A brief tour of the PDP-11, the most influential minicomputer of all time (2022)

https://arstechnica.com/gadgets/2022/03/a-brief-tour-of-the-pdp-11-the-most-influential-minicomputer-of-all-time/

by jensgk • 1 month ago

119 56 1 month ago

gerdesj2• 1 month ago

"PDP-11 with UNIX opened the floodgates for inexpensive interactive computing, which then led to an explosion of office productivity. "

Well before we get too misty eyed: "inexpensive" needs looking at "for inexpensive interactive computing".

I'm not old (55) enough to have really got to grips with a PDP11. I do still own (yes: present tense) a C64 from 1986. The C64 was bought by my dad via the NAAFI in West Germany so I have no idea what it costed. Let's wind forward a bit:

I had a 80286 based PC in 1987ish with 1MB of RAM, 20MB RLL hard disc. The graphics card (ISA) had a whopping 512 bytes of RAM. That thing costed about £1200. I added a 80287 later at about £120 so I could run a pirated copy of AutoCAD.

In 1990ish I had a 80486 with 4Mb RAM and 40MB HD - that costed something like £1600.

Nowadays £1600 will buy quite a decent laptop and 35 years of inflation.

pezezin• 1 month ago

The graphics card (ISA) had a whopping 512 bytes of RAM.

I think you are missing a "kilo" there...

gerdesj2• 1 month ago

The k key on my laptop is a bit shagged. I've just hit it harder twice.

kens2• 1 month ago

I'd suggest the Datapoint 2200 as the most influential minicomputer of all time since half of you are using an instruction set based on it and it is largely responsible for the creation of the microprocessor.

Now mostly forgotten, the Datapoint 2200 was a programmable desktop computer introduced in 1970. It had a processor built from TTL chips, along with shift-register memory from Intel. Datapoint discussed with Intel and Texas Instruments the possibility of building a single-chip processor to replace the board of TTL chips. TI was first with the TMX 1795 processor, followed by Intel's 8008, both copying the Datapoint 2200 instruction set.

Datapoint decided that these chips didn't have enough performance and fatefully gave up rights to them. TI tried to sell the TMX 1795 to Ford, but got nowhere and abandoned the chip. Intel decided to sell the 8008 as a standalong microprocessor, which was used in early personal computers like the Mark-8. Intel improved the 8008 to form the 8080, then made a somewhat compatible 16-bit version, the 8086, which started the x86 architecture. (Because the Datapoint 2200 was little-endian (to use shift-register memory), x86 is little-endian.)

To summarize its influence, without the Datapoint 2200, the microcomputer industry would have been greatly delayed (since the 4004 wasn't suitable for a personal computer) and x86 wouldn't exist.

themafia• 1 month ago

The Internet (mostly), the C language, the Unix Operating System were all developed and deployed on PDPs. The BASIC language /and/ an 8080 emulator was developed on a PDP in order to deploy to the Altair.

I would argue that Intel was so highly influenced by Datapoint due to sheer proximity and early inexperience in the field.

kens2• 1 month ago

I don't think you can count PDP-10 things (Internet, BASIC, 8080 emulator) to support the influence of the PDP-11, since they were completely different computers. The PDP-10 was a 36-bit mainframe, while the PDP-11 was a 16-bit minicomputer.

larsbrinkhoff2• 1 month ago

BASIC was developed on GE mainframes.

throwaway27448• 1 month ago

You can see the influence of the PDP-11 on C more than any other ISA. That alone I think qualifies it as the most influential microprocessor of all time.

smallstepforman3• 1 month ago

The 6809 (and its clone the 6502) would also be strong contendors. And if Motorola were 6 months faster with the successor 68000 compared to Intels 8086, the IBM PC would be 68000 based and Intel would today be known as a memory chip company. We’d be big endian with no legacy in 2026 of segmented memory architecture in mainstream processors. And with big 3 companies using 68000 (IBM, Apple and C=), we definately wouldnt have WinTel.

smallstepforman3• 1 month ago

While I’m on alternative histories, if Jack Tramiel didn’t piss off Chuck Peddle, it would be MOS/C= in Intels pisition today.

rswail3• 1 month ago

That was the 1980s, PDP-11s were the 1970s.

classichasclass2• 1 month ago

The 6502 isn't a 6809 clone; it was largely inspired by the 6800.

chasil• 1 month ago

What is not well-known is that:

a) the PDP-11 CPU was also implemented in TTL logic, like the Datapoint,

b) there were later implementations of the PDP-11 as NMOS microprocessors:

https://en.wikipedia.org/wiki/DEC_T-11

https://en.wikipedia.org/wiki/PDP-11#LSI-11_integrated_circu...

c) the VT100 terminal was based on the Intel 8080 microprocessor

https://en.wikipedia.org/wiki/VT100

There was appreciation and exchange between Intel and DEC.

HarHarVeryFunny• 1 month ago

Apparently Motorola had started work on their PDP 11 inspired 6800 in 1971, same year the Datapoint 2200 terminal was released, and would end up shipping it in 1974, same year Intel released their 8008 successor the 8080 (which the Altair 8800 was based on), so even without Intel's calculator (4004) and terminal (8008) chips we'd still have had single-chip LSI CPUs in a similar time frame.

budman1• 1 month ago

+1 to pdp11. The workhorse of the 1970's. They were everywhere. City Library. Auto parts stores. And reliable; Maytag ain't got nothing on a pdp 11.

Downside, programs are pretty simple that run in 64k. And extended addressing in any form, sucks.

WalterBright• 1 month ago

I had a PDP-11, in the form of the Heathkit H-11.

I loved that computer. Like a fool, I sold it for $25. There's a picture of it on my X profile.

The -11 had an instruction set that fit on one page.

jakzurr3• 1 month ago

ok, that's amazing - glad ya saved the pic at least

wduquette2• 1 month ago

Yes, that’s what we had too. I loved that machine. What was the incantation? 104799L, or something like that?

EvanAnderson• 1 month ago

There's a parallel worldline out there where the PDP-11 made the transition into a desktop PC[0] and the IBM PC didn't take over the world. In that worldline our servers are from the PDP-11 lineage, and not the IBM PC.

[0] https://en.wikipedia.org/wiki/DEC_Professional_(computer)

classichasclass2• 1 month ago

I've got a DEC Pro 380 myself ( https://oldvcr.blogspot.com/2025/03/more-pro-for-dec-profess... ). It's a heck of a desktop computer, and you can run some early Unices on them. The Pro and the DECmate II are some of the best built systems of the era despite their idiosyncrasies, IMHO.

not2b• 1 month ago

It was the preferred lab computer in the mid to late 1970s and into the 80s. I got my first job because I knew PDP-11 assembly language, and worked with both DEC's operating systems for them (RT-11 and RSX-11) and later Unix (the lab I worked with had some machines running Version 6, though Version 7 was the first that I used seriously. It had a very clean and symmetric instruction set that used the program counter as if it were another general purpose register. I had an LSI-11 board (the single-board version of the machine) with 4K 16-bit words of core memory and a paper tape punch with a tiny loader in ROM to read in the tape and peek and poke memory, and I'd sometimes initialize the core memory to a known state by running the one-instruction program

mov -(pc), -(pc) or 014747 in octal. It would fill all of memory with 014747.

getpost• 1 month ago

Ha! The instruction that copies itself. I posted also posted it, 6 years ago. https://news.ycombinator.com/item?id=24820583

amirhirsch2• 1 month ago

014747 makes me smile <3

I built the PDP-11/70 emulator that controls the nuclear reactors in Ontario. That was 20 years ago and I'm probably still the youngest person who can read PDP-11 assembly (and the raw octal)

ErroneousBosh2• 1 month ago

There's a decent chance you're younger than me, but I can still read it. Coincidentally about 20 years ago I got asked to fix a bug in some PDP11 code that ran on real hardware because it was tripping up an emulated system. Nothing wrong with the emulator as such, it was just a logic bug in the program that they no longer wanted to work around.

I made the fateful mistake of saying "Sure, sounds fun, how hard could it possibly be...?"

mrandish5• 1 month ago

It had a very clean and symmetric instruction set

Indeed. Motorola's 68000 CPU took so much inspiration from the PDP-11's ISA, it was almost a spiritual successor. The 68000's 8/16-bit little brother, the 6809, widely considered the most powerful 8-bit CPU ever - was also heavily inspired by the PDP-11.

tralarpa• 1 month ago

Somes lines in PDP-11 assembly from 2.11BSD:

  mov r1,-(sp)
  mov $1024.,-(sp)
  mov $outbuf,-(sp)
  mov fout,-(sp)
  jsr pc,_write
  add $6,sp
  mov (sp)+,r1
  tst r0
  bpl 2f
  jmp wrterr

ggm2• 1 month ago

I'd argue the pdp8 opened the floodgates. That's when the cost of digital computing dropped to the point a research grant or even just discretionary spending in a university department could pay for one, and you didn't need a special room and power supply.

the 11 was when it became more useful. But the 8 was how people realised you could move beyond a calculator to a computer.

ErroneousBosh2• 1 month ago

That's when the cost of digital computing dropped to the point a research grant or even just discretionary spending in a university department could pay for one

I've mentioned this before, but the PDP8 launched in the US about the same time as JCB launched their first (and arguably the first really practical and useful) backhoe loader in the UK, which was about three and a half grand at the time.

Can you imagine the paradigm shift with either of those machines? Not just it's possible to do that work, but you can do that work with *your own one*.

At some point someone has looked at a shiny new PDP8 or a JCB 3C in the showroom and gone "you know what, I'm just going to buy one", and got the chequebook out.

ggm2• 1 month ago

That's an amazing parallel. So much to think about how it changed dynamics of work replacing manual ditch digging and wheelbarrows.

larsbrinkhoff2• 1 month ago

Then maybe also give a nod to the LINC, showing DEC a small 12-bit computer could be made.

rswail3• 1 month ago

My first ever computing experience was in 1977 in year 9 at high school where we were lucky enough to have a PDP-11/10 with 16K(!) and 3 ASR-33s and a VT-52.

Learned how to program in octal on the front panel. I've still got an old front panel (the rest is a rotting collection of wirewrap boards in my garage).

It had a multi-user basic that left out string functions if you went multi-user. You loaded the bootstrap via the front panel, which read the "absolute loader" from paper tape on one of the TTYs, which then read the BASIC interpreter from the same source.

I still have the small reference card with the instruction set and some old paper tapes around somewhere.

The whole structure of the registers with R7 as the PC and R6 as the SP and the various addressing modes was just elegant.

We used to make jokes that your programmed a PDP-11 with 3 fingers (octal) and a VAX with 4 (hex).

nickcw• 1 month ago

I'm not old enough to have used a PDP-11 so I read through the assembly code description with interest.

Wow, it seems so modern. I've used a lot of assembly languages over the years but I'd feel immediately at home here. Nice sensible orthogonal instruction set with enough registers and a stack pointer. It reminded me immediately of ARM assembly - what a breath of fresh air that was when it came it.

I never realised quite how much influence the PDP-11 had on computer architecture. I knew about it's software legacy but I suspect that was enabled by it's ground breaking architecture.

mrandish5• 1 month ago

My first computer had a 6809 with 4K of RAM and I learned assembler on that 6809, which was greatly inspired by the PDP-11 architecture. My second computer was an Amiga 1000 with a 68000, a CPU even more directly a successor to the PDP-11 than any other. As far as I knew, those two PDP-11-like systems were computer architecture. Orthogonal instruction sets, user and system stacks treated just like any other register, layered interrupts, indirect addressing, memory mapped I/O, auto-increment/decrement, relocatable zero page, contiguous flat memory map, relative addressing that made position-independent, re-entrant code trivially easy. Yeah, in the 1980s, I led a sheltered, happy computing life.

It was around the mid-90s by the time I tried x86 assembler. I don't think "shocked" quite captures the experience. It was more like disbelief, then something akin to abject horror, which finally just faded into creeping existential dread. Before the X86, I loved coding in assembler but the nicest thing I can say about x86, is it motivated me to learn C as fast as possible. :-)

TomMasz• 1 month ago

Back in my running days in the 80s, the people who did the timing and race results had a van with a PDP-11 in it that I assume ran off a generator. I wish I had asked them more about it but I was too busy worrying about running.

PopePompus• 1 month ago

What happened if you used auto-decrement addressing with the PC? Did that hang the computer?

Taniwha• 1 month ago

Yup, as mentioned above the famous:

x: mov -(pc), -(pc)

reads an instruction from x, increments the PC to x+2, then decrements the PC back to x and reads the same instruction, then decrements it again to x-2 and writes it, then it executes from x-2 ......

ubermonkey• 1 month ago

There's a concept in history whose name escapes me, but which was evoked often when Harrison Tyler died in 2025. Tyler was a private citizen, but he was of note because he was the grandson of the US's 10th president John Tyler -- who died in 1862.

The gist is how surprising bridges to the past are closer than you realize -- as is the past itself.

At my first corporate job in 1994, we had a machine room. Those weren't uncommon back then. What WAS uncommon was that, over in a corner, sandwiched between racks of shiny new DEC Alphas, was a PDP-11 that was still running production code.

My employer then was TeleCheck, which did point of sale risk analysis for checks. The business had originally been run as independent state-by-state franchises, and back then someone had the bright idea to create an IT company that provided services to these franchisees -- and, occasionally, to other companies, too. By the time I was hired, the franchises AND the IT company had all been brought under one ownership, and all the IT company's external clients had gone elsewhere EXCEPT ONE.

That holdout was perfectly happy with what they got from that ancient computer.

I assume it eventually died, but TeleCheck had a DEEP bench of DEC talent, so it could've kept running a long, long time.

JKCalhoun• 1 month ago

Obsolescence Guaranteed makes a PDP-11 replica [1]. At the heart is a Raspberry Pi running emulation software. All switches, LEDs are physical—all wrapped up in a clever case that looks like the original.

I have bought a couple of their kits and can vouch for the quality of them.

[1] https://obsolescence.dev/pdp11.html

jakzurr3• 1 month ago

Used several of these at school in the 90's. IIRC intro classes used VAX/VMS (was the machine called VAX-1170?). But all the higher undergrad classes used Unix on PDP-11/70's, like in the picture. Of course, I hardly ever saw one up close. :p

jakzurr3• 1 month ago

thank god we weren't still using paper tape & punch cards, everything was old crt terminals, and online storage (though not very much)

allenrb• 1 month ago

I’m not quite old enough to have grown up with the PDP-11. They were gone from university (or at least the parts I had access to) by the time I arrived. But I did see exactly one in the wild.

In 1997-1998, I was working for a small company in Atlanta who did tape backup systems. At a client in maybe Knoxville (?), a hospital had a PDP-11/70 live in their machine room. Amusingly, right next to racks of then super-fancy Cisco gigabit fiber networking.

I was told that the PDP handled payroll. Guess that was important. Wonder how long it lasted there?

hylaride• 1 month ago

PDP-11s and their VAX successors could often still be seen in hospitals as late as the mid 2000s as the back ends to all sorts of expensive equipment (MRIs, etc). At the time I was told that rewriting the medical code was one of those "not worth the risk" endeavours. By now they've probably been virtualized away, but those machines were damned reliable.

hcfman• 1 month ago

Oh man. You are really missing out unless you started with valves and watched transistors come out as the cool kid on the block.

To see all that and still be programming now is one very lucky journey.

TradingPlaces• 1 month ago

My high school had one of them in the early 80s. Me and a few others were the only ones who ever turned it on. Then a friend got an Apple ][ and we never touched the PDP-11 again.

wduquette2• 1 month ago

Heathkit sold an LSI-11 as a kit; my father built one in the mid-70's. I learned BASIC, a bit of Fortran, a bit of PDP-11 assembler, and Pascal on that machine. We started with paper type, and eventually moved up to dual 8" floppies. We started with 16K of memory, IIRC, and ended with 28K.

Been a programmer ever since, me.

aitoh• 1 month ago

The computers that ran Unix were DEC’s (Digital Equipment Corporation) PDP-11 machines. Just then, DEC’s Japanese subsidiary (DEC Japan) was holding its first-ever recruitment of new graduates, and in 1980 I was fortunate enough to be hired.

After joining, however, I ran into one astonishing fact. DEC, then a hardware manufacturer, fully supported its own operating systems (RSX-11, VMS, and the like), but for Unix — a licensed product of Bell Labs (AT&T) — it offered no official service or technical support whatsoever. (It would be added to the official service menu in later years, but in 1980 it was out of scope.) I had joined a manufacturer in order to make Unix my work, only to find that the manufacturer did not support Unix — a historical gap that left me bewildered at the time.

A brief tour of the PDP-11, the most influential minicomputer of all time (2022)

Related Stories