Folding@home: Tips and Advocacy

In 1597, English philosopher Francis Bacon said 'knowledge is power'. Folding@home is proof that 'power (both electrical + computer) produces knowledge'.
I suppose this is evermore true with the rise of Artificial Intelligence and LLM (large language models)

This Wikipedia article is more informative than the following personal effort.
Isaac Asimov

Misfolded proteins have been implicated in numerous diseases. Folding@home is biological research based upon the science of Molecular Dynamics where molecular chemistry and mathematics are combined in computer-based models to predict how protein molecules might fold (or misfold) in space over time. This information is then used to guide scientific and medical research.

When I first heard about this, I recalled the science-fiction magnum opus by Isaac Asimov colloquially known as The Foundation Trilogy which introduced the fictional science of psychohistory (where statistics, history and sociology were combined in computer-based models) to guide humanity's future, in order to minimize a potential dark age. How did Asimov conceive of such a thing? (Answer: "Asimov is to sci-fi" as "Bach is to Baroque classical music")

Years ago I became infected with an Asimov inspired optimism about humanity's future and have since felt the need to promote Asimov's vision. While Folding@home will not cure my "infection of optimism", I am convinced Dr. Isaac Asimov (PhD in Biochemistry from Columbia in 1948, then was employed as a Professor of Biochemistry at the Boston University School of Medicine for 10-years until his publishing workload became too large) would have been fascinated by something like this.

I was considering a financial charitable donation to Folding@home when it occurred to me that my money would be better spent by making a knowledgeable charitable donation to all of humanity by:

  1. Increasing my Folding@home computations (which will advance medical discoveries to increase both the length and quality of human life). I was already folding on a half-dozen computers so all I needed to do was purchase video graphics cards which would increase my computational throughput by a thousand-fold (three orders of magnitude).
  2. Convincing others (like you) to follow my example. My solitary folding efforts will have little effect on humanity's future but together we can make a real difference.
 
Dr. Asimov: I am dedicating this website to you and your work. You have greatly influenced my life.

Quick-Navigation Menu

Protein Folding Overview

Science Problem

Misfolded proteins have been associated with numerous diseases and age-related illnesses. However, proteins are much larger and so much more complicated than smaller molecules that it is not possible to begin a chemical experiment without first providing hints to researchers about where to look and what to look for. Since the behavior of atoms-in-molecules (computational chemistry) as well as atoms-between-molecules (molecular dynamics) can be modeled, it makes more sense to begin with a computer analysis. Once that has been completed, permitted configurations can then be passed on to experimental researchers.

Real-world observation

From a kitchen point of view, chicken eggs are a mix of water, fat (yolk), and protein (albumen). Cooking an egg causes the semi-clear protein to unfold into long strings which now can intertwine into a tangled network which will stiffen then scatter light (now appears white). No chemical change has occurred, but taste, volume and color have been altered.

Click here to read a short "protein article" by Isaac Asimov published in 1993 shortly after his death.

Two (of many) TED-Talk Videos:

Computer Solutions

Single CPU Systems

Using the most powerful single core processor (Pentium-4), simulating the folding possibilities of one large protein molecule for one millisecond of chemical time might require one million days (2737 years) of computational time. However, if the problem is sliced up then assigned to 100,000 personal computers over the internet, the computational time drops to ten days. Convincing friends, relatives, and employers to do the same could reduce the computational time requirement to one day or less.
chemical
time
in
nature		 required simulation time
one
computer		 100,000
computers		 1 million
computers
1 Second 1 billion days
(2.7 million years)		 27 years 2.7 years
1 mS 1 million days
(2,737 years)		 10 days 1 day
1 uS 1 thousand days
(2.73 years)		 14.4
minutes		 1.44
minutes
Additional information for science + technology nerds

Not a nerd? Click skip this section

  1. Special-purpose research computers like IBM's BlueGene employ 10 to 20 thousand processors (CPUs) joined by many kilometers of optical fiber to solve problems. IBM's Roadrunner is a similar technology employing both "CPUs" and "special non-graphic GPUs that IBM refers to as cell processors"
     
    1. The basic terms:
      • Early CPUs were built around integer processing with floating point operations being simulated in software.
      • Later CPUs offered builtin support for floating point operations
      • The combined throughput (fetch floating point data from memory, manipulate it, then write it back) is known as a FLOP (FLoating point OPeration) with the total spec being published as FLOPS (FLoating point OPerations per Second)
    2. As of June 2023, the Folding@home project consists of ~ 40,000 active platforms (some hosting 14,000 GPUs) which yield 26,135 TeraFLOPS (26 PetaFLOPS).
    3. Equivalents:
      • A Pentium-4 was rated at 12 GigaFLOPS
        (26 x 10^15) / (12 x 10^9) = 2,167,000 Pentium-4 processors
      • A Core i7 was rated at 100 GigaFLOPS which is 8-times higher than the Pentium-4).
        (26 x 10^15) / (100 x 10^9) = 260,000 Core i7 processors
    4. This means that the original million-day protein simulation problem could theoretically be completed in (1,000,000 / 2,167,000) 0.46 days (or 11 hours). But since there are many more protein molecules than DNA molecules, humanity could be at this for decades. Adding your computers to Folding@home will permanently advance humanity's progress in protein research and medicine.
       
  2. When the Human Genome Project (to study human DNA) was being planned, it was thought that the task may require 100 years. However, technological change in the areas of computers, robotic sequencers, and the internet after the world-wide-web appeared in 1991 (to coordinate the activities of a large number of universities where each one was assigned a small piece of the problem), allowed the human genome project to publish results after only 15 years (a 660% increase).
     
  3. Distributed computing projects like Folding@home and BOINC have only been possible since 1995:
    1. the world-wide-web (proposed in 1989 to solve a document sharing problem among scientists at CERN in Geneva; then implemented in 1991) began to make the internet both popular and ubiquitous.
    2. CISC was replaced with RISC which further evolved to superscalar RISC then multicore.
    3. Vector processing became ubiquitous (primarily) in the form of video cards.

      Processor technology was traditionally defined like this:

      • Scalar (one data stream per instruction. e.g. CISC CPU)
      • Superscalar (1-6 non-blocking scalar instructions simultaneously in a pipeline. e.g. RISC CPU)
      • See: Flynn's Taxonomy for definitions like SISD (single instruction single data) and SIMD (single instruction multiple data) but remember that Data represents "Data stream"
      • See: Duncan's taxonomy for a more modern twist
        Caveat: these lists purposely omit things like SMP (symmetric multiprocessing) and VAX Clusters

      Then CISC and RISC vendors began adding vector processing instructions to their CPU chips which blurred everything:

      • Vector Processing (multiple data streams per instruction)
        • Terminology from math, science and computer science:
          • scalar: any measurement described by one data point (e.g. 30 km/hour)
          • vector: any measurement described by two data points (e.g. 30 km/hour, North)
            • A collection of vectors is usually referred to as a matrix (although a 2-dimensional data structure created in a computer is also known as a matrix; this includes a single spreadsheet as well as a set of database records where the columns represent field names; note that these examples can contain, scalars, vectors, and tensors) 
          • tensor: any item involving three, or more, data points.
        • Vector processing is a generic name for any kind of multi data point math (vector or tensor) performed on a computer.
        • Technological speed up:
          • While it is possible to do floating point (FP) math on integer-only CPUs, adding specialized logic to support FP and transcendental math can decrease FP processing time by one order of magnitude (x10) or more.
          • Similarly, while it is possible to do vector processing (VP) on a scalar machine, adding specialized logic can decrease VP processing time by 2 to 3 orders of magnitude (x100 to x1000).

      Development over the decades:

      1. Mainframe Computers
      2. Minicomputer / Workstation
        1. 1989: DEC (Digital Equipment Corporation) adds vector processing to their Rigel uVAX chip
        2. 1989: DEC adds optional vector processing to VAX-6000 model 400 minicomputer
        3. 1994: VIS 1 (Visual Instruction Set) was introduced into UltraSPARC processors by Sun Microsystems
          • comment: UltraSPARC was a 64-bit implementation of 32-bit SPARC.
        4. 1996: MDMX (MIPS Digital Media eXtension) is released by MIPS.
        5. 1997: MVI (Motion Video Extension) was implemented on the DEC Alpha 21164. MVI appears again in Alpha 21264 and Alpha 21364.
      3. Microcomputer / Desktop
        1. 1997: MMX was implemented on P55C (a.k.a. Pentium 1) from Intel.
          • The first offering introduced 57 MMX-specific instructions.
        2. 1998: 3DNow! was implemented on AMD K-2.
        3. 1999: AltiVec (also called "VMX" by IBM and "Velocity Engine" by Apple) was implemented on PowerPC 4 from Motorola.
        4. 1999: SSE (Streaming SIMD Extensions) was implemented on Pentium 3 "Katmai" from Intel.
          1. This technology employs 128-bit instructions on eight additional registers.
          2. SSE was Intel's reply to AMD's 3DNow!
          3. SSE replaces MMX (both are SIMD, but SSE uses its own floating-point registers)
        5. 2001: SSE2 was implemented on Pentium 4 from Intel
        6. 2004: SSE3 was implemented on Pentium 4 Prescott on from Intel
        7. 2006: SSE4 was implemented on Intel Core and AMD K10
        8. 2008: AVX (Advanced Vector Instructions) proposed by Intel + AMD but not seen until 2011.
          1. many components extended to 256-bits.
        9. 2012: AVX2 (more components extended to 256-bits)
        10. 2015: AVX-512 (512-bit extensions) first proposed in 2013 but not seen until 2015
          1. many components extended to 512-bits.
        Technology Width Year
        MMX 64 bit 1997
        SSE 128 bit 1999
        AVX 256 bit 2008
        AVX-512 512 bit 2015
      4. Add-on graphics cards
        • GPU (graphics programming unit) take vector processing to a whole new level. Why? A $200.00 graphics card now can equip your system with 1500-2000 streaming processors and 2-4 GB of additional high-speed memory. According to the 2013 book "CUDA Programming", the author provides evidence why any modern high-powered PC equipped with one, or more (if your motherboard supports it), graphics cards can outperform any supercomputer listed 12 years ago on www.top500.org
        • Many companies manufactured graphics cards (I recall seeing them available as purchase options in the IBM-PC back in 1981) but I will only mention two companies here:
          • ATI Technologies (founded in 1985)
            • introduces GPU chipsets in the early 1990s that can do video processing without the need for a CPU.
            • introduces the Radeon line in 2000 specifically targeted at DirectX 7.0 3D acceleration.
            • acquired by AMD in 2006.
          • Nvidia (founded in 1993)
            • introduces the GeForce line in 1999.
            • introduces the Tesla line in 2007; these pure-math video cards have no video connector so cannot be connected to a monitor.
            • CUDA is released in 2007.
      5. The circle of life?
        • specialized mainframe computers from companies like IBM and Cray are built to host many thousands of "non-video video cards" (originally targeted for PCs and workstations). IBM's Roadrunner is one example.

      To learn more:

      I've been in the computer industry for decades but noticed that computers only began to get really interesting again with the releases of CUDA (2007) and OpenCL (2009)
  4. Distributed computing projects like Folding@home and BOINC have only been practical since 2005 when the CPUs in personal computers began to out-perform mini-computers and enterprise servers. Partly because...
    1. AMD added 64-bit support to their x86 processor technology calling it x86-64. (Linux distros still refer to this a 686)
    2. Intel followed suit calling their 64-bit extension technology EM64T
    3. DDR2 memory became popular (this dynamic memory is capable of double-data-rate transfers)
      1. Intel added DDR2 support to their Pentium 4 processor line (2002)
      2. AMD added DDR2 support to their Athlon 64 processor line (2006)
    4. DDR3 memory became popular (this dynamic memory is capable of quadruple-data-rate transfers)
       
  5. Since then, the following list of technological improvements has made computers both faster while less expensive:
    1. Intel's abandonment of NetBurst which meant a return to shorter instruction pipelines starting with Core2
      comment: AMD never went to longer pipelines; a long pipeline is only efficient when running a static CPU benchmark for marketing purposes - not running code in real-world where i/o events interrupt the primary foreground task (science in our case)
    2. multi-core (each core is a fully functional CPU) chips from all manufacturers.
    3. continued development of optional graphic cards where CPUs would off-load much work to a graphics co-processor system (each card appeared as many hundreds to thousands of streaming processors)
      1. ATI Radeon graphics cards (ATI was acquired by AMD in 2009)
      2. NVIDIA GeForce graphics cards
      3. development of high performance "graphics" memory technology (e.g. GDDR3 , GDDR4 , GDDR5) to bypass processing stalls caused when streaming processors are too fast.
      4. Note that GDDR5 is used a main memory in the PlayStation 4 (PS4). While standalone PCs were built to host an optional graphics card, it seems that Sony has flipped things so that their graphics system is hosting an 8-core CPU. These hybrids go by the name APU.
    4. shifting analysis from host CPU cores (usually 2-4) to thousands of streaming processors
    5. Intel replacing 20-year old FSB technology with a proprietary new approach called QuickPath Interconnect (QPI) which is now found in Core-i3, Core-i5, Core i7 and Xeon
      Historical note:
      1. DEC created the 64-bit Alpha processor which was first announced in 1992 (21064 was first, 21164, 21264, 21364, came later)
      2. Compaq bought DEC in 1998.
      3. The DEC division of Compaq created CSI (Common System Interface) for use in their EV8 Alpha processor which was never released.
      4. HP merged with Compaq in 2002.
      5. HP preferred Itanium2 (jointly developed by HP and Intel) so announced their intention to gracefully shut down Alpha.
      6. Alpha technology (which included CSI) was immediately sold to Intel.
      7. approximately 300 Alpha engineers were transferred to Intel between 2002 and 2004.
      8. CSI morphed into QPI (some industry watchers say that Intel ignored CSI until the announcement by AMD to go with a new industry-supported technology known as HyperTransport.
    6. The remainder of the industry went with a non-proprietary technology called HyperTransport which has been described as a multi-point Ethernet for use within a computer system.
       
  6. As is true in any "demand vs. supply" scenario, most consumers didn't need the additional computing power which meant that chip manufacturers had to drop their prices just to keep the computing marketplace moving. This was good news for people setting up "folding farms". Something similar is happening today with computer systems since John-q-public is shifting from "towers and desktops" to "laptops and pads". This is causing the price of towers and graphics cards to plummet ever lower. You just can't beat the price-performance ratio of a Core-i7 motherboard hosting an NVIDIA graphics card.
     
  7. Shifting from brute-force "Chemical Equilibrium" algorithms to techniques involving Bayesian statistics and Markov Models will enable some exponential speedups.
     
  8. Computational Chemistry
    Student Questions:
    1. Using information from the periodic table of the elements you can see that the molecular mass of water (H2O) is ~18 which is lighter than many gases so why is water in a liquid state at room temperature while other slightly heavier molecules take the form of a gas?
    2. Ethanol (a liquid) has one more atom of Oxygen than Ethane (a gas). How can this small difference change the state?
    water molecules
    Liquid Water
    This diagram depicts an
    H2O molecule loosely
    bound to four others by
    Van der Walls forces.
    Substance Molecule Atomic
    Masses    		 Molecular
    Mass    		 State at Room
    Temperature
    Water H2O (1x2) + 16 18 liquid
    Carbon Monoxide CO 12 + 16 28 gas
    Molecular Oxygen O2 (16x2) 32 gas
    Carbon Dioxide CO2 12 + (16x2) 44 gas
    Ozone O3 (16x3) 48 gas
             
    Methane CH4 12 + (1x4) 16 gas
    Ethane C2H6 (12x2) + (1x6) 30 gas
    Ethanol C2H6O (12x2) + (1x6) + 16 46 liquid
    Propane C3H8 (12x3) + (1x8) 44 gas
    Butane C4H10 (12x4) + (1x10) 58 gas
    Pentane C5H12 (12x5) + (1x12) 72 gas
    Hexane C6H14 (12x6) + (1x14) 86 liquid
    Heptane C7H16 (12x7) + (1x16) 100 liquid
    Octane C8H18 (12x8) + (1x18) 114 liquid
     
    Short answers:
    1. In the case of an H20 (water) molecule, even though two hydrogen atoms are covalently bound to one oxygen atom, those same hydrogen atoms are also attracted to each other which causes the water molecule to bend into a Y shape (according to VSEPR Theory). At the mid-point of the bend, a positive electrical charge from the oxygen atom is exposed to the world which allows a weak connection to the hydrogen atom of a neighboring H20 molecule (water molecules weakly sticking to each other form a liquid). These weak connections are called Van der Waals forces
    2. Here are the molecular schematic diagrams of Ethane (symmetrical) and Ethanol (asymmetrical). Notice that Oxygen-Hydrogen kink dangling to the right of Ethanol? That kink is not much different than a similar one associated with water. That is the location where a Van der Waal force weakly connects with an adjacent ethanol molecule (not shown). So it should be no surprise that ethane at STP (Standard Temperature and Pressure) is a gas while Ethanol is a liquid.
       Ethane        Ethanol
(symmetrical) (asymmetrical)

  H H           H H   H
  | |           | |  /
H-C-C-H       H-C-C-O
  | |           | |
  H H           H H
    Van der Waals did all his computations using pencil and paper long before the first computer was invented; and this was only possible because the molecules in question were small and few.
     
    Chemistry Caveat: The Molecular Table above was only meant to get you thinking. Now inspect this LARGER periodic table of the elements where the color of the atomic number indicates whether the natural state is solid or gaseous:
    1. all elements in column 1 (except hydrogen) are naturally solid.
    2. all elements in column 8 (helium to radon) are naturally gaseous.
    3. half the elements in row 2 starting with Lithium (atomic number 3) and ending with Carbon (atomic number 6), as well as two thirds of row 3 starting with Sodium (atomic number 11) and ending with Sulfur (atomic number 16), are naturally solid. I will leave it to you to determine why.
       
  9. Molecular Dynamics
    Proteins come in many shapes and sizes. Here is a very short list:
    Protein Mass Function Notes
    Chlorophyll a 893 facilitates photosynthesis in plants  
    Heme A 852 common ligand for many hemeproteins including hemoglobin and myoglobin  
    Alpha-amylase 56,031 salivary enzyme to digest starch pdbId=1SMD
    hemoglobin 64,458 red blood cell protein  
    DNA polymerase varies from
    50k to 200k    		 enzyme responsible of DNA replication  
    These molecules are so large that modeling their interactions can only be done accurately with a computer.
certificate

FAH Links

https://foldingathome.org cover page
https://foldingathome.org/home/ home page
https://foldingathome.org/start-folding/ download page
https://foldingforum.org technology problems, discussions, news, science, etc.

FAH Targeted Diseases

This "folding knowledge" will be used to develop new drugs for treating diseases such as:

Reference Links: Folding@home - FAQ Diseases

More Information About Proteins and Protein-Folding Science

Online Documents

My Computational Statistics (with help from China)

Folding via GPUs (graphics cards)

  • A single core Pentium-class CPU provides one scalar processor. It also provides one streaming (vector) processor under marketing names like MMX (64-bit) , SSE (128-bit), AVX (256-bit) and AVX-512 (512-bit)
  • Most Core i5 and Core i7 CPU's provide four scalar cores so offer at least four streaming (vector) processors.
  • A single add-on graphics card can provide hundreds to thousands vector processors (the Nvidia RTX-3090 provides 10,496)
Computer technology section moved here

Folding with Linux (2019 - 2024)

click:  Linux tips + real world problems and solutions

Folding with EL7 (2019)

caveat: GPU folding on CentOS-7 failed December-2021 and is no longer possible so: jump to 2022

I found two junk PCs in my basement with 64-bit CPUs that were running 32-bit operating systems (Windows-XP and Windows-Vista). Unfortunately for me, neither were eligible for Microsoft's free upgrade to Windows-10, and I had no intention of buying a new 64-bit OS just for this hobby. So I swapped out Windows with CentOS-7 and was able to get each one folding with very little effort. What follows are some tips for people who are not Linux gurus:
  1. Download a DVD ISO image of CentOS-7 via this link: https://www.centos.org/download/
    • CentOS-7.7 and CentOS-8.0 were released days apart in September 2019 (perhaps due to the invisible hand of IBM?)
    • Software from the top of download page preferentially offers CentOS-8 which is too large (> 4.7 GB) to write to a single-layer writable DVD, but I have has some success with Dual Layer optical media. Linux distros assume you will copy these images to a USB stick, but the BIOS in many older machines does not support booting from a thumb drive.
  2. Transfer to bootable media (choose one of the following)
    • copy the ISO image to a DVD-writer
      -OR-
    • use rufus to format an USB stick then copy the ISO image to the USB
      caveat: newer PCs have transitioned from BIOS to UEFI. Older BIOS-based systems do not support booting from a USB stick (strange because you can connect a USB-based DVD then boot from that)
  3. Boot-install CentOS-7 on the 64-bit CPU
    • Using a larger downloaded image:
      1) burn, boot, install Linux using recipe: "GUI with Server"
      2) reboot; now update via the internet like so: yum update
      3) reboot;
4) add development tools: yum group "Development Tools" install
    • Using smaller downloads (then adding GUI after the fact):
      1) burn, boot, install Linux using recipe: "Server"
      2) reboot; now update via the internet like so: yum update
      3) reboot; optionally enable GUI: yum group "Server with GUI" install
         may also need to type:
            systemctl isolate     graphical.target
            systemctl set-default graphical.target
4) add development tools: yum group "Development Tools" install
  4. My machines hosted NVIDIA graphics cards (GTX-560 and GTX-960 respectively) so these systems required the correct NVIDIA drivers in order to do GPU-based folding. Why? The generic drivers only support video but folding science requires OpenCL and CUDA
    • HARD WAY: If you are a Linux guru and know how to first disable the Nouveau driver, then install the 64-bit Linux driver provided by NVIDIA here: http://www.nvidia.com/Download/index.aspx
    • EASY WAY: First install the nvidia-detect module found at elrepo (Enterprise Linux REPOsitory) here ( http://elrepo.org/tiki/tiki-index.php ) and documented here ( https://elrepo.org/tiki/nvidia-detect ) then use the utility to install another elrepo package. The steps look similar to this if you are logged in as root:

      step-1 (install one/two repos via yum)
      activity Linux command Notes Description
      view available repos yum list \*-release\*
      		backslash escapes the asterisk
      install elrepo yum install elrepo-release required Enterprise Linux REPO
      install epel yum install epel-release optional Extra Packages for Enterprise Linux

      step-2       (install nvidia-detect from elrepo)
      activity Linux command
      install nvidia-detect yum install nvidia-detect
      test nvidia-detect nvidia-detect -v
      install nvidia driver yum install $(nvidia-detect)
      reboot reboot
      If you are not logged in as root then you must prefix every command with sudo (Super User DO).
      Now jump to Linux common

Folding with EL8 (2022)

Both of my CentOS-7 machines stopped folding in Dec-2021. Apparently this is due to several changes by the FAH research team.
  • First off, new FAH downloadable cores require an updated version of Linux library glibc which is not available on CentOS-7, so you need to upgrade to CentOS-8, or any other EL8 (see caveat ~ 6 lines below)
  • Secondly, changes to the FAH GPU cores now require a minimum of OpenCL-1.2. This means that my GTX-560 is no longer useful as a streaming processor. I noticed another blurb about double-precision FP math which definitely rules out my GTX-560 so I replaced it with a GTX-1650
  • On both systems I replaced CentOS-7 with CentOS-8 (via a fresh install) then followed the CentOS-7 instructions just above. However, the Nvidia driver from elrepo-release (EL8) did not contain any OpenCL support so I was forced to install the Linux driver provided by Nvidia (you first need to disable the Nouveau driver which requires some Linux guru voodoo). If doing a fresh reinstall is not possible then you should consider using the ELevate - leapp migration tool published by AlamaLinux
  • For some reason, these CentOS-8 machines seem sluggish, so I jumped to Rocky Linux 8.5 which solved all my problems.
  • Caveat: IBM purchased Red Hat in 2019 for the sum of US$34 billion. In 2020 the transformed Blue Hat announced their intent CentOS-8 minor point updates after 2021-12-31 because too many companies (like Facebook) were using this free software (gotta speed up the ROI on that $34 billion investment). Even though Red Hat begins each OS iteration using open-source software, they announced on 2023-06-21 that they were going to restrict access (close source) their modifications. This will affect all EL (Enterprise Linux) flavors including: AlmaLinux, EuroLinux, Oracle Linux, Rocky Linux, etc. so you might consider moving to any target supported by this migration tool (ELevate - leapp). Note that academics working at Fermilab just outside of Chicago, or CERN (home of the LHC) in Geneva, recommend moving to AlmaLinux. My advice is to always employ a Linux that can be downloaded from a university mirror. Here is one example of 10,000: https://mirror.csclub.uwaterloo.ca/
    Mirror observations:
    • active offerings for AlmaLinux
    • no CentOS updates after CentOS-8.5
    • no offerings for CentOS-stream, Oracle Linux, EuroLinux, Rocky Linux
    Real world concern: you never know when problems (political, commercial, technological) will block you from getting to a single site. But you can always modify "/etc/yum.conf" to point directly to a nearby university mirror.
  1. Updating to a NVIDIA published driver
    • after the initial Linux install, type "sudo yum update" to bring the platform up to the latest level. If a new kernel was installed then you must reboot before you continue.
    • DO NOT use elrepo to update the Nvidia driver (on 2022-01-16 it was missing support for OpenCL and CUDA)
    • read these reference notes:
    • steps:
      1. Only download the desired driver file from NVIDIA into the root account
        1. for GTX960 I used: NVIDIA-Linux-x86_64-470.86.run (will probably be renamed before you read this)
        2. for GTX1650 I used: NVIDIA-Linux-x86_64-525.89.run (will probably be renamed before you read this)
      2. Now disable the Nouveau driver
        • create file /etc/modprobe.d/blacklist-nouveau.conf containing these two lines:
          blacklist nouveau
          options nouveau modeset=0
        • create a new ramdisk for use during system boot: dracut --force
        • reboot
        • caveat: at this point your monitor is no longer capable of displaying small text in character-cell mode but you don't care because you've already downloaded the necessary files from Nvidia corporation. Right?
      3. Install the NVIDIA driver
        • yum group "Development Tools" install
        • chmod 777 NVIDIA-Linux-x86_64-470.86.run
        • ./NVIDIA-Linux-x86_64-470.86.run
          • caveat: the previous command might fail for the following reasons:
            • not an executable file (did you use chmod ?)
            • no build tools (gcc compiler, etc. Did you install "Development Tools"?)
            • nouveau driver is running
        • reboot
    • caveats:
      • kernel updates via "yum update" (CentOS-7) or "dnf update" (CentOS-8) always require a reboot. Ninety percent of the time you will need to reinstall the Nvidia driver after a kernel update (just repeat step 3 above).
      • If your console is blank then type this three-key-combo CTL-ALT-F3 (control alternate F3) then log in as root on the third virtual console.

Common to all Linux installs

  1. Linux software an be downloaded from here:
  2. Installing Folding-at-home on Linux
    • some of the GUI-based software may not work properly (on CentOS-7 or earlier) because Python3 is a prerequisite
      • Python3 can be easily added to EL7 systems like CentOS-7 but do not remove or disable Python2 because that action will break certain system utilities like yum or firewall-cmd to name two of many. 
        sudo yum install epel-release
        sudo yum install python36
      • Python-3.6.8 is the standard offering in EL8 offerings like RHEL-8, AlmaLinux-8, etc.
    • So I recommend:
    • Caveat: just installing the FAH-Client will cause it to be installed as a service then start CPU folding (which is probably what you do not want). If you want to enable GPU-based folding (and/or disable CPU folding) then you will need to stop the client, modify the config file, test the config file, then restart the client. Here are some commands to help out.
      task command
      stop service sudo systemctl stop fahclient
          or
      sudo /etc/init.d/FAHClient stop
      navigate here cd /etc/fahclient
      edit config file

      (nano does not
      require VI or
      VIM skills)      		 nano config.xml
      <config>
       <power v='full'/>
       <user v='neil_rieck'/>
       <team v='10987'/>
       <gpu v='true'/>
       <smp v='true'/>
       <slot id='0' type='CPU'/>
       <slot id='1' type='GPU'/>
      </config>

      notes:
      1) type can be one of: 'cpu', 'smp', 'gpu'
      2) never enable 'smp' unless to have >= 4 cores
      3) with 'gpu' enabled I see no point in folding with 'cpu' or 'smp'
      navigate here cd /var/lib/fahclient
      (optional) see
      other switches      		 /usr/bin/FAHClient --help
      interactive testing /usr/bin/FAHClient --config /etc/fahclient/config.xml -v

      note: GPU folding requires OpenCL-1.2 so if you see errors like
      'cannot find OpenCL' then you might need to rebuild your NVIDIA
      driver (almost always required after any Linux update that
      replaces the kernel
      end interactive test hit: <ctrl-c>
      start the service systemctl start fahclient
          or
      sudo /etc/init.d/FAHClient start &
      monitor step #1
      monitor step #2
      monitor step #3      		 top -d 0.5
      cat      /var/lib/fahclient/logs.txt 
      tail -40 /var/lib/fahclient/logs.txt 

Microsoft Windows Scripting and Programming

  1. MS-DOS/MSDOS Batch Files: Batch File Tutorial and Reference
  2. MS-DOS @wikipedia
  3. Batch file @wikipedia
  4. Microsoft Windows XP - Batch files

Experimental Stuff for Windows Hackers

DOS commands for creating, and starting, a Windows Service to execute a DOS script.

sc create neil369 binpath="cmd /k start c:\folding-0\neil987.bat" type=own type=interact
sc start neil369

Once created, you can stop/start/modify a service graphically from this Windows location:

Start >> Programs >> Administrative Tools >> Services

Stopped services may only be deleted from DOS like so:

sc query neil369
sc delete neil369
 
BOINC

BOINC (Berkeley Open Infrastructure for Network Computing)

BOINC (Berkeley Open Infrastructure for Network Computing) is a science framework in which you can support one, or more, projects of choice.
  • If you are unable to pick a single cause then pick several because the BOINC manager will switch between science clients every hour (this interval is adjustable). In my case I actively support POEM, Rosetta, and Docking.
  • The current BOINC client can be programmed to use one, some, or all cores of a multi-core machine. The BOINC client can also utilize (or not) the streaming processors on your Graphics Card.

Protein / Biology / Medicine Projects

General

POEM@home (via BOINC)

Rosetta@home (via BOINC)

World Community Grid (via BOINC)

Personal Comment: I wonder why HP (Hewlett-Packard) has not followed IBM's lead. Up until now I always thought of IBM as the template of uber-capitalism but it seems that the title of "king of profit by the elimination of seemingly superfluous expenses" goes to HP. Don't they realize that IBM's effort in this area is done under IBM's advertising budgets? Just like IBM's 1990s foray into chess playing systems (e.g. Deep Blue) led to increased sales as well as share prices, one day IBM will be able to say "IBM contributed to a treatments for human diseases including cancer". IBM actions in this area reinforce the public's association with IBM and information processing.

Biology Science Links

Protein Data Bank Links

ENCODE

The Encyclopedia of DNA Elements (ENCODE) Consortium is an international collaboration of research groups funded by the National Human Genome Research Institute (NHGRI). The goal of ENCODE is to build a comprehensive parts list of functional elements in the human genome, including elements that act at the protein and RNA levels, and regulatory elements that control cells and circumstances in which a gene is active.

Local Links

(noteworthy) Remote Links

Recommended Biology Books (I own them all)

Back to Home
Neil Rieck
Waterloo, Ontario, Canada.