I remember it well, back in the very early 2000s, when my own Athlon XP 1800+ (an AGOIA stepping, I do know that for sure :D) met its untimely end at the hands of a device that was designed to prevent the exact misfortune that killed its sister CPU as seen in this submission!
The fragile silicon was a problem the enthusiast/PC-builder community was very aware of, and the market responded with "spacer" kits you could buy for a few $$$ that were cut from solid sheets of thin copper. These spacers would carefully wrap around all the little bumps on the CPU's upper side, and provide a level plane for the heatsink to rest upon, while still establishing direct contact with the die that it was supposed to cool down. So there was, in theory, decidedly less of a chance of killing your CPU by having the heatsink meet it in an unfortunate angle, or something to that effect.
Unfortunately, the spacer that I had bought turned out to be just a wee little bit too thick, and, despite thermal paste (Arctic Silver!), the CPU did not in fact make contact with the heatsink (something shiny from Thermalright or Zalman, I think?) at all, which killed it instantly as soon as I powered up the rig for the first time. That's what it was like back in the days, when we got to school walking ten miles uphill (both ways) barefoot in deep snow.
I had to put in some overtime at the small PC shop I worked after school back then, to afford a measly Duron 800 as its replacement. And still they keep claiming that time heals all wounds... ;)
I accidentally bent some of the pins on mine but I was able to carefully bend them back and luckily it booted. Computer building on those days was much more unfriendly. I remember hearing tons of horror stories of water cooling including leaks and the cpu overheating and dying that kept me far away from it until more recently.
>Computer building on those days was much more unfriendly
Anyone who heard “black-to-black, back-to-back” just a little too late would agree with you (and probably has mild PTSD)
PC building has become MUCH safer over the decades
> Computer building on those days was much more unfriendly.
Has it? CPUs ran much cooler back then. And gpus didn’t need brackets and special power cables and all that. Water cooling was harder. But also way less common or necessary.
CPUs now throttle to keep their temp in check. Back then, if the temp got too high they’d be permanently damaged like in the example above.
> Water cooling was harder. But also way less common or necessary.
Water cooling isn’t necessary today, either. The heat pipe heatsinks we have available are so good that you can cool every consumer CPU with a good air cooler. The automatic throttling means it won’t be a disaster if the temp gets high. Water cooling might get a few 100 more MHz during sustained all core workloads but it’s not necessary.
That's just it. My 12700k has maximum turbo of 190 watts. But it has giant air HSF that was trivial to install and I don't even HEAR the thing. No problems! Same with the 300watt gpu.
My Athlon 1100 was barely a third tdp at 60watts, yet was perilous to install, broke one (almost $1k, in 2001 money!), and then on the replacement had a 6500rpm screamer jet engine under my desk sitting on mobo that barely supported it and which could fail and kill my 1core cpu at any point :0
> CPUs now throttle to keep their temp in check. Back then, if the temp got too high they’d be permanently damaged like in the example above.
Well, it kinda depended on the CPU. And then sometimes the motherboard.
On the Intel side, P3s had a pin on Socket 370 to signal an immediate halt/shutdown. You could probably still damage the CPU, and I'm willing to bet a lot of the cheaper motherboards wouldn't respect it. But plenty did, since S370 needed other changes to support Coppermine anyway.
On the AMD Side, Athlon XPs and Morgan Durons did add a safety mechanism similar to the Pentium 3, but since the Thunderbird and Spitfire Durons did not, it was far far less supported.
Intel had THERMTRIP since the Pentium II (slot 1). This is an output signal, and didn't need any mobo involvement; the CPU automatically shuts down and stays stopped until reset with the temperature below the limit.
I think yes - everything was just more breakable. It feels today's heatsink, CPUs, motherboards, can handle the load and installation they're designed for (rtx 5xxx power connectors notwithstanding). I started building PCs in early 1990s and it was also just easy fun. Early 2000s were this weird time where even doing mainstream things you were absolutely supposed to, such as installing HSF on a CPU, was ridiculously perilous.
Mobos and cases also seem better designed these days. Yes some components need more stuff than before but it all seems to plug in and slot and wrap and screw in nicely if you're patient and methodical. Heck even the cases themselves are somewhat less likely to slot your wrists :).
I feel ~2003 or so was the cusp when I started seeing less builds that were just maddeningly messy and sharp and hyper loud and crazy, and started getting more predictable and sane and workable.
All anecdotal and based on personal perception of course :)
I vaguely remember some builds where setting the DIP switches wrong came with large all-caps text about how it would kill the CPU. Also some Pentium 2 CPUs with cooling shrouds that required duct work the rivaled my house's central HVAC. And yes I remember memory slots on the motherboard that needed to be feathering in or something would snap, and anti-static gloves being critical.
These days I slap the CPU to the motherboard and a giant heatsink with two huge fans and call it a day.
> even the cases themselves are somewhat less likely to slot your wrists
Fair. I have a scar on the knuckle of my index finger from breaking off the cover on a 5 1/4” drive bay. I was trying to push it out and it suddenly gave. My finger slammed into sharp tin and cut deep.
CPUs may run hotter but they have so much more protection. You can even power on and boot a modern Ryzen without even having a heatsink attached and it will not only survive, it will actually go in the BIOS and work for a while! One of these Socket A Athlons would be dead in a microsecond of having pushed the power button if you forgot to put on a heatsink. The Intel PIIIs of the time I believe were one of the first to implement thermal throttling and would crash but probably survive.
I'd say that CPUs run ~about as hot today as they did back then; power efficiency wasn't a primary concern. Regular computer users didn't care much about the thermals (it either worked or it didn't), and CPUs didn't do the thermally-limited clock boosts that are common today.
GPUs could be spooky hot. While the cards themselves didn't seem to care at all, the heat emanating from a 3dfx Voodoo3 card (I've had a 2000 and a 3500TV) could be enough to make other nearby cards stinky. Fan brackets and other hacks, from now long-defunct small companies like 3dfxcool, were pretty common.
We had PCI, then AGP, then PCI Express -- with overlaps. Woe be to those who errantly picked the older standard for their shiny new build; they would forever be stuck with their new hardware on the older bus, or with an even-deeper hole in their wallet.
There were other unfriendly things. RAM compatibility was weird; I remember a time around the turn of the century when double-sided RAM became inexpensive enough that was cheaper to buy 256MB of double-sided RAM that the PC may only use half of, than to buy 128MB of single-sided RAM.
Cables weren't always keyed, and when they were keyed it wasn't always in useful ways. Plugging a floppy drive in backwards was a common experience. A 40-pin IDE cable might have a key pin blocked off, while the hard drive had all 40 pins populated, and the two wouldn't fit together without modification.
Burning CDs was often a slow-moving disaster, involving feeding expensive blanks into an expensive drive installed in the machine every half-hour just to hope that things would work this time so you could finally listen to some of those MP3s in your Ford after you spent hours downloading them.
Unless you were foolish enough to buy an external drive, instead: As USB was attrociously-slow, those burners usually plugged into the printer port in what could be most-charitably described as an awful fucking mess.
And even if we weren't burning CDs, we had that pesky analog cable to deal with -- neither end of which was necessarily standardized -- so we could play Total Annihilation and listen to its (rather excellent) soundtrack at the same time.
We still had things like sound cards and modems and IRQs to deal with. NICs were often separate from motherboards. BIOS flashing felt like a dark and dangerous art. The fan headers often had different shapes, and usually only some of them were capable of ramping fan speeds.
At one point around that time, we rather quietly shifted the main power draw for new motherboards from the 5v rail to the 12v rail. Existing power supplies weren't always ready for this and it led some down expensive and unexpected paths.
I enjoyed working with PC hardware back then. But it was all pretty unfriendly, and I haven't even mentioned yet how much fun SCSI was to deal with.
Nowadays: We buy a fish tank and install a motherboard and power supply into it. The smart builder has already got the NVMe drive, CPU, and cooler fastened down into place in their purposeful mounts before this point. The fan headers all work. If we saved enough pennies, we add a GPU -- and it slots into the same x16 PCI Express socket that we've been using for over 20 years now. The NICs (often a plurality of them!) are built-in. There are no dip switches or configuration jumpers.
All that's left is to plug it in and install an OS. Easy.
Yup, I killed an Athlon XP in a ridiculously (over-)expensive Voodoo Gaming computer in a similar fashion while updating the HSF, as careful as I was. I've been nervous and anxious every time I update a chip or hear sink ever since even though it's a MUCH quieter affair nowadays...
Speaking of uphill every way, I also remember upgrading to a Alpha PAL screaming 6500rpm++ CPU fan+heatsink that sounded like a mix of jet engine or broken vacuum cleaner. Ridiculously loud - this was a few months before we as a society figured out that larger fans spinning more slowly were all around better for everyone :). I think it was something tiny like 60mm diameter industrial fan.
Heh, one of my most prized posessions in my teens was a GlobalWin FOP38 with that PAL6035 you mentioned mounted. An absolutely CRAZY fan that had no objective right to exist, much less in a teenager's Socket A Celeron rig.
I remember the day it arrived in the mail and I installed it - my family had just recently moved into a new house where my own room was located directly above the living room. When I started my PC with the new cooler and fan for the first time, the noise (esp. the obscene vibrations through the floor/ceiling!) actually prompted my mother to visit and check what's up/wrong :D
I ended up rigging together two chopping boards and a few pristine cleaning sponges as a kind of pedestal for my big tower case to rest upon, and was allowed to operate that way for the months to come. If it weren't for the over-ear headphones I wore all day back then, I guess I would have had suffered hearing loss from the noise exposure rather sooner than later...
I had a "small" Mini Super Orb on my thunderbird, more reasonable then the original Super Orb, and yet my neighbors of the floor above came to complain about my computer's noise
Somehow I managed to sleep with it on though xd
Mounting and dismounting it was very scary, and seeing it hang vertically with the cpu attached made me have cold sweats every time O _ o
Hardcore builders will remove the lid of their CPU to get the best possible thermal contact. The performance improvement is very small and not worth it but the bragging rights are fun. Unless you destroy the chip while delidding it, that is.
The article mentions this, but this was before CPUs had lids. When delidding a CPU, you have to be very careful not to crack the silicon die that you are applying force directly to. In the case of this Athlon, even care wasn't enough.
Interestingly, AMD's previous gen processors, the K6/K6-2/K6-III, were all lidded flip chips. A delidded K6-era CPU looks remarkably similar to the early Athlons which did away with the lid.
Your comment reminded me of my K6 (or was it a K7) based PC I had assembled as a high schooler of modest means.
My goto testing protocol was to recompile the Linux kernel and make GNU chess play itself at the highest settings.
I was quite disheartened to find that my rig would crash out repeatedly. Not a single successful compilation of the kernel. Then on a whim I removed the case and pointed my table fan at it. Voila, everything worked.
I had planned to get some more thermal paste and redo the thermals and put the case back. But I never did.
Then one day the dialer of my landline got fried by a thunderstorm. No problem I used to dial using my modem and my skeletonic linux box.
I tried everything for my IBM 6x86 PR166+. Everything. Pretty much every BIOS setup option, fans, reapplying thermal paste… still, the machine would often not survive a Linux kernel build without SIGSEGV, or just randomly hang.
Much later I finally noticed that the PSU, that I had in this same case ever since the PC started out as a 486, was something ridiculous, like 75W or so. I got a new cheap PSU, and finally the spookiness stopped.
Table fan was the way to go in our first LAN party during the summer of the year 2000 in the south of Spain. These computer case side panels were all piled against the wall.
as a kid I once had a PC that would make a very loud siren noise when it was overheating. I don't know why they thought that was a good idea but I did solve it by pointing a fan at it. Somehow my parents never came and asked what's that noise.
The article is mistaken, all the lidded chips (discussed here) are also flip chip/C4. Just with a metal plate on top. That would be in contrast to the earlier lidded chips (like a 68k or i486) with ceramic packages, where the chip is mounted right side up on the ceramic package and then wire bonded, the lid just covers the bond wires.
I learned (can't remember where) that I should try to rotate the cooler a little bit, with the mounting clamp still on, before undoing the clamp and removing it.
I still have a working Athlon XP / KT133A with ISA under my desk.
I've been assembling my own desktop PCs since I was about 15, but I have never gotten comfortable with just how much pressure it takes to affix some heat sinks. I think it was my Thunderbird 1200 where you had to press a slotted-head screwdriver on the lever to force it down and under the clasp.
I'm pretty sure I had exactly that heat sink and man was the pressure crazy (Thermalright IIRC). I'm pretty sure my screwdriver slipped and gashed the board, luckily I didn't notice damage.
Is that actually the silicon that's on top of a cpu?
I've always assumed the silicon is super thin - "wafer" thin. And then some material is added where the bonding wires etc are embedded in and the pads or pins are attached to.
For sure it's like this on newer CPUs, where there is not a single piece of silicon bit rather multiple ones with interconnect between them.
Chips are made from silicon wafers that are ~0.8mm thick. The transistor layer is built on one side of the wafer (because the transistors need to be built from epitaxial silicon, that is, where all the atoms are neatly arranged, and you cannot deposit a perfect epitaxial layer). Then the metal layers are built on top, by depositing copper and dielectric (traditionally silicon dioxide, that is, glass, but these days some more exotic stuff is used too).
Then that chip is flipped around, so the transistors and metal layers face the pcb. Most of the chip on the upside is just bulk silicon, and there for structural and heat spreading purposes, it's not needed for operation.
When chips are built out of multiple stacked layers, typically they thin the chips that go below and leave a thick bulk layer on the topmost one. IIRC, for some AMD chips they needed to use a separate, unprocessed bulk heat spreader layer for bonding reasons.
I would guess that the effective layers are super thin. But such super thin parts do not have much structural integrity. So you leave material under it so that you can handle it. And then just drill to those parts as needed.
On the bottom [0], youtuber/cooling specialist/tuner der8auer did an interview with an intel engineer a few years ago. One of the points they bring up is that it's not a pure gain to have less silicon between the logic and the next interface to another material conducting heat away as the silicon itself acts as a conductor/spreader, i.e. sideways as well as upwards, so there's a balance.
No, that’s the die sawed out of a wafer. Those are wafer-thin, i.e. about 0.8mm. On some newer CPUs the wafer is ground down to improve thermals. In normal C4 packaging like here the active layer is sandwiched between the metal layers (which sit on the interposer PCB) and the bulk silicon (95% of the thickness upwards, to the camera). Silicon is not a great heat conductor.
There was just this video [0] yesterday showing the whole process for a GaN LED, it's fascinating how crude the post processing can be in contrast with the ultra precise etching and layering.
The first computer I bought with my own money was an Athlon XP. Probably got the motherboard and CPU from Comp USA or Office Max or something. I bought it specifically to play The Elder Scrolls: Morrowind! It felt like it was obsolete in a week, the way Moore's Law was rocking at that time.
And yeah, when I opened the box and saw that the CPU had a fragile ceramic/glass top like that, I had a mini panic attack.
These days, of course I have a few in my retro collection, but I'll never remove the heatsinks. The CPU and motherboards might as well be one unit, never to be separated again.
This happened to me several times with Athlons in 2000, the die on these would break exceptionally easily such that it was a known issue at the time. You had to be quite careful.
The first time I built a PC, and the first time I mounted a heatsink, was to one of these bare die AMD CPUs. I heard about the fragility of these, and must have read the directions about ten times, and paused and checked each step about five times while doing it. It came out fine, but it was nerve wracking as hell.
Overclocking an Athlon XP 2500+ to a 3200+ was pretty cool. Easiest overclock ever.
Even as a naive teen, I always found the mechanical forces required to clamp cpu heatsinks and fans brutally high considering the small ceramic package underneath.
No. AMD was taking 3200+ chips and binning/selling them as 2500+ chips. All it took was increasing the FSB speed in the BIOS, and a bit of luck from the silicon lottery. No hardware mods.
My first build was an AMD "Venice" 3000+. Stock 2.8 GHz, I think I got it to 3.4 GHz with a DFI LANParty! motherboard and 512MB of DDR-600. Second CPU was a Pentium D, the first Intel generation of dual-core. Stock 2.66 GHz, I got it to 3.6 GHz. Almost 1GHz of overclock! Prime95 stable!
It's nice that modern CPUs self-overclock well, and my 9850X3D is by far the best CPU I've ever owned. But I miss the tweaking days a little. Fun hobby.
This must be what petrol heads fear about EVs. Sure, they're simpler and less maintenance and healthier. But you can't overbore it, or add a turbo, or change the cams.
> This must be what petrol heads fear about EVs. Sure, they're simpler and less maintenance and healthier. But you can't overbore it, or add a turbo, or change the cams.
I did learn from a Youtube vid not so long ago that you can flash cheaper Taycans to their high performance "turbo" "bios" if you have the nerve!
I've got one sitting next to me that I'm in the process of rejuvenating. The north bridge has a small fan with integrated heatsink/boat that I'm searching for. The CPU fan is still okay. It's a pre-PCIe system, and has a VGA output (no DVI or HDMI).
The main reason I'm keeping it around is that it's got 5-1/4" & 3-1/2" floppy drives, and T1000 and DDR4 tape drives.
the fan is likely a 12vdc 40x40mm, relatively shallow one (not very high airflow and not loud) one so you should be able to get a replacement from mouser or digikey. it's basically the same thing as a 80mm case fan from the same era scaled down.
A month ago, I ordered one from a local vendor who mostly sells parts for old electronic gaming hardware. They substituted a similar 40mm fan, but the one they gave me didn't include a heat sink, or attach to the old heat sink boat, had fewer blades, and spun in the opposite direction. It was only $15 so I didn't bother to return it. I saw some others online that looked promising, but with a very long (October) lead time. Still looking.
Oh man, the nostalgia. I had an Athlon XP too, and I still remember that fear when installing the heatsink. One wrong move and you're basically screwed...
You'd need unlimited money and hyperadvanced future technology. There's just no way to get things lined up and no way to reconnect any of it and no way to reclaim or reattach any of the tiny bits of silicon that preferred to become dust rather than take sides.
If you had enough money to convince AMD to sell you the tapeout, and if you could hire a fab using a similar process, you could make a new CPU. Other than that... probably not.
The next best thing would be buying a replacement.
Maybe, you would have to buy AMD or make them sell you the schema/chip masks for this chip, then buy ASML and ask them to spend the next 10 years figuring out how to regrow silicon on the broken chip layer by layer while also re-exposing the missing part. You would probably also need to buy a chunk of TSMC since this would require advanced chemical processing such that you can re-grow and re-expose a part of the chip without destroying the existing part. You would also need a lot of investment in scanning technology, to measure exactly what the current state of the chip is and what needs rebuilding and where.
Socket 370 and Socket A were very easy to break. You needed a flat head screwdriver to pry off the metal bracket which held the cooler in place. Apply a little too much force (or just repeat the process a few times) and watch the plastic nook break off of the socket, rendering the entire mainboard useless. Zip-ties or super glue might save you, if just for a while. But good luck moving/transporting the computer then.
I always did appreciate the 'nicer' heatsinks that spent the extra bit of cost of materials/machining/etc to provide a proper, safe 'slot' for the screwdriver.
That said, my Socket A system that would move around alot between home/work/school... After the horrors I had seen at work and a particularly unlucky guy at college [0] I made sure that the board I got had the 4 hole setup option (GA-7DXR, awesome board except for the capacitor plague being a thing), and managed to get Alpha Novatech to sell me a discounted (for evaluation) PAL8045 cooler.
For it's era, probably the most durable setup you could have for a Socket A, even though we didn't have Case mounting holes for thermal solutions yet. It also didn't have the torquing risks you would occasionally see on the P4's cam lever setup with cheaper boards.
[0] - Guy tried installing a fancy dragon orb cooler, ground a corner off the CPU, the CPU became a board killer. (I still don't know why he tried another board after seeing the damage...)
Even modern silicon only has some 60 of them right?
Plus don't transformers work in layers, making it natural to put a bunch of layers on top of each other? Regarding cooling, couldn't there be liquid cooling through a plumbing network in the same die?
Our brain takes quite a large volume, and we want chips that have a similar level of parts count (achieved - 80 billion neurons) and interconnectedness (not yet, around 10k synapses per neuron). Why do it in 80 2D layers? Is there no way to etch silicon in 3D?
Ouch, that poor die :(
I remember it well, back in the very early 2000s, when my own Athlon XP 1800+ (an AGOIA stepping, I do know that for sure :D) met its untimely end at the hands of a device that was designed to prevent the exact misfortune that killed its sister CPU as seen in this submission!
The fragile silicon was a problem the enthusiast/PC-builder community was very aware of, and the market responded with "spacer" kits you could buy for a few $$$ that were cut from solid sheets of thin copper. These spacers would carefully wrap around all the little bumps on the CPU's upper side, and provide a level plane for the heatsink to rest upon, while still establishing direct contact with the die that it was supposed to cool down. So there was, in theory, decidedly less of a chance of killing your CPU by having the heatsink meet it in an unfortunate angle, or something to that effect.
Unfortunately, the spacer that I had bought turned out to be just a wee little bit too thick, and, despite thermal paste (Arctic Silver!), the CPU did not in fact make contact with the heatsink (something shiny from Thermalright or Zalman, I think?) at all, which killed it instantly as soon as I powered up the rig for the first time. That's what it was like back in the days, when we got to school walking ten miles uphill (both ways) barefoot in deep snow.
I had to put in some overtime at the small PC shop I worked after school back then, to afford a measly Duron 800 as its replacement. And still they keep claiming that time heals all wounds... ;)
which killed it instantly as soon as I powered up the rig for the first time
For those who haven't seen this infamous video: https://www.youtube.com/watch?v=Xf0VuRG7MN4
At the time, Intel had on-die thermal sensors and automatic throttling while AMD didn't.
I accidentally bent some of the pins on mine but I was able to carefully bend them back and luckily it booted. Computer building on those days was much more unfriendly. I remember hearing tons of horror stories of water cooling including leaks and the cpu overheating and dying that kept me far away from it until more recently.
>Computer building on those days was much more unfriendly Anyone who heard “black-to-black, back-to-back” just a little too late would agree with you (and probably has mild PTSD) PC building has become MUCH safer over the decades
> Computer building on those days was much more unfriendly.
Has it? CPUs ran much cooler back then. And gpus didn’t need brackets and special power cables and all that. Water cooling was harder. But also way less common or necessary.
> Has it? CPUs ran much cooler back then.
CPUs now throttle to keep their temp in check. Back then, if the temp got too high they’d be permanently damaged like in the example above.
> Water cooling was harder. But also way less common or necessary.
Water cooling isn’t necessary today, either. The heat pipe heatsinks we have available are so good that you can cool every consumer CPU with a good air cooler. The automatic throttling means it won’t be a disaster if the temp gets high. Water cooling might get a few 100 more MHz during sustained all core workloads but it’s not necessary.
That's just it. My 12700k has maximum turbo of 190 watts. But it has giant air HSF that was trivial to install and I don't even HEAR the thing. No problems! Same with the 300watt gpu.
My Athlon 1100 was barely a third tdp at 60watts, yet was perilous to install, broke one (almost $1k, in 2001 money!), and then on the replacement had a 6500rpm screamer jet engine under my desk sitting on mobo that barely supported it and which could fail and kill my 1core cpu at any point :0
> CPUs now throttle to keep their temp in check. Back then, if the temp got too high they’d be permanently damaged like in the example above.
Well, it kinda depended on the CPU. And then sometimes the motherboard.
On the Intel side, P3s had a pin on Socket 370 to signal an immediate halt/shutdown. You could probably still damage the CPU, and I'm willing to bet a lot of the cheaper motherboards wouldn't respect it. But plenty did, since S370 needed other changes to support Coppermine anyway.
On the AMD Side, Athlon XPs and Morgan Durons did add a safety mechanism similar to the Pentium 3, but since the Thunderbird and Spitfire Durons did not, it was far far less supported.
Intel had THERMTRIP since the Pentium II (slot 1). This is an output signal, and didn't need any mobo involvement; the CPU automatically shuts down and stays stopped until reset with the temperature below the limit.
I think yes - everything was just more breakable. It feels today's heatsink, CPUs, motherboards, can handle the load and installation they're designed for (rtx 5xxx power connectors notwithstanding). I started building PCs in early 1990s and it was also just easy fun. Early 2000s were this weird time where even doing mainstream things you were absolutely supposed to, such as installing HSF on a CPU, was ridiculously perilous.
Mobos and cases also seem better designed these days. Yes some components need more stuff than before but it all seems to plug in and slot and wrap and screw in nicely if you're patient and methodical. Heck even the cases themselves are somewhat less likely to slot your wrists :).
I feel ~2003 or so was the cusp when I started seeing less builds that were just maddeningly messy and sharp and hyper loud and crazy, and started getting more predictable and sane and workable.
All anecdotal and based on personal perception of course :)
I vaguely remember some builds where setting the DIP switches wrong came with large all-caps text about how it would kill the CPU. Also some Pentium 2 CPUs with cooling shrouds that required duct work the rivaled my house's central HVAC. And yes I remember memory slots on the motherboard that needed to be feathering in or something would snap, and anti-static gloves being critical.
These days I slap the CPU to the motherboard and a giant heatsink with two huge fans and call it a day.
> even the cases themselves are somewhat less likely to slot your wrists
Fair. I have a scar on the knuckle of my index finger from breaking off the cover on a 5 1/4” drive bay. I was trying to push it out and it suddenly gave. My finger slammed into sharp tin and cut deep.
I’m glad they don’t make them like that anymore.
CPUs may run hotter but they have so much more protection. You can even power on and boot a modern Ryzen without even having a heatsink attached and it will not only survive, it will actually go in the BIOS and work for a while! One of these Socket A Athlons would be dead in a microsecond of having pushed the power button if you forgot to put on a heatsink. The Intel PIIIs of the time I believe were one of the first to implement thermal throttling and would crash but probably survive.
I'd say that CPUs run ~about as hot today as they did back then; power efficiency wasn't a primary concern. Regular computer users didn't care much about the thermals (it either worked or it didn't), and CPUs didn't do the thermally-limited clock boosts that are common today.
GPUs could be spooky hot. While the cards themselves didn't seem to care at all, the heat emanating from a 3dfx Voodoo3 card (I've had a 2000 and a 3500TV) could be enough to make other nearby cards stinky. Fan brackets and other hacks, from now long-defunct small companies like 3dfxcool, were pretty common.
We had PCI, then AGP, then PCI Express -- with overlaps. Woe be to those who errantly picked the older standard for their shiny new build; they would forever be stuck with their new hardware on the older bus, or with an even-deeper hole in their wallet.
There were other unfriendly things. RAM compatibility was weird; I remember a time around the turn of the century when double-sided RAM became inexpensive enough that was cheaper to buy 256MB of double-sided RAM that the PC may only use half of, than to buy 128MB of single-sided RAM.
Cables weren't always keyed, and when they were keyed it wasn't always in useful ways. Plugging a floppy drive in backwards was a common experience. A 40-pin IDE cable might have a key pin blocked off, while the hard drive had all 40 pins populated, and the two wouldn't fit together without modification.
Burning CDs was often a slow-moving disaster, involving feeding expensive blanks into an expensive drive installed in the machine every half-hour just to hope that things would work this time so you could finally listen to some of those MP3s in your Ford after you spent hours downloading them.
Unless you were foolish enough to buy an external drive, instead: As USB was attrociously-slow, those burners usually plugged into the printer port in what could be most-charitably described as an awful fucking mess.
And even if we weren't burning CDs, we had that pesky analog cable to deal with -- neither end of which was necessarily standardized -- so we could play Total Annihilation and listen to its (rather excellent) soundtrack at the same time.
We still had things like sound cards and modems and IRQs to deal with. NICs were often separate from motherboards. BIOS flashing felt like a dark and dangerous art. The fan headers often had different shapes, and usually only some of them were capable of ramping fan speeds.
At one point around that time, we rather quietly shifted the main power draw for new motherboards from the 5v rail to the 12v rail. Existing power supplies weren't always ready for this and it led some down expensive and unexpected paths.
I enjoyed working with PC hardware back then. But it was all pretty unfriendly, and I haven't even mentioned yet how much fun SCSI was to deal with.
Nowadays: We buy a fish tank and install a motherboard and power supply into it. The smart builder has already got the NVMe drive, CPU, and cooler fastened down into place in their purposeful mounts before this point. The fan headers all work. If we saved enough pennies, we add a GPU -- and it slots into the same x16 PCI Express socket that we've been using for over 20 years now. The NICs (often a plurality of them!) are built-in. There are no dip switches or configuration jumpers.
All that's left is to plug it in and install an OS. Easy.
Yup, I killed an Athlon XP in a ridiculously (over-)expensive Voodoo Gaming computer in a similar fashion while updating the HSF, as careful as I was. I've been nervous and anxious every time I update a chip or hear sink ever since even though it's a MUCH quieter affair nowadays...
Speaking of uphill every way, I also remember upgrading to a Alpha PAL screaming 6500rpm++ CPU fan+heatsink that sounded like a mix of jet engine or broken vacuum cleaner. Ridiculously loud - this was a few months before we as a society figured out that larger fans spinning more slowly were all around better for everyone :). I think it was something tiny like 60mm diameter industrial fan.
Heh, one of my most prized posessions in my teens was a GlobalWin FOP38 with that PAL6035 you mentioned mounted. An absolutely CRAZY fan that had no objective right to exist, much less in a teenager's Socket A Celeron rig.
I remember the day it arrived in the mail and I installed it - my family had just recently moved into a new house where my own room was located directly above the living room. When I started my PC with the new cooler and fan for the first time, the noise (esp. the obscene vibrations through the floor/ceiling!) actually prompted my mother to visit and check what's up/wrong :D
I ended up rigging together two chopping boards and a few pristine cleaning sponges as a kind of pedestal for my big tower case to rest upon, and was allowed to operate that way for the months to come. If it weren't for the over-ear headphones I wore all day back then, I guess I would have had suffered hearing loss from the noise exposure rather sooner than later...
I had a "small" Mini Super Orb on my thunderbird, more reasonable then the original Super Orb, and yet my neighbors of the floor above came to complain about my computer's noise
Somehow I managed to sleep with it on though xd
Mounting and dismounting it was very scary, and seeing it hang vertically with the cpu attached made me have cold sweats every time O _ o
Hardcore builders will remove the lid of their CPU to get the best possible thermal contact. The performance improvement is very small and not worth it but the bragging rights are fun. Unless you destroy the chip while delidding it, that is.
The article mentions this, but this was before CPUs had lids. When delidding a CPU, you have to be very careful not to crack the silicon die that you are applying force directly to. In the case of this Athlon, even care wasn't enough.
Interestingly, AMD's previous gen processors, the K6/K6-2/K6-III, were all lidded flip chips. A delidded K6-era CPU looks remarkably similar to the early Athlons which did away with the lid.
There are three good pieces of history for you to look at.
First, the NexGen Nx686 prototype... that became the K6 after AMD bought NexGen.
https://web.archive.org/web/19990901212245/http://www.sandpi...
https://web.archive.org/web/20000831030211/http://www.sandpi...
https://web.archive.org/web/20000831030227/http://www.sandpi...
Second, the AMD K6 prototype... which came in a small footprint ceramic package.
https://www.cpushack.com/chippics/AMD/K6/AMDK6-Processor-CA-...
Third, the AMD K7... which also came in that small footprint ceramic package, on a Slot A PCB, with L2 cache chips.
http://www.cpu-galerie.de/html/amdk7athlonslot.html
Enjoy!
Man I sort of miss the cartridge CPU. I get why it's not a thing anymore, but man was it fun to slot in a CPU like you would an old SNES game.
https://news.ycombinator.com/item?id=49407748
Thanks for the nostalgia trip.
Pentium 4 had lids. AMD Athlon XP did not.
Your comment reminded me of my K6 (or was it a K7) based PC I had assembled as a high schooler of modest means.
My goto testing protocol was to recompile the Linux kernel and make GNU chess play itself at the highest settings.
I was quite disheartened to find that my rig would crash out repeatedly. Not a single successful compilation of the kernel. Then on a whim I removed the case and pointed my table fan at it. Voila, everything worked.
I had planned to get some more thermal paste and redo the thermals and put the case back. But I never did.
Then one day the dialer of my landline got fried by a thunderstorm. No problem I used to dial using my modem and my skeletonic linux box.
Great setup when the money is tight. Good times.
I tried everything for my IBM 6x86 PR166+. Everything. Pretty much every BIOS setup option, fans, reapplying thermal paste… still, the machine would often not survive a Linux kernel build without SIGSEGV, or just randomly hang.
Much later I finally noticed that the PSU, that I had in this same case ever since the PC started out as a 486, was something ridiculous, like 75W or so. I got a new cheap PSU, and finally the spookiness stopped.
Table fan was the way to go in our first LAN party during the summer of the year 2000 in the south of Spain. These computer case side panels were all piled against the wall.
as a kid I once had a PC that would make a very loud siren noise when it was overheating. I don't know why they thought that was a good idea but I did solve it by pointing a fan at it. Somehow my parents never came and asked what's that noise.
Flip-chip packaging is still common in laptops. Still a (albeit minor) risk if you're repasting an older laptop.
The article is mistaken, all the lidded chips (discussed here) are also flip chip/C4. Just with a metal plate on top. That would be in contrast to the earlier lidded chips (like a 68k or i486) with ceramic packages, where the chip is mounted right side up on the ceramic package and then wire bonded, the lid just covers the bond wires.
TBH I don’t think I’ve seen a lidded chip on anything other than a CPU.
Thank you for being honest.
I learned (can't remember where) that I should try to rotate the cooler a little bit, with the mounting clamp still on, before undoing the clamp and removing it.
I still have a working Athlon XP / KT133A with ISA under my desk.
I've been assembling my own desktop PCs since I was about 15, but I have never gotten comfortable with just how much pressure it takes to affix some heat sinks. I think it was my Thunderbird 1200 where you had to press a slotted-head screwdriver on the lever to force it down and under the clasp.
I'm pretty sure I had exactly that heat sink and man was the pressure crazy (Thermalright IIRC). I'm pretty sure my screwdriver slipped and gashed the board, luckily I didn't notice damage.
Is that actually the silicon that's on top of a cpu?
I've always assumed the silicon is super thin - "wafer" thin. And then some material is added where the bonding wires etc are embedded in and the pads or pins are attached to.
For sure it's like this on newer CPUs, where there is not a single piece of silicon bit rather multiple ones with interconnect between them.
Chips are made from silicon wafers that are ~0.8mm thick. The transistor layer is built on one side of the wafer (because the transistors need to be built from epitaxial silicon, that is, where all the atoms are neatly arranged, and you cannot deposit a perfect epitaxial layer). Then the metal layers are built on top, by depositing copper and dielectric (traditionally silicon dioxide, that is, glass, but these days some more exotic stuff is used too).
Then that chip is flipped around, so the transistors and metal layers face the pcb. Most of the chip on the upside is just bulk silicon, and there for structural and heat spreading purposes, it's not needed for operation.
When chips are built out of multiple stacked layers, typically they thin the chips that go below and leave a thick bulk layer on the topmost one. IIRC, for some AMD chips they needed to use a separate, unprocessed bulk heat spreader layer for bonding reasons.
I would guess that the effective layers are super thin. But such super thin parts do not have much structural integrity. So you leave material under it so that you can handle it. And then just drill to those parts as needed.
On the bottom [0], youtuber/cooling specialist/tuner der8auer did an interview with an intel engineer a few years ago. One of the points they bring up is that it's not a pure gain to have less silicon between the logic and the next interface to another material conducting heat away as the silicon itself acts as a conductor/spreader, i.e. sideways as well as upwards, so there's a balance.
[0] https://youtu.be/ljZt_TQegHE?t=120
No, that’s the die sawed out of a wafer. Those are wafer-thin, i.e. about 0.8mm. On some newer CPUs the wafer is ground down to improve thermals. In normal C4 packaging like here the active layer is sandwiched between the metal layers (which sit on the interposer PCB) and the bulk silicon (95% of the thickness upwards, to the camera). Silicon is not a great heat conductor.
There was just this video [0] yesterday showing the whole process for a GaN LED, it's fascinating how crude the post processing can be in contrast with the ultra precise etching and layering.
[0] https://www.youtube.com/watch?v=VHyoz8fFpUM
The first computer I bought with my own money was an Athlon XP. Probably got the motherboard and CPU from Comp USA or Office Max or something. I bought it specifically to play The Elder Scrolls: Morrowind! It felt like it was obsolete in a week, the way Moore's Law was rocking at that time.
And yeah, when I opened the box and saw that the CPU had a fragile ceramic/glass top like that, I had a mini panic attack.
These days, of course I have a few in my retro collection, but I'll never remove the heatsinks. The CPU and motherboards might as well be one unit, never to be separated again.
This happened to me several times with Athlons in 2000, the die on these would break exceptionally easily such that it was a known issue at the time. You had to be quite careful.
The first time I built a PC, and the first time I mounted a heatsink, was to one of these bare die AMD CPUs. I heard about the fragility of these, and must have read the directions about ten times, and paused and checked each step about five times while doing it. It came out fine, but it was nerve wracking as hell.
Overclocking an Athlon XP 2500+ to a 3200+ was pretty cool. Easiest overclock ever.
Even as a naive teen, I always found the mechanical forces required to clamp cpu heatsinks and fans brutally high considering the small ceramic package underneath.
Was that the ol’ pencil mod?
No. AMD was taking 3200+ chips and binning/selling them as 2500+ chips. All it took was increasing the FSB speed in the BIOS, and a bit of luck from the silicon lottery. No hardware mods.
Reminds of me of the Celeron 300A to 450 overclock.
I took that old Celeron further and managed 533 stable with a screaming delta fan and a component swap on the mobo, beefier VRMs.
Slap on a 256K stick of Winbond cache and boy did that thing scream for its time.
My first build was an AMD "Venice" 3000+. Stock 2.8 GHz, I think I got it to 3.4 GHz with a DFI LANParty! motherboard and 512MB of DDR-600. Second CPU was a Pentium D, the first Intel generation of dual-core. Stock 2.66 GHz, I got it to 3.6 GHz. Almost 1GHz of overclock! Prime95 stable!
It's nice that modern CPUs self-overclock well, and my 9850X3D is by far the best CPU I've ever owned. But I miss the tweaking days a little. Fun hobby.
This must be what petrol heads fear about EVs. Sure, they're simpler and less maintenance and healthier. But you can't overbore it, or add a turbo, or change the cams.
> This must be what petrol heads fear about EVs. Sure, they're simpler and less maintenance and healthier. But you can't overbore it, or add a turbo, or change the cams.
I did learn from a Youtube vid not so long ago that you can flash cheaper Taycans to their high performance "turbo" "bios" if you have the nerve!
[1] https://www.youtube.com/watch?v=7ygs-vu4D0U
> DFI LANParty
Now that is a name I have not heard in a long time
I've got one sitting next to me that I'm in the process of rejuvenating. The north bridge has a small fan with integrated heatsink/boat that I'm searching for. The CPU fan is still okay. It's a pre-PCIe system, and has a VGA output (no DVI or HDMI).
The main reason I'm keeping it around is that it's got 5-1/4" & 3-1/2" floppy drives, and T1000 and DDR4 tape drives.
the fan is likely a 12vdc 40x40mm, relatively shallow one (not very high airflow and not loud) one so you should be able to get a replacement from mouser or digikey. it's basically the same thing as a 80mm case fan from the same era scaled down.
A month ago, I ordered one from a local vendor who mostly sells parts for old electronic gaming hardware. They substituted a similar 40mm fan, but the one they gave me didn't include a heat sink, or attach to the old heat sink boat, had fewer blades, and spun in the opposite direction. It was only $15 so I didn't bother to return it. I saw some others online that looked promising, but with a very long (October) lead time. Still looking.
Oh man, the nostalgia. I had an Athlon XP too, and I still remember that fear when installing the heatsink. One wrong move and you're basically screwed...
Is it ever fixable assuming one had unlimited money and wants to fix instead of rebuy?
You'd need unlimited money and hyperadvanced future technology. There's just no way to get things lined up and no way to reconnect any of it and no way to reclaim or reattach any of the tiny bits of silicon that preferred to become dust rather than take sides.
If you had enough money to convince AMD to sell you the tapeout, and if you could hire a fab using a similar process, you could make a new CPU. Other than that... probably not.
The next best thing would be buying a replacement.
Yeah the cost of fixing it would get you the technology to make a new one.
Maybe, you would have to buy AMD or make them sell you the schema/chip masks for this chip, then buy ASML and ask them to spend the next 10 years figuring out how to regrow silicon on the broken chip layer by layer while also re-exposing the missing part. You would probably also need to buy a chunk of TSMC since this would require advanced chemical processing such that you can re-grow and re-expose a part of the chip without destroying the existing part. You would also need a lot of investment in scanning technology, to measure exactly what the current state of the chip is and what needs rebuilding and where.
You could change the chip on the substrate, but why, you could just get another cpu...
nope
Just jam the piece back in there, I'm sure it'll be fine. /s
Heat probably cracked it. I had a xp2000+ and even with a heavy copper heatsink and a 9000rpm delta fan, the thing still ran hot.
9000rpm?? And I thought my alpha pal 6035 was a screamer! I think it did 6500-7000rpm though I could be wrong. How did it sound? :)
Loud. Not my video, but the pitch sounds about right... https://youtu.be/-BeTVOgVkAA
Socket 370 and Socket A were very easy to break. You needed a flat head screwdriver to pry off the metal bracket which held the cooler in place. Apply a little too much force (or just repeat the process a few times) and watch the plastic nook break off of the socket, rendering the entire mainboard useless. Zip-ties or super glue might save you, if just for a while. But good luck moving/transporting the computer then.
I always did appreciate the 'nicer' heatsinks that spent the extra bit of cost of materials/machining/etc to provide a proper, safe 'slot' for the screwdriver.
That said, my Socket A system that would move around alot between home/work/school... After the horrors I had seen at work and a particularly unlucky guy at college [0] I made sure that the board I got had the 4 hole setup option (GA-7DXR, awesome board except for the capacitor plague being a thing), and managed to get Alpha Novatech to sell me a discounted (for evaluation) PAL8045 cooler.
For it's era, probably the most durable setup you could have for a Socket A, even though we didn't have Case mounting holes for thermal solutions yet. It also didn't have the torquing risks you would occasionally see on the P4's cam lever setup with cheaper boards.
[0] - Guy tried installing a fancy dragon orb cooler, ground a corner off the CPU, the CPU became a board killer. (I still don't know why he tried another board after seeing the damage...)
Dont forget the fun when screwdriver slipped and gouged delicate CPU data lines near the socket.
More people should think like this
Poor CPU.
My last desktop multimedia PC was powered by Athlon XP.
DIP and PGA packages for the win!
AFLAC!?
We need those microscope guys to take some shots of this.
Why are chips still built in only a few layers?
Even modern silicon only has some 60 of them right?
Plus don't transformers work in layers, making it natural to put a bunch of layers on top of each other? Regarding cooling, couldn't there be liquid cooling through a plumbing network in the same die?
Our brain takes quite a large volume, and we want chips that have a similar level of parts count (achieved - 80 billion neurons) and interconnectedness (not yet, around 10k synapses per neuron). Why do it in 80 2D layers? Is there no way to etch silicon in 3D?
Cooling by pumping water through silicon is an interesting idea. You should invent that.
Thanks for the encouragement, but this has already been invented. Microsoft announced it in September 2025 here:
https://news.microsoft.com/source/features/innovation/microf...
(The article date is included in the source code and referenced elsewhere on the Internet at that time.)