4 ms·

AMD designed a 64-bit architecture. There were many 64-bit architectures around long before x86-64 in 2000.

by Cabal 16y ago

melling 16y ago
It is probably worth noting that AMD brought 64-bit to the PC world. Intel was telling customers they didn't need it. AMD actually took a small bite out of Intel and Intel quickly changed its mind. Breaking the Wintel monopoly will spur a lot of innovation. A third choice like Chrome or Android on tablets or notebooks might do the trick.
- jolan 16y ago
  > Intel was telling customers they didn't need it. They were telling customers if they did need it they should buy Itanium.
- InclinedPlane 16y ago
  Intel was telling everyone they needed 64-bit (more accurately, post-32-bit) technology long before AMD created x86-64. It's called IA64, EPIC, or Itanium. Intel made a huge gamble, betting that the next-generation of CPUs would be founded on a radically different VLIW type architecture. But Intel goofed. Because they didn't factor in how long it would take to fully develop the new architecture (something that takes multiple iterations). The first generation of Itanium was hugely expensive, didn't run x86 code well, and couldn't easily masquerade as a plain Jane x86 machine. (Not to mention the problem of having to modify code so that it could be re-compiled for IA64.) Meanwhile, AMD looked at the same problem and came to a different conclusion. Rather than a revolutionary new 64-bit architecture they thought that an evolutionary architecture was preferable. Simply fix up a few of the most glaring problems of x86 (such as a dearth of registers), widen the bus to 64-bits and call it good. They probably figured that since everything under the hood is all RISC/ILP-ish hidden behind micro-code translation anyway the instruction set isn't super important, as long as its semi-reasonable. They concentrated on maximizing performance for unmodified code and providing good benefits for using actual 64-bit specific code (such as more registers, well-engineered 64-bit memory addressing, etc.) Itanium floundered while AMD64 became hugely popular very quickly. AMD64 processors ran x86 code faster than x86-only processors, so there was no reason not to buy them, even if you never ran any 64-bit code. But once you owned the hardware it was now just a matter of an OS update sometime in the future to upgrade to 64-bit. Allowing the x64 transition to be far more incremental and far less of a leap of faith than it would otherwise have been. Intel quickly realized their mistake and tried to implement a version of x86-64 using the net burst (Pentium 4) architecture. This led to some really sad machines and a very awkward time for Intel as they were hit with multiple setbacks at the same time. Eventually they succeeded with a very substantial redesign of their hardware architecture (the core series) which paid huge dividends in the 2nd iteration as it proved to be far superior even to AMD's offerings. It's not at all that Intel thought that the 32-bit world would last forever, it's just they that screwed up royally in figuring out where to go from there. Note that finally, many, many years too late, Itanium is actually a decent platform now.
  - e40 16y ago
    From a [native compiler] developers point of view, Intel's fatal flaw: the architecture is nearly impossible to write a compiler/assembler for. That's what Intel told my company (that makes native compilers). They said we wouldn't be able to make a high performing compiler because the instruction scheduling, etc were too complex and not documented well enough for us to do it. A port of our compiler, which normally takes a couple of months, was estimated on IA64 to be 1-2 years, for a high-performing result. This is where they failed. Oh, the AMD64 (what it was called when we started the port) took 2 months, and that was fully optimized.
    - InclinedPlane 16y ago
      Yeah, I only mentioned that in passing but it is a serious issue. All the more pressing because a lot of the promise of IA64 hinged on smart compilers. It was a dramatic change of course for CPU architectures since the previous trend had been to make the hardware smart enough to figure out how to take advantage of ILP on the fly, even with unmodified code (thus the superscalar architecture of the P5, the out of order & branch prediction strategies of P6, etc.)
  - Tamerlin 16y ago
    "The first generation of Itanium was hugely expensive, didn't run x86 code well, and couldn't easily masquerade as a plain Jane x86 machine." THAT was its biggest blunder -- the x86 part. It was pointless, and a waste of silicon, and it also throttled the chip back with a lot of extra overhead. On top of everything else, it added NO value whatsoever, but it did make introducing the chip a lot harder, and designing it cost a lot more than it should have. Had they stuck to their original plan of pushing the architecture as far as they could and using a binary translator to provide x86 emulation, they would probably have had a barnstormer that made even IBM quake in its boots. Instead, they hobbled it with the biggest brick they could find. "Meanwhile, AMD looked at the same problem and came to a different conclusion." In addition to the list you provided, AMD started abandoning backward compatibility in Long mode. "Eventually they succeeded with a very substantial redesign of their hardware architecture (the core series) which paid huge dividends in the 2nd iteration as it proved to be far superior even to AMD's offerings." It took Intel less than 2 years to launch Core after AMD hit the market over the head with... er, Sledgehammer. (Pun not intended, but that WAS AMD's code name.) Core was a 5-year project -- obviously, they were already working on it, and turning the cranks almost to the point of melting motherboards was clearly just to extend the life of the Netburst architecture beyond where Intel thought they'd need to. I suspect that their original hope was that they be able to scale up the performance of Itanium far enough to exceed x86 performance even in emulation, and since Itanium was a simpler architecture (without hardware x86 compatibility, it's a FAR simpler architecture with a LOT more floating point throughput per clock cycle even now), it would be cheaper to manufacture. Intel blundered with the hardware x86 silliness and AMD caught Intel with its pants down. I would, however, argue that AMD's ccNUMA architecture was at least as big a deal as the fact that they introduced 64-bit computing to the mainstream -- which was huge.
- rbanffy 16y ago
  > AMD brought 64-bit to the PC world No. For most people, Microsoft brought 64-bits to the PC world. Had Microsoft decided not to support AMD64 in Windows and opted for some other 64-bit x86 architecture, AMD64 would be a curiosity. It's a bit sad because I remember reading AMD's announcement on a 64-bit RISC workstation that was already a bit too old compared to its newer cousins.
  - konad 16y ago
    Rather than mod you down : 64bit servers is where it was at. All those AMD64s in the data centre with >4Gbs RAM XP64 was crippled by bad drivers.
    - rbanffy 16y ago
      I wonder if Alpha would have succeeded if, in addition to NT, Microsoft ported Office, the whole BackOffice family and Visual Studio. It's not that hard to blame them for the fact we still use hypertrophied 8080s.

AMD designed a 64-bit architecture. There were many 64-bit architectures around long before x86-64 in 2000.

by Cabal 16y ago

AMD designed a 64-bit architecture. There were many 64-bit architectures around long before x86-64 in 2000.

It is probably worth noting that AMD brought 64-bit to the PC world. Intel was telling customers they didn't need it. AMD actually took a small bite out of Intel and Intel quickly changed its mind. Breaking the Wintel monopoly will spur a lot of innovation. A third choice like Chrome or Android on tablets or notebooks might do the trick.

jolan 16y ago
> Intel was telling customers they didn't need it. They were telling customers if they did need it they should buy Itanium.
InclinedPlane 16y ago
Intel was telling everyone they needed 64-bit (more accurately, post-32-bit) technology long before AMD created x86-64. It's called IA64, EPIC, or Itanium. Intel made a huge gamble, betting that the next-generation of CPUs would be founded on a radically different VLIW type architecture. But Intel goofed. Because they didn't factor in how long it would take to fully develop the new architecture (something that takes multiple iterations). The first generation of Itanium was hugely expensive, didn't run x86 code well, and couldn't easily masquerade as a plain Jane x86 machine. (Not to mention the problem of having to modify code so that it could be re-compiled for IA64.) Meanwhile, AMD looked at the same problem and came to a different conclusion. Rather than a revolutionary new 64-bit architecture they thought that an evolutionary architecture was preferable. Simply fix up a few of the most glaring problems of x86 (such as a dearth of registers), widen the bus to 64-bits and call it good. They probably figured that since everything under the hood is all RISC/ILP-ish hidden behind micro-code translation anyway the instruction set isn't super important, as long as its semi-reasonable. They concentrated on maximizing performance for unmodified code and providing good benefits for using actual 64-bit specific code (such as more registers, well-engineered 64-bit memory addressing, etc.) Itanium floundered while AMD64 became hugely popular very quickly. AMD64 processors ran x86 code faster than x86-only processors, so there was no reason not to buy them, even if you never ran any 64-bit code. But once you owned the hardware it was now just a matter of an OS update sometime in the future to upgrade to 64-bit. Allowing the x64 transition to be far more incremental and far less of a leap of faith than it would otherwise have been. Intel quickly realized their mistake and tried to implement a version of x86-64 using the net burst (Pentium 4) architecture. This led to some really sad machines and a very awkward time for Intel as they were hit with multiple setbacks at the same time. Eventually they succeeded with a very substantial redesign of their hardware architecture (the core series) which paid huge dividends in the 2nd iteration as it proved to be far superior even to AMD's offerings. It's not at all that Intel thought that the 32-bit world would last forever, it's just they that screwed up royally in figuring out where to go from there. Note that finally, many, many years too late, Itanium is actually a decent platform now.
- e40 16y ago
  From a [native compiler] developers point of view, Intel's fatal flaw: the architecture is nearly impossible to write a compiler/assembler for. That's what Intel told my company (that makes native compilers). They said we wouldn't be able to make a high performing compiler because the instruction scheduling, etc were too complex and not documented well enough for us to do it. A port of our compiler, which normally takes a couple of months, was estimated on IA64 to be 1-2 years, for a high-performing result. This is where they failed. Oh, the AMD64 (what it was called when we started the port) took 2 months, and that was fully optimized.
  - InclinedPlane 16y ago
    Yeah, I only mentioned that in passing but it is a serious issue. All the more pressing because a lot of the promise of IA64 hinged on smart compilers. It was a dramatic change of course for CPU architectures since the previous trend had been to make the hardware smart enough to figure out how to take advantage of ILP on the fly, even with unmodified code (thus the superscalar architecture of the P5, the out of order & branch prediction strategies of P6, etc.)
- Tamerlin 16y ago
  "The first generation of Itanium was hugely expensive, didn't run x86 code well, and couldn't easily masquerade as a plain Jane x86 machine." THAT was its biggest blunder -- the x86 part. It was pointless, and a waste of silicon, and it also throttled the chip back with a lot of extra overhead. On top of everything else, it added NO value whatsoever, but it did make introducing the chip a lot harder, and designing it cost a lot more than it should have. Had they stuck to their original plan of pushing the architecture as far as they could and using a binary translator to provide x86 emulation, they would probably have had a barnstormer that made even IBM quake in its boots. Instead, they hobbled it with the biggest brick they could find. "Meanwhile, AMD looked at the same problem and came to a different conclusion." In addition to the list you provided, AMD started abandoning backward compatibility in Long mode. "Eventually they succeeded with a very substantial redesign of their hardware architecture (the core series) which paid huge dividends in the 2nd iteration as it proved to be far superior even to AMD's offerings." It took Intel less than 2 years to launch Core after AMD hit the market over the head with... er, Sledgehammer. (Pun not intended, but that WAS AMD's code name.) Core was a 5-year project -- obviously, they were already working on it, and turning the cranks almost to the point of melting motherboards was clearly just to extend the life of the Netburst architecture beyond where Intel thought they'd need to. I suspect that their original hope was that they be able to scale up the performance of Itanium far enough to exceed x86 performance even in emulation, and since Itanium was a simpler architecture (without hardware x86 compatibility, it's a FAR simpler architecture with a LOT more floating point throughput per clock cycle even now), it would be cheaper to manufacture. Intel blundered with the hardware x86 silliness and AMD caught Intel with its pants down. I would, however, argue that AMD's ccNUMA architecture was at least as big a deal as the fact that they introduced 64-bit computing to the mainstream -- which was huge.
rbanffy 16y ago
> AMD brought 64-bit to the PC world No. For most people, Microsoft brought 64-bits to the PC world. Had Microsoft decided not to support AMD64 in Windows and opted for some other 64-bit x86 architecture, AMD64 would be a curiosity. It's a bit sad because I remember reading AMD's announcement on a 64-bit RISC workstation that was already a bit too old compared to its newer cousins.
- konad 16y ago
  Rather than mod you down : 64bit servers is where it was at. All those AMD64s in the data centre with >4Gbs RAM XP64 was crippled by bad drivers.
  - rbanffy 16y ago
    I wonder if Alpha would have succeeded if, in addition to NT, Microsoft ported Office, the whole BackOffice family and Visual Studio. It's not that hard to blame them for the fact we still use hypertrophied 8080s.