When downloading the Debian system, there are two options: ADM64 and i386, what is the difference between the two?
I386=intel 80386. In fact, i386 is commonly used as a generic term for Intel (Intel) 32-bit microprocessors.
AMD64, also known as "x86-64" or "x64", is a 64-bit computer processor architecture. It is built on the existing 32-bit x86 architecture, developed by AMD.
The following references:
The
x86 architecture first appeared in 1978 with the Intel 8086 CPU, which was developed from the Intel 8008 processor, while 8008 was developed from Intel 4004. Intel then introduced the 80186, 80286, 80386, and 80486. In design, the instruction set used by these subsequent processors is improved by adding new supported instructions on a 8086 basis, so it can be said to be backwards compatible (backward Compatibie), which means that the program running on 8086 will be able to run on 80486. Since all end in 86 (but 8088 is also x86), this series of instructions is called the x86 instruction set (schema). To be precise, the term x86 does not refer to a specific instruction set (CPU-specific), but to the general term of all instruction sets developed on the basis of 8086 (instruction set).
in fact, most of the CPUs that Intel introduced after 80486 were "x86", including Pentium, Pentium Pro, Pentium MMX, Pentium 2, and Pentium 3, Pentium 4, Pentium D, Core full range, (x86 series) Xeon. In addition, processor manufacturers using the x86 architecture are far from Intel, most notably AMD, and others via,cyrix. Specific products and features can be found in the appendix table. It can be said that the x86 architecture is the standard desktop-level CPU, you can buy the brand or the assembled PC, almost no x86 CPU. However, although all are x86, but can only indicate that the instruction set used is compatible with 8086, except for the 8086 instruction set, the support of other directives is not the same, the internal implementation of the micro-architecture microarchitecture is clearly different, It's like Linux and Unix are all POSIX-compliant OS, and there's a system call to fork, but the implementation of this call is really different.
starting from 1985 80386, Intel implemented a 32-bit CPU with a 32-bit extension to the x86 architecture, which was preceded by 16 bits. Intel names the x86 instruction set architecture that supports the 32-bit ia-32 (Intel Architecture 32bit). In fact, because of the dominance of the 32-bit x86 processor, the term "x86" is almost equal to IA-32, or 32-bit x86 or x86-32, for example, the 32-bit naming of Windows and Linux distributions is x86 (not x86-32 or ia-32). The later "x86-64" name also continued to reinforce this engagement. It is also emphasized here that x86 is a general term of a series of architectures that support 16-bit, 32-bit, and 64-bit directives.
the coming will still come, in 1999, AMD announced the X86-64 architecture. Its implementation is consistent with the previous 80386 ideas, continue to IA-32 extension, add 64-bit General register, certificate budget unit and logical operation, support 64-bit virtual address, forward-compatible ia-32. 2003 the first x86-64 processor was released, AMD operon. AMD also formally named X86-64 as AMD64. This Intel completely 2 B (behind the reasons), was previously amd younger brother followers of the footsteps of Intel, AMD first implemented a 64-bit civil desktop-level CPU. In fact, Intel also has its 64-bit plan, in 2004 IDF, Intel admitted its 64-bit program, named ia-32e, namely ia-32 extension, then changed to em64t,extended Memory technology, The final name is Intel64. In fact EM64T is almost the same as AMD64. Early EM64T was incompatible with AMD64 's small number of instructions, but due to AMD's leading position in 64-bit technology, INTEL2005 had to announce that it would be fully compatible with AMD64. So now the 64-bit version of the operating system has never been said to be for Intel or AMD. In terms of appellation, however, most vendors use x86-64 (x86_64, or x64) to call this architecture, thus maintaining neutrality.
the "i386" and "I486″" aliases for Intel 80386 and 80486 processors are more obvious. Because the number can not be used as a registered trademark, but also need a formal name, from 586 onwards to become Pentium, this time i586 as the most primitive Pentium of the micro-processing architecture "P5" Alias, the same i686 to P6 (Pentium Pro use), i786 corresponds to P68, or NetBurst architecture.
in my opinion, the use of "I-86" in the name of a package should be applied to CPUs that use this generation of micro-architectures. However, the microarchitecture is a hardware thing, so it is still a description of the instruction set applicable to the corresponding CPU usage. For example, i386 corresponds to IA-32, and because compatibility is certainly applicable to i486~i686;i686 not only for Pentium Pro, but also for K7 architecture.
the previous article mentioned that Intel was passive after AMD introduced civil 64-bit technology. In fact, Intel also has 64 x86 CPU plan, but because of the market's lack of attention, the lack of demand, has not been published. In the enterprise server and high performance computing market, where 64-bit demand is stronger, HP partnered with Intel in 1994 to start developing a new IA-64 architecture that is completely irrelevant to the x86 architecture. Obviously IA-64 incompatible x86-64 let alone ia32,ia64 is inherently 64-bit, based on explicit parallel instruction operations (EPIC). 2001 the first IA-64 architecture of the processor was released by Itanium itanium®. As can be seen, IA-64 and IA-32 in the implementation of nothing, and x86-64 is also fully parallel to the two architectures, x86 architecture of Intel Xeon is one of the competitors of Intel Itanium. Later, Intel also simply changed IA-64 to be called the Intel Itanium architecture.
The difference between AMD64 and i386
