Dedicated rig High end GPU Low end CPU?

[JonasTheMovie]

Yeah, that is why I keep a spreadsheet based on submission date and overall points, to see the real performance, and also extrapolate expected dates of milestones
So far typical WU estimates and measured real performance conform (350-360k), but the WUs with everyother 420k/390k frames intrigued me.
It´s actually quite interresting to review the systems temp and load graphs. WUs are very different from eachother, in how GPU load varies, GPU clock goes up and down (or stay maximized) and how the CPU is used.
Doing some tweaks to the GPU current-limit, temp goals and fan curve to keep it from throttling, while keeping a steady fan speed.

[JonasTheMovie]

Guess what! I have thrown the rig into the originally planned Phantech Entoo Evolv mITX case, making it really really really quiet. Now I can just forget about it sitting under the elevated bookshelf behind the sofa. I was able to cram dual water cooling into the Elite 110, but was not happy with the noise and the placement under the shelf created some air looping. But now it´s mint. Also a HDMI dummy arrived in the mail, so headless VNC use is not a walk in molasses anymore.

Here is my stat xls, a bit hard to read but as you kan see shortterm I have had 360kppd, but longturn with downtime due to rebuilds I average 291kppd.

Start date Points Goal
2016-01-21 06:12 1 646 202 5 000 000
Milestone date Points PPD latest PPD longterm Estimated date
2016-01-22 08:14 2 036 757 360 166 360 166 2016-01-30 13:41
2016-01-22 22:11 2 251 378 368 925 363 224 2016-01-30 11:48
2016-01-24 10:12 2 620 301 245 905 307 637 2016-02-01 03:51
2016-01-25 07:14 2 909 961 330 550 312 604 2016-01-31 23:41
2016-01-25 10:13 2 946 353 291 811 311 982 2016-02-01 00:12
2016-01-26 14:15 3 272 818 279 536 304 879 2016-02-01 06:13
2016-01-27 05:12 3 469 624 315 890 306 031 2016-02-01 05:13
2016-01-28 08:13 3 725 430 227 231 293 508 2016-02-01 16:26
2016-01-29 02:14 3 897 525 229 241 287 350 2016-02-01 22:19
2016-01-29 22:12 4 175 455 334 254 291 851 2016-02-01 18:00

[mmonnin]

A dummy plug is no longer needed.

[JonasTheMovie]

A dummy plug is no longer needed.

To use the computer over VNC without too much aggravation it is....

[7im]

No longer needed by the client to run on multiple GPUs.

[JonasTheMovie]

I have some ground loop problems, or maybe some other cause, with my tv. Get noise from my bass drivers (<120Hz) when the HDMI is connected.
So I use the dummy to keep VNC usable and stereo silent. With the Fury X 3xDP 1xHDMI a HDMI dummy was the ticket.

Hmm, my performance has been really low last two days, ~250-270kppd is not really great.

[7im]

I'll write that down. Quarter Million PPD was not that great.

[JonasTheMovie]

Well, I would prefer to see the GPU pushing a Third at least

[7im]

Ya, me too. Sorry, my sarcasm may be lost on newer contributors to this worthy project. I started when work units were worth a couple points a day.

[muziqaz]

Well, I would prefer to see the GPU pushing a Third at least

You might have stumbled upon certain low PPD project which just came out of beta But still it should be 300k PPD average on a bad day

[JonasTheMovie]

Ya, me too. Sorry, my sarcasm may be lost on newer contributors to this worthy project. I started when work units were worth a couple points a day.

The points are fun, climbing ranks is fun, but the real return is knowing the load is 100% 24/7.

[Mstenholm]

You did it despite all the advises against it. Wtg. Please ask again.

[JonasTheMovie]

You did it despite all the advises against it. Wtg. Please ask again.

?

[3D_Now!!]

I'm sure many great things have been achieved despite naysayers saying it can't be done.

“Whether you think you can, or you think you can't--you're right.” ― Henry Ford

[Mstenholm]

You did it despite all the advises against it. Wtg. Please ask again.

?

Re-read page one. All but none say it is a poor idea to fold on a Fury. That's why I say please ask again.