwin NT

Can you use RAC as a measure?

Seti RAC deprecates with increases in computing power, while Einstein with its fixed credits gives an RAC over 20X greater than Seti.

Unfortunately we don't have any other metric... I'm still using the Cobblestone definition where 1GF over a day gives a RAC of 200, so one credit is 432 billion floating-point operations.

Also need to consider that over half our computation power disappears due to quorum.

Edit: When I moved my farm over to Einstein I was getting about 2-4x the RAC depending on the workload here and there, nowhere close to 20x.

Looking at my valid tasks.
Einstein 3465 credits in ~780 seconds = 4.44cr/s
Seti 90 credits in ~315 seconds = 0.28cr/s

4.44 / 0.028 = 15.85.

So my mental arithmetic was a little off, lets call it 16X more credit at Einstein.

As Einstein decided to do fixed credits when DA decided Seti would use Credit New, then Einsteins credits are probably close to cobblestone standard.

edit - That is using optimized apps here, using normally issued app then my 20X is not out of the question.

Can you use RAC as a measure?

Seti RAC deprecates with increases in computing power, while Einstein with its fixed credits gives an RAC over 20X greater than Seti.

Unfortunately we don't have any other metric... I'm still using the Cobblestone definition where 1GF over a day gives a RAC of 200, so one credit is 432 billion floating-point operations.

Also need to consider that over half our computation power disappears due to quorum.

Edit: When I moved my farm over to Einstein I was getting about 2-4x the RAC depending on the workload here and there, nowhere close to 20x.

Looking at my valid tasks.
Einstein 3465 credits in ~780 seconds = 4.44cr/s
Seti 90 credits in ~315 seconds = 0.28cr/s

4.44 / 0.028 = 15.85.

So my mental arithmetic was a little off, lets call it 16X more credit at Einstein.

As Einstein decided to do fixed credits when DA decided Seti would use Credit New, then Einsteins credits are probably close to cobblestone standard.

edit - That is using optimized apps here, using normally issued app then my 20X is not out of the question.

That is a long time discussion

Einstein paid a fixed rate per each crunched WU, so faster is your crunching machine, more credit you gain, directly proportional.
From it's perspective each crunched WU received the same number of credits no matter who crunch the WU.

But in SETI is exactly the inverse, faster your host and more optimized is your app, less credit you receive per WU. That is because something weird called "Credit Screw". From Credit Screw perspective, the credit given to a determinate crunched WU depends on the host who crunch it. With one even worst point the credit is always smaller with the pass of the years. So if you could crunch a WU crunched 5 years ago probably you will receive 50-60% of the credit received years ago.

IIRC Maybe because that the only project who uses Credit Screw is SETI, all others runs from it. LOL

---

Einstein paid a fixed rate per each crunched WU, so faster is your crunching machine, more credit you gain, directly proportional.
From it's perspective each crunched WU received the same number of credits no matter who crunch the WU.
But in SETI is exactly the inverse, faster your host and more optimized is your app, less credit you receive per WU. That is because something weird called "Credit Screw". From Credit Screw perspective, the credit given to a determinate crunched WU depends on the host who crunch it. With one even worst point the credit is always smaller with the pass of the years. So if you could crunch a WU crunched 5 years ago probably you will receive 50-60% of the credit received years ago.
IIRC Maybe because that the only project who uses Credit Screw is SETI, all others runs from it. LOL

. . And I cannot blame them ... for running away from Credit Screw ...

Stephen

< shrug >

They gave us some more data to split. Nothing too exciting just some more blc64 for the day previously mentioned and the next day.

Hopefully they've fix the Sunday morning bug so we can uneventfully process this data and maybe, just maybe find a signal of interest.

They gave us some more data to split. Nothing too exciting just some more blc64 for the day previously mentioned and the next day.
Hopefully they've fix the Sunday morning bug so we can uneventfully process this data and maybe, just maybe find a signal of interest.

. . Sorry Unixchick but you need to take another look at the new data. It is 50 days newer than groundhog day II. It is not 58642/58643 data, it is 58692 data like the Blc11 tasks already mounted on the splitters.

Stephen

< waving at Unixchick with a sheepish grin >

They gave us some more data to split. Nothing too exciting just some more blc64 for the day previously mentioned and the next day.
Hopefully they've fix the Sunday morning bug so we can uneventfully process this data and maybe, just maybe find a signal of interest.

. . Sorry Unixchick but you need to take another look at the new data. It is 50 days newer than groundhog day II. It is not 58642/58643 data, it is 58692 data like the Blc11 tasks already mounted on the splitters.

Stephen

< waving at Unixchick with a sheepish grin >

Thanks for the correction. Time for me to have a nap or more coffee...or both.

< waving at Unixchick with a sheepish grin >

Thanks for the correction. Time for me to have a nap or more coffee...or both.

. . Just finish the coffee before taking your nap, otherwise it could get messy :)

Stephen ;)

They found some more Aricebo tapes. So tell me where do you think they found these?? 11my15ad 19au10ac 21dc08af

In that unlabeled box that someone has been using as a monitor stand to make a standing desk.

what do you think??

They found some more Aricebo tapes. So tell me where do you think they found these?? 11my15ad 19au10ac 21dc08af

In that unlabeled box that someone has been using as a monitor stand to make a standing desk.

what do you think??

They were likely files that had errors in them at the time (there were plenty of those) and were pulled until someone went through them to either correct or salvage good data from. ;-)

Cheers.

Must be the weekend- looks like a big bunch of shorties coming through to hammer the servers.

---

Grant
Darwin NT

We are at 145k results returned per hour. This is a bit high, but the system seems to be handling it just fine, so no panic.

Are the WUs just short, but valid or garbage that is a problem??