HCC GPU Launched.

I'm running a GTX 580 SuperOverclock which runs at 855 MHz instead of the stock 772 MHz, and I'm averaging 3m14 per WU. This task definitely favours AMD.

Quote:

---

Originally Posted by [XC] Oj101

I'm running a GTX 580 SuperOverclock which runs at 855 MHz instead of the stock 772 MHz, and I'm averaging 3m14 per WU. This task definitely favours AMD.

---

Seems like it does. Or at lest the 7xxx series. Still, only 3 minutes for a WU is crazy good and a lot more work will get done. In the end this is what is most exciting to me, the huge amount of extra science this brings. About time too! I've always kept my eye on F@H specifically for the GPU support.

Quote:

---

Originally Posted by Rob_B

Use at your own risk. app_info only usable for HCC GPU only.

It works for my 5850, but now it only uses the GPU. No work for the CPU. And it went up from 3 minutes per WU to over 5 minutes.

---

Depending on how much over 5 minutes you are still getting a pretty good increase in overall throughput. Each GPU + CPU (including HT or not) + OS configuration needs to be assessed to find the overall benefit.

Quote:

---

Originally Posted by 0ne.shot

Something seems a bit off with some of those numbers. The 7870 has effectively double the SPs as the 7770 and 25% more than the 7850 and half of the 7970, yet the numbers don't appear to be reflect this, or there are variables we are unaware of.

---

Right now that 7870 card is temporarily in my Work PC (see sig below for specs). Not sure of that matters, but it will be going in my 2600k rig this weekend hopefully. Now, I am doing WU anywhere from 1:32 to 1:57, so much better than the 2:06 I was using yesterday. Also, it never goes above 84% GPU utilization.

So we a looking at a range of 172-217K WCG PPD at that rate depended on where you look. I say split the diff and call it 196K PPD WCG.

I'll have to see if putting it in my main 2600k rig improves things.

Anyone know if the 285 works on this?

Quote:

---

Originally Posted by Mydog

Anyone know if the 285 works on this?

---

You mean GTX 285? It should as my GTX 260 works fine.

Just RMA'd a 5870 and got a 7870 OC back. Might have to put it through its paces instead of selling now :D

In the meantime.. let's see where this gets me:
Attachment 130577

Quote:

---

Originally Posted by artemm

Just RMA'd a 5870 and got a 7870 OC back. Might have to put it through its paces instead of selling now :D

In the meantime.. let's see where this gets me:
Attachment 130577

---

Personally I think you should send it to your good buddy Dave who REALLY needs a card...:wasntme:

Something for the stats:

GTX 260 (600/1300/1100 core/shader/mem) takes roughly 7 minutes per WU, GPU load between 93 and 98%, 1 dedicated CPU thread full load (Intel Hex @ 3.6Ghz), BOINC 6.10.58

Quote:

---

Originally Posted by Movieman

Personally I think you should send it to your good buddy Dave who REALLY needs a card...:wasntme:

---

Says the guy who just recently got the next to largest slice of the pie!! :p:

Quote:

---

Originally Posted by Sparky

Says the guy who just recently got the next to largest slice of the pie!! :p:

---

yea, BUT assembling a 3770K and 2600K and only have one 5870 card ..I do NEED another card so hush your mouth :D

Hey jcool, nice to see you..
Ok, I'm off, just got a call for some work!

Im using a 5850 and a 5870 averaging 3 minutes a WU. after like 2 days, my output has double so far and increasing by the minute

I have a 4850x2 and a 5830 lying around i could add to the mix if my PSU's could handle that kind of load

Quote:

---

Originally Posted by artemm

Just RMA'd a 5870 and got a 7870 OC back. Might have to put it through its paces instead of selling now :D

In the meantime.. let's see where this gets me:
Attachment 130577

---

Since I made that post, I have 4 pages of Pending WUs. Completion times vary between 3.5 and 4 minutes. The cards clock themselves up to 3D clocks when in use. What really surprises me is the low heat output and power consumption. I KNOW what these cards will do when pushed to the absolute max.... instead of breaking 1KW of total system draw, I'm seeing 600W from the wall with a fully loaded 990X. Also, despite what others reported, I do notice GUI lag, but nothing serious. Aero effects are a little laggy, but the input never freezes fully.

EDIT: Not sure why the board software keeps resizing my image for me even though it's hosted elsewhere, so I'm just going to link to it.

Quote:

---

Originally Posted by Snow Crash

Depending on how much over 5 minutes you are still getting a pretty good increase in overall throughput. Each GPU + CPU (including HT or not) + OS configuration needs to be assessed to find the overall benefit.

---

At the moment it is taking between 5:10 and 5:30 for a WU to complete. This afternoon I did see some below 4:30. At 5:30 per WU that is 17k boinc per day.


Is every one getting what they claimed for a completed WU?

I'm getting two 460's and one 7970 today from a buddy, he also has 10 GTX285's that I can take but should I??

Need to check my PSU status here I guess, power consumption will go through the roof here :p:

I have been playing around with cpu/gpu settings (3770k/HD 7770) and have these results,

2:19 to 2:55 min/wu - at stock cpu/gpu settings
2:10 to 2:40 min/wu - with gpu oc'ed 1100/1200 from 1000/1125 (core/mem) [144 watts total]
1:52 to 2:34 min/wu - with gpu oc and cpu oc'ed to 4.2 [154 watts total]

@herndermd...what driver you using again (I'm on 12.8)? And what program for OCing? Have you had any "grey screens" with your OCs? I've been trying to OC my 7870 and keep getting solid gray screens eventually even if the OC is a little, like 1100 core/1300 mem. Stock is at 1050/1250.

Quote:

---

Originally Posted by Conquistador SW

I calculated the average CPU time for all the WU's that have been validated by now (46 WU's) for my GTX 260 (192SP) + 3770K at 4.8GHz.

.........

Average CPU time was 0.1208 which would mean a WU on average takes 7 min and 40 sec.
It can do slightly more than 8 WU's per hour which means it would do ~45000 WCG points per day.

Oldchap, how did you calculate from time per WU to WU's per hour? I seem to be getting slightly different numbers and probably forgetting something stupid :shakes:

---

I am taking the runtime from the machine.... I watch it run for 20 seconds then it starts to ramp up through the percentages to 99.415% then run on cpu for maybe 50 is seconds on this q6600 that I am testing on. The time I take is that showing whilst it is uploading. I then take an hour and divide by the average time figure of a number of wu's to get wu's an hour etc.

The last few have seen a variance in completion times between 3m 01s and 3m 48s with most toward the lower figure so I choose to say 3m 20s for mine as a reasonable average figure to give folks a good feel for what this card is capable of.

Looking at the numbers produced by these I hope we will be able to decide if it is worth selling some cards and buying others or sticking with what we have as well as seeing which new purchases might be the most cost effective

As for running cards at full speed with 2 wu's running.... I don't think I want to go there purely for power cost reasons just now.

Chart update soon ;)

Quote:

---

Originally Posted by bluestang

@herndermd...what driver you using again (I'm on 12.8)? And what program for OCing? Have you had any "grey screens" with your OCs? I've been trying to OC my 7870 and keep getting solid gray screens eventually even if the OC is a little, like 1100 core/1300 mem. Stock is at 1050/1250.

---

I have catalyst 12.9 beta and using sapphire Trixx to oc. Currently 1200 / 1400 at XFX 7870. Running fine no "grey screens" or other problems. CPU 3930K ocd to 4900Mhz. Average WU time 1.30 - 1.40 minutes :)

First WU took 4:01 minutes and the second GPU WU took 3:43 minutes... Dual 5870's are pretty darn fast at this!

Quote:

---

Originally Posted by bluestang

@herndermd...what driver you using again (I'm on 12.8)? And what program for OCing? Have you had any "grey screens" with your OCs? I've been trying to OC my 7870 and keep getting solid gray screens eventually even if the OC is a little, like 1100 core/1300 mem. Stock is at 1050/1250.

---

I'm on 12.8 as well, my card is a HIS HD 7770 and I used the iTurbo program that came with.

Ok, here are my findings so far, I'm sure many of you have noticed the same:

There is a "GPU-part" and a "CPU-part" in current HCC GPU WUs.
When the WU starts, GPU usage goes up (90% in my case), and the WU computes to 99,5% finished with no real CPU load (getting 2% on a HT-enabled Dual Core). Then, the GPU usage stops completely, and the CPU usage starts to hit max. for 1 core/thread until the WU shows up as completed, and the GPU starts its next WU. It almost seems like the CPU has to convert the GPUs result to a CPU result, or maybe check it for errors, or something.

What this means:

A faster CPU will improve total computational time of GPU WUs!

-------------------------------

Results for single 5870 (900/1200) and L5630 @ 2.8Ghz using Catalyst 12.8 and BOINC 7.10.28 on Win7 x64:

GPU time 2:15-2:30 + 45s CPU time = total time around 3 minutes 15 seconds

i would say definitely don't purchase new gtx 460 for the project.. know limitaion in the arch makes them only 1/2 or 2/3 of the shaders useable for CUDA tasks (can't remember exactly, I bet snow crash knows..)

edit: boinc knows that the GPU tasks will take 1 cpu core and reserves a core for them, so you can still set 100% of cores. notice the task says "1 CPUS + 1 NVIDIA GPU"

Quote:

---

Originally Posted by werdwerdus

edit: boinc knows that the GPU tasks will take 1 cpu core and reserves a core for them, so you can still set 100% of cores. notice the task says "1 CPUS + 1 NVIDIA GPU"

---

Didn't realize that, but you are right - scratch that then.
For now it looks like the HD7770 is the card to buy, although I dont get how it can be faster than a 5870. In every other application, be it GPGPU or not, the 5870 is twice as fast. Bitcoin, which is probably the best OpenCL benchmark there is, has the 5870 at ~80% higher performance. Hopefully with future optimizations, the performance of present and past top cards will be put to better use (7970 apparently doesn't go over 60% usage either)

MicroCenter has Gigabyte HD 7770 1GB cards for $109.99 right now before a $20 MIR...

http://www.microcenter.com/product/3...x16_Video_Card

Update: still checking info.... If you are listed please help me out and post rig info :up:

http://imageshack.us/a/img708/598/hccgpu.jpg

Thanks for the chart OC!

Maybe add core/mem clocks as well???

OC, here's some data for you:
GTX470 @ 740MHz Core: 3 minutes 10 seconds on a Core i7-2700k @ 4.4GHz
GTX460 @ 846MHz Core: 6 minutes 5 seconds on a Core i7-860 @ 3.4GHz

OC, my GTX 260 is on an Intel Hex (6c/12t) @ 3.6Ghz as I said, it has no CPUID since it's an A0 :P
The Info for the 5870 is here - thanks :)

[Ion]: if you have some spare time at the weekend you could provide some great info about the importance of the rig by swapping those over for an hour or two and getting new data.

bluestang: because tables is still disabled on xs the whole exercise is somewhat manual. If we are going to go into that sort of depth then maybe later once we have a clearer picture of the hierarchy of the cards.

I will run this same card in my MVG 3770K rig tomorrow... that will show me a good deal of info and I will post results on rig importance/clock speed a lot easier on that platform.

jcool: got it!

Quote:

---

Originally Posted by OldChap

[Ion]: if you have some spare time at the weekend you could provide some great info about the importance of the rig by swapping those over for an hour or two and getting new data.

bluestang: because tables is still disabled on xs the whole exercise is somewhat manual. If we are going to go into that sort of depth then maybe later once we have a clearer picture of the hierarchy of the cards.

I will run this same card in my MVG 3770K rig tomorrow... that will show me a good deal of info and I will post results on rig importance/clock speed a lot easier on that platform.

---

I'll see what I can do, I might have time for some good tests :up:

I also have an i7-920 system that I'm building Tuesday, that will run with a GTX470. I'm going to try it at 3.4GHz and 4GHz and see what the differences are in terms of time/GPU WU on that vs the 3.4GHz i7/GTX460 rig and vs the 4.4GHz i7/GTX470 rig.

And I should be able to swap my 7870 from a E7500 Dual Core @ 3.8 GHz to a 2600k @ 4.5 GHz this weekend.

I'll also try and do some OCing of the core/mem clocks and chart that as well once its in the better rig.

I'll let you know what improvements all this makes (I suspect pretty good).

I'm still seeing lots of differences in the time it takes to finish a WU. I haven't changed anything on this system. But the first 4 hours or so it took up to 5:30 per WU and now the last 2 hours it is down to 4:40-5:00.

I suspect it may depend on the CPU work units in progress at the time...There are so many variables.

Changed to a pic at top of this page.... easier if everybody can still read it

I just put my ATI 6870 on this. It goes quite fast. Each WU takes around 3 minutes.

Maybe i'll bring my desktop back to college for free electricity.

Quote:

---

Originally Posted by OldChap

I suspect it may depend on the CPU work units in progress at the time...There are so many variables.

---

It is running 2 WU at a time on the GPU and no CPU WU's (haven't figured out how to do that yet)
Down to 4:10 on the last 10 WU's. I'm not complaining :) But it will be difficult to see performance differences after changing the CPU and or GPU settings this way.

Can anyone provide the power draw on a 7770 HCC only load? I'd like to see the difference in KWh/WU between that and my 570 on this project.

Quote:

---

Originally Posted by kjeldoran

I think we can do some more tweaking to get more WUs in one day. Unlike other projects, knocking off seconds is a huge percentage of overall time.

For example, I saw no noticeable decrease in time increasing core clocks of my 7850 from 900mhz to 950mhz. This most likely has to deal with the fact my GPU usage is pegged at a max of 87%. Inefficiencies in coding or maybe shader use?

However, changing the percentage of CPU usage in boinc to 87% (freeing one thread) decreased my time between 7 and 18 seconds. Yes, you lose one thread for other CPU projects, but you increase the amount of WUs being done by the GPU.

Previous time per WU 2:06, with 87% cpu usage being used it dropped it by an average 10 seconds. A drop in 10 seconds produces and extra 59 WUs. One thread of missing CPU HCC is only about 24WUs on a 4.5ghz 2600k. Losing a CPU thread provides an overall increase in WU completion :).

Of course, I only watched this for like 10 minutes. With some daily stats and more tweaking we could maybe drop the overall time of GPU WUs with the right settings.

---

IIRC changing your desktop graphics theme from Aero to Classic will free up some graphics "horsepower". That should make WUs crunch a bit quicker. I have no real evidence of this in WCG but with F@H I could get a few more PPD by making the switch.

Need advice here.
Can I run different GPU's like GTX680L and GTX480 in the same rig?
Seem to me they use the same driver or am I way out there :p:

Edit:
Do I really need an SLI bridge when running two GTX460's?

Quote:

---

Originally Posted by Mydog

Need advice here.
Can I run different GPU's like 680 and 480 in the same rig?
Seem to me they use the same driver or am I way out there :p:

---

My instinct is: yes. I will try soon with mixed ATI cards. I expect it to work with a dummy plug installed:
http://www.overclock.net/t/384733/th...ond-dummy-plug

Quote:

---

Originally Posted by artemm

My instinct is: yes. I will try soon with mixed ATI cards. I expect it to work with a dummy plug installed:
http://www.overclock.net/t/384733/th...ond-dummy-plug

---

Why do you need a dummy plug?

Same reason bitcoin needed a dummy plug. I just don't know that reason :)

Quote:

---

Originally Posted by Mydog

Why do you need a dummy plug?

---

Without SLI/crossfire enabled and with no "monitor" connected, the GPU won't be able to crunch. This is an old school trick from F@H people had to do. I don't frequent the WCG forums long enough to see if this has been tried yet, but I suspect we may need to dummy plug the odd GPU out...

Feel free to correct me if I'm wrong here, guys, I've yet to test this myself.

You can also just attach a monitor to each card and then once it's crunching remove the monitors. I did this for bitcoin since i didn't want to mess with a dummy plug.

Quote:

---

Originally Posted by shoota

You can also just attach a monitor to each card and then once it's crunching remove the monitors. I did this for bitcoin since i didn't want to mess with a dummy plug.

---

That I'll do, thanks :)

For what it's worth, I pulled the Crossfire cable on the cards and it's still crunching along fine on all 4 cores. You should be fine mixing and matching cards, mydog.

Quote:

---

Originally Posted by artemm

For what it's worth, I pulled the Crossfire cable on the cards and it's still crunching along fine on all 4 cores. You should be fine mixing and matching cards, mydog.

---

That's good news, just got a bag full of different ones :up:

Just installed a HD7970 on one of the 3770K rigs and now trying to get a GTX480 to run on a 3960X rig together with a 680L.

Putting two GTX460's in on an SR-2 rig later, no need for an SLI bridge there if I understand correctly.

Attachment 130582

Cant get the 480 to start any ideas fellas?

So is HCC going to put XS at the top of the WCG charts?

Quote:

---

Originally Posted by Mydog

Attachment 130582

Cant get the 480 to start any ideas fellas?

---

Interesting... plug in a monitor or a dummy plug to the 480 and reboot...

Quote:

---

Originally Posted by artemm

Interesting... plug in a monitor or a dummy plug and reboot...

---

Monitor pluged in and rebooted twice and still nothing :shrug:

Quote:

---

Originally Posted by Mydog

Monitor pluged in and rebooted twice and still nothing :shrug:

---

Windows is detecting both and they're displaying something?

Quote:

---

Originally Posted by artemm

Windows is detecting both and they're displaying something?

---

Yes and yes, both montiors got picture.
Could reinstalling the driver help?

Quote:

---

Originally Posted by 0ne.shot

Can anyone provide the power draw on a 7770 HCC only load? I'd like to see the difference in KWh/WU between that and my 570 on this project.

---

1:52 to 2:34 min/wu with gpu oc and cpu oc'ed to 4.2 [154 watts total] 123 watts before gpu kicks in, so 31 watts.

Quote:

---

Originally Posted by artemm

Without SLI/crossfire enabled and with no "monitor" connected, the GPU won't be able to crunch. This is an old school trick from F@H people had to do. I don't frequent the WCG forums long enough to see if this has been tried yet, but I suspect we may need to dummy plug the odd GPU out...

Feel free to correct me if I'm wrong here, guys, I've yet to test this myself.

---

That's actually wrong. My Core i7-860/GTX460 setup has a pair of GTX460s. There is a SLI bridge on the cards (I enable SLI when I game on it), but SLI is disabled, and only the first card has a monitor hooked up to it. Both crunch at the same speed :up:

Quote:

---

Originally Posted by [Ion]

That's actually wrong. My Core i7-860/GTX460 setup has a pair of GTX460s. There is a SLI bridge on the cards (I enable SLI when I game on it), but SLI is disabled, and only the first card has a monitor hooked up to it. Both crunch at the same speed :up:

---

I used a GTX 570 and 2 GTX 260 C216s without an SLI bridge and 1 monitor and never had any problems running PrimeGrid on them. There was an issue years ago with F@H, but as far as I recall it was fixed in 2009. I ran both 260s on F@H without any issues with 1 monitor.

Quote:

---

Originally Posted by Mydog

Yes and yes, both montiors got picture.
Could reinstalling the driver help?

---

I think I ran across this when I was dabbling with Nvidia and crunching GPU Grid a while back.

Try this in your cc_config.xml file:

Code:

---

<cc_config>
<options>
<use_all_gpus>1</use_all_gpus>
</options>
</cc_config>

---

Quote:

---

Originally Posted by artemm

I think I ran across this when I was dabbling with Nvidia and crunching GPU Grid a while back.

Try this in your cc_config.xml file:

Code:

---

<cc_config>
<options>
<use_all_gpus>1</use_all_gpus>
</options>
</cc_config>

---

---

That did it :D
Thanks a lot :up:

nice one artemm

Just looking at the stats over at WCG. Might be time for the GPUGrid boys to come home coz to take on the next step up in the dailies will need the team output to more than double. Easynews, however, appears to be history at history point.

Quote:

---

Originally Posted by D_A

Just looking at the stats over at WCG. Might be time for the GPUGrid boys to come home coz to take on the next step up in the dailies will need the team output to more than double. Easynews, however, appears to be history at history point.

---

no linux gpu crunching yet and that is what our big gun (Stoneageman) is running.

Quote:

---

Originally Posted by fallwind

Just for another data point, my GTX 460's are averaging about 6 minutes per workunit and claiming 33 credits each. That calculates out to around 55000 WCG/day. I recall during the beta I was getting 55 credits/WU so something has changed there.

If the Radeon 7770 really can do 150k/day then that's likely the card to go with. I'm seeing them on sale here in Canada for as low as $99!

---

Don't beta WU typically credit higher than normal?

Also an update on mine. It seems my initial estimate was low. The last WU just finished at 3:12

Quote:

---

Originally Posted by jspace

Don't beta WU typically credit higher than normal?

Also an update on mine. It seems my initial estimate was low. The last WU just finished at 3:12

---

Never noticed that before with beta's. :shrug:

Quote:

---

Originally Posted by Moo

no linux gpu crunching yet and that is what our big gun (Stoneageman) is running.

---

Thanks for pointing this out! Now I know that I won't want to use Mint on my new i7-920/GTX470 build. Windows 7 x64 "SE-W-E" it is :up:

Quote:

---

Originally Posted by jspace

Don't beta WU typically credit higher than normal?

Also an update on mine. It seems my initial estimate was low. The last WU just finished at 3:12

---

Yes. When the GPU WUs were in beta, they were worth ~56 points each, now they're worth ~34. Still amazing, but a bit less so ;)

We need more ATI GPU's running on our team fast :yepp:

I've got an HD6850 I'll get online soonish. I have to reinstall win7 first :(

Quote:

---

Originally Posted by Moo

no linux gpu crunching yet and that is what our big gun (Stoneageman) is running.

---

It'll come, but when is another question entirely.

I will have to do the same ...changing 1 rig to win 7 then changing the accounts around some is how I feel I will do this just now However that is before I run the 775 ddr2-800 rig @3.3 vs the 1155 ddr3-2000 @4.8 test. I still have not put this on a Kill a Watt either but on that at least I am hopeful. I think I have 5 hd 5870's all ready to go...... Should be good to add 70K to my current 25K ppd... I have the slots.... and probably the psu's just need a few more cards if I can afford the power.

It will be the power use that determines if I need to sell these cards and buy newer faster better here

This was a rig (Work PC in sig) that normally would only run 12-14k WCG PPD going 24/7:

Statistics Date -- 10/12/2012
Total Run Time -- 0:001:23:50:53

Points Generated -- 173,573 :D

Results Returned -- 728


EDIT:

Quote:

---

Originally Posted by OldChap

I think I have 5 hd 5870's all ready to go...... Should be good to add 70K to my current 25K ppd... I have the slots.... and probably the psu's just need a few more cards if I can afford the power.

It will be the power use that determines if I need to sell these cards and buy newer faster better here

---

Let me know if you are going to sell. I can put one in my Work PC for with free electricity.

So my question on this -how long are the HCC WUs going to last? Are we going to eat them all up then have a bunch of left over GPUs?

Not trying to rain on the parade - just want to know before I start making plans!

And thanks for all the GPU numbers guys! :up: :up: :up:

old_chap, shame your so far away.. I could use another 5870..
460GTX..6.5 mins
5870..4 mins
Oh, and I won't take away anyones joy but Xs did over 20 million yesterday and is looking at 23 million today..
I might even hit a million..Now that is funny! :D
Off to help out a friend..later guys..:up:

Quote:

---

Originally Posted by Mydog

We need more ATI GPU's running on our team fast :yepp:

---

I want to join you guys again so much. Just have to wait out for a little while longer until the asic's get here and I'll be on full force. with ~1.8 million wcg a day. Very thankful to see WCG find a use for my soon to be otherwise useless 59/870's.

I am wondering how the 7750 would do as it uses the same chip as the 7770, only 512SP vs 640SP and 800 MHz vs 1000 MHz compare to the 7770. I could buy 3 (I have an empty socket 1366 board) for the price of 2 7770's and they use almost half the power at full load per card, so 3 7750's should use less power than 2 7770.

I found a post on the WCG boards reporting about 4 min/wu for a 7750.
https://secure.worldcommunitygrid.or...ad?post=395655

It's nice what a shiny new GPU will do.
10/13/2012 0:000:19:34:31 92,855 404 (averaging ~85 seconds per)
<<<replaced a 5450 w/ a 7870>>>
10/12/2012 0:000:14:42:36 4,166 27 (averaging ~30 minutes per)
10/11/2012 0:001:02:47:07 5,325 23
10/10/2012 0:000:01:43:41 438 2
10/02/2012 0:000:02:11:26 1,591 4

The best part is that, from what I remember, it "only" adds about 60W to power usage while crunching. Like a dumb***, I forgot to check the wattage before the switch.

Quote:

---

Originally Posted by hendermd

I found a post on the WCG boards reporting about 4 min/wu for a 7750.
https://secure.worldcommunitygrid.or...ad?post=395655

---

Thanks!

If that is true, that would also make the 7750 about 2 times slower in which case the 7770 has the edge in points/watts. Looking at the specs I was thinking it would do better.

Edit: according to another post he made later on in that thread the GPU takes 3 minutes and his athon x4 adds another minute, so maybe with a faster CPU you can do these units in less than 4 minutes with the 7750, but I guess the 7770 will still be a better pick.

I got another kill-a-watt like device to check the power usage of the GPU.

I was running this rig without CPU work and 2 GPU WU's at a time. I was using 205 watts with 2 WU's on the GPU. And with one WU's suspended it was using 193 watts.

Now I have switched back to do CPU work and 1 GPU WU at a time. With the CPU WU's suspended (only GPU work being done) it uses 186 watts and with the GPU idle it drops to 131 watts.

The power consumption between 1 or 2 WU's is just 20 watts and between 1 WU and idle is 55 watts.
When running 2 GPU WU's at a time it takes 4:40-5:00 to complete a WU. And with 1 WU at a time it takes 3:00-3:15 to complete a WU.

This is on a i7-860 @ 4 GHz (no HT) with a ATI 5850 @900/1000 MHz.

Now food for thought, what would an SR-X with two 15 core IB Xeons paird with four 7990 dual GPU cards do in WCG and Boinc? One mill WCG PPD?

I have a

q9550 + 5850

3770k + 5870

e3 1235

this normally nets me 12k boinc a day.

I threw down 32k yesterday >: )

and just added another 5830 to my e3 1235

will an ati 6670 with HF acetal extreme should it have any problem to run 24/7 ? I've had 2 or 3 gpu cards die before

If you can keep the vrm's cool I think you are good.

A couple of hours of playing around with these cards and I get the following info:

As seen before the 5870 on q6600 @ 3.3 gives a runtime of somewhere around 3m 20s. the cpu time at end of wu is maybe 53s plus

running on 3770K @ 4.8 the average time for 20 wu's was 2m 57s and the second cpu time is around 41s

Conclusion: GPU will breath new life into an old rig and make it useful for crunching but if you can use a newer, faster rig you will definitely see benefits. My spreadsheet suggests nearly 2k ppd between these two.

Power use:

MVG, 3770K at 4.8, 8G samsung green @ 2000, Samsung spinpoint hj103 plus 1 fan running together with an asus hd5870 Idles in windows at 100W

Running 8 x GFAM takes this to 210 watts and would net 10K+ ppd in linux.

Running 1 x GPU plus 7 x GFAM takes the power use to 270 watts where I predict it will produce around 15K extra ppd

Running 2 x GPU plus 6 x GFAM takes power use to 360 watts so a single card would seem to use 90 watts to produce 15K. This figure takes into account the losses expected by running less cpu tasks

This setup is running on a 650w seasonic x ..... 55% of rated power now which is right in the sweet spot

Conclusion: this single rig with simple old cards running mildly overclocked on cores to 900 seems likely to produce in the order of 32K GPU + 7K5 CPU.

My new processors arrived today... the new board I have my eye on has 7 slots. If the 7770's are about as power hungry as these cards I could end up with all my crunching on one machine that does 125K ppd with current cards, 160Kppd with mid range cards or 230K ppd with top end cards :eek:

EDIT:Tomorrow I will investigate the implications of running HT off.... With 2 cards running it may be break even but with more ...we will see

I wonder how long the WUs will last?

Quote:

---

Originally Posted by retro77

I wonder how long the WUs will last?

---

I hope they last a long time but with the speed GPU's are chewing through them they might not. :shrug:

Okay, so I did the math. If I run my GTX 570 with a 400W load, or a 7770 at 200W load, it'll take (6000 hours)250 days of 24/7 constant running to equal the cost of running a GTX 570. The cost is factored in with $0.10/KWh. However, the 7770 will also produce many more points/KWh, too, so that has to be factored in.

I see a lot 5000 and 7000 series ATI cards but where are the numbers for the 6000 series?

Are your calculations based on measured load running HCC GPU?

Manufacturers ratings are no good here just now .... I think mine are rated to pull around 200w here it is only around 90w

With power use remember too that a proportion of the wu runtime is not with GPU under any load so again that is less cost.


I've yet to see any power use figures for 6000's but looking at jspace's results the 6970 should fall near the 7770 for performance

IMO the workunits need to be about 10x longer (more work done) than they current are. For comparison the GPUGrid tasks take 6-12 hours on decent hardware.

Quote:

---

Originally Posted by werdwerdus

IMO the workunits need to be about 10x longer (more work done) than they current are. For comparison the GPUGrid tasks take 6-12 hours on decent hardware.

---

I tend to agree with you but not to that degree..10 mins would be fine BUT they did this thinking about heat and that is understandable also.
They think worst case, IE: Mrs Jones Dell that hasn't been cleaned inside since new..:D

My 6870 takes around 2:30 per WU, but my stock i7 920 adds ~30 seconds in the beginning and ~1:30 at the end :(

Too bad the IGPU in the 2600k is not supported....:down:

Quote:

---

Originally Posted by OldChap

Are your calculations based on measured load running HCC GPU?

Manufacturers ratings are no good here just now .... I think mine are rated to pull around 200w here it is only around 90w

With power use remember too that a proportion of the wu runtime is not with GPU under any load so again that is less cost.


I've yet to see any power use figures for 6000's but looking at jspace's results the 6970 should fall near the 7770 for performance

---

I was making an estimation based off of this: http://www.anandtech.com/bench/Product/536?vs=518

At the bottom it shows load temps with OCCT on both GPUs. A 2700k @ 4.7GHz and GTX 570 load is right at 400W - sometimes just under or over. The 7770 looks to be right around 200W. I'm guessing HCC GPU isn't as stressful as OCCT, but the power consumption numbers still should be relative to the HCC GPU tasks. I have an APC UPS that tells me current power draw. Of course, that is from the wall and I'm not certain of the PSU Efficiency @ 400W with a Corsair TX750.

I have no idea what the manufacturer's power rating are.

I just finished a HCC GPU WU on my A-3650 APU with Radeon 6530D (320SPs) @ a total of 13:27. It takes quite awhile and the UPS on that PC says 90W with a HCC GPU task and 3 HCMD 2 tasks running.

Quote:

---

Originally Posted by retro77

Too bad the IGPU in the 2600k is not supported....:down:

---

It is a bummer. The iGPU on my A6 3650 works but not on this 2700k. :shrug:

Quote:

---

Originally Posted by Rob_B

Taken from the WCG forum.

Code:

---

<app_info>
<app>
 <name>hcc1</name>
 <user_friendly_name>Help Conquer Cancer</user_friendly_name>
</app>
<file_info>
 <name>wcg_hcc1_img_6.56_windows_intelx86__ati_hcc1</name>
 <executable/>
</file_info>
<file_info>
 <name>hcckernel.cl.6.56</name>
</file_info>
<app_version>
 <app_name>hcc1</app_name>
 <version_num>656</version_num>
 <platform>windows_intelx86</platform>
 <avg_ncpus>1.000000</avg_ncpus>
 <max_ncpus>1.000000</max_ncpus>
 <flops>32282641180.000000</flops>
 <plan_class>ati_hcc1</plan_class>
 <api_version>7.1.0</api_version>
 <file_ref>
  <file_name>wcg_hcc1_img_6.56_windows_intelx86__ati_hcc1</file_name>
  <main_program/>
 </file_ref>
 <file_ref>
  <file_name>hcckernel.cl.6.56</file_name>
  <open_name>hcckernel.cl</open_name>
 </file_ref>
 <coproc>
  <type>ATI</type>
  <count>0.500000</count>
 </coproc>
 <gpu_ram>262144000.000000</gpu_ram>
</app_version>
</app_info>

---

Use at your own risk. app_info only usable for HCC GPU only.

It works for my 5850, but now it only uses the GPU. No work for the CPU. And it went up from 3 minutes per WU to over 5 minutes.

---

I started to use this app_info this evening. Yes, you lose your work for your CPU, but if you are focusing only on HCC then you will complete more WUs than using both your CPU and GPU based on my observations with my setup.

Previously, I was doing GPU WUs around 1:56 with only one WU and 7 CPU threads. Using the above app_info, I lose all of my CPU threads but I work on 2 GPU WUs at once. Completion times around around 2:06 to 2:20 per WU. Yes, it is slower but I am completing two WUs at once. Even if I had the max of 2:20 per WU that is over 1200 WUs in one day. Of course, there will be variations and even if it was only 1100 WUs it is still more than what I would get with 7 CPU threads and 1 GPU thread running HCC (which is around 900 WUs in one day at 1:56 per GPU WU). I'd really only recommend the above app_info if you run HCC only and your GPU is able to handle the work.

You can also do more than 2 GPU WUs at once if you change <count>0.500000</count> to .33 (3 at once) or .25 (4 at once). Info was taken from nanoprobe on WCG forums for increasing the number of WUs beyond 2.

Pros:
Decreased power consumption since only 2 threads are being used on the CPU
More WUs completed

Cons:
Only able to run HCC GPU at this point in time, maybe there can be code added to run another project later on.
More stress on the video card
Some video cards may not benefit from this app_info

I am still in the testing phase of this and I will see how the results are in the next few days. I might see if changing the count to 3 or 4 can increase overall WU completion.

Last thing, I would imagine using this tweak will show a greater effect when the core clock is increased since the GPU is being utilized more.

Quote:

---

Originally Posted by Conquistador SW

Thanks!

If that is true, that would also make the 7750 about 2 times slower in which case the 7770 has the edge in points/watts. Looking at the specs I was thinking it would do better.

Edit: according to another post he made later on in that thread the GPU takes 3 minutes and his athon x4 adds another minute, so maybe with a faster CPU you can do these units in less than 4 minutes with the 7750, but I guess the 7770 will still be a better pick.

---

Gentlemen. Would you all be kind enough to publish your CPU times. I have a feeling that we will see that regardless of GPU used we can get a clear picture of the effect of platform. So, what is needed is from observation ...

primary cpu time, secondary cpu time, cpu type, cpu speed and whether ht is on or off.

EDIT: HT results ;)

http://imageshack.us/a/img407/9268/htoff.jpg


Given that with ht on there are 4 threads of cpu running then there is no good reason to move to ht off with a single card of this type, however when running 2 cards the advantage seems to be in running ht off

I suspect this may also have an effect on the decision to run multiple instances on certain cards.

Quote:

---

Originally Posted by 0ne.shot

...
I just finished a HCC GPU WU on my A-3650 APU with Radeon 6530D (320SPs) @ a total of 13:27. It takes quite awhile and the UPS on that PC says 90W with a HCC GPU task and 3 HCMD 2 tasks running.

---

Quick question: is that with the iGPU only or in combination with a discrete card?
Would love to see more APU numbers, since I'm researching currently in this direction for a new HTPC build.

Does anyone know how to read exact times out of the results? I'm guessing it's in the tail end.

Quote:

---

cl_khr_d3d10_sharing
Estimated kernel execution time = 0.31736 [sec]
Starting analysis of ../../projects/www.worldcommunitygrid.org/X0960069160298200605241517_X0960069160298200605241 517.jp2...
Extracting GLCM features...
Total kernel time: 98.204834 (1026 kernel executions)
Total memory transfer time: 3.162154
Average kernel time: 0.095716
Min kernel time: 0.087317 (dx=24 dy=7 sample_dist=24 )
Max kernel time: 0.104350 dx=2 dy=2 sample_dist=2
INFO: GPU calculations complete.
Total time for ../../projects/www.worldcommunitygrid.org/X0960069160298200605241517_X0960069160298200605241 517.jp2: 197 seconds
Finished Image #0, pctComplete = 1.000000
CPU time used = 88.998571
08:32:57 (4316): called boinc_finish

</stderr_txt>
]]>

---

Edit: Let's add a few more results. I only have one machine capable of running theses atm. No power consumption figures.

Quote:

---

cl_khr_d3d10_sharing
Estimated kernel execution time = 0.33044 [sec]
Starting analysis of ../../projects/www.worldcommunitygrid.org/X0960069161475200605171506_X0960069161475200605171 506.jp2...
Extracting GLCM features...
Total kernel time: 116.657501 (1026 kernel executions)
Total memory transfer time: 1.940834
Average kernel time: 0.113701
Min kernel time: 0.104511 (dx=23 dy=11 sample_dist=24 )
Max kernel time: 0.122929 dx=2 dy=0 sample_dist=1
INFO: GPU calculations complete.
Total time for ../../projects/www.worldcommunitygrid.org/X0960069161475200605171506_X0960069161475200605171 506.jp2: 213 seconds
Finished Image #0, pctComplete = 1.000000
CPU time used = 92.024990
08:22:55 (12696): called boinc_finish

</stderr_txt>
]]>

---

Quote:

---

cl_khr_d3d10_sharing
Estimated kernel execution time = 0.32184 [sec]
Starting analysis of ../../projects/www.worldcommunitygrid.org/X0930069660332200606081307_X0930069660332200606081 307.jp2...
Extracting GLCM features...
Total kernel time: 92.362144 (1026 kernel executions)
Total memory transfer time: 1.823963
Average kernel time: 0.090022
Min kernel time: 0.082192 (dx=25 dy=4 sample_dist=24 )
Max kernel time: 0.098286 dx=2 dy=2 sample_dist=2
INFO: GPU calculations complete.
Total time for ../../projects/www.worldcommunitygrid.org/X0930069660332200606081307_X0930069660332200606081 307.jp2: 188 seconds
Finished Image #0, pctComplete = 1.000000
CPU time used = 91.135784
05:46:52 (12684): called boinc_finish

</stderr_txt>
]]>

---

I'm finishing a wu about every 2.5 minutes w/ gtx580 sli + 2600k. All clocks are stock. 6 cpu threads (any project) & 2 gpu threads + 2 cpu threads (HCC gpu).

With a KILL A WATT....my Home PC setup in sig is drawing 186W full load CPU only (running GFAM and HCC GPU).
When GPU kicks in for HCC GPU its at 249W
And running with 22" display and 2.1 speakers I'm at 289W.

2600k @ 4.5GHz, 7870 @ 1050/1250

I'm running anywhere from 1:31 to 1:53 with most of them in the 1:3x range. On the E7500 machine it was 1:32 to 1:57 with most in the 1:5x range. So an improvement using the 2600k.

When do a common 1:32 WU...

15 sec CPU
37 sec GPU
40 sec CPU

Now, I can get them in the 1.2x range if I clock my 7870 at 1200/1400. I let in run like that overnight and was amazed. However, I started using my pc this morning and after a few minutes a solid grey screen appeared and locked that way. Had to reset to restart PC. May have a bum card. Now running at 1050/1250 stk -- not very Xtreme.

I want to try 2 GPU WU at once. How do I accomplish that and still run 6 WU GFAM on CPU?

Quote:

---

Originally Posted by mreuter80

Quick question: is that with the iGPU only or in combination with a discrete card?
Would love to see more APU numbers, since I'm researching currently in this direction for a new HTPC build.

---

That is with the iGPU only. The APU is not overclocked and stock CPU Clock speed is 2600 MHz. Youtube videos lag badly when running HCC GPU and make the PC unusable. I'm going to check the CPU/GPU/CPU times on the project and I'll post them up shortly.

Update: I watched a few WUs at the start and toward the end, and it looks like the first ~40 seconds and last ~60 seconds are CPU time and the rest is GPU. This was observed on 2 WUs that completed in ~12:40 seconds. It's hard to gauge when the CPU started and GPU stopped as the CPU utilization barely changes through the entire run time of the WU.

bluestang: I believe that it is not possible to run 2 wu at once and have cpu wu's running too