The cpu is less prone to fail yet has a comparable level of integration.

I'm pretty much on the software side these days, so it's no big problem at the moment. Unfortunately, it looks like Microsoft is killing off Legacy Windows functionality in favor of the newer and cross-platform friendly software Windows design that was introduced with Windows 8. That means that Win32, the foundation of retro PC gaming for many, is going away. This is rumored to be the case with the next version of Windows (12). The good news is that the Linux side of things is fairly usable with WINE and OpenAL, but it's a difficult task to get many "enthusiast" cards working for audio. MIDI is also an issue, with no real options outside of FluidSynth and SoundFont usage for newer hardware, as there aren't things like the VST MIDI Driver for using the Windows soft-synths.

chris2021 wrote on 2022-02-22, 11:02:

The cpu is less prone to fail yet has a comparable level of integration.

The latest Ryzen is 10 billion transistors. 16GB of RAM is 137 billion.

And with ram you're generally dealing with multiple chips. People are free.to disgree. This conversation is going nowhere.

Again, that is entirely my point.

Ok, if my point needs to be made plain and clear - high transistor count devices, by any reasoning, should be more prone to failure. The highest density devices are cpu's and ram. In my experience, ram is way more likely to fail then a cpu. Both have high levels of integration. But one seems to be far more reliable.

Do you actually believe memories have the same longevity as cpu's? If so, your
experience is way different from mine. I guess.

In a typical desktop computer, there are going to be at least 10x more transistors in the RAM than in the CPU. It doesn't matter if they are on one chip or multiple chips. If one of them fails, the RAM fails. RAM is generally less reliable because there are more opportunities for failure. Certainly there can be low quality RAM, but even the highest quality RAM will never be as reliable as the accompanying CPU.

The same holds true for a motherboard vs CPU, but it's not an apples to apples transistor count comparison because motherboards have many other types of components. Even the very best electrolytic capacitor is highly unlikely to outlast silicon. This makes motherboards inherently less reliable. Not because they are universally built with lower standards.

RAM is typically more bleeding edge than CPUs too, gets on the smaller process nodes sooner.

Plasma wrote on 2022-02-23, 01:19:

In a typical desktop computer, there are going to be at least 10x more transistors in the RAM than in the CPU. It doesn't matter if they are on one chip or multiple chips. If one of them fails, the RAM fails. RAM is generally less reliable because there are more opportunities for failure. Certainly there can be low quality RAM, but even the highest quality RAM will never be as reliable as the accompanying CPU.

The same holds true for a motherboard vs CPU, but it's not an apples to apples transistor count comparison because motherboards have many other types of components. Even the very best electrolytic capacitor is highly unlikely to outlast silicon. This makes motherboards inherently less reliable. Not because they are universally built with lower standards.

My point exactly. Cpus are more reliable then ram and generally every other component on a mobo. The cpu is also constantly "working". Other devices not as much.

From a troubleshooting standpoint, cpus should be built more shoddy 😀. This way when the box craps the bed, you know where the problem is. Instead of having to fish around all day.

You originally claimed motherboard components were built to lower standards, that was my disagreement. Intel makes high quality server motherboards that will still never be as reliable as their CPUs, for reasons I previously explained.

What I say here has nothing to do with reliability per se, but: As a small @home enthusiast, I tend to have accrued over time slightly more RAM than CPUs, and certainly than motherboards. I surmise in hindsight that one of the key reasons is the tiny, low profile storage of RAM, c0mpared to almost any other PC part. Figurative reams of it can sit on the shelf for decades and take up barely a small box.

I'll also mention that electrical durability is only a present factor when duty cycle is high. Read: a significant proportion of stuff hasn't spent a lot of time switched on.

I wonder how much of the recent surge in prices of vintage computer equipment is really from decreasing supply due to inevitable failure of components vs some kind of pandemic fueled increase in interest in the hobby.

No, I don't think so. There isn't a lot of availability of certain items I've noticed, as people are probably holding on to what they've got. Not just vintage computers. As prices go higher, you're not going to be as inclined to sell something. And if you paid a lot, it proves you're a fanatic.

I remember when you could barely give away Mindsets. No one really even knew what they were. Now they're a grand if they're a nickel. I almost sold an inoperative IBM 5175 a while ago. An inoperative 1 went for over 1200$. More people are getting into this, some appatently have money to burn.

The sad fact is, unless.you're into obscure inoperative techno bric a brac, that's probably what you'll wind up owning before long.

There is and will continue to be a market for faithful reproductions. People will overcome their desire to own an original, as the original is or will be an expensive paper weight before long. Some reproductions aren't worthwhile (think of comic books, but even those command some significant value). But an Amiga reproduction sitting on display in front of you will likely cause your eyes to light up as much as an original.

Some of we enthusiasts are already partly bridging that retro-modern gap.

Just to name a couple of the many examples: I'd wager there are probably more active Vogoners running mt32-pi now, than are running original MT-32's. And those us who use a MiSTer are now getting a great deal of potentially-replaces-an-entire-arcade-and-console-collection enjoyment out of a tiny little $few-hundred thing. While still/also owning a ton of original gear.

It doesn't have to be one or the other.

There's no doubt that the trend of rising vintage prices happens IN TANDEM with the trend of enthusiasts replacing old functionality with fresh modern incarnations. They're both undeniably, gradually happening.

No of course not. But as new people get into the hobby, and can't find or can't afford (also means can't justify the expense) what they want may have no choice but resort to reproductions.

Remember all this stuff was overwhelmingly regarded as worthless crappola. People choose to arbitrarily assign value to items, which they also could assign to reproductions. Granted people need to be reimbursed for their efforts. But when reproductions of collectibled become collectible, the world just may end at that point.

It's all more-or-less standard economics, supply and demand, economies of scale, all that jazz.

A lot of people just want the hardware for the (gaming, or whatever) experience, they don't care if it's original or not. At that point, the thing that's rolling off the production line today handily beats out the thing that stopped being made 20 years ago. And probably contains more features and enhancements, too.

I maintain that collectors pale in numbers compared to people who just want to play games. Collectors suck up the last dregs of any given vintage market, but for that to make an appreciable difference (to prices, rarity, etc), there needs be only dregs remaining in the first place.

Clearly there is a limited amount of any number of items. But in actuality you can thank collectors and /or those who actually still use this stuff, for it's preservation to begin with. I was surfing ebay back in 2001-2002. Someone was selling a Tandy 2000. Wow, I had 1 of those. It sold 1$ + shipping. Something got me interested in obtaining old copies of PC Magazine, Byte, etc. Found there were other people who were similarly interested in old tech. Started chatting. Got into loads of fights. Before long I obtained my own Tandy 2000. The rest is history.

Now my.purpose was to do a lot more with that Tandy then I wound.up doing. And in reality I already had some old computers I had picked up. A TI Pro bought for 10$ at a thrift store blocks from me in early 1998. A Commodore Sx64 I had a guy in CT ship me, probably 96 or 97. Several Zenith portables and a PC Peanut at a computer show around that same time. But I wasn't collecting back then. They were just interesting, maybe not a significant difference.to some, but a marked difference.to me. This hobby didn't become manic until ebay came around. And as a consequence I've owned over 100 different items. Driven across multiple states to obtain them.

Truly vintage items, roughly from prehistoric times until say 1990 is a different category from most of the dicussions on this board. People don't just want gear from 1995.- 2005. They want to overwhelmingly operate it. Or find ways to run those wares on newer hardware. I may mispeak at times, but I don't consider something vintage just because it's 20-25 years old. It's just obsolete.

chris2021 wrote on 2022-03-18, 21:31:

But in actuality you can thank collectors and /or those who actually still use this stuff

Thanks, ALL VOGONERS. 😀

And thanks for your interesting hardware history, you are clearly a long term dedicated enthusiast and collector. Truly, one of us!

I may mispeak at times, but I don't consider something vintage just because it's 20-25 years old. It's just obsolete.

I tend to think that way too, but then I remember that kids who grew up in the nineties probably feel the same way about stuff from that era as I do about my Apple IIe and 8086 stuff.

I really need to liquidate much, easily 80% of what I currently own. Too much for too long. It always ends the same. What's left of my stash will be coddled.and appreciated, at long last. And your welcome