PiDP-8/I Software

Artifact [fc81a76931]
Log In

Artifact fc81a76931c1de0725570c9a377900215ac7876dc718641d43d6ff71501c4263:

Wiki page [TD8E vs TC08] by tangent 2019-02-12 05:36:57.
D 2019-02-12T05:36:57.509
L TD8E\svs\sTC08
N text/x-markdown
U tangent
W 1707
The `os8.tu56` file shipped with early releases of the PiDP-8/I software was an image created from a DECtape found running on someone's PDP-8 system. That system had a TD8E "Simple DECtape" controller that required the CPU to monitor the tape as it sped past the heads, and transferred data through the AC, not by using Direct Memory Access.

Most surviving PDP-8.e hardware has that single-card DECtape controller.  However, there is another controller available for the PDP-8, the TC08 which offloads to hardware more of the tape motion and uses DMA to transfer data. Before the advent of the PDP-8 OMNIBUS, the TC08 was the only controller option available, but it was expensive.

Here's where the interesting tradeoff comes in: emulating a DMA device in SIMH is quick and easy.  Your driver gets the pointer to the blocks and copies them. The TC08 emulator for SIMH is simple and runs as fast as a byte copy.

Emulating the TD8E is harder.  You listen for the commands to start the tape moving, and then for the desired block number. Then you wait for each 12 bit word to go into that block to come to you through the emulation of a programmed I/O instruction that put that word in the PDP-8 AC.

If all you ever want to do is play with DECtape under SIMH, [configure the PiDP-8/I software distribution](/doc/trunk/README.md] in the default way, which now uses the TC08 device. (This since release 201902xx.) The TC08 simulation is a LOT faster. 

However, if you are using SIMH and the PiDP-8/I to create TU56 images for use on real hardware, you may want to configure it to use the TD8E, because it's the more commonly found DECtape controller:

     $ ./configure --boot-tape-config=td12k
Z 645e3f2cca6c15c078feebe3aa4dcedd