Settling into the #6 spot.

[don256us]

We still have a slight downhill trend in our production but as of late, we have picked up production considerably. With summer fighting to stay active here, Fall temps will soon be here. I know that my production will increase as will that of several other teammates in addition to members of other teams. This in the face of EVGA trending up but recently seeing a reduction in production. We are set to overtake EVGA in 7.7 years. I'm interested in how much of that we can chew up this winter. Taiwan Team is still chasing us. 2.5 years. Thier trending line shows a steep increase in production but as of late, they are leveling off. Finally, EPFL IC-IT has jumped from 0 points to 850 million ppd with a single user called ICCluster. If memory serves me right (and it often doesn't), EPFL IC-IT is an energy company in France? Whoever they are, they are slated to overtake us in 3.5 years.

T32 in the world top 100/200:
@HayesK #17
@WhitehawkEQ #28
@Holdolin #45
@HayesK #55
@dfonda #58

Top 200 round out: @torin3 , @P4EE, @Farwalker2u , @orion456 , @Macaholic and @don256us .

The team is up to 37 active folders this week. Average ppd for the team is 17 million. Our top folder this week is the subgroup of @folding_monkeys. They have been struggling with power outages, but so too has HayesK been struggling to keep production up. End of summer thunderstorms have really been creating a lot of fun throughout the country.

Top 18 are earning 1 million ppd or more.
Top 8 are earning 10 million ppd or more.
Top 7 are earning 20 million ppd or more.
Top 6 are earning 30 million ppd or more.
Top 4 are earning 50 million ppd or more.
@WhitehawkEQ is earning more than 78 million.
@HayesK is earning more than 146 million.
@folding_monkeys are earning more than 188 million.

@KenSavage has earned 600 million.
@Blastomania has earned 400 million.
@AndrewThomasThomas has earned 300 million.

@Zuzzz has earned 50 million.
@Is_907 has earned 20 million.

@freakdiablo has earned 4 million.
@ec052002 has earned 4 million.

Great milestones guys. Keep it up.

As I keep mentioning the cooler temps of fall and winter, now is probably a great time to ensure that your rigs are ready for the prime point making time. The goal is to run our machines as hard as possible for as long as possible. Winter is a great time to do this as we can use the waste heat in our homes. So take a look at your folding gear. Clean out the dust and replace any noisy fans as needed.

Finally, for us in the U.S. Labor Day weekend is here. Do NOT drink and drive. Or smoke and drive if that's your thing. No judgement. Just don't combine those things. We lost @equinox (still @ # 63 on the team) a few years back because someone else lost control of their vehicle and rolled into our late friend. So be a defensive driver not an offensive one and be sober while driving. Live to fold another day. Oh, and your family might miss you or something too. You never know.

Okay. With that somber thought:

Until next week:
Build, Borg, Recruit.

 

[the garynator]

We monkeys have been down a bit in ppd over the past couple weeks. I'm currently on the tail end of troubleshooting one of my rigs that was having some stability issues. It's a 5900x with a 4070 super and 4070 Ti that was folding in ubuntu.

I think I've got it figured out and finally got it folding with the 4070 super in Ubuntu yesterday (easier to troubleshoot the stability issues in windows so was running on the windows install for the last couple weeks).

The 4070Ti is in my old 2500k rig in windows currently so it's doing about 2/3rds or less of it's normal PPD. If everything keeps going smoothly I'll probably toss the Ti back in the 5900x system this weekend so we should get a nice little bump in addition to the bump we got since mid yesterday.

Also, @KeeperOfTheButch might have a lil somethin somethin up his sleeve. (Probably a banana, but we'll see lol).

Hopefully once I have the 5900x system back at full chooch, I'll be able to update the Ubuntu how to with more detail. I'm sure there's quite a bit of PPD on the table if some of our guys switch over!

Also, I'm snagging some data from the 4070ti while it's in the 2500k system. Preliminary results show a significant decrease in PPD vs pcie 3/4 systems with faster cpus. Not sure how close these differences translate compared to older cards but I've seen some people running 30xx and 40xx cards in older systems. There could be a significant amount of efficiency and PPD on the table with fairly cheap upgrades.

I'll have to quantify it based on the data I've collected, but for example, my overclocked 4070Ti in my 2700x @4GHZ (pcie 3.0 x16) is pretty much on par ppd-wise with a stock 4070 super in the 5900x (pcie 4.0 x16) in Ubuntu. There's normally about a 20% difference between a stock Ti vs Super. The stock Ti in the 5900x, even with the super in it (so pcie 4.0 x8) in ubuntu, out performs the overclocked Ti in the 2700x system by a couple million PPD.

A Pcie 4.0 board, CPU and ram can usually be had for like $350 or less on Facebook marketplace. Heck, the Dell 5820s Kyle has can be snagged for as little as $100 (add a ~$60 CPU upgrade) and you have room for 2 cards at full pcie 3.0 x16. Removing or minimizing bottlenecks on newer cards can give big gains!

 

[dfonda]

Also, @KeeperOfTheButch might have a lil somethin somethin up his sleeve. (Probably a banana, but we'll see lol).

Congrats to the hardworking T32 mates. Banana up a sleeve ...Shaped like a 4090.

 

[the garynator]

Here's an example of platform efficiency (These aren't especially good PPD WUs, but both running at the same time is good for illustration purposes). A 4070 Super, in Ubuntu at stock clocks, stock 5900x, and PCIE 4.0 x16 slightly out-performing a 4070Ti, overclocked, 2700x overclocked to 4ghz, and PCIE 3.0 x16. From our tests, given the same platform, a 4070Ti is around 20% faster. The 4070 super is a bit slower in this platform comparison on some WUs, but it's within spitting distance across the board. The ppd differences are even more stark when you run even older systems and multiple GPUs.


I would venture to guess that this would apply to 30x0 cards and above.

 

[EarthDog]

Thanks for the info.

Does the CPU really matter much for F@H with single cards?
What's the difference between the SAME cards/system at PCIe 4.0 to 3.0 on these cards?
How does the 4070 Ti perform in the 5900X system?

At least for gaming, the difference between the 4070S and 4070 Ti is 6%/8% (1080p/4K)...I know this is folding but, that's a significant difference, 20%.

I would venture to guess that this would apply to 30x0 cards and above.

Sorry, what would apply to those cards? I'm not sure what to walk away with (my fault)...

 

[the garynator]

Thanks for the info.

Does the CPU really matter much for F@H with single cards?
What's the difference between the SAME cards/system at PCIe 4.0 to 3.0 on these cards?
How does the 4070 Ti perform in the 5900X system?

At least for gaming, the difference between the 4070S and 4070 Ti is 6%/8% (1080p/4K)...I know this is folding but, that's a significant difference, 20%.


Sorry, what would apply to those cards? I'm not sure what to walk away with (my fault)...

Yeah, CPU does matter, mainly in speeding up checkpoints. Even 1 sec diff per checkpoint adds up to a lot of bonus points. Some WUs it makes a big difference, some it doesn't. A good rule of thumb is that if your CPU spikes to near 100% for a second during checkpoints, there's ppd on the table. We saw a few million ppd difference when upgrading the CPU in the Dell 5820s @KeeperOfTheButch has down in FL (dual 4070 supers). He probably has a better idea of differences as well as which cpus he's running. If it's a single card, yes, it can matter, iirc the 5820s were hanging at 100% cpu for like 2 seconds during checkpoints on some WUs (checkpoints were staggered so would translate to single cards).

I don't have data on the 5900x with just a single Ti in it, but I can compare the 2700x vs an I5 2500k @4GHZ and PCIE 2.0 x16 with a single Ti in it, both on Windows 10:
I5 2500k @ 4ghz:


2700x @4GHZ PCIE 3.0 x16


Haven't had time to do direct comparisons of WUs and whatnot, but you should be able to see the difference pretty easily.

Each thing (I.e. CPU, PCIe, Ram Speed, Linux etc) makes a difference, but not a huge difference. When you add them all together they DO make a huge difference though. I have to get some time to go through the data and quantify the differences between each change, but overall we've seen that to maximize PPD within reason you'll want at least PCIe 3.0 x16 per card, and an 8 core CPU or better. Anything less starts to drop output by the millions.

As far as what applies to the 30x0 cards, I would guess that CPU and PCIe (Ram to a lesser extent) will make similar percentage gains.

As far as Ti vs Super %, IIRC we saw 20% gain on some WUs, but 10-15% on others iirc. I still have to go through the data to get exact numbers but there's quite a noticable jump in PPD.

 

[freakdiablo]

I've actually been offline over the holiday weekend, traveled and didn't want to leave things running, but should be able to get my meager contributions back online later this evening.

 

[the garynator]

I've actually been offline over the holiday weekend, traveled and didn't want to leave things running, but should be able to get my meager contributions back online later this evening.

Awesome! Every point counts!!

 

[Norcalsteve]

False fall should be coming through here in the next 3-6 weeks in the FL panhandle… looking to churn the home PC for a bit… work PCs been steaming away at least

Keep up the good work yall!