 | This articleneeds additional citations forverification. Please helpimprove this article byadding citations to reliable sources. Unsourced material may be challenged and removed. Find sources: "64-bit computing" – news ·newspapers ·books ·scholar ·JSTOR(April 2023) (Learn how and when to remove this message) |
| Computer architecture bit widths |
|---|
| Bit |
| Application |
| Binary floating-pointprecision |
| Decimal floating-pointprecision |
Incomputer architecture,64-bitintegers,memory addresses, or otherdata units[a] are those that are 64bits wide. Also, 64-bitcentral processing units (CPU) andarithmetic logic units (ALU) are those that are based onprocessor registers,address buses, ordata buses of that size. Acomputer that uses such a processor is a 64-bit computer.
From the software perspective, 64-bit computing means the use ofmachine code with 64-bitvirtual memory addresses. However, not all 64-bit instruction sets support full 64-bit virtual memory addresses;x86-64 andAArch64, for example, support only 48 bits of virtual address, with the remaining 16 bits of the virtual address required to be all zeros (000...) or all ones (111...), and several 64-bit instruction sets support fewer than 64 bits of physical memory address.
The term64-bit also describes a generation of computers in which 64-bit processors are the norm. 64 bits is aword size that defines certain classes of computer architecture, buses, memory, and CPUs and, by extension, the software that runs on them. 64-bit CPUs have been used insupercomputers since the 1970s (Cray-1, 1975) and inreduced instruction set computers (RISC) basedworkstations andservers since the early 1990s. In 2003, 64-bit CPUs were introduced to the mainstreamPC market in the form of x86-64 processors and thePowerPC G5.
A 64-bit register can hold any of 264 (over 18quintillion or 1.8×1019) different values. The range ofinteger values that can be stored in 64 bits depends on theinteger representation used. With the two most common representations, the range is 0 through 18,446,744,073,709,551,615 (equal to 264 − 1) for representation as an (unsigned)binary number, and −9,223,372,036,854,775,808 (−263) through 9,223,372,036,854,775,807 (263 − 1) for representation astwo's complement. Hence, a processor with 64-bit memory addresses can directly access 264 bytes (16exabytes or EB) ofbyte-addressable memory.
With no further qualification, a64-bit computer architecture generally has integer and addressingregisters that are 64 bits wide, allowing direct support for 64-bit data types and addresses. However, a CPU might have externaldata buses oraddress buses with different sizes from the registers, even larger (the 32-bitPentium had a 64-bit data bus, for instance).[1]
| This sectiondoes notcite anysources. Please helpimprove this section byadding citations to reliable sources. Unsourced material may be challenged andremoved.(April 2023) (Learn how and when to remove this message) |
Processor registers are typically divided into several groups:integer,floating-point,single instruction, multiple data (SIMD),control, and often special registers for address arithmetic which may have various uses and names such asaddress,index, orbase registers. However, in modern designs, these functions are often performed by more general purposeinteger registers. In most processors, only integer or address-registers can be used to address data in memory; the other types of registers cannot. The size of these registers therefore normally limits the amount of directly addressable memory, even if there are registers, such as floating-point registers, that are wider.
Most high performance 32-bit and 64-bit processors (some notable exceptions are older or embeddedARM architecture (ARM) and 32-bitMIPS architecture (MIPS) CPUs) have integrated floating point hardware, which is often, but not always, based on 64-bit units of data. For example, although thex86/x87 architecture has instructions able to load and store 64-bit (and 32-bit) floating-point values in memory, the internal floating-point data and register format is 80 bits wide, while the general-purpose registers are 32 bits wide. In contrast, the 64-bitAlpha family uses a 64-bit floating-point data and register format, and 64-bit integer registers.
Many computerinstruction sets are designed so that a single integer register can store thememory address to any location in the computer's physical orvirtual memory. Therefore, the total number of addresses to memory is often determined by the width of these registers. TheIBMSystem/360 of the 1960s was an early 32-bit computer; it had 32-bit integer registers, although it only used the low order 24 bits of a word for addresses, resulting in a 16 MiB (16 × 10242 bytes) address space. 32-bitsuperminicomputers, such as theDECVAX, became common in the 1970s, and 32-bit microprocessors, such as theMotorola 68000 family and the32-bit members of the x86 family starting with theIntel 80386, appeared in the mid-1980s, making 32 bits something of ade facto consensus as a convenient register size.
A 32-bitaddress register meant that 232 addresses, or 4 GB ofrandom-access memory (RAM), could be referenced. When these architectures were devised, 4 GB of memory was so far beyond the typical amounts (4 MiB) in installations, that this was considered to be enoughheadroom for addressing. 4.29 billion addresses were considered an appropriate size to work with for another important reason: 4.29 billion integers are enough to assign unique references to most entities in applications likedatabases.
Somesupercomputer architectures of the 1970s and 1980s, such as theCray-1,[2] used registers up to 64 bits wide, and supported 64-bit integer arithmetic, although they did not support 64-bit addressing. In the mid-1980s,Intel i860[3] development began culminating in a 1989 release; the i860 had 32-bit integer registers and 32-bit addressing, so it was not a fully 64-bit processor, although its graphics unit supported 64-bit integer arithmetic.[4] However, 32 bits remained the norm until the early 1990s, when the continual reductions in the cost of memory led to installations with amounts of RAM approaching 4 GB, and the use of virtual memory spaces exceeding the 4 GB ceiling became desirable for handling certain types of problems. In response, MIPS and DEC developed 64-bit microprocessor architectures, initially for high-endworkstation andserver machines. By the mid-1990s,HAL Computer Systems,Sun Microsystems,IBM,Silicon Graphics, andHewlett-Packard had developed 64-bit architectures for their workstation and server systems. A notable exception to this trend weremainframes from IBM, which then used 32-bit data and 31-bit address sizes; the IBM mainframes did not include 64-bit processors until 2000. During the 1990s, several low-cost 64-bit microprocessors were used in consumer electronics and embedded applications. Notably, theNintendo 64[5] and thePlayStation 2 had 64-bit microprocessors before their introduction in personal computers. High-end printers, network equipment, and industrial computers also used 64-bit microprocessors, such as theQuantum Effect DevicesR5000.[6] 64-bit computing started to trickle down to the personal computer desktop from 2003 onward, when some models inApple's Macintosh lines switched toPowerPC 970 processors (termedG5 by Apple), andAdvanced Micro Devices (AMD) released its first 64-bitx86-64 processor. Physical memory eventually caught up with 32-bit limits. In 2023, laptop computers were commonly equipped with 16GB and servers starting from 64 GB of memory,[7] greatly exceeding the 4 GB address capacity of 32 bits.
In principle, a 64-bit microprocessor can address 16 EB (16 × 10246 = 264 = 18,446,744,073,709,551,616 bytes) of memory. However, not all instruction sets, and not all processors implementing those instruction sets, support a full 64-bit virtual or physical address space.
Thex86-64 architecture (as of March 2024[update]) allows 48 bits for virtual memory and, for any given processor, up to 52 bits for physical memory.[31][32] These limits allow memory sizes of 256 TB (256 × 10244 bytes) and 4 PB (4 × 10245 bytes), respectively. A PC cannot currently contain 4 petabytes of memory (due to the physical size of the memory chips), but AMD envisioned large servers, shared memory clusters, and other uses of physical address space that might approach this in the foreseeable future. Thus the 52-bit physical address provides ample room for expansion while not incurring the cost of implementing full 64-bit physical addresses. Similarly, the 48-bit virtual address space was designed to provide 65,536 (216) times the 32-bit limit of 4 GB (4 × 10243 bytes), allowing room for later expansion and incurring no overhead of translating full 64-bit addresses.
ThePower ISA v3.0 allows 64 bits for an effective address, mapped to a segmented address with between 65 and 78 bits allowed, for virtual memory, and, for any given processor, up to 60 bits for physical memory.[33]
The OracleSPARC Architecture 2015 allows 64 bits for virtual memory and, for any given processor, between 40 and 56 bits for physical memory.[34]
The ARMAArch64 Virtual Memory System Architecture allows from 48 to 56 bits for virtual memory and, for any given processor, from 32 to 56 bits for physical memory.[35]
TheDEC Alpha specification requires minimum of 43 bits of virtual memory address space (8 TB) to be supported, and hardware need to check and trap if the remaining unsupported bits are zero (to support compatibility on future processors).Alpha 21064 supported 43 bits of virtual memory address space (8 TB) and 34 bits of physical memory address space (16 GB).Alpha 21164 supported 43 bits of virtual memory address space (8 TB) and 40 bits of physical memory address space (1 TB).Alpha 21264 supported user-configurable 43 or 48 bits of virtual memory address space (8 TB or 256 TB) and 44 bits of physical memory address space (16 TB).
A change from a32-bit to a 64-bit architecture is a fundamental alteration, as mostoperating systems must be extensively modified to take advantage of the new architecture, because that software has to manage the actual memory addressing hardware.[36] Other software must also beported to use the new abilities; older 32-bit software may be supported either by virtue of the 64-bit instruction set being a superset of the 32-bit instruction set, so that processors that support the 64-bit instruction set can also run code for the 32-bit instruction set, or through softwareemulation, or by the actual implementation of a 32-bit processor core within the 64-bit processor, as with someItanium processors from Intel, which included anIA-32 processor core to run 32-bitx86 applications. The operating systems for those 64-bit architectures generally support both 32-bit and 64-bit applications.[37]
One significant exception to this is theIBM AS/400, software for which is compiled into a virtualinstruction set architecture (ISA) calledTechnology Independent Machine Interface (TIMI); TIMI code is then translated to native machine code by low-level software before being executed. The translation software is all that must be rewritten to move the full OS and all software to a new platform, as when IBM transitioned the native instruction set for AS/400 from the older 32/48-bitIMPI to the newer 64-bitPowerPC-AS, codenamedAmazon. The IMPI instruction set was quite different from even 32-bit PowerPC, so this transition was even bigger than moving a given instruction set from 32 to 64 bits.
On 64-bit hardware withx86-64 architecture (AMD64), most 32-bit operating systems and applications can run with no compatibility issues. While the larger address space of 64-bit architectures makes working with large data sets in applications such asdigital video, scientific computing, and largedatabases easier, there has been considerable debate on whether they or their 32-bitcompatibility modes will be faster than comparably priced 32-bit systems for other tasks.
A compiled Java program can run on a 32- or 64-bit Java virtual machine with no modification. The lengths and precision of all the built-in types, such aschar,short,int,long,float, anddouble, and the types that can be used as array indices, are specified by the standard and are not dependent on the underlying architecture. Java programs that run on a 64-bit Java virtual machine have access to a larger address space.[38]
Speed is not the only factor to consider in comparing 32-bit and 64-bit processors. Applications such as multi-tasking, stress testing, and clustering – forhigh-performance computing (HPC) – may be more suited to a 64-bit architecture when deployed appropriately. For this reason, 64-bit clusters have been widely deployed in large organizations, such as IBM, HP, and Microsoft.
Summary:
A common misconception is that 64-bit architectures are no better than 32-bit architectures unless the computer has more than 4 GB ofrandom-access memory.[39] This is not entirely true:
inta,b,c,d,e;for(a=0;a<100;a++){b=a;c=b;d=c;e=d;}
The main disadvantage of 64-bit architectures is that, relative to 32-bit architectures, the same data occupies more space in memory (due to longer pointers and possibly other types, and alignment padding). This increases the memory requirements of a given process and can have implications for efficient processor cache use. Maintaining a partial 32-bit model is one way to handle this, and is in general reasonably effective. For example, thez/OS operating system takes this approach, requiring program code to reside in 31-bit address spaces (the high order bit is not used in address calculation on the underlying hardware platform) while data objects can optionally reside in 64-bit regions. Not all such applications require a large address space or manipulate 64-bit data items, so these applications do not benefit from these features.
x86-based 64-bit systems sometimes lack equivalents ofsoftware that is written for 32-bit architectures. The most severe problem in Microsoft Windows is incompatibledevice drivers for obsolete hardware. Most 32-bit application software can run on a 64-bit operating system in acompatibility mode, also termed anemulation mode, e.g., MicrosoftWoW64 Technology for IA-64 and AMD64. The 64-bit Windows Native Mode[42] driver environment runs atop 64-bitNTDLL.DLL, which cannot call 32-bit Win32 subsystem code (often devices whose actual hardware function is emulated in user mode software, like Winprinters). Because 64-bit drivers for most devices were unavailable until early 2007 (Vista x64), using a 64-bit version of Windows was considered a challenge. However, the trend has since moved toward 64-bit computing, more so as memory prices dropped and the use of more than 4 GB of RAM increased. Most manufacturers started to provide both 32-bit and 64-bit drivers for new devices, so unavailability of 64-bit drivers ceased to be a problem. 64-bit drivers were not provided for many older devices, which could consequently not be used in 64-bit systems.
Driver compatibility was less of a problem with open-source drivers, as 32-bit ones could be modified for 64-bit use. Support for hardware made before early 2007, was problematic for open-source platforms,[citation needed] due to the relatively small number of users.
64-bit versions of Windows cannot run16-bit software. However, most 32-bit applications will work well. 64-bit users are forced to install avirtual machine of a 16- or 32-bit operating system to run 16-bit applications or use one of the alternatives forNTVDM.[43]
Mac OS X 10.4 "Tiger" andMac OS X 10.5 "Leopard" had only a 32-bit kernel, but they can run 64-bit user-mode code on 64-bit processors.Mac OS X 10.6 "Snow Leopard" had both 32- and 64-bit kernels, and, on most Macs, used the 32-bit kernel even on 64-bit processors. This allowed those Macs to support 64-bit processes while still supporting 32-bit device drivers; although not 64-bit drivers and performance advantages that can come with them.Mac OS X 10.7 "Lion" ran with a 64-bit kernel on more Macs, andOS X 10.8 "Mountain Lion" and latermacOS releases only have a 64-bit kernel. On systems with 64-bit processors, both the 32- and 64-bit macOS kernels can run 32-bit user-mode code, and all versions of macOS up to macOS Mojave (10.14) include 32-bit versions of libraries that 32-bit applications would use, so 32-bit user-mode software for macOS will run on those systems. The 32-bit versions of libraries have been removed by Apple in macOS Catalina (10.15).
Linux and most otherUnix-like operating systems, and theC andC++toolchains for them, have supported 64-bit processors for many years. Many applications and libraries for those platforms areopen-source software, written in C and C++, so that if they are 64-bit-safe, they can be compiled into 64-bit versions. This source-based distribution model, with an emphasis on frequent releases, makes availability of application software for those operating systems less of an issue.
In 32-bit programs,pointers and data types such as integers generally have the same length. This is not necessarily true on 64-bit machines.[44][45][46] Mixing data types in programming languages such asC and its descendants such asC++ andObjective-C may thus work on 32-bit implementations but not on 64-bit implementations.
In many programming environments for C and C-derived languages on 64-bit machines,int variables are still 32 bits wide, but long integers and pointers are 64 bits wide. These are described as having anLP64data model, which is an abbreviation of "Long, Pointer, 64".[47][48] Other models are theILP64 data model in which all three data types are 64 bits wide,[49][48] and even theSILP64 model whereshort integers are also 64 bits wide.[50][51] However, in most cases the modifications required are relatively minor and straightforward, and many well-written programs can simply be recompiled for the new environment with no changes. Another alternative is theLLP64 model, which maintains compatibility with 32-bit code by leaving bothint andlong as 32-bit.[52][48]LL refers to thelong long integer type, which is at least 64 bits on all platforms, including 32-bit environments.
There are also systems with 64-bit processors using anILP32 data model, with the addition of 64-bit long long integers; this is also used on many platforms with 32-bit processors. This model reduces code size and the size of data structures containing pointers, at the cost of a much smaller address space, a good choice for some embedded systems. For instruction sets such as x86 and ARM in which the 64-bit version of the instruction set has more registers than does the 32-bit version, it provides access to the additional registers without the space penalty. It is common in 64-bit RISC machines,[citation needed] explored in x86 asx32 ABI, and has recently been used in theApple Watch Series 4 and 5.[53][54]
| Data model | short int | int | long int | long long | Pointer, size_t | Sample operating systems |
|---|---|---|---|---|---|---|
| ILP32 | 16 | 32 | 32 | 64 | 32 | x32 andarm64ilp32 ABIs on Linux systems; MIPS N32 ABI. |
| LLP64 | 16 | 32 | 32 | 64 | 64 | Microsoft Windows (x86-64, IA-64, and ARM64) usingVisual C++; andMinGW |
| LP64 | 16 | 32 | 64 | 64 | 64 | MostUnix andUnix-like systems, e.g.,Solaris,Linux,BSD,macOS.Windows when usingCygwin;z/OS |
| ILP64 | 16 | 64 | 64 | 64 | 64 | HAL Computer Systems port of Solaris to theSPARC64 |
| SILP64 | 64 | 64 | 64 | 64 | 64 | ClassicUNICOS[50][51] (versus UNICOS/mp, etc.) |
Many 64-bit platforms today use anLP64 model (including Solaris,AIX,HP-UX, Linux, macOS, BSD, and IBM z/OS). Microsoft Windows uses anLLP64 model. The disadvantage of the LP64 model is that storing along into anint truncates. On the other hand, converting a pointer to along will "work" in LP64. In the LLP64 model, the reverse is true. These are not problems which affect fully standard-compliant code, but code is often written with implicit assumptions about the widths of data types. C code should prefer (u)intptr_t instead oflong when casting pointers into integer objects.
A programming model is a choice made to suit a given compiler, and several can coexist on the same OS. However, the programming model chosen as the primary model for the OSapplication programming interface (API) typically dominates.
Another consideration is the data model used fordevice drivers. Drivers make up the majority of the operating system code in most modern operating systems[citation needed] (although many may not be loaded when the operating system is running). Many drivers use pointers heavily to manipulate data, and in some cases have to load pointers of a certain size into the hardware they support fordirect memory access (DMA). As an example, a driver for a 32-bit PCI device asking the device to DMA data into upper areas of a 64-bit machine's memory could not satisfy requests from the operating system to load data from the device to memory above the 4gigabyte barrier, because the pointers for those addresses would not fit into the DMA registers of the device. This problem is solved by having the OS take the memory restrictions of the device into account when generating requests to drivers for DMA, or by using aninput–output memory management unit (IOMMU).
| This sectiondoes notcite anysources. Please helpimprove this section byadding citations to reliable sources. Unsourced material may be challenged andremoved.(April 2023) (Learn how and when to remove this message) |
As of August 2023[update], 64-bit architectures for which processors were manufactured included:
Most architectures of 64 bits that are derived from the same architecture of 32 bits can execute code written for the 32-bit versions natively, with no performance penalty. For example, processors with AMD64 and Intel 64 can run 32-bit applications at full speed.[55] This kind of support is commonly calledbi-arch support or more generallymulti-arch support.
Versions of the VR4300 processor are widely used in consumer and office automation applications, including the popular Nintendo 64™ video game and advanced laser printers such as the recently announced, award-winning Hewlett-Packard LaserJet 4000 printer family.
Status: The kernel, compiler, tool chain work. The kernel boots and work on simulator and is used for porting of userland and running programs