3 ms·

Ken Shirriff Tracks Down Intel's Infamous Pentium FDIV Bug in the Silicon Itself

53 points by voxadam 2y ago

rwmj 2y ago
Blog spam for https://oldbytes.space/@kenshirriff/113606898880486330 https://oldbytes.space/@kenshirriff/113606898880486330
- kens 2y ago
  For anyone who prefers Bluesky, my thread is also at https://bsky.app/profile/did:plc:svh6dgjnpkdl4dhxahj4xvkv/post/3lcnn6na52s2d https://bsky.app/profile/did:plc:svh6dgjnpkdl4dhxahj4xvkv/po... I'm still working on a blog post, which will go into much more detail.
taylodl 2y ago
The Pentium FDIV Bug is what put AMD on the map. Until that time, AMD was struggling with getting traction in the lucrative market of corporate sales. AMD was seen as being the cheap, and inferior, alternative to Intel and was primarily used by people building their own systems on the cheap. This bug, and Intel's gross mishandling of the crises, severely damaged the Intel brand. People who would never have considered AMD processors before were now buying AMD processors. In fact, ADM went on the offensive and started producing processors that smoked Intel's Pentium lineup - and were cheaper. Thinking about it now 30 years after the fact, I'm not sure Intel ever fully recovered from this event.
- dboreham 2y ago
  Severe revisionism here from my perspective. AMD was a respected vendor of CPUs and other complex chips since the 1970s. Intel was known to be flaky since their exit of the memory business. The Intel "brand" is a figment of Intel's marketing department's imagination.
  - taylodl 2y ago
    AMD was licensed by Intel to produce x86 processors so that IBM would have a second source for their PCs. That licensing arrangement ended with the 386 (well, Intel ended it, AMD claimed it was still valid). AMD was forced to produce their own 386 which then prompted a lawsuit from Intel claiming their intellectual copyrights had been violated since AMD had to reverse-engineer the 386. That lawsuit was settled in AMD's favor just two years before this bug. During that time before the lawsuit was settled, AMD was on the ropes, and it didn't look like they were going to make it. Intel's PR fiasco with this bug put AMD back in the market and to their credit, AMD went on the offensive and went on to make outstanding products. As you probably know, there are even more stories of AMD besting Intel throughout the 90s and even on to present-day. I'd like to see someone put together a timeline of Intel missteps, starting with the 286.
- chasil 2y ago
  AMD actually sold the most popular RISC processor at this time, the AMD 29k. "They were, for a time, the most popular RISC chips on the market, [citation needed] widely used in laser printers from a variety of manufacturers. "Developed since 1984–1985, announced in March 1987 and released in May 1988, the initial Am29000 was followed by several versions, ending with the Am29040 in 1995." https://en.wikipedia.org/wiki/AMD_Am29000 https://en.wikipedia.org/wiki/AMD_Am29000
- tverbeure 2y ago
  Thats not how I remember it. Back then you already had the Am486 and the K5. They were fine but didn’t set the world on fire. It was the K6 that made a huge impact and that was a couple of years later (and a design they acquired externally.) Intel mishandled the situation initially, but the whole episode was quickly put to rest when they reversed course and replaced all units (though I never returned mine.)
  - chasil 2y ago
    I think AMD's purchase of NexGen was the external design for the K6. https://en.wikipedia.org/wiki/NexGen https://en.wikipedia.org/wiki/NexGen
analog31 2y ago
I wonder if the Pentium Navajo Rug has the same errors. ;-)
- kens 2y ago
  That's an interesting question. The Navajo rug that Intel commissioned shows the Pentium P54C. This is the higher-frequency version of the original Pentium (P5). The FDIV bug had been fixed by this point, so the rug does not have the FDIV bug. But you can see in the rug exactly where the FDIV bug was. I wrote about the Pentium Navajo rug here, by the way: https://www.righto.com/2024/08/pentium-navajo-fairchild-shiprock.html https://www.righto.com/2024/08/pentium-navajo-fairchild-ship...
  - rvba 2y ago
    Fascinaging stuff. This should be a separate submission!
    - kens 2y ago
      It was on HN a few months ago: https://news.ycombinator.com/item?id=41418301 https://news.ycombinator.com/item?id=41418301
glimshe 2y ago
"Intel inside, can't divide" as they used to say.
- KerrAvon 2y ago
  my favorite was letterman's non-sequitur: "How 'bout some defective Pentium salsa to go with those defective Pentium chips?"

53 points by voxadam 2y ago

rwmj 2y ago

Blog spam for https://oldbytes.space/@kenshirriff/113606898880486330 https://oldbytes.space/@kenshirriff/113606898880486330

kens 2y ago
For anyone who prefers Bluesky, my thread is also at https://bsky.app/profile/did:plc:svh6dgjnpkdl4dhxahj4xvkv/post/3lcnn6na52s2d https://bsky.app/profile/did:plc:svh6dgjnpkdl4dhxahj4xvkv/po... I'm still working on a blog post, which will go into much more detail.

taylodl 2y ago

The Pentium FDIV Bug is what put AMD on the map. Until that time, AMD was struggling with getting traction in the lucrative market of corporate sales. AMD was seen as being the cheap, and inferior, alternative to Intel and was primarily used by people building their own systems on the cheap. This bug, and Intel's gross mishandling of the crises, severely damaged the Intel brand. People who would never have considered AMD processors before were now buying AMD processors. In fact, ADM went on the offensive and started producing processors that smoked Intel's Pentium lineup - and were cheaper. Thinking about it now 30 years after the fact, I'm not sure Intel ever fully recovered from this event.

dboreham 2y ago
Severe revisionism here from my perspective. AMD was a respected vendor of CPUs and other complex chips since the 1970s. Intel was known to be flaky since their exit of the memory business. The Intel "brand" is a figment of Intel's marketing department's imagination.
- taylodl 2y ago
  AMD was licensed by Intel to produce x86 processors so that IBM would have a second source for their PCs. That licensing arrangement ended with the 386 (well, Intel ended it, AMD claimed it was still valid). AMD was forced to produce their own 386 which then prompted a lawsuit from Intel claiming their intellectual copyrights had been violated since AMD had to reverse-engineer the 386. That lawsuit was settled in AMD's favor just two years before this bug. During that time before the lawsuit was settled, AMD was on the ropes, and it didn't look like they were going to make it. Intel's PR fiasco with this bug put AMD back in the market and to their credit, AMD went on the offensive and went on to make outstanding products. As you probably know, there are even more stories of AMD besting Intel throughout the 90s and even on to present-day. I'd like to see someone put together a timeline of Intel missteps, starting with the 286.
chasil 2y ago
AMD actually sold the most popular RISC processor at this time, the AMD 29k. "They were, for a time, the most popular RISC chips on the market, [citation needed] widely used in laser printers from a variety of manufacturers. "Developed since 1984–1985, announced in March 1987 and released in May 1988, the initial Am29000 was followed by several versions, ending with the Am29040 in 1995." https://en.wikipedia.org/wiki/AMD_Am29000 https://en.wikipedia.org/wiki/AMD_Am29000
tverbeure 2y ago
Thats not how I remember it. Back then you already had the Am486 and the K5. They were fine but didn’t set the world on fire. It was the K6 that made a huge impact and that was a couple of years later (and a design they acquired externally.) Intel mishandled the situation initially, but the whole episode was quickly put to rest when they reversed course and replaced all units (though I never returned mine.)
- chasil 2y ago
  I think AMD's purchase of NexGen was the external design for the K6. https://en.wikipedia.org/wiki/NexGen https://en.wikipedia.org/wiki/NexGen

analog31 2y ago

I wonder if the Pentium Navajo Rug has the same errors. ;-)

kens 2y ago
That's an interesting question. The Navajo rug that Intel commissioned shows the Pentium P54C. This is the higher-frequency version of the original Pentium (P5). The FDIV bug had been fixed by this point, so the rug does not have the FDIV bug. But you can see in the rug exactly where the FDIV bug was. I wrote about the Pentium Navajo rug here, by the way: https://www.righto.com/2024/08/pentium-navajo-fairchild-shiprock.html https://www.righto.com/2024/08/pentium-navajo-fairchild-ship...
- rvba 2y ago
  Fascinaging stuff. This should be a separate submission!
  - kens 2y ago
    It was on HN a few months ago: https://news.ycombinator.com/item?id=41418301 https://news.ycombinator.com/item?id=41418301

glimshe 2y ago

"Intel inside, can't divide" as they used to say.

KerrAvon 2y ago
my favorite was letterman's non-sequitur: "How 'bout some defective Pentium salsa to go with those defective Pentium chips?"