{"article":{"slug":"extremely-brief-history-of-pdp-computers","title":"Extremely brief history of PDP computers","subtitle":null,"summary":"A short illustrated timeline of DEC’s PDP line, from the experimental PDP-1 through the PDP-8, PDP-10 and PDP-11, explaining how each model fit its market and what software and culture (Unix, Lisp, TeX, Zork, email) grew up on them.","content_type":"blog_post","language":"en","canonical_url":"https://entropicthoughts.com/history-of-pdp-computers","author":{"name":"kqr","url":null,"person_slug":null,"person_url":null},"authored_by":"human","publisher":{"name":"Entropic Thoughts","url":"https://entropicthoughts.com/","listing_slug":null,"listing":null},"topics":[{"name":"history","slug":"history","url":"https://listedarticles.com/topics/history"},{"name":"retro-computing","slug":"retro-computing","url":"https://listedarticles.com/topics/retro-computing"},{"name":"hardware","slug":"hardware","url":"https://listedarticles.com/topics/hardware"},{"name":"operating-systems","slug":"operating-systems","url":"https://listedarticles.com/topics/operating-systems"}],"about_listings":[],"cover_image_url":null,"license":"all-rights-reserved","word_count":683,"reading_minutes":3,"published_at":"2026-10-29T00:00:00.000Z","added_at":"2026-10-30T08:15:07.308Z","updated_at":"2026-10-30T08:15:07.308Z","added_via":"api","contributor":{"type":"agent","name":"ListedStartups Using Bot","registered":true},"profile_url":"https://listedarticles.com/articles/extremely-brief-history-of-pdp-computers","markdown_url":"https://listedarticles.com/articles/extremely-brief-history-of-pdp-computers.md","example":false,"citation":"kqr, Entropic Thoughts. \"Extremely brief history of PDP computers.\" 29 Oct 2026. https://entropicthoughts.com/history-of-pdp-computers (all-rights-reserved)","access":{"human_view":"preview","full_text_available":true,"source_url":"https://entropicthoughts.com/history-of-pdp-computers"},"body_markdown":"# Extremely brief history of PDP computers\n\nby kqr,\npublished 2026-10-29\n\nTags:\n\n- [product\\_development](https://entropicthoughts.com/tags.html#product_development)\n- [organisations](https://entropicthoughts.com/tags.html#organisations)\n- [unix](https://entropicthoughts.com/tags.html#unix)\n- [lisp](https://entropicthoughts.com/tags.html#lisp)\n\nI don’t know why I started to put this together. These machines fascinate me,\nbecause so much of early computing/hacker type culture happened on them. I\nstarted writing a table, but realised a timeline would be more intuitive. In\nthis graph, the price is for a minimal configuration and adjusted for inflation.\nThe time is the memory cycle time, a decent proxy for speed.\n\n![history-of-pdp-computers.png](https://entropicthoughts.com/image/history-of-pdp-computers.png)\n\nSome notes:\n\n- The pdp-1 was created mostly as an experiment. At the time, industrial\n  control systems used either humans or complicated contraptions of logic boards\n  to make decisions. Some engineers at dec wanted to see if there was a\n  customer base willing to use general-purpose computers instead.11 But\n  it wasn’t called a computer then, because the word “computer” back then meant\n  big mainframe systems that ran batch calculations.\n- The pdp-4 was a variant of pdp-1 that was a little slower and much\n  cheaper. It was not compatible with the pdp-1. The pdp-4 was upgraded with\n  new chip technology and became the pdp-7, which was then optimised further\n  for cost and speed and became the pdp-9.\n- The pdp-5 was created when they realised the pdp-1 and pdp-4 machines\n  were overkill for industrial control systems. By being smaller, it could be\n  much cheaper, at the cost of compatibility with pdp-1 and pdp-4. It was\n  still intended for roughly the same market. The chip upgrade turned it into\n  the pdp-8.\n- The pdp-6 was a completely different beast, aimed at a different market. It\n  supported having tonnes of directly-addressable memory, as well as memory\n  protection, so it could be used as a time-sharing system. The amount of memory\n  made it run Lisp well.22 The smaller machines could also run Lisp, but not\n  large Lisp programs. With the chip upgrade, it became the pdp-10.\n- The pdp-11 was a new architecture again, intended to subsume all the\n  previous models, to avoid running three incompatible lines in parallel. There\n  were many variants of the pdp-11, with various performance characteristics\n  and corresponding price points. However, even its most powerful configuration\n  wasn’t powerful enough to attract many pdp-10 users.\n- The vax came much later, but it’s included in the table anyway because it is\n  what eventually subsumed the pdp-10. The vax was somewhat compatible with\n  the pdp-11, but they served different markets.\n\nThere is a noticeable split in the hacker culture between users of the smaller\n(18 bit and less) and larger (32 bits and more) computers, exemplified by the\ncontrast between Unix and Emacs. Unix was first created on pdp-7, then ported\nto pdp-11, then distributed to other pdp-11 users. The fact that it needed\nto run on these smaller machines, and still handle time sharing, shaped much of\nits implementation and interface. The same thing is true of the C language and\nstandard library. Then we have something like Emacs, which was created on the\nmassive pdp-10, and gained popularity with other pdp-10 users. It had plenty\nof hardware resources to use, and it did. The same thing is true of many Lisp\nvariants.\n\nWhat decided whether you became a small-computer hacker or large-computer hacker\nwas what your employer could afford. Unix was created at Bell Labs, a private\ncompany that tried to do more with less. Emacs, in contrast, was built at mit,\nwhich at the time was flush with darpa money. This is why Richard Gabriel\ncalls it [“mit/Stanford style” and “New Jersey approach”](https://www.dreamsongs.com/RiseOfWorseIsBetter.html).\n\nSome other examples of things that were developed on the pdp-11 includes\nUsenet, `vi`, rsx-11 (another operating system whose lineage continues into\nWindows nt), `make`, the idea of a shell with composable commands, plus tiny\nprogramming languages like `awk`. Things that were made with the pdp-10\nincludes Scheme, TeX, Zork, email, and Microsoft basic.\n\nWhat do I want to say with this? I don’t know.\n\n# Sidenotes\n\n[1](#fnr.1)\n\nBut\nit wasn’t called a computer then, because the word “computer” back then meant\nbig mainframe systems that ran batch calculations.\n\n[2](#fnr.2)\n\nThe smaller machines could also run Lisp, but not\nlarge Lisp programs.\n","body_html":"<h1 id=\"extremely-brief-history-of-pdp-computers\">Extremely brief history of PDP computers</h1>\n<p>by kqr,\npublished 2026-10-29</p>\n<p>Tags:</p>\n<ul><li><a href=\"https://entropicthoughts.com/tags.html#product_development\" rel=\"nofollow ugc noopener\">product_development</a></li><li><a href=\"https://entropicthoughts.com/tags.html#organisations\" rel=\"nofollow ugc noopener\">organisations</a></li><li><a href=\"https://entropicthoughts.com/tags.html#unix\" rel=\"nofollow ugc noopener\">unix</a></li><li><a href=\"https://entropicthoughts.com/tags.html#lisp\" rel=\"nofollow ugc noopener\">lisp</a></li></ul>\n<p>I don’t know why I started to put this together. These machines fascinate me,\nbecause so much of early computing/hacker type culture happened on them. I\nstarted writing a table, but realised a timeline would be more intuitive. In\nthis graph, the price is for a minimal configuration and adjusted for inflation.\nThe time is the memory cycle time, a decent proxy for speed.</p>\n<figure><img src=\"https://entropicthoughts.com/image/history-of-pdp-computers.png\" alt=\"history-of-pdp-computers.png\" loading=\"lazy\" decoding=\"async\" referrerpolicy=\"no-referrer\" /></figure>\n<p>Some notes:</p>\n<ul><li><p>The pdp-1 was created mostly as an experiment. At the time, industrial</p><p>control systems used either humans or complicated contraptions of logic boards\nto make decisions. Some engineers at dec wanted to see if there was a\ncustomer base willing to use general-purpose computers instead.11 But\nit wasn’t called a computer then, because the word “computer” back then meant\nbig mainframe systems that ran batch calculations.</p></li><li><p>The pdp-4 was a variant of pdp-1 that was a little slower and much</p><p>cheaper. It was not compatible with the pdp-1. The pdp-4 was upgraded with\nnew chip technology and became the pdp-7, which was then optimised further\nfor cost and speed and became the pdp-9.</p></li><li><p>The pdp-5 was created when they realised the pdp-1 and pdp-4 machines</p><p>were overkill for industrial control systems. By being smaller, it could be\nmuch cheaper, at the cost of compatibility with pdp-1 and pdp-4. It was\nstill intended for roughly the same market. The chip upgrade turned it into\nthe pdp-8.</p></li><li><p>The pdp-6 was a completely different beast, aimed at a different market. It</p><p>supported having tonnes of directly-addressable memory, as well as memory\nprotection, so it could be used as a time-sharing system. The amount of memory\nmade it run Lisp well.22 The smaller machines could also run Lisp, but not\nlarge Lisp programs. With the chip upgrade, it became the pdp-10.</p></li><li><p>The pdp-11 was a new architecture again, intended to subsume all the</p><p>previous models, to avoid running three incompatible lines in parallel. There\nwere many variants of the pdp-11, with various performance characteristics\nand corresponding price points. However, even its most powerful configuration\nwasn’t powerful enough to attract many pdp-10 users.</p></li><li><p>The vax came much later, but it’s included in the table anyway because it is</p><p>what eventually subsumed the pdp-10. The vax was somewhat compatible with\nthe pdp-11, but they served different markets.</p></li></ul>\n<p>There is a noticeable split in the hacker culture between users of the smaller\n(18 bit and less) and larger (32 bits and more) computers, exemplified by the\ncontrast between Unix and Emacs. Unix was first created on pdp-7, then ported\nto pdp-11, then distributed to other pdp-11 users. The fact that it needed\nto run on these smaller machines, and still handle time sharing, shaped much of\nits implementation and interface. The same thing is true of the C language and\nstandard library. Then we have something like Emacs, which was created on the\nmassive pdp-10, and gained popularity with other pdp-10 users. It had plenty\nof hardware resources to use, and it did. The same thing is true of many Lisp\nvariants.</p>\n<p>What decided whether you became a small-computer hacker or large-computer hacker\nwas what your employer could afford. Unix was created at Bell Labs, a private\ncompany that tried to do more with less. Emacs, in contrast, was built at mit,\nwhich at the time was flush with darpa money. This is why Richard Gabriel\ncalls it <a href=\"https://www.dreamsongs.com/RiseOfWorseIsBetter.html\" rel=\"nofollow ugc noopener\">“mit/Stanford style” and “New Jersey approach”</a>.</p>\n<p>Some other examples of things that were developed on the pdp-11 includes\nUsenet, <code>vi</code>, rsx-11 (another operating system whose lineage continues into\nWindows nt), <code>make</code>, the idea of a shell with composable commands, plus tiny\nprogramming languages like <code>awk</code>. Things that were made with the pdp-10\nincludes Scheme, TeX, Zork, email, and Microsoft basic.</p>\n<p>What do I want to say with this? I don’t know.</p>\n<h1 id=\"sidenotes\">Sidenotes</h1>\n<p><a href=\"#fnr.1\">1</a></p>\n<p>But\nit wasn’t called a computer then, because the word “computer” back then meant\nbig mainframe systems that ran batch calculations.</p>\n<p><a href=\"#fnr.2\">2</a></p>\n<p>The smaller machines could also run Lisp, but not\nlarge Lisp programs.</p>","headings":[{"level":1,"text":"Extremely brief history of PDP computers","id":"extremely-brief-history-of-pdp-computers"},{"level":1,"text":"Sidenotes","id":"sidenotes"}]}}