Is it impossible to produce old electronics in new factories?

I'm working on my PhD about emulation as preservation method. I did my bachelor's thesis and master's thesis from this same subject. Many times I've found myself thinking this question: Is it impossible to produce old electronics in present factories? If, let's say Intel would get an order to produce one 8080 processor, just as "clumsy" (from present day perspective) as it was back then, with the same technology and same architecture. 100% the same without using any modern techniques?

I have asked this question from few engineers and usually they avoid the question by counterquestion or then they say this is a wrong question to ask, because "who would want to have old ones?"

I have started to believe that with this counterquestion or correcting my question they are trying to hide the fact that they do not know the answer.

In digital preservation this question is interesting because, if we don't have a single working machine available, where the program was supposed to run, could we produce one? (Let's play now we don't have to care about the costs).

From my own studies I have come to conclusion that producing replicas from old ones that would be 100% the same as 30 years ago is theoretically possible but in practice impossibility, because machines that make machines in factories have been made to produce new technology and the old technology is scrapped. These new machines that make machines could make a machine that logically functions just like the old one, but they would do it in a new, more developed way.

To produce a new series of old 8-bit computers would require whole factories and assembly lines that are similar to the ones that was used in 80's and rebuilding these would be too expensive even for the most gigantic companies.

Is my conclusion right or wrong or partially right/wrong?

Samuel Ranta

• digital
• preservation
• technology
• emulation
• computer

Comments

• Paul Wheatley: Why is this theoretical question of relevance to digital preservation? I think you're accepting that even if what you're asking was possible, the high costs of such an approach would make it an unrealistic approach for digital preservation. If that is the case, this really is a theoretical discussion and hence, I think, out of scope for DP Stack.
• mopennock: The research question of how we can reproduce old tech is probably a valid one, eg to access v old content or to validate emulations/migrations (thus my answer below); the problem here lies with the theoretical framework and the somewhat strawman argument around using factories. Could it be re-phrased?
• Robert Cartaino: Even if this were possible (or not), you're asking the wrong folks. Expertise in the subject of digital preservation doesn't really involve manufacturing process or the economics of such business ventures. I'm sorry, but I have to close this as outside the scope of *this* site.
• Bill Lefurgy: Don't believe this question has broad applicability, as it is difficult to imagine manufactures undertaking the task of remaking obsolete equipment. Question could ask about custom production of old equipment using 3d printers--legal matters, specs needed, etc.

Answer by mopennock

I suspect it isn't impossible, more that from a cost perspective it would be an undesirable venture for a commercial company to undertake. To 'rollback' the technology in a factory so that it produced hardware in exactly the same way as in previous decades would be phenomenally expensive. But is it really necessary to produce the technology in exactly the same way, or is it acceptable to use new production methods to create it? That would depend on your requirements and the importance of the process.

The Computer Conservation Society does an excellent job in restoring and/or recreating old machine. CCS volunteers recently restored Flossie (also known as CT 1301), though how they validated the functionality of the restoration is unclear. From what I've seen though, they do not attempt to do so in modern factories.

Comments

Answer by lechlukasz

We have tanks and airplanes, but there are still people producing swords. But to produce a sword you need a skilled blacksmith and a bit of tools. To produce, say, 286 computer, you need a whole factory.

When someone wants to sacrifice his free time and a few thousand \$ for his blacksmithing hobby, it's just a hobby like anything else. But if someone was to sacrifice a few million \$ and tens (when not hundreds) of people's full-time job to produce old electronics with no practical meaning but only for hobby, this is a social problem. Such huge resources could be used for more rational and practical aim.

For data and software preservation, software emulation is fully enough. The aim is to be able to run the software or use the data, not run them on exactly the same device as 20 or 30 years ago.

Comments

• db48x: Indeed, emulation of hardware old hardware is often not that difficult. You can still buy some of the more popular old cpus, such as the 6502/6510, although they're manufactured with modern process sizes. I think the most interesting way to go is to use FPGAs though; a single FPGA can be reconfigured to emulate any number of old chips.
• lechlukasz: Hmmm in fact, although it's still emulation