If you want to look at a newly upgraded to Win10, clean install on new M2 SSD, host running Lunatics apps on 670, then my HostID 8048155 is available.
Nothing go faster, like o/clocking or cmd line instructions. one task at a time.

It used to be host HostID 7006214

@Shaggie76 Just wanted to say thank you for this work.

---

I ended up using the latest version of the X 16 software,

What about MSI Afterburner?
https://gaming.msi.com/features/afterburner
http://www.guru3d.com/files-details/msi-afterburner-beta-download.html

---

 


- ALF - "Find out what you don't do well ..... then don't do it!" :)
 

I ended up using the latest version of the X 16 software,

What about MSI Afterburner?
https://gaming.msi.com/features/afterburner
http://www.guru3d.com/files-details/msi-afterburner-beta-download.html

Regardless of which software is used, here is some useful info on Pascal based GPUs how to OC & otherwise tweak them.

I ended up using the latest version of the X 16 software,

What about MSI Afterburner?
https://gaming.msi.com/features/afterburner
http://www.guru3d.com/files-details/msi-afterburner-beta-download.html

I think that EVGA had mentioned something about that when we were having our conversation, which I thought was pretty weird, suggesting to your hard earned customer to go to a competitors support site and download a program that is similar to yours, and designed for your products, but is basically broken, hence the suggestion. Crazy, if you ask me, and shows them off in a pretty poor light.

I might give it a try, but it will be a week or 2 before I get the opportunity. Anyone else blazed the trail before me? If so did you end up running both products concurrently, and how has it been working for you? Or not working if that is the case?

---

Thanks for the links, I will check them out.

---

For those of you who want to put the "Monster" in MONSTER RIG.

Nvidia announces 12GB Pascal Titan X

Available on August 2, for a paltry $1200.  At the end of the article, there is a link to a video.

Personally I've been disappointed in the performance of the Fury cards compared to my R9 390X. With less than 70% of the shaders of a Fury Nano or Fury X it still manages to churn through MB tasks in ~6 minutes. With the fans set to auto it does run up to 68ºC but they are still silent at ~40%. THe only config settings I use are -hp -cpu_lock & I think that -cpu_lock might be depreciated in the current version app. So it might not be doing anything.

What's your RAC with your R9 390X?[/quote]

Nvidia announces 12GB Pascal Titan X

I was excited at first but then I did the back of the envelope calculation:

3 x GTX 1080 = 26.6 TF / 540W for $2100
2 x Titan X = 22 TF / 500W for $2400

So 1080's would probably be cheaper to build, cheaper to run, and more compute power.

Nvidia announces 12GB Pascal Titan X

I was excited at first but then I did the back of the envelope calculation:

3 x GTX 1080 = 26.6 TF / 540W for $2100
2 x Titan X = 22 TF / 500W for $2400

So 1080's would probably be cheaper to build, cheaper to run, and more compute power.

That's generally the case.
When they finally come up with a GTX 1080Ti, and a GTX 1050Ti or 1060Ti and the prices (eventually) stabilise then those will be the ones to get.
The GTX 1080Ti for performance, the GTX 1050Ti/1060Ti for performance per watt per $.


As it is I like your graphs showing credit per watt- the GTX 750Ti may not have the greatest performance, but when it comes to running costs, it is king. And I expect it will still be so even with the improved OpenCL_SoG & CUDA applications. Even greater performance, and only a slight increase in power usage.
:-)

---

Grant
Darwin NT

This thread has made me curious about my own GPUs benchmarks. What are the main benchmarks people use? WU/hr? How do I find this?

This thread has made me curious about my own GPUs benchmarks. What are the main benchmarks people use? WU/hr? How do I find this?

Well I think the best bench is RAC but sadly that doesn't take into account the split between GPU/CPU like you're looking for.

My perl scripts listed below can help give you a sense of the breakdown but they're easily tricked. If you give me your host ID and tell me how many concurrent GPU tasks you're running I can scan your results and give you some numbers if you like.

I updated my graphs with a fresh scan last night; I also filtered out any hosts running an anonymous app for their work-units (although this seems generally to be the higher-performance route it is usually consistent with people running multiple GPU tasks concurrently which skews the stats).

This thread has made me curious about my own GPUs benchmarks. What are the main benchmarks people use? WU/hr? How do I find this?

Well I think the best bench is RAC but sadly that doesn't take into account the split between GPU/CPU like you're looking for.

My perl scripts listed below can help give you a sense of the breakdown but they're easily tricked. If you give me your host ID and tell me how many concurrent GPU tasks you're running I can scan your results and give you some numbers if you like.

http://setiathome.berkeley.edu/results.php?userid=9584707

This is my task list. Is that what you need? (Just started crunching a week or two ago)

This is my task list. Is that what you need? (Just started crunching a week or two ago)

Host, Device, Credit/Hour, Work Units
8045176, AMD Hawaii, 564.526776433762, 128

I don't see any pure-CPU tasks for you -- I'm not familiar with how AMD tasks show up but that's the average you're getting with OpenCL.

Interesting. Smack dab in the middle of your chart for Hawaii processors.

Updated with fresh scan this weekend. New: Ellesmere (Rx480) finally has enough data to show up! Also surprising GTX 1070's showing VERY strong -- presumably because stock users aren't likely to run multiple tasks on their GPU and without more work the extra 5 SMs on the 1080 are idle. I'm curious how the GTX 1060 will compare when it starts showing up.

I am using a single 1080.
The computer info from my account:

Computer ID: 8053987
Name:Micky-PC
Avg. credit: 2,679.88
Total credit: 31,373
BOINC version: 7.6.22
CPU: Genuine Intel Intel(R) Core(TM) i7-3770S CPU @ 3.10GHz [Family 6 Model 58 Stepping 9] (8 processors)
GPU: NVIDIA GeForce GTX 1080 (4095MB) driver: 368.81 OpenCL: 1.2
Operating System: Microsoft Windows 10 Professional x64 Edition, (10.00.10586.00)
Last contact: 31 Jul 2016, 1:49:48 UTC

Device, Credit/Hour, Work Units
Intel Core i7-3770S @ 3.10GHz, 284.326803544063, 77
NVIDIA GeForce GTX 1080, 659.177888540607, 72

It's a little low compared to the average for a 1080 -- this may be natural given how my scripts takes the average of the top half of the hosts for a card.

Device, Credit/Hour, Work Units
Intel Core i7-3770S @ 3.10GHz, 284.326803544063, 77
NVIDIA GeForce GTX 1080, 659.177888540607, 72

It's a little low compared to the average for a 1080 -- this may be natural given how my scripts takes the average of the top half of the hosts for a card.

Also they've only just started crunching, there are a couple of errored out WUs and several abandoned, and the stock application is still trying out all the different applications to determine which one is fastest. And depending on the work mix at the time work is being dished out, it is possible to end up with the slowest, not the fastest.

---

Grant
Darwin NT