Roll on the day when the Nvidia app that's been under test on Beta is rolled out on main.....

I want that Nvidia app for VLARs. I can crunch those puppies in 20 minutes, 30 if I do 2 at a time.

Roll on the day when the Nvidia app that's been under test on Beta is rolled out on main.....

I want that Nvidia app for VLARs. I can crunch those puppies in 20 minutes, 30 if I do 2 at a time.

Yes, the OpenCL app gets some nice time improvement on those VLARs. It'll be a little while before X-Branch Cuda development can compete on those (Switching development methods), so any help the VLARs can get is well worth it IMO.

---

"Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions.

After a manual restart of BOINC on my computer, it started immediately to download MB VLARs for my CPU.

Magic... :?

---

Aloha, Uli

There is a Mac CPU App just sitting at Beta that can complete these VLARs twice as fast as the current Mac CPU App. It's just sitting there...

The first number is the size of cache. The second is effectively how often your system will check to see if there is any work available.
So having "2+2" means you have a two day cache, and your system will only check every two days, and will then try and get two days worth of data. If you had say "2+0.1" your system would have a two day cache, but would check every 2.4hours to see if you need work, and, provided the servers have work would try to get enough work to fill your cache to 2 days worth of work. Obviously if you reduce the "0.2" to a smaller figure the check will be more frequent, and your cache will be more consistently topped up

Awesome ! I will make that change now. I thought I had the 2 preference pages all figured out but I am glad to be set straight on the additional cache setting.

Thanx,
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

There is a Mac CPU App just sitting at Beta that can complete these VLARs twice as fast as the current Mac CPU App. It's just sitting there...

I was actually looking at Macs trying to work out what I'd need for basic development (preferably with some kindof GPU, not necessarily Cuda), and got pretty confused. Moving to a continuous integration path would mean automatic deployment cross platform (eventually). What's the deal with hardly any models with decent GPUs of any brand ?

---

"Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions.

I remove BOINC from my daily user Mac Laptop (Beta Seti) as I wasn't sure why it keep crashing all the time. The last time was so severe I almost couldn't recover the data on it. I was looking for the sad mac but even he wasn't around........

Pure luck I was able to recover it. So now only my back up mac is crunching any Seti..

There is a Mac CPU App just sitting at Beta that can complete these VLARs twice as fast as the current Mac CPU App. It's just sitting there...

I was actually looking at Macs trying to work out what I'd need for basic development (preferably with some kindof GPU, not necessarily Cuda), and got pretty confused. Moving to a continuous integration path would mean automatic deployment cross platform (eventually). What's the deal with hardly any models with decent GPUs of any brand ?

Strange, the GPU list at SETI has more Mac GPUs than Windows does, http://setiathome.berkeley.edu/gpu_list.php, for AMD/ATI cards. Apple has Always favored ATI. I have a 1997 PowerMac with an ATI chip in it, Most iMacs came with ATI video. In fact I've been using Macs for 20 years and never saw an nVidia card until I bought a NV8800 back in 2008 for my last MacPro. So, if you are looking for CUDA cards in Macs you might have a hard time, there are a few. You would do better looking for the twin FirePros in all the new MacPros though. Personally I think a twin FirePro D700 machine would be nice.

Oh wait, now that I think about it, back when I was flashing PC ATI video cards to work in PMG4s I did buy a nVidia 5500. The flash didn't work so well though, the DVI didn't work and you had limited VGA resolutions, if I remember correctly. I had much better luck reducing the Mac ROM size down to 64k and flashing it onto a PC ATI card. I had a really nice ATI 850 in my PMG4...until the fan died...

There is a Mac CPU App just sitting at Beta that can complete these VLARs twice as fast as the current Mac CPU App. It's just sitting there...

I was actually looking at Macs trying to work out what I'd need for basic development (preferably with some kindof GPU, not necessarily Cuda), and got pretty confused. Moving to a continuous integration path would mean automatic deployment cross platform (eventually). What's the deal with hardly any models with decent GPUs of any brand ?

Strange, the GPU list at SETI has more Mac GPUs than Windows does, http://setiathome.berkeley.edu/gpu_list.php, for AMD/ATI cards. Apple has Always favored ATI. I have a 1997 PowerMac with an ATI chip in it, Most iMacs came with ATI video. In fact I've been using Macs for 20 years and never saw an nVidia card until I bought a NV8800 back in 2008 for my last MacPro. So, if you are looking for CUDA cards in Macs you might have a hard time, there are a few. You would do better looking for the twin FirePros in all the new MacPros though. Personally I think a twin FirePro D700 machine would be nice.

What about running a small GPU off a thunderbolt 2 chassis ? Would it, or would it even be meant to work ? For development the flexibility is a bit more critical than the device integration.

[Edit:] ah I see that's a whole thing... found this: http://forum.techinferno.com/diy-e-gpu-projects/6578-implementations-hub-tb-ec-mpcie.html#Thunderbolt

---

"Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions.

What about running a small GPU off a thunderbolt 2 chassis ? Would it, or would it even be meant to work ? For development the flexibility is a bit more critical than the device integration.

[Edit:] ah I see that's a whole thing... found this: http://forum.techinferno.com/diy-e-gpu-projects/6578-implementations-hub-tb-ec-mpcie.html#Thunderbolt

Yes, there are a few of those devices around. I was looking at a breakout box that would hold 4 double wide cards and connected via a PCIe2 slot. It looked just like a second smaller MacPro that would set next to the real deal. Everything about it looked great, except the price.


Here's another,


Nothing like Search, pcie expansion chassis mac

Both PCs are almost out of work!!! :-( What's happened???!!!??? :-O Did something break, again??? :-/


TL

---

TimeLord04
Have TARDIS, will travel...
Come along K-9!
Join Calm Chaos

Think it's a VLAR storm. No GPU work but my CPUs are filled with VLARs

After suggestions on adjusting my personal preferences to 2 days of work and .3 of cache I got my GPU filled with max 100 GPU WUs. I changed it also to no CPU work as I had a full cache of that. My GPU is is an ATI R9 and uses CL, so I don't know if that makes any difference. I was totally out of GPU WUs until i made those changes.

Crunch On
Bob
------------------

Some days your the windshield, some days the bug.

---

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

I doubt it. I have a R7 260X and 10/.01 cache. I am low on WUs. I have all the VLARs I could ever want. Just need some GPU.

I know I have done something odd. All my GPU WUs are ati5_nocal unless it is a AP.

---

I doubt it. I have a R7 260X and 10/.01 cache. I am low on WUs. I have all the VLARs I could ever want. Just need some GPU.

I know I have done something odd. All my GPU WUs are ati5_nocal unless it is a AP.

Take a look at my tasks Admiral I have close to 100 opencl_ati_cat132 for my R9 Not sure why yours are different. I was out of GPU task before that.

Bob
------------------

Some days you are the windshield, some days the bug.

---

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

Here's another,


Nothing like Search, pcie expansion chassis mac

I like that unit. Horizontal mounting with spacing between cards...bonus handle on top.

---

I doubt it. I have a R7 260X and 10/.01 cache. I am low on WUs. I have all the VLARs I could ever want. Just need some GPU.

I know I have done something odd. All my GPU WUs are ati5_nocal unless it is a AP.

Yup - NO GPU, CPU only, all VLAR all the time...

Sucks!!!!

---

MB splitters have fallen off the pace again.
Luckily with all the VLARs about it hasn't affected the ready-to-send buffer. But if the splitter output remains this low, it won't take long to run out of work once the VLAR storm is over.

---

Grant
Darwin NT

I Think that VLAR's are gone now, My Nvidia GPU's are full of MB now...

I Think that VLAR's are gone now, My Nvidia GPU's are full of MB now...

Nope.
Almost all bar a couple of my CPU WUs are VLARs, all of those in the last 6 hours are.

---

Grant
Darwin NT