Older Intel Processor Diagnostic Tool runs on XP, might be worth a try. If its not daft it should test all instructions and processor modules.

H3nrik V! wrote on 2023-03-12, 12:52:

Interesting. That proves that something's wrong with the first one. Puzzles me though, how it can run so much, and still be defective

I've never known an Intel CPU of that era to fail but because the Pentium III was known to be easy to OC, it's possible someone went a little overboard and damaged it enough for it to shut down the system when it receives certain instruction but not enough to kill it.

rasz_pl wrote on 2023-03-12, 16:44:

Older Intel Processor Diagnostic Tool runs on XP, might be worth a try. If its not daft it should test all instructions and processor modules.

That's interesting to know - for now I'll be returning the first CPU since the second is very stable so far. After 5-6 hours of gaming & ripping CDs I've had zero crashes with the new one.

red_avatar wrote on 2023-03-14, 07:46:

H3nrik V! wrote on 2023-03-12, 12:52:

Interesting. That proves that something's wrong with the first one. Puzzles me though, how it can run so much, and still be defective

I've never known an Intel CPU of that era to fail but because the Pentium III was known to be easy to OC, it's possible someone went a little overboard and damaged it enough for it to shut down the system when it receives certain instruction but not enough to kill it.

I would actually rather suspect it being an ESD damage the 1GHz wasn't that overclocking friendly AFAIK. And actually, the only actual damages due to long term overclocking, I remember having heard of, is electromigration or "sudden Northwood dead syndrome" ...

H3nrik V! wrote on 2023-03-14, 17:33:

I would actually rather suspect it being an ESD damage the 1GHz wasn't that overclocking friendly AFAIK. And actually, the only actual damages due to long term overclocking, I remember having heard of, is electromigration or "sudden Northwood dead syndrome" ...

Interesting ... I've now installed Windows XP and done a bunch more testing with 100% stability. I'm not very eager to put the old one back though - I'll just ship it back to the seller.

Since we're on the topic of partially dead CPUs, let me tell you guys about a Pentium D 945 that's absolutely fine under XP, but will not run Windows 7, 10, or Ubuntu.

It's the weirdest thing. Windows 7 and newer operating systems (Ubuntu 20 included) won't even install, usually rebooting before even making it to the setup screen. It's definitely the CPU, as moving it to a known good motherboard with known good RAM and SSD results in the same failure modes.

The only way I can get this CPU to boot Windows 7 or 10 is to install the OS using a different CPU, then slap in the Pentium D once the installation is complete. But even then, it's horribly unstable. Around 5 minutes after booting to desktop, the mouse cursor will begin moving rather choppily, then bam: blue screen.

Under Windows XP, the Pentium D is completely stable and will even loop CPU benchmarks all day. No idea what's wrong with it.

Standard Def Steve wrote on 2023-03-14, 20:59:

Since we're on the topic of partially dead CPUs, let me tell you guys about a Pentium D 945 that's absolutely fine under XP, but […]
Show full quote

Since we're on the topic of partially dead CPUs, let me tell you guys about a Pentium D 945 that's absolutely fine under XP, but will not run Windows 7, 10, or Ubuntu.

It's the weirdest thing. Windows 7 and newer operating systems (Ubuntu 20 included) won't even install, usually rebooting before even making it to the setup screen. It's definitely the CPU, as moving it to a known good motherboard with known good RAM and SSD results in the same failure modes.

The only way I can get this CPU to boot Windows 7 or 10 is to install the OS using a different CPU, then slap in the Pentium D once the installation is complete. But even then, it's horribly unstable. Around 5 minutes after booting to desktop, the mouse cursor will begin moving rather choppily, then bam: blue screen.

Under Windows XP, the Pentium D is completely stable and will even loop CPU benchmarks all day. No idea what's wrong with it.

Well since Windows XP is 32 bit and the CPU is 64 bit, perhaps the damage is only limited to specific 64 bit instructions? I'll assume you're installing 64 bit Windows 7 and 32 bit Windows XP.

red_avatar wrote on 2023-03-15, 07:26:

Standard Def Steve wrote on 2023-03-14, 20:59:

Since we're on the topic of partially dead CPUs, let me tell you guys about a Pentium D 945 that's absolutely fine under XP, but […]
Show full quote

Since we're on the topic of partially dead CPUs, let me tell you guys about a Pentium D 945 that's absolutely fine under XP, but will not run Windows 7, 10, or Ubuntu.

It's the weirdest thing. Windows 7 and newer operating systems (Ubuntu 20 included) won't even install, usually rebooting before even making it to the setup screen. It's definitely the CPU, as moving it to a known good motherboard with known good RAM and SSD results in the same failure modes.

The only way I can get this CPU to boot Windows 7 or 10 is to install the OS using a different CPU, then slap in the Pentium D once the installation is complete. But even then, it's horribly unstable. Around 5 minutes after booting to desktop, the mouse cursor will begin moving rather choppily, then bam: blue screen.

Under Windows XP, the Pentium D is completely stable and will even loop CPU benchmarks all day. No idea what's wrong with it.

Well since Windows XP is 32 bit and the CPU is 64 bit, perhaps the damage is only limited to specific 64 bit instructions? I'll assume you're installing 64 bit Windows 7 and 32 bit Windows XP.

Its certainly odd that it doesn't like 64bit operating systems, would be interesting to find out if its isolated to one of the CPUs or both are affected. Since the Pentium D is pretty much two separate CPUs sandwiched together on the same substrate it possible only one is defective and doesnt fully come into play till a 64bit aware OS tries to use the instructions. Though I suspect something else is going on here, possibly a hardware/resource conflict somewhere.

red_avatar wrote on 2023-03-15, 07:26:

Standard Def Steve wrote on 2023-03-14, 20:59:

Since we're on the topic of partially dead CPUs, let me tell you guys about a Pentium D 945 that's absolutely fine under XP, but […]
Show full quote

Since we're on the topic of partially dead CPUs, let me tell you guys about a Pentium D 945 that's absolutely fine under XP, but will not run Windows 7, 10, or Ubuntu.

It's the weirdest thing. Windows 7 and newer operating systems (Ubuntu 20 included) won't even install, usually rebooting before even making it to the setup screen. It's definitely the CPU, as moving it to a known good motherboard with known good RAM and SSD results in the same failure modes.

The only way I can get this CPU to boot Windows 7 or 10 is to install the OS using a different CPU, then slap in the Pentium D once the installation is complete. But even then, it's horribly unstable. Around 5 minutes after booting to desktop, the mouse cursor will begin moving rather choppily, then bam: blue screen.

Under Windows XP, the Pentium D is completely stable and will even loop CPU benchmarks all day. No idea what's wrong with it.

Well since Windows XP is 32 bit and the CPU is 64 bit, perhaps the damage is only limited to specific 64 bit instructions? I'll assume you're installing 64 bit Windows 7 and 32 bit Windows XP.

Hmm, you know, that could totally be it! Yep, the newer OSes are 64-bit. One of these days, I will try Win7-32 on that chip.

@TrashPanda: I don't believe it's a resource conflict, since moving the CPU to a different motherboard results in the same error. Also, using a different CPU in the original system results in smooth sailing. I, too wondered if one of the cores had gone bad. I tried disabling one, but the system would still crash running a newer OS. It's definitely possible that the system was disabling the working core, however.

red_avatar wrote on 2023-03-07, 21:01:

I figure someone probably had their fun OCing this CPU and now it's unreliable. I'll ask for a refund. The second one is on the way anyway.

Have you _ever_ seen such mythical processor? and I mean saw and not read about it on forums.
[/quote]

When I worked in a small mom and pop PC store in the early/mid 2000's (first IT job) we built maybe 20 to 30 Tyan board based custom systems, using Pentium 4 chips. I can't remember the exact model, but they 2.2Ghz, maybe 2.4 socket 478 machines. We built them up for a school I think, so they were fairly basic but good machines.
In the build we had two Pentium 4 chips turn up faulty.
They would both post/boot and seemingly operate normally until you tried to put a load or run any kind of a load (ie; installing an OS) they would then BSOD or power off all together.
We tested RAM other known working chips, PC Check preboot diagnostics the works, turns out the PC check software was correct in identifying the cache on both CPUs as faulty.

These were bought new from a supplier, I can't recall the supplier but it was a large OEM wholesaler. Both were from the same sku/stepping, and both replaced immediately.
I've never seen that happen before or since.