The disassembly continues, with intentions still uncertain. But we’ve arrived at a surprising point.
At one time, I addressed the taking in of my close friend Sean’s old high school era computer, the most remarkable thing throughout its existence being its case – a venerable all aluminum Lian Li whose model number is lost to time.
Largely running out of obvious tasks to take care of around the basement (not that there isn’t still a nearly endless amount of cleanup and organization to do, it’s just that it’s … well it’s not obvious what else to do immediately) I happened to lay eyes on one of the systems yet to be re-evaluated and probably disassembled: Sean’s old PC.
I had recently considered again using the case as a pseudo sleeper build, but again stopped myself due to airflow concerns. Certainly I could just put in a lower power system, I have components for it, but I just found again, uncertainty, with what to do and/or what the point would really be. So again and again the system remained shelved.
This time I realized at some point I’d removed the cpu’s heatsink/fan unit and never put it back, so I thought that would at least be worth doing if only from a sheerly organizational standpoint. I air dusted it as carefully as I could, with a surprising amount of the ancient dust ending up in the bin. I applied some cheap thermal grease since, frankly it isn’t unfair to say that, at this point it doesn’t really matter as long as there’s something – although a note on that almost immediately. I reapplied the cooling unit and took a good long look inside the system which I hadn’t done in some time.
The system is based around a socket 939 AMD Athlon 64 3000+ on a mainboard by MSI – the K8T Neo2-F, which runs the VIA chipset K8T800 Pro. This ran the industry standard DDR RAM, at 266, 333, AND 400 MHz, topping out at 4 GB of that crazy fast stuff.
This is some old ish.
We have just at the outset, a northbridge and a southbridge and HyperTransport(™) based interconnects. Features include an AGP 3.0 8x slot, which I think meant a whopping 8x the speed of whatever AGP 1.0 was versus a standard 32-bit PCI slot. In my experience, both VIA chipsets AND AGP slots over 4x were just frankly unstable – software issues which were never resolved – but when they worked, they worked and as long as no one saw the crashes you got serious bragging rights.
Assuming you had an 8x capable GPU… some things don’t really change. Or the more they stay the same the more they change. Or something.
PCIe 4.0 currently that’s what I’m talking about.
Incredibly it features SATA ports (that’s SATA 1.0) plus two IDE channels AND a whopping 8 USB 2.0 ports on board, four by header. There are also other standard features on board but suffice to say this was, to my memory, the king of the budget boards, and I mean that in a positive way.
The system has remained the same for an obscenely long time. It must be 20 years by now, impossible as that seems. Yet it is complete. He never altered it from is assembled parts so long ago, and I didn’t either.
Realizing just an hour ago that that means it should be bootable, I set it down on the bench and took one last look inside at what passed for a reasonably competent build. And that isn’t to cast aspersions on either of us – clean layouts just weren’t as important in the DIY space. (Boutique Alienwares, and the like, of the era were wholly separate beasts indeed).
But as I was getting that aforementioned heatsink back in place, I spotted the issue. At least one of the capacitors I assume governs power delivery to directly to the CPU had blown, black innards peeking out of the silvery top of the old cylindrical components, and at least two more were bulging.
Far as I know, I can’t power this thing on without risking a potentially dangerous “catastrophic” failure. I won’t even plug it in.
It’s really dead.
And this isn’t truly any form of surprise since it is so old, and consisted of the value components of the day. It’s just vaguely sad versus the nostalgia of the system that it would be a worse idea than a good one to even try to turn it on. And nothing can be swapped around to other systems in any meaningful way. Drives sure, but they’ve been set back to factory installs ages ago.
It’s just an unexpected, should have been expected, abrupt final farewell to this legacy age system. Yes the case could go on, but airflow viability remains in question.
Two front 80 mm fan with thick, poorly designed grills. One top 80 mm exhaust also probably a third blocked by bulky “meshing”. And One piddly rear 80 mm exhaust out the back.
It would probably work out for a run of the mill modern system.
Who knows if it would be worth doing.
It’s just weird that it’s a truly failed system now.
It makes me triple question having the Pentium III and II era systems still shelved down here. Maybe there’s truly no point to them any more either.
Maybe I’ll investigate slightly sooner than I might have.
Maybe… …maybe I’ll be taking a trip to the dump…
When you get sentimental attachment to every damn thing in your life, that’s actually a pretty sad statement.
Sean Legacy is, as mentioned, built in a top of its day Lian Li aluminum enclosure with a not yet identified top mounted ATX power supply. It is an AMD Athlon 3000 on the socket 939 with the stock AMD heatsink fan combo which happens to be just a solid aluminum block, no copper to be found. It seems to be running two gigabytes of DDR400 RAM, ran in a dual channel configuration – how about that! Gigabit LAN is also on board as is middle of the road VIA audio including S/PDIF out on RCA. Storage utilizes two of the SATA/150 ports with the primary drive being a 160 GB Seagate at 7200 rpm and secondary storage being a massive 320 GB Seagate at 7200 rpm. The graphics card is an AGP 8x device – a Radeon 9600 sporting a simply miniscule passive heatsink on the GPU and no other cooling on its 64 MB of DDR RAM. I mean, picking out the details which I freely admit I had to look up on Google, that’s a Radeon 9600 non-pro sku on the Rage 8 architecture and a 130 nm manufacturing process.
As of the writing of this article, the state of the art manufacturing process is on 7 nm transistors and the other competitor in the field is still producing devices on highly refined 14 nm processes.
130 nm process GPU.
Tiny passive heatsink.
I don’t even know how things truly worked 20 years ago.
And there’s multiple levels of truth to that statement.
I digress.
Other specs? Two front USB 2.0 ports like WOW at the time. 16x DVD+/-RW ATA based drive for optical media. And a fully functioning 1.44 MB 3.5 inch floppy disk drive on the floppy drive controller channel, since sometimes someone would still need to update a bios or load a storage controller driver when installing Windows XP.
Incredible.
It was honestly totally different building computers back then in so very many ways.
Can’t even say it was a simpler time because it was WAY harder, especially if you were determined to stay in AMD’s camp.
Sean: Legacy build.
…
I mean I won’t even miss it. Not yet. It’s just going back up on the damn shelf because there’s so many things wrong in my brain. It’s not going anywhere. Not yet.