I am running Rosetta@home tasks. I have read on "Nature" that the Rosetta software has been ported on quantum computers, described in a long "Nature" article.
Tullio

---

I think the goodwill was always there. I personally am quite proud to have donated $17,000 to buy the Centurion Server, in honour of my dad that died aged 101.

. . I am sure it still is for a great many of us, both oldies and newbies like me.

The First Nitpicker idea was to take a chunk of the database, second sift them and feed the best candidates out for further processing, and then another chunk. But as Richard confirmed, the later idea was "near real time", which just wasn't practical. So life contunued as usual. It was DA's report that made people think again about the whole concept.


I would love to see S&H return in some format, I think we all would, even if it might be a bit of a shadow of it's former self. And as Wiggo said, it might attract some more funding. The dedication and the bonhomie is still there amongst us oldies.

. . Amen brother ...

And yes, I would come back with S@H Mk 2, but not with the current Mod team. They are the real reason that I deliberately deleted my account with over 20 years worth of memories and a billion Boinc credits.

. . I know nothing of "the current mod team" so no comment.

Keep on crunching as Wiggo says ...

Stephen

In December 2019 / January 2020, users raised nearly $60,000, primarily for a new storage array. Those figures can still be seen on the donations page.

Two years earlier (December 2017), there was another special fundraiser to extend the processing range to include data from the Parkes telescope in Australia. That's fallen off the official record, but my note from the time says that nearly $25,000 was received from users in a matter of days. In addition, the fundraiser was backed by matched funding, so the final total will have been well over $100,000. In the end, the Parkes processing channel was never implemented, but the goodwill was there.

But having said that, I think the main reason Nitpicker was abandoned was because the whole darn concept was too slow. It could never have kept up with the ideal of 'near real time' (the original Nitpicker concept), even with oodles more money: hence the switch to a batched snapshot schema with 'Nebula'.

Hi Richard,

It's been a while since I last checked out the site. But yes, every funds drive proved again and again there was strong genuine support from so many eager volunteers. I am still sad that after the excitement over the installation of the collection equipment at Parkes that we never saw any data in the processing queue. :(

It's nice that some people still use the forums.

Keep well

Stephen

World Community Grid has asked to increase the number of running together AfricaRainfall tasks. I brought it to 2, since it was 1. They take about 15 hours.
Tullio

---

Apart from World Community Grid which has limited GPU work. Curious to know where is everybody putting their Nvidia GPU power these days? Would also be interested to hear how the "M Queens" subproject on GPU is going over at yoyo at home? Unfortunately I do not have a GPU that supports 2.0 I have an RTX 2070

---

Curious to know where is everybody putting their Nvidia GPU power these days?

E@h.
GW search on 1050Ti and pulsars on older one.

---

SETI apps news
We're not gonna fight them. We're gonna transcend them.

Apart from World Community Grid which has limited GPU work. Curious to know where is everybody putting their Nvidia GPU power these days? Would also be interested to hear how the "M Queens" subproject on GPU is going over at yoyo at home? Unfortunately I do not have a GPU that supports 2.0 I have an RTX 2070

Current Nvidia drivers support OpenCL 3.0 as do all modern Nvidia generations such as your RTX 2070.
I wasn't even aware yoyo had any gpu applications as their applications page shows no gpu applications.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

Apart from World Community Grid which has limited GPU work. Curious to know where is everybody putting their Nvidia GPU power these days? Would also be interested to hear how the "M Queens" subproject on GPU is going over at yoyo at home? Unfortunately I do not have a GPU that supports 2.0 I have an RTX 2070

Current Nvidia drivers support OpenCL 3.0 as do all modern Nvidia generations such as your RTX 2070.
I wasn't even aware yoyo had any gpu applications as their applications page shows no gpu applications.

I have copied the news item from yoyo at home front page. I forgot to say that this is only for AMD & manual installation. I completely agree with you Keith about us being up to version 3.0 drivers

16 December 2020 GPU version for M Queens
We have now a first GPU version for the M Queens project for Windows. It requires an OpenCL 2.0 GPU. The version can be downloaded from github.com. The zip contains an installation instruction, the app and the app_info.xml.

Near the bottom of the readme file it says the following

This has been tested and confirmed working with:
AMD Radeon RX550
AMD Radeon RX570
AMD Radeon RX580

cards will not work since they don't support OpenCL 2.x

---

Well that post is wrong as Nvidia cards do in fact support OpenCL 2.0 and 3.0. The hardware always did.

It's just that the software drivers did not until the 465 drivers were released. That case did in fact support the statement when using drivers << 465.

Using the OpenCL 3.0 capabilities on my Pascal, Turing and Ampere cards for Einstein gpu applications.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

I sent you a PM Keith

---

I sent you a PM Keith

Got it and sent a reply.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

World Community Grid has credited me with a Sapphire Badge for running Mapping Cancer Markers. . Since QuChem was stopped for electrical problems in the Angers University Campus I am runnng Einstein@home GPU tasks both on opencl-nvidia and opencl-ati.
Tullio

---

Well that post is wrong as Nvidia cards do in fact support OpenCL 2.0 and 3.0. The hardware always did.

It's just that the software drivers did not until the 465 drivers were released. That case did in fact support the statement when using drivers << 465.

Using the OpenCL 3.0 capabilities on my Pascal, Turing and Ampere cards for Einstein gpu applications.

. . I'm almost tempted to fire up my machines and see if I can get OCL3.0 working on my poor old Pascal cards...

Stephen

:)

Well that post is wrong as Nvidia cards do in fact support OpenCL 2.0 and 3.0. The hardware always did.

It's just that the software drivers did not until the 465 drivers were released. That case did in fact support the statement when using drivers << 465.

Using the OpenCL 3.0 capabilities on my Pascal, Turing and Ampere cards for Einstein gpu applications.

. . I'm almost tempted to fire up my machines and see if I can get OCL3.0 working on my poor old Pascal cards...

Stephen

:)

If you want to have a go by all means feel free. My understanding is the code is hardwired to look for OCL 2.0 I tried on my machine with no success :( I don't expect there to be a OCL 3.0 version anytime soon as the code has not been touched in over a year.

---

Well that post is wrong as Nvidia cards do in fact support OpenCL 2.0 and 3.0. The hardware always did.

It's just that the software drivers did not until the 465 drivers were released. That case did in fact support the statement when using drivers << 465.

Using the OpenCL 3.0 capabilities on my Pascal, Turing and Ampere cards for Einstein gpu applications.

. . I'm almost tempted to fire up my machines and see if I can get OCL3.0 working on my poor old Pascal cards...

Stephen

:)

The new beta Linux GRP 1.28 application just needs OpenCL 2.0 to work. So any driver version >> 465 will work. The new application is using the performance tweaks developed by Petri33 who you might remember from the Seti special sauce application.

Speeds up Pascal cards by 40-60%. 100% on Ampere cards.

Give it a try.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

Well that post is wrong as Nvidia cards do in fact support OpenCL 2.0 and 3.0. The hardware always did.

It's just that the software drivers did not until the 465 drivers were released. That case did in fact support the statement when using drivers << 465.

Using the OpenCL 3.0 capabilities on my Pascal, Turing and Ampere cards for Einstein gpu applications.

. . I'm almost tempted to fire up my machines and see if I can get OCL3.0 working on my poor old Pascal cards...

Stephen

:)

If you want to have a go by all means feel free. My understanding is the code is hardwired to look for OCL 2.0 I tried on my machine with no success :( I don't expect there to be a OCL 3.0 version anytime soon as the code has not been touched in over a year.

. . The problem is not the hardware. At the top of this thread is the issue. It comes down to the drivers supporting what the hardware can do. So I would have to fire up the machine/s (if they still will) and updating the drivers to see if it works :)

Stephen

Well that post is wrong as Nvidia cards do in fact support OpenCL 2.0 and 3.0. The hardware always did.

It's just that the software drivers did not until the 465 drivers were released. That case did in fact support the statement when using drivers << 465.

Using the OpenCL 3.0 capabilities on my Pascal, Turing and Ampere cards for Einstein gpu applications.

. . I'm almost tempted to fire up my machines and see if I can get OCL3.0 working on my poor old Pascal cards...

Stephen

:)

The new beta Linux GRP 1.28 application just needs OpenCL 2.0 to work. So any driver version >> 465 will work. The new application is using the performance tweaks developed by Petri33 who you might remember from the Seti special sauce application.

Speeds up Pascal cards by 40-60%. 100% on Ampere cards.

Give it a try.

. . [bowing low] O' great and glorious Petri ...

.. OK now I am getting excited :)

. . But then comes the pain of getting the old things to work. They have been shut down for nearly one whole year ...

. . I think I will soon forget lockdown because I will be absorbed with getting this working ... )

Stephen

X8 - }

. . And the dance begins ...

. . To my surprise the big machine fired up AOK, even loaded the correct time and date. It went through lots of updates (as expected) and it is still running 460 drivers. But when I tried to update to the only additional driver I can see over 460 which is 470 it takes a while to accept it, but as soon as I click on "apply changes" it hesitates then returns to 460 drivers. :(

. . So then I went to the BOINC drive to checkout access and as soon as I double click on boincmgr.exe it throws it to archive manager and tells me it is an archive :( what the .... ??

Stephen

PS I do recall I had to jump through hoops by manually starting several things when it was last running but I cannot recall just what they were ... :(

. . I'm almost tempted to fire up my machines and see if I can get OCL3.0 working on my poor old Pascal cards...

Stephen

:)

The new beta Linux GRP 1.28 application just needs OpenCL 2.0 to work. So any driver version >> 465 will work. The new application is using the performance tweaks developed by Petri33 who you might remember from the Seti special sauce application.

Speeds up Pascal cards by 40-60%. 100% on Ampere cards.

Give it a try.

. . I can confirm this statement. My GTX1060's went from 16.4 mins per task to 10.5 min/task, a pretty good improvement, just by updating to driver package version 470. Way to go. Now I have to fire up the No.2 machine and see if the 970's can get a similar level of improvement ... :)

Stephen

. . I'm almost tempted to fire up my machines and see if I can get OCL3.0 working on my poor old Pascal cards...

Stephen

:)

The new beta Linux GRP 1.28 application just needs OpenCL 2.0 to work. So any driver version >> 465 will work. The new application is using the performance tweaks developed by Petri33 who you might remember from the Seti special sauce application.

Speeds up Pascal cards by 40-60%. 100% on Ampere cards.

Give it a try.

. . I can confirm this statement. My GTX1060's went from 16.4 mins per task to 10.5 min/task, a pretty good improvement, just by updating to driver package version 470. Way to go. Now I have to fire up the No.2 machine and see if the 970's can get a similar level of improvement ... :)

Stephen

Iffy improvement on the 970. I didn't see much of any improvement in compute times with my 970 when I was testing the code. Maybe a few seconds. Not worth it in my opinion. Save the power.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)