{"id":7997,"date":"2026-07-08T20:08:08","date_gmt":"2026-07-09T01:08:08","guid":{"rendered":"https:\/\/justinparrtech.com\/JustinParr-Tech\/?p=7997"},"modified":"2026-07-08T22:03:33","modified_gmt":"2026-07-09T03:03:33","slug":"data-centers-in-space-is-a-stupid-idea","status":"publish","type":"post","link":"https:\/\/justinparrtech.com\/JustinParr-Tech\/data-centers-in-space-is-a-stupid-idea\/","title":{"rendered":"Data Centers in Space is a STUPID Idea"},"content":{"rendered":"<p>Building Data Centers in space is stupid.\u00a0 Here&#8217;s why&#8230;<\/p>\n<p><!--more--><\/p>\n<p>&nbsp;<\/p>\n\n<p>&nbsp;<\/p>\n<div id=\"ez-toc-container\" class=\"ez-toc-v2_0_85 counter-hierarchy ez-toc-counter ez-toc-custom ez-toc-container-direction\">\n<div class=\"ez-toc-title-container\"><p class=\"ez-toc-title\" style=\"cursor:inherit\">Table of Contents<\/p>\n<\/div><nav><ul class='ez-toc-list ez-toc-list-level-1 ' ><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-1\" href=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/data-centers-in-space-is-a-stupid-idea\/#what-is-a-data-center\" >What is a Data Center?<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-2\" href=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/data-centers-in-space-is-a-stupid-idea\/#designing-a-data-center\" >Designing a Data Center<\/a><ul class='ez-toc-list-level-3' ><li class='ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-3\" href=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/data-centers-in-space-is-a-stupid-idea\/#location\" >Location<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-4\" href=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/data-centers-in-space-is-a-stupid-idea\/#property-orientation-and-layout\" >Property Orientation and Layout<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-5\" href=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/data-centers-in-space-is-a-stupid-idea\/#main-building-layout\" >Main Building Layout<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-6\" href=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/data-centers-in-space-is-a-stupid-idea\/#all-about-racks\" >All About Racks<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-7\" href=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/data-centers-in-space-is-a-stupid-idea\/#anatomy-of-a-rack-mount-server\" >Anatomy of a Rack-Mount Server<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-8\" href=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/data-centers-in-space-is-a-stupid-idea\/#high-density-compute\" >High-Density Compute<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-9\" href=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/data-centers-in-space-is-a-stupid-idea\/#data-center-floor-layout\" >Data Center Floor Layout<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-10\" href=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/data-centers-in-space-is-a-stupid-idea\/#data-center-tiers\" >Data Center Tiers<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-11\" href=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/data-centers-in-space-is-a-stupid-idea\/#nut-shell\" >Nut Shell<\/a><\/li><\/ul><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-12\" href=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/data-centers-in-space-is-a-stupid-idea\/#space-data-centers\" >Space Data Centers<\/a><ul class='ez-toc-list-level-3' ><li class='ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-13\" href=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/data-centers-in-space-is-a-stupid-idea\/#a-quick-note-on-direct-current-power-distribution\" >A Quick Note on Direct Current Power Distribution<\/a><\/li><\/ul><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-14\" href=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/data-centers-in-space-is-a-stupid-idea\/#problems-with-space-data-centers\" >Problems With Space Data Centers<\/a><ul class='ez-toc-list-level-3' ><li class='ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-15\" href=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/data-centers-in-space-is-a-stupid-idea\/#cost-is-a-problem\" >Cost is a Problem<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-16\" href=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/data-centers-in-space-is-a-stupid-idea\/#life-span-and-life-cycle-are-both-huge-problems\" >Life Span and Life Cycle are both HUGE Problems<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-17\" href=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/data-centers-in-space-is-a-stupid-idea\/#form-factor-is-a-design-problem\" >Form Factor is a Design Problem<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-18\" href=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/data-centers-in-space-is-a-stupid-idea\/#cooling-is-a-huge-problem\" >Cooling is a HUGE Problem<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-19\" href=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/data-centers-in-space-is-a-stupid-idea\/#atmosphere-and-humidity-lack-thereof-is-a-problem\" >Atmosphere and Humidity (Lack Thereof) is a Problem<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-20\" href=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/data-centers-in-space-is-a-stupid-idea\/#some-equipment-wont-survive-launch\" >Some Equipment Won&#8217;t Survive Launch<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-21\" href=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/data-centers-in-space-is-a-stupid-idea\/#maintenance-is-a-huge-problem\" >Maintenance is a Huge Problem<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-22\" href=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/data-centers-in-space-is-a-stupid-idea\/#physical-resiliency-is-a-huge-problem\" >Physical Resiliency is a HUGE Problem<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-23\" href=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/data-centers-in-space-is-a-stupid-idea\/#high-energy-particles-%e2%80%9ccosmic-radiation%e2%80%9d-are-a-huge-problem\" >High-Energy Particles (&#8220;Cosmic Radiation&#8221;) are a HUGE Problem<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-24\" href=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/data-centers-in-space-is-a-stupid-idea\/#bandwidth-is-a-problem\" >Bandwidth is a Problem<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-25\" href=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/data-centers-in-space-is-a-stupid-idea\/#power-is-not-a-problem\" >Power is NOT a Problem<\/a><\/li><\/ul><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-26\" href=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/data-centers-in-space-is-a-stupid-idea\/#conclusion\" >Conclusion<\/a><\/li><\/ul><\/nav><\/div>\n<h2><span class=\"ez-toc-section\" id=\"what-is-a-data-center\"><\/span>What is a Data Center?<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p>Since &#8220;Launching AI data centers in to space&#8221; is an idea, we need to discuss why it&#8217;s a terrible idea.<\/p>\n<p>First, let&#8217;s talk about what a data center is.\u00a0 I think most people understand the concept, but here is a brief explanation.<\/p>\n<p>A data center is the place where servers &#8220;live&#8221; and run.<\/p>\n<p>What is a &#8220;server&#8221;?<\/p>\n<p>A server is a heavy-duty personal computer (PC) with lots of memory and processors.\u00a0 When referring to a PC, some people call the &#8220;box&#8221; a &#8220;CPU&#8221; or &#8220;processor&#8221;, but that&#8217;s not correct.\u00a0 &#8220;The box&#8221; is PC, and the CPU (Central Processing Unit) is the main chip inside the PC.\u00a0 In fact, there&#8217;s lots of stuff inside a PC:<\/p>\n<ul>\n<li>CPU<\/li>\n<li>Memory<\/li>\n<li>Hard Drive (storage)<\/li>\n<li>Power Supply<\/li>\n<li>Graphics card (maybe)<\/li>\n<li>Network \/ wireless card (maybe)<\/li>\n<\/ul>\n<p>All of this stuff (and more) is connected to the motherboard.<\/p>\n<p>A typical PC might have:<\/p>\n<ul>\n<li>1 CPU<\/li>\n<li>16 gig of RAM<\/li>\n<li>Maybe 2 hard drives &#8211; a Solid State Drive (SSD &#8211; basically a big memory stick) to boot and run, and perhaps a larger drive for archival storage<\/li>\n<li>PC cost:\u00a0 $500 to $1,500 per unit.<\/li>\n<\/ul>\n<p>However a typical server has much higher specs:<\/p>\n<ul>\n<li>2, 4, or 8 physical CPUs<\/li>\n<li>256 gig of RAM<\/li>\n<li>Multiple drives in a RAID configuration.<\/li>\n<li>Multiple network controllers<\/li>\n<li>Server cost:\u00a0 $10,000 to $30,000 per unit.<\/li>\n<\/ul>\n<p>RAID (Redundant Array of Inexpensive Disks) allows multiple hard drives to act as a single drive while providing fault tolerance.<\/p>\n<p>In a server, network controllers are used to communicate on the network, but can also be used to connect the server to a Storage Area Network (SAN), which is a networked disk array.<\/p>\n<p>Everything in a server is highly fault-tolerant through redundancy.\u00a0 For example, servers can be configured with 2 to 3 power supplies, multiple CPU boards, and multiple network controllers.\u00a0 This redundancy allows a server to lose one or more components due to failure and still operate, but perhaps in a degraded state.\u00a0 For example, if a power supply fails, the remaining power supply (or power supplies) continues to power the system.\u00a0 In addition, most servers have hot-swappable drives and power supplies, allowing the server to be repaired using replacement parts without ever shutting it down.<\/p>\n<p>As you might imagine, extra CPUs, memory, and hard drives, as well as the fans that are required for extra cooling, all use way more power than a normal PC.\u00a0 For example, a laptop (which is a PC) might use 40 Watts of power, and a high-end gaming PC might use 1000 Watts, but a typical server can easily suck down 1300 Watts or more.\u00a0 In fact, a single high-end server can consume more Watts than an electric dryer.<\/p>\n<p>As the number of servers increases, the problem is compounded.\u00a0 You need LOTS of power, and you need a way to carry the resulting heat away from every server so that they don&#8217;t fry themselves.<\/p>\n<p>Which is where a data center enters the scene.<\/p>\n<p>Data centers are designed to house hundreds or even thousands of servers, while supplying power, cooling, and network access.\u00a0 Like a server, the data center itself is designed to be highly-redundant, and therefore fault-tolerant.\u00a0 Data centers typically have multiple, redundant paths for network access (internet or private network, or both), redundant power sources, usually with battery and generator backup, and redundant cooling throughout.<\/p>\n<ul>\n<li>If the data center loses &#8220;street power&#8221; from the power company, the Automatic Transfer Switch (ATS) automatically starts the generator or generators, and then switches from street power to generator.\u00a0 Meanwhile, the servers continue running on battery power during the entire event.\u00a0 Later, when street power is restored, the ATS switches back and shuts down the generator(s).<\/li>\n<li>Data centers typically have what is known as &#8220;diverse egress&#8221; for network connections.\u00a0 If one of them is accidentally cut by road construction (happens more often than you might think), the data center still has its other connection(s) to rely on.<\/li>\n<li>In order to cool hundreds or thousands of servers, data centers have multiple HVAC cooling systems, so that if a few of them fail, the remaining cooling capacity is sufficient to prevent what&#8217;s known as &#8220;thermal cascade&#8221;, where the other HVAC units can&#8217;t keep up with demand, heat builds dramatically, and servers end up failing due to thermal damage.<\/li>\n<\/ul>\n<p>Beyond network, power, and cooling, data centers also provide:<\/p>\n<ul>\n<li><strong>Physical security<\/strong>.\u00a0 Data centers are typically staffed with armed guards to prevent physical intrusion, but a good data center is also built like a bunker, with no or minimal windows, concrete and steel walls, perimeter fences with security gates, and internal security controls such as <a href=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/how-to-kill-a-terminator\/#man-traps\" target=\"_blank\" rel=\"noopener\">man-traps<\/a> (not as cool as it sounds), scanners, sensors, lasers, cameras, electronic door locks, keypads, fingerprint readers, and some of the other cool stuff you see in movies.<\/li>\n<li><strong>Physically resiliency<\/strong>.\u00a0 Although most data centers are NOT designed to withstand a missile attack, MOST are designed to be tornado-resistant, fire-resistant (think California wildfires), flood-resistant, hurricane-resistant, and freeze\/blizzard-resistant.<\/li>\n<li><strong>Fire suppression<\/strong>.\u00a0 If something catches fire inside a data center, which CAN and HAS happened, it must be addressed quickly and effectively.\u00a0 Data centers have special fire suppression systems that are designed to detect and extinguish a fire with minimal damage to the surrounding equipment.\u00a0 Consider, if a server or power supply were to catch on fire, you can&#8217;t just hose it down with water &#8211; this would be catastrophic to nearby equipment.\u00a0 Therefore specialized sensors and fire suppression gear is required.<\/li>\n<\/ul>\n<p>At a nominal $20k per server, it would suck to have someone drive up a truck in the middle of the night, and steal a bunch of them.\u00a0 Even worse, the real reason data centers have such tight security, is to protect the <em>data<\/em> on the hard drives.\u00a0 If someone steals a bag full of hard drives, the potential damage is WAY worse than stealing the equivalent cost in servers, because this could result in the disclosure of bank records, healthcare data, customer lists, trade secrets, and who knows what else.<\/p>\n<p>Likewise, if the data is lost or destroyed, perhaps due to theft, fire, attack, or natural disaster, there could be irreparable harm to the data&#8217;s owners.<\/p>\n<p>Therefore, the main reason that data centers are hardened is to protect the\u00a0<em>data<\/em> rather than the servers which process the data.<\/p>\n<p>&nbsp;<\/p>\n<h2><span class=\"ez-toc-section\" id=\"designing-a-data-center\"><\/span>Designing a Data Center<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p>A LOT goes in to planning a data center build, and in order to really understand space data centers, we need to understand those complexities and challenges.<\/p>\n<p>Although everything is based on requirements which vary by use case, anything in Financial Services or Healthcare sectors have some fairly stringent requirements, and I will assume those as a baseline when describing data centers and data center operations.\u00a0 A &#8220;general-purpose&#8221; data center should be PCI certified, and compliant with GLBA as well as HIPPA.<\/p>\n<ul>\n<li>PCI (Payment Card Industry) is a credit card industry standard which defines and requires physical hardening as well as strong protections for payment card and transaction data.<\/li>\n<li>GLBA (Graham-Leach-Bliley Act) is a law which applies to the Financial Services industry, and requires physical hardening as well as strong protections for bank data.<\/li>\n<li>HIPPA (Health Information Portability and Privacy Act) is a law which applies to the Healthcare industry, and requires strong protections for healthcare data.<\/li>\n<\/ul>\n<p>If any tenant within a specified data center zone plans to process credit card transactions, that entire data center zone must be PCI certified.\u00a0 Likewise, if a tenant gathers, stores, or processes bank data, there are data center-wide requirements for physical hardening and other forms of physical security.\u00a0 And, if a tenant gathers, stores, or processes healthcare data, some or all of the equipment within the corresponding zone may fall in to scope for the corresponding HIPPA security requirements.<\/p>\n<p>If the data center is intended to host mission-critical applications such as telemetry and support for space missions, or real-time monitoring of patient vitals, the availability requirements are way more stringent.\u00a0 Other use cases, such as hosting government-classified data may require much more hardening.\u00a0 Likewise, a data center hosting only &#8220;my pictures of cats dot com&#8221;, which has no impact to personal, financial, or healthcare data, and certainly doesn&#8217;t host any trade secrets nor government secrets might have much less stringent availability and hardening requirements.\u00a0\u00a0And, obviously, lower-end requirements have a much lower associated cost, while higher-end requirements have a significantly higher cost.<\/p>\n<h3><span class=\"ez-toc-section\" id=\"location\"><\/span>Location<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<p><strong>But, as they say, it all starts with location, and location is everything.<\/strong><\/p>\n<p>These are <em>some<\/em> of the planning considerations when picking a data center location:<\/p>\n<ul>\n<li>Geographical threat analysis.\n<ul>\n<li>Is the proposed location near any airports, factories, refineries, or other industrial facilities that might cause a hazard of fire, explosion, or chemical release?<\/li>\n<li>Is the proposed location near any natural hazards, such as a forest that could burn, or a lake or river that could flood?<\/li>\n<li>Is the proposed location near any tourist attractions, arenas, concert halls, amusement parks, or similar?\u00a0 The thought process is that these could be targets for mass shooting and \/ or other forms of terrorism.\u00a0 (Not kidding &#8211; this is ACTUALLY a question that has come up during a DC risk-assessment)<\/li>\n<li>Is the proposed location prone to natural disasters (flood, tornado, hurricane, blizzard) and in-land enough to survive a tidal wave?<\/li>\n<\/ul>\n<\/li>\n<li>Should be close enough to one or more major roads which facilitates delivery of generator fuel and data center equipment, but far enough from all major roads that it wouldn&#8217;t be affected by a major crash or accident.\u00a0 Many data centers are built on a slight plateau or employ dirt berms if they are too close to a freeway, for example.<\/li>\n<li>Needs to be near one or more power substations.\u00a0 If not, the power company will build a substation near your data center, and charge you for it.<\/li>\n<li>Needs to accommodate network access from various providers, as needed.\n<ul>\n<li>The data center can act as a peering point for multiple providers<\/li>\n<li>The data center can act as a junction on one or more providers&#8217; networks<\/li>\n<li>The data center can be geographically near a peering point or provider junction, allowing for the high bandwidth \/ connectivity requirements<\/li>\n<li>If the proposed location is too far away from an existing colo\/peering point or junction, obtaining fully-redundant, high-bandwidth access may be challenging, and may require that the provider &#8220;back haul&#8221; a fiber-optic circuit, or preferably multiple fiber-optic circuits from their closest junction to the data center&#8217;s location.\u00a0 The longer the back-haul, the more it&#8217;s going to cost.\u00a0 And the same is true for each provider you add to the data center.<\/li>\n<\/ul>\n<\/li>\n<li>Local \/ City \/ State zoning, permits, and regulations.<\/li>\n<li>A data center should be built on a larger piece of land, compared to other commercial real-estate, because you need room for other logistical considerations:\n<ul>\n<li>Perimeter walls and fences<\/li>\n<li>Berms for resiliency and ditches for water drainage<\/li>\n<li>Generator pads<\/li>\n<li>Security gates and barriers<\/li>\n<li>Secure loading area near the loading dock<\/li>\n<li>Access for water and refueling trucks<\/li>\n<li>Segregated parking<\/li>\n<li>Emergency vehicle access<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n<p>Despite being designed with resiliency in mind, stuff happens.\u00a0 What special equipment and training are required for data center staff responding to things like:<\/p>\n<ul>\n<li><strong>Exterior fire<\/strong>: What if a generator fuel tank catches on fire?\u00a0 What if there is an uncontrolled wilderness fire (A.K.A. California)?\u00a0 What if a car in the parking lot catches fire?<\/li>\n<li><strong>Interior fire<\/strong>: What if there is a fire in the battery room (happens more often than you would think)?\u00a0 What if a piece of equipment catches on fire?\u00a0 What if a fire starts in the break room?<\/li>\n<li><strong>Flood<\/strong>:\u00a0 Despite geographic location, water tanks can leak, and hurricanes are a thing.\u00a0 How do you keep water out of the building?\u00a0 After water enters the building, how do you keep it away from the equipment?<\/li>\n<li><strong>Inclement Weather \/ Extended Freeze<\/strong>:\u00a0 What if the roads are buried in snow or covered in ice?\u00a0 How do you rotate staff?\u00a0 How do you get fueling trucks in and out?\u00a0 Do you have cots, provisions, and showers for trapped data center staff?\u00a0 Do you have enough generator fuel for an extended freeze?\u00a0 Is there a way for Emergency Response units to reach the data center?\u00a0 For example, could a helicopter land on the roof, in the event that staff evacuation is required?<\/li>\n<li><strong>Medical Emergency<\/strong>:\u00a0 Some aspects of data center operations can be dangerous.\u00a0 In a worst-case scenario, data center operations staff could get cut, burned, electrocuted, poisoned, asphyxiated, crushed, pinched, stabbed, or suffer a fall or have limbs broken.\u00a0 Addressing these types of injuries go WAY beyond the limits of a typical first-aide kit.\u00a0 Does your data center have AED&#8217;s, burn kits, oxygen tanks, and trauma kits?\u00a0 Are staff trained to use them?\u00a0 Is there express entry for fire and medical?<\/li>\n<\/ul>\n<p>Any special equipment or gear, and some aspects of the building itself may require extra land and special permits.<\/p>\n<h3><span class=\"ez-toc-section\" id=\"property-orientation-and-layout\"><\/span>Property Orientation and Layout<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<p>Once you have enough property in the correct location, the next thing to consider is <strong>how that property is situated and oriented near a main road<\/strong>.<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-large wp-image-8042\" src=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Datacenter-property-orientation-600x412.png\" alt=\"\" width=\"600\" height=\"412\" srcset=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Datacenter-property-orientation-600x412.png 600w, https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Datacenter-property-orientation-300x206.png 300w, https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Datacenter-property-orientation.png 706w\" sizes=\"auto, (max-width: 600px) 100vw, 600px\" \/><\/p>\n<p>The data center should be located about 1\/2 mile from the nearest main road, with its front facing the side road.\u00a0 If there are nearby risks, such as a busy road or active businesses, a dirt berm outside the main wall is recommended.\u00a0 A dirt berm is a good way to mitigate damage due to crashes, accidents, fires, bullets, and also deters trespassing.<\/p>\n<p>The property layout, as well as the data center itself, is designed around &#8220;defense-in-depth&#8221;, which means that the data center as well as the property around it implements many layers of security, rather than one single &#8220;impenetrable&#8221; security layer.<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter wp-image-8041 size-full\" src=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Datacenter-property-layout.png\" alt=\"\" width=\"710\" height=\"619\" srcset=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Datacenter-property-layout.png 710w, https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Datacenter-property-layout-300x262.png 300w, https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Datacenter-property-layout-600x523.png 600w\" sizes=\"auto, (max-width: 710px) 100vw, 710px\" \/><\/p>\n<p>Data centers may implement some or all of the following features:<\/p>\n<ul>\n<li><strong>Exterior, steel-reinforced masonry wall<\/strong>:\u00a0 A 8&#8242; to 10&#8242; wall around the sides and back of the main building provides significant exterior resilience against crashes, fires, bullets, etc.\u00a0 If allowed by local law, this wall is often topped with razor wire in order to deter trespassers and thieves.<\/li>\n<li><strong>Wrought-iron \/\/ Steel fences<\/strong>:\u00a0 A 8&#8242; iron or steel fence is typical for the front portion of the property.\u00a0 Although some data centers simply implement a masonry wall on all four sides, the theory behind a see-through fence is that it deters snooping and trespassing.\u00a0 If all you can see from the street is a wall, natural curiosity leads you to see what&#8217;s behind it.\u00a0 However, if you can SEE through the fence, and all that you can see is a parking lot and a building behind it, you are less likely to want to snoop.\u00a0 By showing you that there is literally nothing to see, you are less likely to look.<\/li>\n<li><strong>Line of trees near the road<\/strong>:\u00a0 Trees are a natural, resilient, physical barricade, but also offer partial privacy.\u00a0 By adding street appeal while allowing line of sight from the road, psychologically, it makes the building less-interesting.<\/li>\n<li><strong>Two property entrances<\/strong>:\u00a0 Sometimes, there is a front and a rear, or in other cases both entrances connect to the same road.\u00a0 There is usually a &#8220;U-shaped&#8221; driveway surrounding the main building, connecting both entrances.\u00a0 In some cases, both entrances are used for both ingress and egress, but my opinion is that one should be primarily ingress, and the other should be primarily egress.\u00a0 However, in the case of a gate failure or other blockage, both entrances can serve both purposes.<\/li>\n<li><strong>Encircling driveway<\/strong>:\u00a0 A driveway encircles the building, allowing access to the loading dock(s), as well as allows fuel, water, and maintenance trucks to access the generator pads.\u00a0 The driveway is far enough from the building that a vehicle fire won&#8217;t damage the building or equipment, and such that a disabled or parked vehicle can be bypassed.<\/li>\n<li><strong>Sliding gates<\/strong>:\u00a0 A sliding gate is implemented at each entrance.\u00a0 Access is granted via a security kiosk for both ingress and egress.\u00a0 The security kiosks allow badged entry for regular employees, and allows visitors to contact the security desk at the entrance vestibule.\u00a0 In order to get in that gate, you must have a badge or you must be on the visitor list.\u00a0 Otherwise, you are turned away.\u00a0 If you refuse to leave, they lock the gate down and call the cops.\u00a0 Most data centers require that you also must swipe your badge to leave the property, and guests to contact the security desk again.\u00a0 This is done to prevent an attacker from triggering a &#8220;REX&#8221; (Request to Exit) sensor, then using the exit as an entrance.<\/li>\n<li><strong>Gate Arms<\/strong>:\u00a0 Once you get past the perimeter gates, gate arms allow you access to the parking lot, loading dock, and generator pads.\u00a0 All of this is monitored by the security desk, where a guard raises the arm(s) when needed.<\/li>\n<li><strong>Spur fences or concrete bollards<\/strong>:\u00a0 Spur fences and \/ or concrete bollards block traffic from the driveway to the entrance, and from the parking lot to the entrance.\u00a0 Concrete bollards can be disguised as decorative planters.<\/li>\n<li><strong>Emergency Staging Area<\/strong>:\u00a0 I&#8217;ve been to two data centers that have part of the parking lot sectioned off&#8230; 40&#215;40 foot to be used as an emergency helicopter landing pad, or police \/ fire \/ medical emergency staging.\u00a0 A &#8220;true&#8221; landing pad needs an additional clearance of 40 feet on each side (120 x 120 feet), not depicted.<\/li>\n<li><strong>Generator Pads<\/strong>:\u00a0 Toward the back of the building, generator pads house diesel generators used for power redundancy, along with fuel tanks, and also water tanks depending on the type of cooling employed inside the data center.\u00a0 Each generator pad is surrounded by an additional steel-reinforced masonry wall to protect against fire \/ explosion.<\/li>\n<li><strong>There are very few windows<\/strong>:\u00a0 The security vestibule may have glass doors, and there might be some offices at the front of the building with small windows.\u00a0 Other than that, there are no windows anywhere else.\u00a0 Not only is UV light bad for most electronic equipment, but a window is a hole in an otherwise-secure wall.\u00a0 In many cases, data centers use clear acrylic resin rather than glass.<\/li>\n<\/ul>\n<h3><span class=\"ez-toc-section\" id=\"main-building-layout\"><\/span>Main Building Layout<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<p>As you\u00a0 arrive at the data center, you turn in to the property and pull up to the security kiosk.\u00a0 You press the button, and talk to the guard.\u00a0 You state your name and your reason for being there.\u00a0 After a few seconds, the gate opens, and the guard tells you to turn left in to the parking lot.\u00a0 You pull forward, and the guard raises the gate arm, allowing you to turn left in to the parking lot.\u00a0 You find a parking spot, and walk toward the data center entrance, passing a giant &#8220;H&#8221; painted inside a 40&#215;40 foot square in the center of the parking lot. As you open the door to the security vestibule, you notice that the doors and windows are thick, but not heavy.\u00a0 On your left is the security desk behind thick glass, which you assume to be bullet-resistant, as well as the security guard that you spoke to a few moments ago.<\/p>\n<p>Here is the layout of a typical 20,000 to 40,000 square foot data center.<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter wp-image-8040 size-full\" src=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Datacenter-layout.png\" alt=\"\" width=\"739\" height=\"634\" srcset=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Datacenter-layout.png 739w, https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Datacenter-layout-300x257.png 300w, https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Datacenter-layout-600x515.png 600w\" sizes=\"auto, (max-width: 739px) 100vw, 739px\" \/><\/p>\n<p>Generally, the building is designed with the least-secure areas toward the front of the building, with increasing levels of security as you move toward the back.<\/p>\n<p>At the front, there is the security vestibule, where you must sign in and state the reason for access, your finger print will be scanned, and you will receive a guest badge that must be returned at the end of your visit.\u00a0 You enter the outer hallway through a man trap.\u00a0 Sometimes the vestibule itself is a man trap, in other cases, the man trap sits between the vestibule and the hallway.\u00a0 A &#8220;man trap&#8221; is not as cool as it sounds.\u00a0 It&#8217;s simply a set of interlocking doors, such that only one door an be opened at a time.<\/p>\n<p>Past the man trap is a common hallway.\u00a0 Every room beyond this point requires some combination of PIN, password, badge, and fingerprint in order to gain access.<\/p>\n<p>The front of the building usually houses low-security areas, such as:<\/p>\n<ul>\n<li>Offices<\/li>\n<li>Break room<\/li>\n<li>Storage<\/li>\n<li>Equipment staging areas<\/li>\n<li>Access to the loading dock<\/li>\n<\/ul>\n<p>Despite being a very secure area, Data center operations (DC Ops) is also usually near the front of the building.\u00a0 DC Ops is a very secure room where the workers conduct monitoring and operations of the data center.\u00a0 DC Ops is usually very secure, because there are monitors and controls inside that operate:<\/p>\n<ul>\n<li>All HVAC units<\/li>\n<li>Street power, generators, power switch gear, batteries, inverters, and power distribution<\/li>\n<li>Heat detectors, fire detection and suppression gear, which includes life\/safety devices such as strobe lights and Emergency Power Off (EPO) switches.<\/li>\n<li>Cameras, door sensors, motion sensors, etc. (note that the security desk usually has access to these as well)<\/li>\n<li>Network and infrastructure equipment<\/li>\n<\/ul>\n<p>DC Ops is basically responsible for keeping everything running.\u00a0 In addition to monitoring for and responding to events, DC Ops also:<\/p>\n<ul>\n<li>Racks \/ Un-racks equipment (install \/ remove equipment)<\/li>\n<li>Perform module-level repairs, such as replacing hard drives, power supplies, network interfaces, etc<\/li>\n<li>Perform hardware repairs, such as replacing CPU \/ memory and other internal server components<\/li>\n<li>Assist data center clients with various tasks, such as troubleshooting, power equipment on\/off, and cabling adjustments<\/li>\n<\/ul>\n<p>As you move inward, the data center <strong>floor space is segmented in to &#8220;zones&#8221; or &#8220;data halls&#8221;<\/strong> that are each like a building within a building.\u00a0 Each zone:<\/p>\n<ul>\n<li>Has it&#8217;s own HVAC and air handlers, essentially making each zone air-isolated from all other zones.\u00a0 This is done so that smoke, dust, or other particulate contaminants originating from one zone can&#8217;t affect the other zones.<\/li>\n<li>Has its own power distribution.\u00a0 Although every zone leverages facility-wide redundant power, each zone has feeds from both redundant rails, and power distribution within the zone provides every rack with both rails, and then further distribution within each rack supplies redundant power to each piece of equipment.<\/li>\n<li>Has isolated heat monitoring, fire detection and suppression.\u00a0 Heat monitoring allows DC Ops to monitor hot spots before they elevate to the level of a fire risk.\u00a0 If needed, DC Ops can remotely power off an entire rack and proactively engage fire prevention or suppression measures.<\/li>\n<li>Each zone is physically hardened, where the inner walls are brick or masonry with embedded wire mesh.\u00a0 Not only are the walls designed to resist ingress by force, but they are internally and externally fire-resistant.<\/li>\n<li>Doors are heavy-duty wood or steel, fire resistant with electronic strikes.\u00a0 In a normal office environment, you might notice that secure doors use a big electromagnet, and when you badge in you hear a &#8220;ping&#8221; when the electromagnet releases the door.\u00a0 These can be forced open.\u00a0 Data doors use electronic strikes which are normally locked, where the strike completely blocks the door latch.\u00a0 When the user badges in, the strike is retracted by a solenoid.\u00a0 Therefore, the doors stay locked even during a power failure.<\/li>\n<li>Every door has access control, and requires a badge swipe.\u00a0 Least-secure areas such as offices and break room may require ONLY a badge swipe.\u00a0 Other areas may require badge+PIN, and secure areas require badge+PIN+fingerprint.\u00a0 Egress from any area usually requires a badge swipe as well.<\/li>\n<li>In modern data centers, your badge is also used for RFID-based location tracking.\u00a0 This is done to ensure that visitors stay out of unauthorized areas, but also so that you can be located in an emergency.<\/li>\n<\/ul>\n<p>There are typically\u00a0<strong>four types of zones<\/strong>:<\/p>\n<ul>\n<li><strong>Staging zones<\/strong> are usually located near the loading dock.\u00a0 These are low-security zones designed to offload and assemble equipment on its way to being installed in the data center, or temporarily store equipment waiting to leave the data center.<\/li>\n<li><strong>Common zones<\/strong> are usually multi-tenant zones.\u00a0 In the case where clients own their own hardware, sometimes these spaces are partitioned in to cages, which prevent one client from accessing another client&#8217;s hardware.<\/li>\n<li><strong>Critical zones<\/strong> have unusual high-availability requirements, such as hospital systems, or data center-wide infrastructure.\u00a0 These zones are more secure and prioritized over common zones.<\/li>\n<li><strong>Shared \/ Colocation zones<\/strong> are areas where network providers can house network switches and cross-connect with other network providers.\u00a0 Don&#8217;t pay attention to the blue box at the bottom of some of the racks.\u00a0 That&#8217;s the alleged NSA internet spying device.<\/li>\n<\/ul>\n<p>At the back of the facility are:<\/p>\n<ul>\n<li><strong>Battery rooms<\/strong>.\u00a0 Each side of the building has a battery room and corresponding generator pad.\u00a0 Equipment runs from the battery bank, basically a giant stack of car batteries, which is connected to an inverter that simulates 220V AC line power.\u00a0 Feeding the batteries is a transfer switch (ATS) that connects either grid power (&#8220;street power&#8221;) or generator power.\u00a0 If both are down, the batteries can supply power for a few hours.\u00a0 Incoming power is conditioned via a power conditioner that sits between the batteries and the ATS.\u00a0 Each battery room feeds a &#8220;power rail&#8221;, and there are two battery rooms, thus two power rails.\u00a0 Every piece of equipment in the facility is connected to both rails.<\/li>\n<li><strong>Telecom demarcation<\/strong>.\u00a0 \u00a0 The demarcation room is where all network connections &#8220;terminate&#8221; inside the data center facility.\u00a0 From there, patch panels and cross-connects are used to route various network connections throughout the facility.\u00a0 Inside the demarcation room are equipment such as ONT (Optical Network Terminator), distribution frames, patch panels, and similar.\u00a0 Network connections run underneath the floor in cabling trenches, or over the ceiling on &#8220;ladders&#8221; that carry cables from one zone to another.<\/li>\n<\/ul>\n<h3><span class=\"ez-toc-section\" id=\"all-about-racks\"><\/span>All About Racks<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<p>Inside the data center floor space, equipment is stored in 19&#8243; equipment racks.<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-full wp-image-8047\" src=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Racks-features.png\" alt=\"\" width=\"595\" height=\"542\" srcset=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Racks-features.png 595w, https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Racks-features-300x273.png 300w\" sizes=\"auto, (max-width: 595px) 100vw, 595px\" \/><\/p>\n<p>Racks are 2 feet wide, 3 feet deep, and about 8 feet tall.<\/p>\n<p>Equipment racks, along with all rack-mount equipment are designed for front-to-back cooling, where cool air is ingested in the front, cools the equipment inside, and exhausted from the rear.<\/p>\n<p>Equipment racks have four mounting posts, 2 front and 2 rear, that are divided in to 42 &#8220;rack units&#8221;.\u00a0 Each piece of equipment is like a kitchen sliding drawer that mounts inside.<\/p>\n<p>Racks have removable sides, as well as removable front and rear doors.<\/p>\n<p>There are punch-out panels for wire management on the sides and top.<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-full wp-image-8048\" src=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Racks-holes.png\" alt=\"\" width=\"566\" height=\"370\" srcset=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Racks-holes.png 566w, https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Racks-holes-300x196.png 300w\" sizes=\"auto, (max-width: 566px) 100vw, 566px\" \/><\/p>\n<p>An equipment rack is divided in to &#8220;Rack Units&#8221; (RU) that are 1.75&#8243; tall, separated by 0.5&#8243;.\u00a0 Each RU has a top and bottom hole, and a center hole.\u00a0 The top and bottom holes are used for rails, while the center hole is used to secure the actual device.<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-full wp-image-8049\" src=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Racks-mounting-a-server.png\" alt=\"\" width=\"595\" height=\"254\" srcset=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Racks-mounting-a-server.png 595w, https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Racks-mounting-a-server-300x128.png 300w\" sizes=\"auto, (max-width: 595px) 100vw, 595px\" \/><\/p>\n<p>To rack-mount ( &#8220;rack&#8221; ) a server, you install the rails first, between the front and rear mounting posts.\u00a0 Then you install the server itself.<\/p>\n<ul>\n<li>Install the rails via the top\/bottom holes<\/li>\n<li>Install the server in to the rails, secured by the middle hole<\/li>\n<li>Install drives and power supplies<\/li>\n<li>Connect power and network<\/li>\n<\/ul>\n<p>Most servers use rails that look like a kitchen drawer slider, but heavy-duty.\u00a0 When maintenance is required, you unscrew the center screw, and slide the server forward.\u00a0 You lift the top panel to gain access to the server&#8217;s interior.<\/p>\n<p>Some servers offer rails that &#8220;snap in&#8221; to an equipment rack, requiring no tools.<\/p>\n<p>Small devices such as routers and firewalls have &#8220;mounting wings&#8221; that bolt to the side of the device, and use the top and bottom mounting holes on the front posts<\/p>\n<blockquote><p><em>Anecdote:<\/em><br \/>\nWhen I was a consultant in the 90&#8217;s, many of my clients had &#8220;sagging&#8221; equipment, where they mounted a 1U switch, router, or firewall using the center mounting hole, and therefore the top or bottom mounting holes on the bracket didn&#8217;t line up.\u00a0 When using the top holes of the mounting tabs, this allows downward force to deflect the mounting bracket.\u00a0 The fix?\u00a0 Move the device up 1\/2 RU.\u00a0 And, when using only two mounting holes, mount from the bottom.\u00a0 When forced down by gravity, the top holes are cantilevered against the rack&#8217;s mounting posts.<\/p><\/blockquote>\n<h3><span class=\"ez-toc-section\" id=\"anatomy-of-a-rack-mount-server\"><\/span>Anatomy of a Rack-Mount Server<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<p>In a &#8220;standard compute&#8221; model, a standard 42U rack contains 30 to 40 individual servers, along with network, power distribution, and anything else a healthy, growing server might need.<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-large wp-image-8046\" src=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Rack-mount-server-600x405.png\" alt=\"\" width=\"600\" height=\"405\" srcset=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Rack-mount-server-600x405.png 600w, https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Rack-mount-server-300x202.png 300w, https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Rack-mount-server-768x518.png 768w, https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Rack-mount-server.png 811w\" sizes=\"auto, (max-width: 600px) 100vw, 600px\" \/><\/p>\n<p>A rack-mount server&#8217;s case is like a drawer that slides in and out of the rack once it&#8217;s rails are mounted.\u00a0 The server has a sheet metal lid that can be removed once the server slides forward, out of the rack.\u00a0 At the front of the server chassis is where hot-pluggable drives are installed.\u00a0 At the rear are hot-pluggable power supplies, plus network, USB and other ports that are directly attached to the backplane.\u00a0 A server&#8217;s &#8220;backplane&#8221; is like the motherboard for a normal PC, but a backplane has no CPU or memory.\u00a0 The backplane has connectors for CPU boards as well as the PCI riser, where normal PCI cards can be installed.\u00a0 Each CPU board has a physical CPU, plus its associated memory.\u00a0 For AI workloads (which is why we are eventually going to discuss space data centers), the PCI slots would have GPU boards installed.<\/p>\n<p>So, conservatively, one whole rack of &#8220;standard compute&#8221; has around 70 CPUs and about 140 GPUs.<\/p>\n<h3><span class=\"ez-toc-section\" id=\"high-density-compute\"><\/span>High-Density Compute<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<p>In order to get compute densities above &#8220;standard&#8221;, you have to use &#8220;server blades&#8221;.<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-large wp-image-8051\" src=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Server-vs-Blade-600x393.png\" alt=\"\" width=\"600\" height=\"393\" srcset=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Server-vs-Blade-600x393.png 600w, https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Server-vs-Blade-300x196.png 300w, https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Server-vs-Blade.png 681w\" sizes=\"auto, (max-width: 600px) 100vw, 600px\" \/><\/p>\n<p>A server &#8220;blade&#8221; is a giant card that has CPUs, Memory, and GPU all in one package.\u00a0 Multiple blades &#8211; around 10 or so &#8211; slide in to a blade chassis, and the blade chassis slides in to the rack, the same as a single server.<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-large wp-image-8043\" src=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/High-Density-Compute-600x459.png\" alt=\"\" width=\"600\" height=\"459\" srcset=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/High-Density-Compute-600x459.png 600w, https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/High-Density-Compute-300x230.png 300w, https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/High-Density-Compute.png 677w\" sizes=\"auto, (max-width: 600px) 100vw, 600px\" \/><\/p>\n<p>Networking and power are handled by dedicated modules elsewhere in the rack, and storage is handled via Storage Area Network (SAN).\u00a0 In a high-density compute model, one rack houses 200 or more CPUs and 400 or more GPUs, or nearly triple the density of a &#8220;standard&#8221; rack.<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-large wp-image-8050\" src=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/SAN-600x390.png\" alt=\"\" width=\"600\" height=\"390\" srcset=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/SAN-600x390.png 600w, https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/SAN-300x195.png 300w, https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/SAN-768x500.png 768w, https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/SAN.png 864w\" sizes=\"auto, (max-width: 600px) 100vw, 600px\" \/>&#8216;<\/p>\n<p>As mentioned, all of these servers use Storage Area Network (SAN), rather than local hard drives.\u00a0 Each SAN array has many hard drives that are grouped in to RAID arrays.\u00a0 These RAID volumes are further partitioned in to logical volumes, and logical volumes are connected to each server via the SAN.\u00a0 Each server might have a small boot drive, but once running, stores all of its data to the SAN.<\/p>\n<p>The SAN is also capable of managing &#8220;storage snapshots&#8221;, which is a way of creating a point-in-time backup of any volume, and is also a convenient way for the backup system to back up each server:<\/p>\n<ul>\n<li>A snapshot is created<\/li>\n<li>The backup system copies the snapshot<\/li>\n<li>The snapshot is removed<\/li>\n<\/ul>\n<h3><span class=\"ez-toc-section\" id=\"data-center-floor-layout\"><\/span>Data Center Floor Layout<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<p>On the data center floor, racks are arranged very specifically.<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-large wp-image-8052\" src=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Typical-Rack-Layout-600x331.png\" alt=\"\" width=\"600\" height=\"331\" srcset=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Typical-Rack-Layout-600x331.png 600w, https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Typical-Rack-Layout-300x165.png 300w, https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Typical-Rack-Layout.png 619w\" sizes=\"auto, (max-width: 600px) 100vw, 600px\" \/><\/p>\n<p>Alternate rows of racks face eachother, while the back of each row faces the back of the next row.\u00a0 Since every rack cools from front to back, this forms cold aisles where the fronts face eachother, and hot aisles where the backs face eachother.\u00a0 This scheme depends on the data center&#8217;s cooling ability, to remove heat from all of the hot aisles.\u00a0 This can be accomplished in a number of ways, such as HVAC return vents above hot aisles.<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-large wp-image-8045\" src=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Rack-Layout-WRONG-600x196.png\" alt=\"\" width=\"600\" height=\"196\" srcset=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Rack-Layout-WRONG-600x196.png 600w, https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Rack-Layout-WRONG-300x98.png 300w, https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Rack-Layout-WRONG.png 619w\" sizes=\"auto, (max-width: 600px) 100vw, 600px\" \/><\/p>\n<p>If all racks faced front to back, this causes thermal cascade, where the air is heated more as it passes through each rack, and the last rack in line has insufficient cooling to prevent hardware damage.<\/p>\n<p>In addition to thermal considerations, every piece of equipment in the data center has what&#8217;s known as a &#8220;maintenance clearance&#8221;, which is the area around it required in order to provide normal maintenance.\u00a0 Typically maintenance clearances are required for removable panels, or to provide room to slide a component out of the main housing.\u00a0 Racks typically have a 3 foot clearance front and back, which corresponds to the rack depth but is also ADA compliant.<\/p>\n<h3><span class=\"ez-toc-section\" id=\"data-center-tiers\"><\/span>Data Center Tiers<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<p>Data Centers are divided up in to tiers based on redundancy and recovery capabilities.<\/p>\n<p>Normally, we would think of &#8220;tier 1&#8221; as the best, but actually it&#8217;s the worst.\u00a0 A tier 1 data center only has one data path and one power path with no redundancy.<\/p>\n<p>Data centers go up to tier 4, where tier &#8220;n&#8221; is &#8220;n&#8221; redundant power and network paths.\u00a0 Most data centers are tier 2 or tier 3.\u00a0 I&#8217;ve been in a tier 4 data center ONCE.<\/p>\n<h3><span class=\"ez-toc-section\" id=\"nut-shell\"><\/span>Nut Shell<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<p>In a nut shell, everything discussed above is a factor or consideration when designing and operating a data center.<\/p>\n<p>&nbsp;<\/p>\n<h2><span class=\"ez-toc-section\" id=\"space-data-centers\"><\/span>Space Data Centers<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p>Now that we know how Earth data centers work, let&#8217;s talk about how &#8220;space data centers&#8221; would work, especially in the context of AI.<\/p>\n<p>The first limiting factor is payload.\u00a0 You can only send as much stuff in to space at a time that your rocket can carry in a single launch.<\/p>\n<p>Looking at the latest Falcon 9 spec sheet, it can carry 25 tons to Low Earth Orbit (LEO) or about 10 tons to Geo Transfer Orbit (requires higher velocity, thus more acceleration, thus way more fuel). Depending on what you are trying to accomplish, LEO is designed for a &#8220;constellation&#8221; of satellites that make rapid passes across the Earth&#8217;s surface, multiple times per hour.\u00a0 For example, GPS is deployed in LEO because it doesn&#8217;t matter WHICH satellites you can see at any given time.\u00a0 We have to assume that the goal of a data center would be geosynchronous orbit, because a data center that&#8217;s only accessible part of the time is only useful part of the time.\u00a0 So that limits our payload to about 10 tons.<\/p>\n<p>Standard compute weighs about 100lbs per RU, which is 4500 to 5000 lbs for a fully-populated rack, or about 2.5 tons.\u00a0 However, the high-density compute used by AI is like 2.5 times that!\u00a0 If we conservatively guess about 2x weight, that would be about 10,000lbs or about 5 tons per rack for high-density compute, meaning that each trip can carry about 2 racks&#8217; worth of equipment.\u00a0 Batteries, also, are quite dense and would require multiple trips.<\/p>\n<p>Therefore, the best approach would be to build the shell based on a modules.\u00a0 Each &#8220;module&#8221; would be based on 5m diameter x 15m tall (about the payload size of the Falcon 9), and might possibly unfold or some slight assembly may be required once in orbit.\u00a0 Subsequent trips would populate racks within the modules, and it&#8217;s assumed that each rack is pre-populated with server equipment.<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-large wp-image-8098\" src=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Space-Module-Rack-Layout-600x211.png\" alt=\"\" width=\"600\" height=\"211\" srcset=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Space-Module-Rack-Layout-600x211.png 600w, https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Space-Module-Rack-Layout-300x105.png 300w, https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Space-Module-Rack-Layout-768x270.png 768w, https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Space-Module-Rack-Layout.png 888w\" sizes=\"auto, (max-width: 600px) 100vw, 600px\" \/><\/p>\n<p>In a round module, assuming maintenance clearance for both the front and back of each rack, you could stack 4 racks on &#8220;top&#8221; of eachother (depicted side-by-side above), in a circular configuration.\u00a0 This would yield about 12 racks per tier, 4 tiers deep, or about 48 racks.<\/p>\n<p>One of the problems with stacking square things inside a round container is the massive amount of wasted space.\u00a0 Therefore, in contrast, a square (ish) module could easily hold double the amount of racks, and therefore double the equipment in about the same volume.\u00a0 Again, the corners could easily unfold during deployment.<\/p>\n<p>Guessing about the configuration, a &#8220;space data center&#8221; might have multiple modules tied together.<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-large wp-image-8100\" src=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Space-Module-Layout-600x310.png\" alt=\"\" width=\"600\" height=\"310\" srcset=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Space-Module-Layout-600x310.png 600w, https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Space-Module-Layout-300x155.png 300w, https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Space-Module-Layout.png 624w\" sizes=\"auto, (max-width: 600px) 100vw, 600px\" \/><\/p>\n<p>Power modules in the center would be dedicated to batteries and other power-switching equipment, with &#8220;compute&#8221; modules attached on the outsides.<\/p>\n<p>I think the objective would be tier 2, with full redundancy for every component and module.<\/p>\n<h3><span class=\"ez-toc-section\" id=\"a-quick-note-on-direct-current-power-distribution\"><\/span>A Quick Note on Direct Current Power Distribution<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<p>Most equipment in a data center runs on Alternating Current (AC 120V to 240V).\u00a0 However, power in a data center comes in from the power company (or generator) as AC, then passes through a rectifier\/conditioner, which converts it to Direct Current (DC), where it&#8217;s stored in batteries.\u00a0 On demand, electricity is pulled from the batteries (DC), where it passes through an inverter which converts it back to AC.\u00a0 As the AC power enters each device, it&#8217;s power supply converts it BACK to DC!<\/p>\n<p>We could just run all of the equipment on direct (conditioned) DC!<\/p>\n<p>This depends on distance from the power distribution gear, because DC voltage drops as a function of cable length.\u00a0 However, in a relatively compact configuration, everything can be driven from direct DC.\u00a0 In fact, some data centers do this today, and some server \/ equipment manufacturers even offer 48VDC power supplies as an option.<\/p>\n<p>This is like 15% to 20% more efficient than converting back and forth multiple times, and creating unwanted heat in the process.<\/p>\n<p>&nbsp;<\/p>\n<h2><span class=\"ez-toc-section\" id=\"problems-with-space-data-centers\"><\/span>Problems With Space Data Centers<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p>Now that we&#8217;ve looked at what an &#8220;Earth&#8221; data center is, and can speculate about how a &#8220;space&#8221; data center would work, let&#8217;s look at all the problems that need to be solved.<\/p>\n<h3><span class=\"ez-toc-section\" id=\"cost-is-a-problem\"><\/span>Cost is a Problem<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<p>The first and largest problem is cost.<\/p>\n<p>For the purposes of putting a stake in the ground, let&#8217;s assume each compute module would be equivalent to about 1,000 square feet.\u00a0 Here&#8217;s how we arrive at this:<\/p>\n<ul>\n<li>In our model above, we deployed 96 racks per compute module.<\/li>\n<li>Each rack is 2&#8242; x 3&#8242;, plus another 2&#8242; x 3&#8242; (average) for maintenance clearances (which are also cooling clearances, by the way)<\/li>\n<li>Therefore 96 racks plus their maintenance clearances would require 1,152 (or about 1000) square feet<\/li>\n<\/ul>\n<p>A modern, 20,000 sq-ft data center costs $20M to $40M, and would be the equivalent of about 20 &#8220;space data center&#8221; compute modules.<\/p>\n<p>ONE Falcon 9 launch costs $75M.<\/p>\n<p>We can make some assumptions and then extrapolate:<\/p>\n<ul>\n<li>Each compute module requires the shell, plus 96 racks of equipment<\/li>\n<li>We can assume that other supporting gear, such as power distribution, network connectivity, and cooling is baked in to the racks, the module, or both<\/li>\n<li>Assuming that the max weight load is accurate (10 tons) and around 4 tons per rack, that&#8217;s 2.5 racks per trip<\/li>\n<li>The per-module total would be around 40 launches for the equivalent of 1,000 square feet.<\/li>\n<\/ul>\n<p>Additional launches&#8230;let&#8217;s be conservative and say 100 additional launches would be required for batteries (being VERY generous)<\/p>\n<p>Grand total:<\/p>\n<ul>\n<li>40 launches per module x 20 modules = 800 launches.<\/li>\n<li>100 additional launches for batteries<\/li>\n<li>900 TOTAL launches times 75 million = $67 BILLION<\/li>\n<\/ul>\n<p>$20 MILLION on Earth = $67 BILLION in space.\u00a0 Even if we double the Earth cost and cut the space cost in half, that&#8217;s still 3 orders of magnitude of cost difference.<\/p>\n<h3><span class=\"ez-toc-section\" id=\"life-span-and-life-cycle-are-both-huge-problems\"><\/span>Life Span and Life Cycle are both HUGE Problems<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<p>An Earth-based data center facility&#8230; the building itself&#8230; practically lasts forever, but we can put a number on it, for example 100 years.\u00a0 The building itself is made out of concrete that will last for as long as it&#8217;s properly maintained.\u00a0 Buildings such as the Roman Colloseum are proof of that.\u00a0 And during its useful life, a data center building can be completely gutted and retrofitted with modern technology as often as necessary without incurring 100% renovation cost.\u00a0 It might be like 20% renovation cost.<\/p>\n<p>$40M plus three $10M renovations = $70M total cost over 100 years (or whatever).<\/p>\n<p>In contrast, a giant, glorified metal shack in space is going to degrade over time.\u00a0 Every bump, ding, or micro-meteorite reduces its useful life, just like owning a car.\u00a0 Likewise, any data center facility is one technological innovation away from obsolescence, but unlike an Earth-based data center, it would be cheaper to rebuild than re-fit.<\/p>\n<p>So let&#8217;s be generous and cap the space facility itself at 20 years&#8217; life span.\u00a0 This would bring us to the point where aggregate damage plus some technological innovation makes the old facility obsolete, and it would now be cheaper to build new than re-fit \/ retrofit.<\/p>\n<p>The next issue is equipment life cycle.\u00a0 Any server equipment ages in dog years.\u00a0 7 is OLD, 10 isn&#8217;t unheard of, but unlikely.\u00a0 And yes, if well maintained, and assuming you can still find parts for it, you can run a server way beyond 10 years.\u00a0 However, because technological innovation continuously improves the efficiency of new devices and equipment, you can always get double the RAM, double the CPU, double the GPU every 3 years (Moore&#8217;s Law).\u00a0 So a 10-year-old machine provides 1\/8 the computing capability of an equivalent modern machine.\u00a0 Eventually, it COSTS TOO MUCH to run the old machine, assuming you can even find parts to keep it running that long.<\/p>\n<p>However, with AI using high-performance compute, servers age out at about 3 years.\u00a0 It&#8217;s not that the hardware is no good, it&#8217;s just too expensive to operate.<\/p>\n<p>Therefore, you have to have a plan for 100% equipment replacement over 3 to 4 years.\u00a0 So that&#8217;s another ~200 launches per year, just to keep the equipment current, whether that&#8217;s launching new modules or replacing server racks with newer ones.<\/p>\n<h3><span class=\"ez-toc-section\" id=\"form-factor-is-a-design-problem\"><\/span>Form Factor is a Design Problem<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<p>Unless you design your own equipment and a custom rack system, EVERY piece of data center equipment is designed to be mounted within a 19&#8243; rack.\u00a0 This means that your data center MUST be designed around a standard 3&#8242; x 2&#8242; x 8&#8242; form factor.<\/p>\n<p>Further, the racks and the devices within them are designed to have maintenance clearance in front and back to perform installation \/ repair \/ service \/ removal.<\/p>\n<h3><span class=\"ez-toc-section\" id=\"cooling-is-a-huge-problem\"><\/span>Cooling is a HUGE Problem<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<p>On Earth, data centers are cooled by some combination of air and water.\u00a0 Maintaining pressurization in an unmanned data center through the course of its useful life is extremely difficult.\u00a0 Therefore, we have to assume that the cooling strategy is NOT going to use atmosphere.<\/p>\n<p>On Earth, air and water carry heat, which is known as convection.\u00a0 All modern servers and other data center equipment are cooled by either air or liquid convection.<\/p>\n<ul>\n<li>CPUs, memory, and other components turn electricity in to heat, minus the actual work done.<\/li>\n<li>Conduction transfers heat energy from the surface of the component in to the surrounding air (or liquid, if liquid-cooled).<\/li>\n<li>Convection carries hot air \/ liquid out of the device by pumping cool air in to the front of the device, resulting in hot air exiting the rear of the device.<\/li>\n<\/ul>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-full wp-image-8039\" src=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Cooling-in-Space-Questionmark.png\" alt=\"\" width=\"595\" height=\"389\" srcset=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Cooling-in-Space-Questionmark.png 595w, https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Cooling-in-Space-Questionmark-300x196.png 300w\" sizes=\"auto, (max-width: 595px) 100vw, 595px\" \/><\/p>\n<p>However, in space, without atmosphere, the only way to eliminate heat is for that heat to be radiated as infrared light.\u00a0 And without air flow, that heat energy just bounces around inside the server&#8217;s case.<\/p>\n<p>Moreover, racks can develop &#8220;hot spots&#8221; if one piece of equipment is &#8220;boxed in&#8221;.<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-large wp-image-8044\" src=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Hot-Spots-600x482.png\" alt=\"\" width=\"600\" height=\"482\" srcset=\"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Hot-Spots-600x482.png 600w, https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Hot-Spots-300x241.png 300w, https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-content\/uploads\/Hot-Spots.png 621w\" sizes=\"auto, (max-width: 600px) 100vw, 600px\" \/><\/p>\n<p>Hot spots can propagate to nearby equipment, and equipment in nearby racks.\u00a0 Therefore, cooling via direct radiation isn&#8217;t feasible.<\/p>\n<p>Liquid cooling is more thermally-efficient than air, but can only cool specific components unless the entire server is submerged in oil or something.<\/p>\n<p>Heat pipes are always an option &#8211; heat pipes are sealed and use a combination of oil and inert gas to &#8220;move&#8221; the heat from components such as a CPU to the exterior of a device, where a heat sink could radiate some of the heat.\u00a0 However, even heat pipes are dependent upon some kind of exterior cooling in order to work efficiently.<\/p>\n<p>So&#8230;how do you cool stuff in space without atmosphere?\u00a0 Or even worse, how do you MAINTAIN atmosphere for 10 to 20 years?<\/p>\n<h3><span class=\"ez-toc-section\" id=\"atmosphere-and-humidity-lack-thereof-is-a-problem\"><\/span>Atmosphere and Humidity (Lack Thereof) is a Problem<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<p>This is not commonly known, but most Earth-based electronics are not designed to run in the vacuum of space, and might not run at all.\u00a0 Most devices don&#8217;t give a crap about gravity, despite what you saw in the movie, &#8220;Hail Mary&#8221;.\u00a0 HOWEVER, air is a dielectric.\u00a0 On earth, air permeates every device, providing an insulating barrier that prevents arcing and shorting between electrical traces on circuit boards.\u00a0 To run reliably in a vacuum, an electronic device has to be specially designed to prevent this.<\/p>\n<p>Further, many high-speed \/ high-frequency devices can&#8217;t run properly or reliably in extremely low humidity.\u00a0 You would think that data centers are kept as dry as possible, but the reality is that most equipment won&#8217;t operate reliably under about 5% humidity.\u00a0 In fact, most data centers are kept at around 15%, and some have to employ humidifiers if the regional, ambient air is too dry.<\/p>\n<h3><span class=\"ez-toc-section\" id=\"some-equipment-wont-survive-launch\"><\/span>Some Equipment Won&#8217;t Survive Launch<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<p>Launching in to orbit is a violent process.\u00a0 It&#8217;s pretty much the opposite of being driven somewhere in the back of a Mercedes.\u00a0 Everything is under extreme acceleration for a long duration, and subject to extreme bouncing and shaking, which is a byproduct of travelling at high-velocity through the atmosphere.<\/p>\n<p>Conversely, data center equipment is designed to be delivered by truck, unpacked, racked, and sit there motionless (more or less) until it gets decommissioned at some future point.<\/p>\n<p>I&#8217;ve done data center moves where we ship the equipment on a truck &#8211; nothing remotely similar to a rocket hurtling through the atmosphere at 9g &#8211; and servers had loose memory, loose PCI cards, and even loose CPU boards.\u00a0 Hard drives are held in by clips, but I&#8217;ve seen those shake loose.<\/p>\n<p>What&#8217;s the moral of the story?<\/p>\n<p>Either fabricate custom mounting clips for every component, or be sure to re-seat every component in every server after entering a stable orbit, and prior to deployment.<\/p>\n<p>Further, Earth data centers don&#8217;t move (relative to the equipment).\u00a0 Space data centers may need corrective trajectory adjustments, which means acceleration, which means that components can be jostled, which means that jostled components could become unseated, even after deployment.<\/p>\n<h3><span class=\"ez-toc-section\" id=\"maintenance-is-a-huge-problem\"><\/span>Maintenance is a Huge Problem<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<p>Every component inside a server will fail eventually.<\/p>\n<p>Although there is no specific timeline for 100% failure for each component, there also isn&#8217;t a guaranteed minimum service life.\u00a0 Instead, each component has sort of a &#8220;half life&#8221; called the Mean Time Between Failures (MTBF) which is a &#8220;good guess&#8221; about the maximum service life.\u00a0 Mechanical components, such as fans and power supplies have a much lower MTBF than solid-state components such as CPUs and RAM chips.<\/p>\n<p>So a brand new hard drive with a MTBF of around 3 years could fail within a couple days, or it could last for 6, and there is no way to determine which one you have.\u00a0 A brand new power supply with a MTBF of 24 months could fail tomorrow, or it could run just fine for 4 years.\u00a0 And, there is no\u00a0<em>guaranty<\/em> of failure.\u00a0 It just becomes more likely as time progresses.<\/p>\n<p>In fact, the best practice for installing new equipment is to establish a &#8220;burn-in&#8221; period &#8211; an amount of time between the first cold boot until the new equipment is processing production workloads &#8211; to ensure that there are no early component failures, or to mitigate them before processing production workloads.<\/p>\n<p>And, there are many factors that affects service life:<\/p>\n<ul>\n<li>Workload is a huge factor.\u00a0 Servers that process more data require more CPU and memory utilization, and therefore consume more power, and generate more heat.<\/li>\n<li>Environment is a huge factor.\u00a0 Servers that operate in higher temperatures are more likely to fail.<\/li>\n<li>Frequent power cycling.\u00a0 Frequent power cycling can degrade power supplies, but can also degrade other components due to &#8220;surge voltage&#8221; that occurs when a circuit is first energized.\u00a0 This is such as significant consideration that often, experienced IT staff are reluctant to power-cycle old equipment, opting instead for a soft reboot if possible.\u00a0 You sit there biting your nails, wondering if this was the device&#8217;s last power cycle.<\/li>\n<li>Hot \/ Cold cycles.\u00a0 \u00a0It was common in the 80&#8217;s and 90&#8217;s to find that components held by friction (at the time, BIOS, RAM, CPU, and ISA\/EISA boards) could migrate over time due to heat expansion followed by cooling contraction.\u00a0 During this period &#8220;re-seating&#8221; (making sure everything was seated properly) was a common troubleshooting method.\u00a0 So to make this clear, some percentage of the time, you could fix a PC or server just by re-seating all of its components.\u00a0 However, even today, in the age of clips and fasteners, hot\/cold cycles can still cause components to migrate.\u00a0 Or, cooling cycles can cause corrosion between components, which can cause &#8220;dead spots&#8221;, where re-seating the component (such as a stick of RAM) can bypass the corrosion.\u00a0 This is why most standby hardware is kept in a &#8220;warm&#8221; state &#8211; powered on, but not in use.<\/li>\n<li>Age.\u00a0 As components age, broken solder joints, migration, and internal heat damage become more likely.\u00a0 The longer something runs, the more likely it is to fail.<\/li>\n<\/ul>\n<p>Understanding all of this, the reality is that components can fail at any time.\u00a0 In fact, the more components you have, the more likely that a random failure will occur.<\/p>\n<blockquote><p><em>Anecdote:<\/em><\/p>\n<p>I once toured a data center.\u00a0 I won&#8217;t share when, which one, which company owned it, or where it was located.\u00a0 The company who owned it BRAGGED for months about how resilient their data center is, and how if we let them host our workloads, we would never have to worry about anything.\u00a0 They cited power resiliency, and even that their data center free-floated on a pad of diesel fuel, providing isolation from seismic events.\u00a0 But when we actually toured the data hall, we saw rack after rack with random orange lights.\u00a0 At the time, and for the particular brand of equipment in use, orange = failure.\u00a0 A failed drive had an orange light.\u00a0 A server with a failed component had an orange light.\u00a0 What this meant was that there was a pattern of negligence, where no one was monitoring the hardware, and no one was &#8220;walking&#8221; the data center.<\/p>\n<p>One of the first policies that I established, the first time I managed a data center, was to have a specific person whose responsibility was to walk through the data center every day looking for failures or anomalies.\u00a0 This accounts for situations where you have a potential hardware issue or environmental issue that monitoring didn&#8217;t catch for some reason.\u00a0 In addition, that person&#8217;s responsibility included removing trash, looking for stray cables (midnight cable changes), unlocked cabinets, and any other anomalous situation that might result in a power, network, or equipment failure.<\/p>\n<p>Over the years, I had several clients who adopted this practice after taking a tour of my data center(s).<\/p><\/blockquote>\n<p>To some extent, most equipment can be configured with redundancy:<\/p>\n<ul>\n<li>Power supplies can be redundant.\u00a0 If one fails, the other(s) take over the load.<\/li>\n<li>CPU boards can be redundant.\u00a0 If one fails, the workload is shifted to the remaining CPU board(s), but perhaps in a degraded state.<\/li>\n<li>Network and SAN connections can be redundant, supporting a failed, link, interface, board or switch.<\/li>\n<li>Firewalls, routers, and load balancers can be deployed in redundant pairs or &#8220;availability groups&#8221;, where an entire failed device is automatically mitigated by the remaining devices.<\/li>\n<li>Storage can be deployed in RAID groups, where a failed drive or an entire RAID array is redundant and therefore fault-tolerant.<\/li>\n<\/ul>\n<p>However, over time, failures add up.\u00a0 Without human intervention to replace failed modules, servers become less reliable over time.<\/p>\n<p>In space, there are no daily walk-throughs, nor operations staff to replace failed equipment.<\/p>\n<p>Beyond servers and other rack equipment, what about:<\/p>\n<ul>\n<li>Failed batteries?<\/li>\n<li>Failed power regulators \/ conditioners \/ distribution nodes?<\/li>\n<li>Interfaces and cables?\u00a0 (yes, even a cable can fail)<\/li>\n<li>Cooling?<\/li>\n<\/ul>\n<p>What about a simple hole in the wall caused by a micro-meteor?\u00a0 Or a degraded door seal?<\/p>\n<p>And, keep in mind, repairs depend on stocking spare parts.\u00a0 Where do you store the spare parts, and how do you access \/ deploy them?<\/p>\n<h3><span class=\"ez-toc-section\" id=\"physical-resiliency-is-a-huge-problem\"><\/span>Physical Resiliency is a HUGE Problem<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<p>In the comedy show &#8220;Space Force&#8221;, a chinese satellite is depicted using deployable arms to sever the solar panels off of an American satellite.<\/p>\n<p>Although this isn&#8217;t realistic, what IS realistic is that an Earthbound data center and its air space is protected by the entire US military.\u00a0 A spacebound data center is a tin can that can be taken out by a proverbial BB gun.<\/p>\n<p>Because china&#8217;s industry intersects their government, essentially there is no difference between industrial espionage and governmental espionage.\u00a0 If china decides to &#8220;accidentally&#8221; fire a missile in to your $30 BILLION space data center, there isn&#8217;t shit you can do about it.<\/p>\n<h3><span class=\"ez-toc-section\" id=\"high-energy-particles-%e2%80%9ccosmic-radiation%e2%80%9d-are-a-huge-problem\"><\/span>High-Energy Particles (&#8220;Cosmic Radiation&#8221;) are a HUGE Problem<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<p>HOOO boy.<\/p>\n<p>What is &#8220;radiation&#8221;?<\/p>\n<p>In essence, any particle.<\/p>\n<p>Photons, neutrons, electrons, protons, neutrinos, muons, leptons, sesame-street-ons.<\/p>\n<p>In empty space, free particles zip around at nearly the speed of light, and can erode metal or disrupt electrical circuits.\u00a0 Even if you employ electrical shielding, non-charged particles such as neutrons or neutrinos can randomly interact with unprotected electrical circuits, disrupting logic and producing irrational and\/or inconsistent computational outcomes.<\/p>\n<p>On Earth, we have tremendous shielding against this, called &#8220;the atmosphere&#8221;.<\/p>\n<p>There are strategies to block cosmic radiation, such as to employ exterior water jackets, where water flows through the exterior wall, and the density of water helps absorb unwanted cosmic particles.<\/p>\n<p>But&#8230;water is heavy.\u00a0 \u00a0Weight = cost.<\/p>\n<h3><span class=\"ez-toc-section\" id=\"bandwidth-is-a-problem\"><\/span>Bandwidth is a Problem<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<p>Data centers process a lot of data.\u00a0 In order to do that effectively, you have to move a lot of data in and out of the data center.\u00a0 In order to do this effectively, you need to move data\u00a0<em>quickly<\/em>, which means more bandwidth = more efficiency.<\/p>\n<p>Earth data centers have OODLEBITS (not a real word) of bandwidth.\u00a0 HUNDREDS of gigabits of bandwidth.<\/p>\n<p>If space data centers are depending on Starlink, the raw bandwidth isn&#8217;t even going to be close to what you can deliver to an Earthbound data center via fiber optic cables.<\/p>\n<h3><span class=\"ez-toc-section\" id=\"power-is-not-a-problem\"><\/span>Power is NOT a Problem<span class=\"ez-toc-section-end\"><\/span><\/h3>\n<p>Assuming solar is the primary power source, and adequate battery capacity to run at night, power kind of solves itself.<\/p>\n<p>Solar -&gt; Panel -&gt; Regulator -&gt; Battery -&gt; Conditioner -&gt; PDU -&gt; Server, DCV the entire trip.<\/p>\n<p>Panels and batteries have to be regularly replaced, but&#8230; if you just spent $67 BILLION on around 900 launches, what&#8217;s a couple $bil more?<\/p>\n<p>&nbsp;<\/p>\n<h2><span class=\"ez-toc-section\" id=\"conclusion\"><\/span>Conclusion<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p>Data centers in space is stupid.<\/p>\n<ul>\n<li>1,000 times the cost, no benefit<\/li>\n<li>The only way to address life span of the facility is to replace the facility<\/li>\n<li>The only way to address equipment life cycle is to continuously replace equipment<\/li>\n<li>ALL modern data center equipment is designed for 19&#8243; racks, which presents a significant design \/ cost constraint.<\/li>\n<li>How do you cool equipment in space without heavy (and therefore expensive) liquid cooling, or pressurized atmosphere?<\/li>\n<li>Without atmosphere and humidity control, how do you ensure that electronic components operate consistently and reliably?<\/li>\n<li>How do you ensure that equipment designed to work on Earth won&#8217;t be damaged during launch?<\/li>\n<li>How do you harden a space data center against physical attacks, especially from space-faring, hostile countries?\u00a0 Like, china?<\/li>\n<li>How do you perform routine hardware maintenance?<\/li>\n<li>How do you block cosmic radiation?<\/li>\n<li>How do you provide consistent, latency-free, bandwidth?<\/li>\n<\/ul>\n<p>Data centers in space is a great plot for a campy Muppets sci-fi movie, but not a good solution for the real world with real workloads.<\/p>\n<p>&nbsp;<\/p>\n<p>&nbsp;<\/p>\n<p>&nbsp;<\/p>\n<p>&nbsp;<\/p>\n","protected":false},"excerpt":{"rendered":"<p>Building Data Centers in space is stupid.\u00a0 Here&#8217;s why&#8230;<\/p>\n","protected":false},"author":16,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[],"class_list":["post-7997","post","type-post","status-publish","format-standard","hentry","category-other-stuff"],"_links":{"self":[{"href":"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-json\/wp\/v2\/posts\/7997","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-json\/wp\/v2\/users\/16"}],"replies":[{"embeddable":true,"href":"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-json\/wp\/v2\/comments?post=7997"}],"version-history":[{"count":10,"href":"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-json\/wp\/v2\/posts\/7997\/revisions"}],"predecessor-version":[{"id":8128,"href":"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-json\/wp\/v2\/posts\/7997\/revisions\/8128"}],"wp:attachment":[{"href":"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-json\/wp\/v2\/media?parent=7997"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-json\/wp\/v2\/categories?post=7997"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/justinparrtech.com\/JustinParr-Tech\/wp-json\/wp\/v2\/tags?post=7997"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}