[tictoc]

I can throw some 1070 results up later. I have a ton of them in my HFM WU History, at stock clocks and OC'd running in Linux.

 

[Simmons572]

Fantastic, I really appreciate it bud

I am now negotiating for a Vega 64 on ebay, so hopefully this works out.

 

[Simmons572]

I ordered a Vega 64 Powercolor GPU over the weekend for what I believe is a decent price. Once that shows up, I can start collecting WU information, and start tinkering with overclocks

 

[notyettoday]

That change was made when the Ryzen CPUs initially came out. Then the 8700k was allowed in the fold.

I figured that by lifting the CPU limit to 16c/32t, we can get some old 2p systems or some modern beefy CPUs a workout

EDIT: Wait a second, you are right. The previous limit was 6c/12t, that is my bad. Good catch @notyettoday

I am tempted to reduce the thread count to 8c/16t, but at the same time, I feel a higher cap of 16c/32t would actually make the CPU category competitive. What do you all think?

It seems to me, with 8c/16t being common on mainstream, that capping it there may be able to score us more folders. I know 1st gen TR is coming down pretty quickly, but the boards are not, and ddr4 is not, so it's still going to be expensive to jump to that platform to begin with.

 

[Simmons572]

@tictoc @mmonnin @CynicalUnicorn @WonderMutt @NBrock @msgclb @navjack27 @u3b3rg33k @CptAsian @axipher @PimpSkyline @notyettoday (I wish there was a more graceful way to mass-notify...)

I would like to get a tally on what you all think we should do with the CPU core cap. https://www.strawpoll.me/16657228

I can't figure out how to retroactively add a poll to the thread, so this will have to do.

 

[mmonnin]

Maxing it out is fine with me. By a quick glance it looks like everyone could be squished down to 4 teams (at least before re-balance). It's not like we should be that picky.

 

[WonderMutt]

Maxing it out is fine with me. By a quick glance it looks like everyone could be squished down to 4 teams (at least before re-balance). It's not like we should be that picky.

This, this, and this! We should be as open as possible and we should be encouraging people who have top end hardware to donate to the cause. In the end, it is all about the science, so I say, we should have a spot for anyone who wants to join, for the highest core counts, to the lowest graphics card, all should be able to find a home in the TC if they are willing to donate hardware to the cause.

 

[msgclb]

There’s one thing for sure, I’ll never have the cash for one of these TR’s.
I’d sure would like to see how one of these 12c/24t or 16c/32t puppies will fold but I have no problem with adding them to the CPU category

 

[tictoc]

I'm also cool with bumping it up to 16/32.

 

[Simmons572]

Welp, the vote seems unanimous at this point. We will be sticking with the proposed 16C/32T limit

 

[axipher]

I would say max it out, arguably, the more cores you have, the slower each core is so it's not like you absolutely take off.


Correct me if I'm wrong, but [email protected] CPU projects don't scale linearly with more threads, especially SMT/HT threads; keep in mind, I haven't folded on a CPU in a long time.

 

[mmonnin]

It might not be half the TPF for double the cores. But I bet 8 real cores is lower PPD than 8c/16HT even though it can be the same hardware.

If the limit is 8c16t then someone with a 1950x can still run the CPU w/o SMT enabled or just at 16t and still have an advantage over someone with a 1900x at 8c16t.

 

[tictoc]

Someone should test what the ppd increase is as you go up in core count. With the SMP bonus it is possible that going from 4 to 8 cores (HT/SMT off) is more than double the ppd of 4 cores. I don't think many people have played around with CPU folding since Stanford killed big advantage and started focusing on GPU folding.

 

[Simmons572]

The last big hardcore testing that was done on optimizing CPU folding was when @navjack27 did a super deep dive into optimizing Ryzen folding.

But absolutely, I would love to see research done on this category.

 

[PimpSkyline]

It's not cut and dry no more....

I say 8c/16t max. That's still an increase but not a crazy one.

 

[Simmons572]

It's not cut and dry no more....

I say 8c/16t max. That's still an increase but not a crazy one.

I understand where you are coming from. My biggest concern is the fact that CPU's are barely coming close to matching the performance of GPU-W's performance. My thought is by allowing users to fold on CPU however they want, with basically no [typical] Core cap, users will have the ability to tinker and push their CPUs to the max. From what I can remember, @navjack27 's custom built linux-based OS netted a 30k gain in PPD on 6C/12T Ryzen.

I can throw some 1070 results up later. I have a ton of them in my HFM WU History, at stock clocks and OC'd running in Linux.

I'd like to start the discussion with you about the Vega 64. I just submitted my first to WUs to the database a bit ago.

Spoiler

As of right now, I am averaging 730-750k PPD. We were talking before about potentailly adding Vega 64 to AMD, and lowering the 980ti (and Titan X) to the nVidia category. What are your thoughts on this?


---


EDIT: I have been talking with @CptAsian for quite a bit this morning, and we are coming to the conclusion that if Vega64 numbers remain this high, we should put it in GPU-L. If we were to put Vega64 in the AMD category, it would make sense to put the 980ti and 1070 into the nVidia category.

 

[CptAsian]

I understand where you are coming from. My biggest concern is the fact that CPU's are barely coming close to matching the performance of GPU-W's performance. My thought is by allowing users to fold on CPU however they want, with basically no [typical] Core cap, users will have the ability to tinker and push their CPUs to the max. From what I can remember, @navjack27 's custom built linux-based OS netted a 30k gain in PPD on 6C/12T Ryzen.



I'd like to start the discussion with you about the Vega 64. I just submitted my first to WUs to the database a bit ago.

Spoiler

As of right now, I am averaging 730-750k PPD. We were talking before about potentailly adding Vega 64 to AMD, and lowering the 980ti (and Titan X) to the nVidia category. What are your thoughts on this?


---


EDIT: I have been talking with @CptAsian for quite a bit this morning, and we are coming to the conclusion that if Vega64 numbers remain this high, we should put it in GPU-L. If we were to put Vega64 in the AMD category, it would make sense to put the 980ti and 1070 into the nVidia category.

To continue that thought, here are the four big GPU TC categories, with PPD averages from the OCN database.


GPU-O
Titan Xp - 1.5 M
Titan X(P) - 1.4 M
1080 TI - 1.3 M
2080 (Contingent on Submitted WU's) - 1.65 M
Any GPU-L GPU


GPU-L
1080 - 814 K
1070 TI - 870 K
1070 - 698 K
980 TI - 657 K
Any nVidia GPU
Any AMD GPU


nVidia
980 - 443 K
1060 - 337 K
970 - 345 K
780 TI - 268 K
Any nVidia GPU-W GPU


AMD
Fury X - 469 K
Fury - 402 K
RX 580 - 344 K
RX 480 - 301 K
R9 390X - 372 K
Any AMD GPU-W GPU


Dropping the Vega 64 into the GPU-L category would make it right at home, PPD-wise, and like @Simmons572 said, moving the 980 Ti down to Nvidia to match the Vega in AMD would also probably result in pulling the 1070 down as well, since they're so close in PPD. Then, we would be left with just the 1080 and 1070 Ti in GPU-L, and that wouldn't be a terribly interesting category. Additionally, having the 980 Ti in Nvidia and the Vega in AMD would spice things up a bit, but it would make those two cards practically mandatory to lead in those categories, and I personally find that no single card should be the obvious choice in a category (though that becomes much harder to achieve at the highest level of GPU).


On another note, if someone has interest in the Vega 56, that could be a bit more interesting.

 

[tictoc]

Vega 64 in GPU-L makes sense. It crushes my Fury X.

I think the 980 at the top of NVIDIA is fine. That follows what we have now, more or less, and just updates it by one generation of GPUs.

 

[Simmons572]

I have updated the OP, as well as the TC Manual to reflect the upcoming changes.

I suppose, since we have 2 weeks left in the month, we need to begin working on getting the changes staged for November.


Brass Bottom Boys: @makr @gowanlock @DarthBaggins
Big Bang Theorists: @superericla @mmonnin @Ithanul
Explosm: @BWG @IXcrispyXI @JayKthnx
Full Auto: @WonderMutt @utparatrooper @maestro0428 @u3b3rg33k
Infinity: @tictoc @jcharlesr75 @ikem @k4m1k4z3 @emoga
MLP Folding is Magic: @Fir3Chi3f @mbmumford @lanofsong
Power Rangers: @PimpSkyline @Klue22 @Danbeme32 @josephimports
The PPD Police: @irda123 @Tex1954 @Erick Silver @msgclb
The Royal Navy: @Simmons572 @jarble @navjack27 @CynicalUnicorn @kabj06
Still in Beta: @notyettoday @CptAsian

Please begin contacting your captains to start working out how you want to distribute hardware.

Captains, once you have everything worked out, please PM lanofsong, axipher and myself so we can coordinate the mass changes.

 

[CptAsian]

I have updated the OP, as well as the TC Manual to reflect the upcoming changes.

I suppose, since we have 2 weeks left in the month, we need to begin working on getting the changes staged for November.


Brass Bottom Boys: @makr @gowanlock @DarthBaggins
Big Bang Theorists: @superericla @mmonnin @Ithanul
Explosm: @BWG @IXcrispyXI @JayKthnx
Full Auto: @WonderMutt @utparatrooper @maestro0428 @u3b3rg33k
Infinity: @tictoc @jcharlesr75 @ikem @k4m1k4z3 @emoga
MLP Folding is Magic: @Fir3Chi3f @mbmumford @lanofsong
Power Rangers: @PimpSkyline @Klue22 @Danbeme32 @josephimports
The PPD Police: @irda123 @Tex1954 @Erick Silver @msgclb
The Royal Navy: @Simmons572 @jarble @navjack27 @CynicalUnicorn @kabj06
Still in Beta: @notyettoday @CptAsian

Please begin contacting your captains to start working out how you want to distribute hardware.

Captains, once you have everything worked out, please PM lanofsong, axipher and myself so we can coordinate the mass changes.

Do you have my proposed change noted from when I mentioned it earlier? Just let me know when I should change passkeys.

@notyettoday may have some changes closer to the end of the month, yes? There are some more details that you may have seen in the SiB thread.