Folding@homelab - Inspiring podcasts about the global vCommunity, with nerd stats on TinkerTry's 4 F@h VMs on VMware ESXi 7.0 on 4 Xeon D Servers

Posted by Paul Braren on Apr 15 2020 (updated on May 5 2020) in
  • Efficiency
  • ESXi
  • Virtualization
  • HowTo
  • HomeLab
  • vSphere7
  • TinkerTry-folding-consoles-2020-04-16

    If you haven't already come across my article with a walk-through video about how easy it is to get started with Folding@home VMs, have a look!


    It's now been about 3 and a half weeks since I started my own Folding at my home project, and I thought I'd pull together some podcasts and articles that have recently come out about the enormous growth that the Folding@home program has enjoyed lately, along with a progress update in my own home.


    Let's start yet another inspiring Virtually Speaking Podcast, featuring special guest William Lam talking about how the whole Folding@home VM came to be:

    I made a couple of enhancements, and so, depending on how much virtual CPUs you can allocate your VMs, it will go ahead and set that up and optimize it for you, there's a lot of folks that want to help out, they've got the computing resources, but not necessarily the background on the's supported on vSphere obviously, but we also made it work on stand-alone ESXi hosts, so if you don't have vCenter,we fully support that, it also works on Mac Fusion...

    A little later on, this great moment, not to be missed!

    [When talking about how quickly it went through the approval process, John says:] Since you're on a first name basis with legal already, I'm sure that went pretty quick.
    William: (playing the lawyer role) "Oh, you, you're actually doing something that is supported!"


    Next up, we've got Steven Gibson on Security Now #760 "Folding Proteins"

    In the intro, Steve sets the stage, with Leo expressing skepticism with some of the same concerns I initially had. I'm glad I'm not burning too many watts doing this, I'll let you know exactly how many in a bit.

    Later on, Steve dives deeper, with a good discussion about how it works:

    It's a happy coincidence that just as the SETI project decided that it didn't need any more raw signal processing number crunching, a crucial new need for distributed computing at massive scale would arise... because our unused CPU cycles can now help to increase our understanding of the structure and function of the COVID-19 virus.
    Molecular modeling can be used to identify therapeutic drugs that might be of use in preventing the COVID-19 spike protein from binding to the ACE2 receptors of the cells in our lungs. In other words, our medical science and molecular modeling technology has progressed to the point that the field of "Computational Biology" is now a thing. It's possible to run simulations -- entirely in very sophisticated math -- to predict the 3D molecular shape and thus the interactions of biological compounds.
    WikiPedia has a terrific fact-filled and up-to-the-minute introduction:
    The project has pioneered the utilization of central processing units (CPUs), graphics processing units (GPUs), PlayStation 3s, Message Passing Interface (used for computing on multi-core processors), and some Sony Xperia smartphones for distributed computing and scientific research. The project uses statistical simulation methodology that is a paradigm shift from traditional computing methods.[6] As part of the client–server model network architecture, the volunteered machines each receive pieces of a simulation (work units), complete them, and return them to the project's database servers, where the units are compiled into an overall simulation. Volunteers can track their contributions on the Folding@home website, which makes volunteers' participation competitive and encourages long-term involvement.



    With that, I'll transition to giving a little status update on my own progress as a "Donor" of CPU cycles, giving to both the DellTechnologies Team ID 236797 and VMware Team ID 52737. Here's my latest stats for my donor name of


      Total score 2,424,341
      Total WUs 427
      Overall rank (if points are combined) 90,511 of 2,599,598
      Feel free to share how you've been doing in your home lab by dropping a comment below!



    Using a combination of my Ubiquiti mPower Pro power strip's measurements of my TinkerTry workbench's 3 Xeon D systems that I already detailed, I've got a screenshot showing you how many watts it's taking to keep these systems running at max watt burn. Since I normally leave these systems off when I'm not tinkering with them, it's fair to simply add all 3 systems up, then subtract the network switch.

    147 watts minus 120 watts = 27 watts, not a bad cost of donating 20 vCPUs!

    Next, we'll have a look at my 12 core Supermicro SuperServer Xeon D Bundle running vSphere 7.0 and the vCenter Server Appliance. This system runs 24x7 and it's also my Windows 10 workstation. So in this case, I used my Edimax switch to show power used when running my fully loaded mini-tower (1 PGU, 4 3.5" drives, 2 2.5" SSDs, and 1 M.2 NVMe drive!) at idle versus under heavy Folding@home load.

    April 2 2020 - NASA Satellite Data Show 30 Percent Drop In Air Pollution Over Northeast U.S. My town in central Connecticut went from orange to green.

    Time for some basic math. I sure wish I had solar, because I know full well that a bit of fossil fuel is being burned to power my home. Let's figure out what it costs me personally in electricity, given it costs 16 cents per kilowatt hour in my area of Connecticut served by Eversource.

    Take the TinkerTry 3-node-lab pictured below, remove the network switch's 2 watts, and we have 262 watts. Add to that the additional 27 watts my 12 core workstation/datacenter is using, we get 289 watts total. Given I don't have AC on yet, we'll set the cost of cooling my home to offset the heat for now, and gladly, I keep most of the gear in my unconditioned basement anyway, with our hybrid water heater using some of that waste heat to warm my home's water.

    Plunk those numbers into an online cost calculator and we have ourselves $7.77 per week, $31.07 per month, and $403.95 per year. I can safely say that with over 1,000 folks with the same Bundle I use, I'm sure glad those Xeon D systems are very efficient. In my small way, I'm trying to contribute to a greener future for all of us, while near-term helping with the fight against this pandemic.

    So with that, I'll return to quoting another gem from Virtually Speaking Podcast, this one from VMware CEO Pat Gelsinger, at this spot:


    I think we'll actually find ways to be more productive, also, maybe people don't need to commute 4 hours, 2 hours in and 2 hours out, maybe we can lower the carbon footprint as a result. Maybe we don't need as many office spots...I think there's just a lot of these things where we're actually going to say it's not just different, it's better.


    May 05 2020 Update


    See also at TinkerTry




    See also