Einstein@Home seems to have a LOAD of arecibo WU's since they have a new PhD there now crunching the data

He thought of a new way to analyse radio data. Arecibo was already there, but they've added Meerkat, and soon FAST. Coupled with gravity finishing, and the GPU part of gamma finishing, more radio is being done.

It's always difficult to hold people to timelines on a project that's always been at the cutting edge, so while I'd hope we'll see something in November, it might be more realistic to expect something this side of the new year. It gets here when it gets here

I hope they hurry up, they're doing very important stuff. My Aunt and her neighbour have Coeliac disease and chronic pain (one each), both of which David Baker has mentioned they work on.

"This parrot is dead"
Errm . .
Oh .
sorry wrong forum .

I have 19 (was 20) parrots and that sketch is not very nice. The rest of Monty Python is funny though.

They're now back to tasks for at least three sub-projects, at the cost of much less total work.

The ARP1 tasks now seem to be only retries for tasks originally sent out about a week ago.

Download problems again.

They ought to put Covid just on GPU and let the CPUs do the other ones.

The GPU Covid work takes almost the same downloads as the CPU version, and is therefore subject to the same download problems. It runs faster once it's downloaded, but that's not the limiting problem.

They may have taken my suggestion that they provide mostly tasks with small sets of input files while the download server is especially busy, and leave the tasks with large sets of input files for times when the download server is not very busy.

The GPU Covid work takes almost the same downloads as the CPU version, and is therefore subject to the same download problems. It runs faster once it's downloaded, but that's not the limiting problem.

They may have taken my suggestion that they provide mostly tasks with small sets of input files while the download server is especially busy, and leave the tasks with large sets of input files for times when the download server is not very busy.

It may not improve server throughput, but why waste our CPUs on something a GPU could run? I refuse to do CPU work for them until they get their act together. I could be using far less of my own resources on covid work there.

And I doubt the money needed to upgrade the server is much compared to running a huge institute with several departments full of expensive professors.

A new batch of units downloaded yesterday, and all are completing with 'Error while computing'.

Is there an issue with these units?

A new batch of units downloaded yesterday, and all are completing with 'Error while computing'.

Is there an issue with these units?

I checked what happened when computers for other users tried those workunits. All of them that have completed them also gave 'Error while computing', so I'd say that batch were all defective.

The GPU Covid work takes almost the same downloads as the CPU version, and is therefore subject to the same download problems. It runs faster once it's downloaded, but that's not the limiting problem.

They may have taken my suggestion that they provide mostly tasks with small sets of input files while the download server is especially busy, and leave the tasks with large sets of input files for times when the download server is not very busy.

It may not improve server throughput, but why waste our CPUs on something a GPU could run? I refuse to do CPU work for them until they get their act together. I could be using far less of my own resources on covid work there.

In the last batch of work I got from them, nearly all of the Covid work was GPU type. So they may be trying your suggestions also.

Better simply set WCG to NNT and enable other projects until they get it to work properly, running scripts like that makes everything only worse and no additional science is going to be done that way, as instead of yours some other, maybe unattended computers are going to be idle. WCG has currently more than enough resources, no need to force more in there.

Some of us like the stats, as do you judging by your sig.

Well, I like stats, however I don't mind if I increase my credits for Einstein instead of WCG or even add a new project, there are few more doing valuable scientific work, I just usually try to stick to the ones I already participate in as long as they have enough work for everyone. But for example TN-Grid is on my list in case I need more CPU work.
Actually I thought I could add some credits to my Rosetta stats while WCG isn't able to supply enough work to everyone, but Rosetta doesn't have enough work for everyone either. So I do Einstein instead, they have enough FGRP5 work for everyone (and servers capable of sending it out to everyone, who requests it) and that's CPU-only work, so it seems like a good idea to let my CPU do some of those. I simply don't see the point of investing much time into "stealing" some of the rarely available WUs here from perhaps unattended PCs without backup projects, which will than sit idle or blocking others from downloading work from WCG as long as there are many other projects, that can use my help. Rosetta was btw. my first backup project, from where I used to grab a few WUs, when SETI servers were overloaded, long time before the backup project feature was introduced in BOINC, so I did that manually.

---

.

It may not improve server throughput, but why waste our CPUs on something a GPU could run? I refuse to do CPU work for them until they get their act together. I could be using far less of my own resources on covid work there.

Once they have enough work for the projects you want to run on your CPUs, you can limit OPN to OPNG by selecting just the projects you want on CPU and enabling "If there is no work available for the project(s) I have selected above, please send me work from another project.".

And I doubt the money needed to upgrade the server is much compared to running a huge institute with several departments full of expensive professors.

The issue in such huge institutes isn't usually the amount of money, it's just complicated and needs time to get it approved to spend them on something, in particullar if you can't use the money you've already got for something else and want additional money. If you just want to change the use of money you already got approved, it's usually easier. So it might depend on wether they can repurpose some "IT money" for this or if they need additional money.

---

.

In the last batch of work I got from them, nearly all of the Covid work was GPU type. So they may be trying your suggestions also.

I'll turn it back on, thanks.

So I do Einstein instead, they have enough FGRP5 work for everyone (and servers capable of sending it out to everyone, who requests it) and that's CPU-only work, so it seems like a good idea to let my CPU do some of those.

I'm doing the radio stuff from them (as it's a new thing a PhD student came up with), it runs on smartphones, CPUs, internal graphics, and discrete graphics, so everything I own can do radio waves at once.

Once they have enough work for the projects you want to run on your CPUs, you can limit OPN to OPNG by selecting just the projects you want on CPU and enabling "If there is no work available for the project(s) I have selected above, please send me work from another project.".

I like all their projects, I just want to avoid Covid CPU. So I've set as you said, which should give the CPU anything but Covid and make the GPU resort to Covid. We'll see if their server actually does that....

I've stuck Rosetta and WCG on a high weight and Einstein on low, so Einstein can fill in when I can't get enough Biology stuff.

The issue in such huge institutes isn't usually the amount of money, it's just complicated and needs time to get it approved to spend them on something, in particular if you can't use the money you've already got for something else and want additional money. If you just want to change the use of money you already got approved, it's usually easier. So it might depend on whether they can repurpose some "IT money" for this or if they need additional money.

I used to work in such a place, there was no problem just buying things. I chose the budget it came from and as long as the boss was ok with that money being spent on that thing, no problem, no delay.

Me: "Mr Smith needs 5 computers for his project, but his grant isn't in yet"
Boss: "Just take it out of budget X, he can pay it back."

So I do Einstein instead, they have enough FGRP5 work for everyone (and servers capable of sending it out to everyone, who requests it) and that's CPU-only work, so it seems like a good idea to let my CPU do some of those.

I'm doing the radio stuff from them (as it's a new thing a PhD student came up with), it runs on smartphones, CPUs, internal graphics, and discrete graphics, so everything I own can do radio waves at once.

Isn't "the new thing" BRP7? BRP4 is old and when reporting issues about it, it's always "low prority". But yes, my Android device is running that as the only project (with Moo as backup), nothing else runs on it, it has just Android 4.0 and almost no storage available.

---

.

I like all their projects, I just want to avoid Covid CPU. So I've set as you said, which should give the CPU anything but Covid and make the GPU resort to Covid. We'll see if their server actually does that....

It will as long as there's work from other projects available, but sometimes you might get OPN1 in case that's the only thing in the feeder. Unfortunately they don't have separate preferences for OPN1 and OPNG.

---

.

Isn't "the new thing" BRP7? BRP4 is old and when reporting issues about it, it's always "low priority".

I assumed all the radio stuff was for the same guy. BRP4 is Arecibo telescope data, and BRP7 is Meerkat telescope data. They created BRP4G to allow it to run more efficiently on CPU and Internal Graphics, so I assumed they were just as interested in it.

But yes, my Android device is running that as the only project (with Moo as backup), nothing else runs on it, it has just Android 4.0 and almost no storage available.

I had three Android devices, 4, 7, and 9.

The 4 I gave up on, the OS was so outdated it kept going wrong or refusing to run most projects. It ended up I couldn't get Boinc to work at all, so I threw it away. It was a free phone anyway, I bought it from Ebay China for £20, then got a full refund as it's a fake, they claimed it was Android 6! And it's full of adware in the firmware.

The 7 I've set to run WCG, and Einstein as a backup.

The 9 I recently accidentally discovered I could upgrade to 11 [1], so I did. But newer Androids (from 8 or 9 upwards) don't allow half the projects to run for security reasons (I'm gonna have a go at Google/Samsung/whoever writes Android for that), plus it's got a stupidly small amount of RAM - 3GB for 8 cores, so running more than 4 Einsteins causes some to stick (you'd think Boinc would manage the RAM but it doesn't). WCG crashes on it. So that's doing the only thing it can, 4 Einstein + 4 Universe (I could also do 8 Universe).

[1] I was visiting my Aunt and tried to connect my phone wirelessly to her TV to show photos and it failed, and a google search suggested upgrading the OS. I didn't realise that was possible and thought it would do it automatically if it could.

It will as long as there's work from other projects available, but sometimes you might get OPN1 in case that's the only thing in the feeder. Unfortunately they don't have separate preferences for OPN1 and OPNG.

Cancer has loads to do, that will probably always be available. What I meant was up until recently the server ignored any changes you made to your account, and someone recently (in here?) mentioned it still doesn't work for them. Most things seem to to work for me now though.

It's working. I just got a Covid GPU task:

3184	World Community Grid	04-11-2022 04:47 PM	Sending scheduler request: To fetch work.	
3185	World Community Grid	04-11-2022 04:47 PM	Requesting new tasks for AMD/ATI GPU	
3186	World Community Grid	04-11-2022 04:47 PM	Scheduler request completed: got 1 new tasks	
3187	World Community Grid	04-11-2022 04:47 PM	This project seems to have changed its URL.  When convenient, remove the project, then add http://www.worldcommunitygrid.org/	
3188	World Community Grid	04-11-2022 04:47 PM	No tasks are available for the applications you have selected	
3189	World Community Grid	04-11-2022 04:47 PM	No tasks are available for Africa Rainfall Project	
3190	World Community Grid	04-11-2022 04:47 PM	No tasks are available for Help Stop TB	
3191	World Community Grid	04-11-2022 04:47 PM	No tasks are available for Mapping Cancer Markers	
3192	World Community Grid	04-11-2022 04:47 PM	Your preferences allow tasks from applications other than those selected	
3193	World Community Grid	04-11-2022 04:47 PM	Sending tasks from other applications	

The 4 I gave up on, the OS was so outdated it kept going wrong or refusing to run most projects.

My works with Einstein BRP4, Moo! and WUProp. But I'm using NativeBOINC and not the official client on it.

---

.

My works with Einstein BRP4, Moo! and WUProp. But I'm using NativeBOINC and not the official client on it.

I had all sorts of projects on mine, but then it stopped communicating with my PC so I reformatted and reinstalled Boinc and everything went wrong. I'm using the Boinc official client, and before that was invented, something from Fred at Boinctasks. I think the problem is more the adware than the OS version. It was free and only two very slow cores, so it won't be missed. My working phones have 4 and 8 faster cores.