Winter is Coming

Dave

Dave

Well-known member
Staff member
Administrator
Winter is coming and I've decided I'll be heating my apartment with my PC folding@home. :)

7820X @ 4.6GHz and a MSI Gaming X 1080Ti running @ 1975Mhz.

9NeTBXX.png


https://folding.extremeoverclocking.com/user_summary.php?s=&u=15574
 
Decided to stop folding on the CPU. Only putting out 150k-200k at best, pales in comparison to the 1080 Ti @ 1.1M
 
Considering its probably consuming 350-400watts of power compared to the 1080 Ti at 300 watts @ 1.1M PPD, not so much.
 
That's a large increase over the TDP for that processor, but yeah even if the power consumption was half, the GPU would still give much better points per watt. I guess if you need more heat then the CPU would come in handy :lol:
 
You'll see hit and miss points as I try and find a stable middle ground for my Vega card. Tried 1,800Mhz @ 1.100V today. Seems it only got 1 or 2 Work Units in. Honestly, I dislike how it handles its voltages. I want it to hit 1800Mhz and HOLD 1.100V while its working. Instead, it holds 1800Mhz and the voltages fluctuate based on load. Voltage goes up as load goes up but in that transition, it crashes. If it would hold that 1.100V, it'd be stable as can be.

At 1800Mhz with 1.1V it pulls 148-155W while running F@H.

Ok, I am back to producing a bit. 1.185V @ 1,800Mhz is where its 100% stable. lol.
 
Last edited:
Flyordie said:
You'll see hit and miss points as I try and find a stable middle ground for my Vega card. Tried 1,800Mhz @ 1.100V today. Seems it only got 1 or 2 Work Units in. Honestly, I dislike how it handles its voltages. I want it to hit 1800Mhz and HOLD 1.100V while its working. Instead, it holds 1800Mhz and the voltages fluctuate based on load. Voltage goes up as load goes up but in that transition, it crashes. If it would hold that 1.100V, it'd be stable as can be.

At 1800Mhz with 1.1V it pulls 148-155W while running F@H.

Ok, I am back to producing a bit. 1.185V @ 1,800Mhz is where its 100% stable. lol.
Click to expand...

Excellent! Glad you got the stability issues worked out.

I noticed that we got Turbo, VW Factor, and DarkFoss back as well, so 8MPPD here we come! :D :drool:

:fold:
 
absolutefunk said:
Excellent! Glad you got the stability issues worked out.

I noticed that we got Turbo, VW Factor, and DarkFoss back as well, so 8MPPD here we come! :D :drool:

:fold:
Click to expand...

10.3 drivers hurt production though. Gonna update to 11.1 and see if it improves again. I downloaded 11.1 2 times though. 1 on the 2nd of November and again on the 10th.. File sizes are different. Nov 10th= 408MB, Nov 2nd = 318MB.
 
Just a heads up for the CPU folders, some of the WUs submitted are not being counted yet for credit, so if anyone notices that their PPD is a bit lower than usual, that would most likely be why. It's been this way for the past week and a half, and I'm not sure when the issue will be fixed (though when it does, we should get the points, cause that's what normally happens :lol: )

BTW, ironically our PPD is a bit higher than it should be... that's due to the stats server outage last week, and the subsequent backlog of points we got is skewing the 7-day PPD average.
 
Well, Kill-A-Watt tells me I am pulling around 365W (after 10m, when the HDDs drop into power saving mode) at the wall with my FX-8320 @ 4.0Ghz (1.308V) folding with 6 threads and the Vega card running at 1,800Mhz @ 1.150V. Plan on adding a 7th thread soon to the CPU side. The only reason I kept the CPU at 6 was because Nvidia GPUs require more CPU usage than AMD/ATI GPUs. Like right now, the GPU slot is pulling 2-3% CPU usage. So, will probably add a single thread later this week.

Temps are actually really good.

VRM temps after I started cooling the back of the mainboard with a small low RPM fan dropped from 85-90C to 55-60C. Case manufacturers have a MAJOR opportunity there when it comes to directing some air behind the mainboard to help cool the VRMs.
 
Flyordie said:
Well, Kill-A-Watt tells me I am pulling around 365W (after 10m, when the HDDs drop into power saving mode) at the wall with my FX-8320 @ 4.0Ghz (1.308V) folding with 6 threads and the Vega card running at 1,800Mhz @ 1.150V. Plan on adding a 7th thread soon to the CPU side. The only reason I kept the CPU at 6 was because Nvidia GPUs require more CPU usage than AMD/ATI GPUs. Like right now, the GPU slot is pulling 2-3% CPU usage. So, will probably add a single thread later this week.
Click to expand...

Awesome! With my 1070 rig I have two threads set aside, one for the GPU and one for OS (doing that is debatable, but the CPU is only an i3, and testing having 2 or 3 threads available for CPU folding resulted in slightly higher output from the GPU when I had a spare thread available for the OS). Nvidia cards will always use 100% of a thread as it uses a loop to continually check the GPU tasks - AMD uses interrupts. Pros and cons to both approaches.

Flyordie said:
VRM temps after I started cooling the back of the mainboard with a small low RPM fan dropped from 85-90C to 55-60C. Case manufacturers have a MAJOR opportunity there when it comes to directing some air behind the mainboard to help cool the VRMs.
Click to expand...

I think many mobo/case manufacturers are still stuck in the mindset that mobo components are cooled by the CPU HSF fan, which won't work if your CPU is on water :lol: That's a great idea! I employ a 40mm fan in one of my micro rigs for a similar purpose.
 
I'm coming for you Turbo! :evil:

Also, I learned that on FX-8xxx series CPUs, its not a good idea to have more than 6 threads going on CPU stuff. Even if the CPU isn't at 100% load.. the FPUs are pushing it. It kicked my output down about 50-60K PPD adding a single CPU thread.

Anyone else here running an RX Vega?

What are your GPU usages at when running F@H?

Also... My HBM2 does 1100. lol. Been slowly ramping it up over the past few days. Soon as I cross 1100, my SoC clock jumps to 1,207Mhz. So the SoC on Vega must feed the HBM2.
 
Flyordie said:
I'm coming for you Turbo! :evil:

Also, I learned that on FX-8xxx series CPUs, its not a good idea to have more than 6 threads going on CPU stuff. Even if the CPU isn't at 100% load.. the FPUs are pushing it. It kicked my output down about 50-60K PPD adding a single CPU thread.

Anyone else here running an RX Vega?

What are your GPU usages at when running F@H?

Also... My HBM2 does 1100. lol. Been slowly ramping it up over the past few days. Soon as I cross 1100, my SoC clock jumps to 1,207Mhz. So the SoC on Vega must feed the HBM2.
Click to expand...

I fold on a 1070, but in general I leave one thread for GPU and then one for misc. It does depend on the hardware though... I also have a R9 Nano which folds sometimes (like tonight since it's cold and my apartment heater stopped working), and I find that Nano's output is cut in half if I CPU fold on even 2 of 4 threads (so no CPU folding while the GPU is active).

My 1070 is overclocked, and I find that folding doesn't benefit much from memory clock as it does for core clock. Some WUs seemed to have no benefit from the memory being overclocked. So I'd focus more on core and save the energy usage from overclocked memory.

Btw, your card I've seen for over $2k now. Crazy times!
 
Yea.... Its hard not to throw it on Ebay and skimp along with an HD5770.

Paid for it in September... $525.

Could sell it for on E-Bay- $1,250. (I've seen bid wars on an XFX one go that much... +25.86 for shipping. lol.)
 
Flyordie said:
Yea.... Its hard not to throw it on Ebay and skimp along with an HD5770.

Paid for it in September... $525.

Could sell it for on E-Bay- $1,250. (I've seen bid wars on an XFX one go that much... +25.86 for shipping. lol.)
Click to expand...

I noticed you haven't been folding lately. Did you give into temptation and sell your Vega to a miner? :lol:
 
You must log in or register to reply here.
Back
Top