The complex history of the Intel i960 RISC processor

The Intel i960 was a remarkable 32-bit processor of the 1990s with a confusing set of versions. Although it is now mostly forgotten (outside the many people who used it as an embedded processor), it has a complex history. It had a shot at being Intel's flagship processor until x86 overshadowed it. Later, it was the world's best-selling RISC processor. One variant was a 33-bit processor with a decidedly non-RISC object-oriented instruction set; it became a military standard and was used in the F-22 fighter plane. Another version powered Intel's short-lived Unix servers. In this blog post, I'll take a look at the history of the i960, explain its different variants, and examine silicon dies. This chip has a lot of mythology and confusion (especially on Wikipedia), so I'll try to clear things up.

Roots: the iAPX 432

"Intel 432": Cover detail from Introduction to the iAPX 432 Architecture.

"Intel 432": Cover detail from Introduction to the iAPX 432 Architecture.

The ancestry of the i960 starts in 1975, when Intel set out to design a "micro-mainframe", a revolutionary processor that would bring the power of mainframe computers to microprocessors. This project, eventually called the iAPX 432, was a huge leap in features and complexity. Intel had just released the popular 8080 processor in 1974, an 8-bit processor that kicked off the hobbyist computer era with computers such as the Altair and IMSAI. However, 8-bit microprocessors were toys compared to 16-bit minicomputers like the PDP-11, let alone mainframes like the 32-bit IBM System/370. Most companies were gradually taking minicomputer and mainframe features and putting them into microprocessors, but Intel wanted to leapfrog to a mainframe-class 32-bit processor. The processor would make programmers much more productive by bridging the "semantic gap" between high-level languages and simple processors, implementing many features directly into the processor.

The 432 processor included memory management, process management, and interprocess communication. These features were traditionally part of the operating system, but Intel built them in the processor, calling this the "Silicon Operating System". The processor was also one of the first to implement the new IEEE 754 floating-point standard, still in use by most processors. The 432 also had support for fault tolerance and multi-processor systems. One of the most unusual features of the 432 was that instructions weren't byte aligned. Instead, instructions were between 6 and 321 bits long, and you could jump into the middle of a byte. Another unusual feature was that the 432 was a stack-based machine, pushing and popping values on an in-memory stack, rather than using general-purpose registers.

The 432 provided hardware support for object-oriented programming, built around an unforgeable object pointer called an Access Descriptor. Almost every structure in a 432 program and in the system itself is a separate object. The processor provided fine-grain security and access control by checking every object access to ensure that the user had permission and was not exceeding the bounds of the object. This made buffer overruns and related classes of bugs impossible, unlike modern processors.

This photo from the Intel 1981 annual report shows Intel's 432-based development computer and three of the engineers.

This photo from the Intel 1981 annual report shows Intel's 432-based development computer and three of the engineers.

The new, object-oriented Ada language was the primary programming language for the 432. The US Department of Defense developed the Ada language in the late 1970s and early 1980s to provide a common language for embedded systems, using the latest ideas from object-oriented programming. Proponents expected Ada to become the dominant computer language for the 1980s and beyond. In 1979, Intel realized that Ada was a good target for the iAPX 432, since they had similar object and task models. Intel decided to "establish itself as an early center of Ada technology by using the language as the primary development and application language for the new iAPX 432 architecture." The iAPX 432's operating system (iMAX 432) and other software were written in Ada, using one of the first Ada compilers.

Unfortunately, iAPX 432 project was way too ambitious for its time. After a couple of years of slow progress, Intel realized that they needed a stopgap processor to counter competitors such as Zilog and Motorola. Intel quickly designed a 16-bit processor that they could sell until the 432 was ready. This processor was the Intel 8086 (1978), which lives on in the x86 architecture used by most computers today. Critically, the importance of the 8086 was not recognized at the time. In 1981, IBM selected Intel's 8088 processor (a version of the 8086 with an 8-bit bus) for the IBM PC. In time, the success of the IBM PC and compatible systems led to Intel's dominance of the microprocessor market, but in 1981 Intel viewed the IBM PC as just another design win. As Intel VP Bill Davidow later said, "We knew it was an important win. We didn't realize it was the only win."

Caption: IBM chose Intel's high performance 8088 microprocessor as the central processing unit for the IBM Personal Computer, introduced in 1981. Seven Intel peripheral components are also integrated into the IBM Personal Computer.  From Intel's 1981 annual report.

Caption: IBM chose Intel's high performance 8088 microprocessor as the central processing unit for the IBM Personal Computer, introduced in 1981. Seven Intel peripheral components are also integrated into the IBM Personal Computer. From Intel's 1981 annual report.

Intel finally released the iAPX 432 in 1981. Intel's 1981 annual report shows the importance of the 432 to Intel. A section titled "The Micromainframe™ Arrives" enthusiastically described the iAPX 432 and how it would "open the door to applications not previously feasible". To Intel's surprise, the iAPX 432 ended up as "one of the great disaster stories of modern computing" as the New York Times put it. The processor was so complicated that it was split across two very large chips:1 one to decode instructions and a second to execute them Delivered years behind schedule, the micro-mainframe's performance was dismal, much worse than competitors and even the stopgap 8086.2 Sales were minimal and the 432 quietly dropped out of sight.

My die photos of the two chips that make up the iAPX 432 General Data Processor. Click for a larger version.

My die photos of the two chips that make up the iAPX 432 General Data Processor. Click for a larger version.

Intel picks a 32-bit architecture (or two, or three)

In 1982, Intel still didn't realize the importance of the x86 architecture. The follow-on 186 and 286 processors were released but without much success at first.3 Intel was working on the 386, a 32-bit successor to the 286, but their main customer IBM was very unenthusiastic.4 Support for the 386 was so weak that the 386 team worried that the project might be dead.5 Meanwhile, the 432 team continued their work. Intel also had a third processor design in the works, a 32-bit VAX-like processor codenamed P4.6

Intel recognized that developing three unrelated 32-bit processors was impractical and formed a task force to develop a Single High-End Architecture (SHEA). The task force didn't achieve a single architecture, but they decided to merge the 432 and the P4 into a processor codenamed the P7, which would become the i960. They also decided to continue the 386 project. (Ironically, in 1986, Intel started yet another 32-bit processor, the unrelated i860, bringing the number of 32-bit architectures back to three.)

At the time, the 386 team felt that they were treated as the "stepchild" while the P7 project was the focus of Intel's attention. This would change as the sales of x86-based personal computers climbed and money poured into Intel. The 386 team would soon transform from stepchild to king.5

The first release of the i960 processor

Meanwhile, the 1980 paper The case for the Reduced Instruction Set Computer proposed a revolutionary new approach for computer architecture: building Reduced Instruction Set Computers (RISC) instead of Complex Instruction Set Computers (CISC). The paper argued that the trend toward increasing complexity was doing more harm than good. Instead, since "every transistor is precious" on a VLSI chip, the instruction set should be simplified, only adding features that quantitatively improved performance.

The RISC approach became very popular in the 1980s. Processors that followed the RISC philosophy generally converged on an approach with 32-bit easy-to-decode instructions, a load-store architecture (separating computation instructions from instructions that accessed memory), straightforward instructions that executed in one clock cycle, and implementing instructions directly rather than through microcode.

The P7 project combined the RISC philosophy and the ideas from the 432 to create Intel's first RISC chip, originally called the 809607 and later the i960. The chip, announced in 1988, was significant enough for coverage in the New York Times. Analysts said that the chip was marketed as an embedded controller to avoid stealing sales from the 80386. However, Intel's claimed motivation was the size of the embedded market; Intel chip designer Steve McGeady said at the time, "I'd rather put an 80960 in every antiskid braking system than in every Sun workstation.” Nonetheless, Intel also used the i960 as a workstation processor, as will be described in the next section.

The block diagram below shows the microarchitecture of the original i960 processors. The microarchitecture of the i960 followed most (but not all) of the common RISC design: a large register set, mostly one-cycle instructions, a load/store architecture, simple instruction formats, and a pipelined architecture. The Local Register Cache contains four sets of the 16 local registers. These "register windows" allow the registers to be switched during function calls without the delay of saving registers to the stack. The micro-instruction ROM and sequencer hold microcode for complex instructions; microcode is highly unusual for a RISC processor. The chip's Floating Point Unit8 and Memory Management Unit are advanced features for the time.

The microarchitecture of the i960 XA. FPU is Floating Point Unit. IEU is Instruction Execution Unit. MMU is Memory Management Unit. From the 80960 datasheet.

The microarchitecture of the i960 XA. FPU is Floating Point Unit. IEU is Instruction Execution Unit. MMU is Memory Management Unit. From the 80960 datasheet.

It's interesting to compare the i960 to the 432: the programmer-visible architectures are completely different, while the instruction sets are almost identical.9 Architecturally, the 432 is a stack-based machine with no registers, while the i960 is a load-store machine with many registers. Moreover, the 432 had complex variable-length instructions, while the i960 uses simple fixed-length load-store instructions. At the low level, the instructions are different due to the extreme architectural differences between the processors, but otherwise, the instructions are remarkably similar, modulo some name changes.

The key to understanding the i960 family is that there are four architectures, ranging from a straightforward RISC processor to a 33-bit processor implementing the 432's complex instruction set and object model.10 Each architecture adds additional functionality to the previous one:

  • The Core architecture consists of a "RISC-like" core.
  • The Numerics architecture extends Core with floating-point.
  • The Protected architecture extends Numerics with paged memory management, Supervisor/User protection, string instructions, process scheduling, interprocess communication for OS, and symmetric multiprocessing.
  • The Extended architecture extends Protected with object addressing/protection and interprocess communication for applications. This architecture used an extra tag bit, so registers, the bus, and memory were 33 bits wide instead of 32.

These four versions were sold as the KA (Core), KB (Numerics), MC (Protected), and XA (Extended). The KA chip cost $174 and the KB version cost $333 while MC was aimed at the military market and cost a whopping $2400. The most advanced chip (XA) was, at first, kept proprietary for use by BiiN (discussed below), but was later sold to the military. The military versions weren't secret, but it is very hard to find documentation on them.11

The strangest thing about these four architectures is that the chips were identical, using the same die. In other words, the simple Core chip included all the circuitry for floating point, memory management, and objects; these features just weren't used.12 The die photo below shows the die, with the main functional units labeled. Around the edge of the die are the bond pads that connect the die to the external pins. Note that the right half of the chip has almost no bond pads. As a result, the packaged IC had many unused pins.13

The i960 KA/KB/MC/XA with the main functional blocks labeled. Click this image (or any other) for a larger version. Die image courtesy of Antoine Bercovici.  Floorplan from The 80960 microprocessor architecture.

The i960 KA/KB/MC/XA with the main functional blocks labeled. Click this image (or any other) for a larger version. Die image courtesy of Antoine Bercovici. Floorplan from The 80960 microprocessor architecture.

One advanced feature of the i960 is register scoreboarding, visible in the upper-left corner of the die. The idea is that loading a register from memory is slow, so to improve performance, the processor executes the following instructions while the load completes, rather than waiting. Of course, an instruction can't be executed if it uses a register that is being loaded, since the value isn't there. The solution is a "scoreboard" that tracks which registers are valid and which are still being loaded, and blocks an instruction if the register isn't ready. The i960 could handle up to three outstanding reads, providing a significant performance gain.

The most complex i960 architecture is the Extended architecture, which provides the object-oriented system. This architecture is designed around an unforgeable pointer called an Access Descriptor that provides protected access to an object. What makes the pointer unforgeable is that it is 33 bits long with an extra bit that indicates an Access Descriptor. You can't set this bit with a regular 32-bit instruction. Instead, an Access Descriptor can only be created with a special privileged instruction, "Create AD".14

An Access Descriptor is a pointer to an object table. From BiiN Object Computing.

An Access Descriptor is a pointer to an object table. From BiiN Object Computing.

The diagram above shows how objects work. The 33-bit Access Descriptor (AD) has its tag bit set to 1, indicating that it is a valid Access Descriptor. The Rights field controls what actions can be performed by this object reference. The AD's Object Index references the Object Table that holds information about each object. In particular, the Base Address and Size define the object's location in memory and ensure that an access cannot exceed the bounds of the object. The Type Definition defines the various operations that can be performed on the object. Since this is all implemented by the processor at the instruction level, it provides strict security.

Gemini and BiiN

The i960 was heavily influenced by a partnership called Gemini and then BiiN. In 1983, near the start of the i960 project, Intel formed a partnership with Siemens to build high-performance fault-tolerant servers. In this partnership, Intel would provide the hardware while Siemens developed the software. This partnership allowed Intel to move beyond the chip market to the potentially-lucrative systems market, while adding powerful systems to Siemens' product line. The Gemini team contained many of the people from the 432 project and wanted to continue the 432's architecture. Gemini worked closely with the developers of the i960 to ensure the new processor would meet their needs; both teams worked in the same building at Intel's Jones Farm site in Oregon.

The BiiN 60 system. From BiiN 60 Technical Overview.

The BiiN 60 system. From BiiN 60 Technical Overview.

In 1988, shortly after the announcement of the i960 chips, the Intel/Siemens partnership was spun off into a company called BiiN.15 BiiN announced two high-performance, fault-tolerant, multiprocessor systems. These systems used the i960 XA processor16 and took full advantage of the object-oriented model and other features provided by its Extended architecture. The BiiN 20 was designed for departmental computing and cost $43,000 to $80,000. It supported 50 users (connected by terminals) on one 5.5-MIPS i960 processor. The larger BiiN 60 handled up to 1000 terminals and cost $345,000 to $815,000. The Unix-compatible BiiN operating system (BiiN/OS) and utilities were written in 2 million lines of Ada code.

BiiN described many potential markets for these systems: government, factory automation, financial services, on-line transaction processing, manufacturing, and health care. Unfortunately, as ExtremeTech put it, "the market for fault-tolerant Unix workstations was approximately nil." BiiN was shut down in 1989, just 15 months after its creation as profitability kept becoming more distant. BiiN earned the nickname "Billions invested in Nothing"; the actual investment was 1700 person-years and $430 million.

The superscalar i960 CA

One year after the first i960, Intel released the groundbreaking i960 CA. This chip was the world's first superscalar microprocessor, able to execute more than one instruction per clock cycle. The chip had three execution units that could operate in parallel: an integer execution unit, a multiply/divide unit, and an address generation unit that could also do integer arithmetic.17 To keep the execution units busy, the i960 CA's instruction sequencer examined four instructions at once and determined which ones could be issued in parallel without conflict. It could issue two instructions and a branch each clock cycle, using branch prediction to speculatively execute branches out of order.

The i960 CA die, with functional blocks labeled. Photo courtesy of Antoine Bercovici. Functional blocks from the datasheet.

The i960 CA die, with functional blocks labeled. Photo courtesy of Antoine Bercovici. Functional blocks from the datasheet.

Following the CA, several other superscalar variants were produced: the CF had more cache, the military MM implemented the Protected architecture (memory management and a floating point unit), and the military MX implemented the Extended architecture (object-oriented).

The image below shows the 960 MX die with the main functional blocks labeled. (I think the MM and MX used the same die but I'm not sure.18) Like the i960 CA, this chip has multiple functional units that can be operated in parallel for its superscalar execution. Note the wide buses between various blocks, allowing high internal bandwidth. The die was too large for the optical projection of the mask, with the result that the corners of the circuitry needed to be rounded off.

The i960MX die with the main functional blocks labeled. This is a die photo I took, with labels based on my reverse engineering.

The i960MX die with the main functional blocks labeled. This is a die photo I took, with labels based on my reverse engineering.

The block diagram of the i960 MX shows the complexity of the chip and how it is designed for parallelism. The register file is the heart of the chip. It is multi-ported so up to 6 registers can be accessed at the same time. Note the multiple, 256-bit wide buses between the register file and the various functional units. The chip has two buses: a high-bandwidth Backside Bus between the chip and its external cache and private memory; and a New Local Bus, which runs at half the speed and connects the chip to main memory and I/O. For highest performance, the chip's software would access its private memory over the high-speed bus, while using the slower bus for I/O and shared memory accesses.

A functional block diagram of the i960 MX. From Intel Military and Special Projects Handbook, 1993.

A functional block diagram of the i960 MX. From Intel Military and Special Projects Handbook, 1993.

Military use and the JIAWG standard

The i960 had a special role in the US military. In 1987 the military mandated the use of Ada as the single, common computer programming language for Defense computer resources in most cases.19 In 1989, the military created the JIAWG standard, which selected two 32-bit instruction set architectures for military avionics. These architectures were the i960's Extended architecture (implemented by the i960 XA) and the MIPS architecture (based on a RISC project at Stanford).20 The superscalar i960 MX processor described earlier soon became a popular JIAWG-compliant processor, since it had higher performance than the XA.

Hughes designed a modular avionics processor that used the i960 XA and later the MX. A dense module called the HAC-32 contained two i960 MX processors, 2 MB of RAM, and an I/O controller in a 2"×4" multi-chip module, slightly bigger than a credit card. This module had bare dies bonded to the substrate, maximizing the density. In the photo below, the two largest dies are the i960 MX while the numerous gray rectangles are memory chips. This module was used in F-22's Common Integrated Processor, the RAH-66 Comanche helicopter (which was canceled), the F/A-18's Stores Management Processor (the computer that controls attached weapons), and the AN/ALR-67 radar computer.

The Hughes HAC-32. From Avionics Systems Design.

The Hughes HAC-32. From Avionics Systems Design.

The military market is difficult due to the long timelines of military projects, unpredictable volumes, and the risk of cancellations. In the case of the F-22 fighter plane, the project started in 1985 when the Air Force sent out proposals for a new Advanced Tactical Fighter. Lockheed built a YF-22 prototype, first flying it in 1990. The Air Force selected the YF-22 over the competing YF-23 in 1991 and the project moved to full-scale development. During this time, at least three generations of processors became obsolete. In particular, the i960MX was out of production by the time the F-22 first flew in 1997. At one point, the military had to pay Intel $22 million to restart the i960 production line. In 2001, the Air Force started a switch to the PowerPC processor, and finally the plane entered military service in 2005. The F-22 illustrates how the fast-paced obsolescence of processors is a big problem for decades-long military projects.

The Common Integrated Processor for the F-22, presumably with i960 MX chips inside. It is the equivalent of two Cray supercomputers and was the world's most advanced, high-speed computer system for a fighter aircraft. Source: NARA/Hughes Aircraft Co./T.W. Goosman.

The Common Integrated Processor for the F-22, presumably with i960 MX chips inside. It is the equivalent of two Cray supercomputers and was the world's most advanced, high-speed computer system for a fighter aircraft. Source: NARA/Hughes Aircraft Co./T.W. Goosman.

Intel charged thousands of dollars for each i960 MX and each F-22 contained a cluster of 35 i960 MX processors, so the military market was potentially lucrative. The Air Force originally planned to buy 750 planes, but cut this down to just 187, which must have been a blow to Intel. As for the Comanche helicopter, the Army planned to buy 1200 of them, but the program was canceled entirely after building two prototypes. The point is that the military market is risky and low volume even in the best circumstances.21 In 1998, Intel decided to leave the military business entirely, joining AMD and Motorola.

Foreign militaries also made use of the i960. In 2008 a businessman was sentenced to 35 months in prison for illegally exporting hundreds of i960 chips into India for use in the radar for the Tejas Light Combat Aircraft.

i960: the later years

By 1990, the i960 was selling well, but the landscape at Intel had changed. The 386 processor was enormously successful, due to the Compaq Deskpro 386 and other systems, leading to Intel's first billion-dollar quarter. The 8086 had started as a stopgap processor to fill a temporary marketing need, but now the x86 was Intel's moneymaking engine. As part of a reorganization, the i960 project was transferred to Chandler, Arizona. Much of the i960 team in Oregon moved to the newly-formed Pentium Pro team, while others ended up on the 486 DX2 processor. This wasn't the end of the i960, but the intensity had reduced.

To reduce system cost, Intel produced versions of the i960 that had a 16-bit bus, although the processor was 32 bits internally. (This is the same approach that Intel used with the 8088 processor, a version of the 8086 processor with an 8-bit bus instead of 16.) The i960 SB had the "Numerics" architecture, that is, with a floating-point unit. Looking at the die below, we can see that the SB design is rather "lazy", simply the previous die (KA/KB/MC/XA) with a thin layer of circuitry around the border to implement the 16-bit bus. Even though the SB didn't support memory management or objects, Intel didn't remove that circuitry. The process was reportedly moved from 1.5 microns to 1 micron, shrinking the die to 270 mils square.

Comparison of the original i960 die and the i960 SB. Photos courtesy of Antoine Bercovici.

Comparison of the original i960 die and the i960 SB. Photos courtesy of Antoine Bercovici.

The next chip, the i960 SA, was the 16-bit-bus "Core" architecture, without floating point. The SA was based on the SB but Intel finally removed unused functionality from the die, making the die about 24% smaller. The diagram below shows how the address translation, translation lookaside buffer, and floating point unit were removed, along with much of the microcode (yellow). The instruction cache tags (purple), registers (orange), and execution unit (green) were moved to fit into the available space. The left half of the chip remained unchanged. The driver circuitry around the edges of the chip was also tightened up, saving a bit of space.

This diagram compares the SB and SA chips. Photos courtesy of Antoine Bercovici.

This diagram compares the SB and SA chips. Photos courtesy of Antoine Bercovici.

Intel introduced the high-performance Hx family around 1994. This family was superscalar like the CA/CF, but the Hx chips also had a faster clock, had much more cache, and included additional functionality such as timers and a guarded memory unit. The Jx family was introduced as the midrange, cost-effective line, faster and better than the original chips but not superscalar like the Hx. Intel attempted to move the i960 into the I/O controller market with the Rx family and the VH.23 This was part of Intel's Intelligent Input/Output specification (I2O), which was a failure overall.

For a while, the i960 was a big success in the marketplace and was used in many products. Laser printers and graphical terminals were key applications, both taking advantage of the i960's high speed to move pixels. The i960 was the world's best-selling RISC chip in 1994. However, without focused development, the performance of the i960 fell behind the competition, and its market share rapidly dropped.

Market share of embedded RISC processors. From ExtremeTech.

Market share of embedded RISC processors. From ExtremeTech.

By the late 1990s, the i960 was described with terms such as "aging", "venerable", and "medieval". In 1999, Microprocessor Report described the situation: "The i960 survived on cast-off semiconductor processes two to three generations old; the i960CA is still built in a 1.0-micron process (perhaps by little old ladies with X-Acto knives)."22

One of the strongest competitors was DEC's powerful StrongARM processor design, a descendant of the ARM chip. Even Intel's top-of-the-line i960HT fared pitifully against the StrongARM, with worse cost, performance, and power consumption. In 1997, DEC sued Intel, claiming that the Pentium infringed ten of DEC's patents. As part of the complex but mutually-beneficial 1997 settlement, Intel obtained rights to the StrongARM chip. As Intel turned its embedded focus from i960 to StrongARM, one writer wrote, "Things are looking somewhat bleak for Intel Corp's ten-year-old i960 processor." The i960 limped on for another decade until Intel officially ended production in 2007.

RISC or CISC?

The i960 challenges the definitions of RISC and CISC processors.24 It is generally considered a RISC processor, but its architect says "RISC techniques were used for high performance, CISC techniques for ease of use."25 John Mashey of MIPS described it as on the RISC/CISC border26 while Steve Furber (co-creator of ARM) wrote that it "includes many RISC ideas, but it is not a simple chip" with "many complex instructions which make recourse to microcode" and a design that "is more reminiscent of a complex, mainframe architecture than a simple, pipelined RISC." And they were talking about the i960 KB with the simple Numerics architecture, not the complicated Extended architecture!

Even the basic Core architecture has many non-RISC-like features. It has microcoded instructions that take multiple cycles (such as integer multiplication), numerous addressing modes27, and unnecessary instructions (e.g. AND NOT as well as NOT AND). It also has a large variety of datatypes, even more than the 432: integer (8, 16, 32, or 64 bit), ordinal (8, 16, 32, or 64 bit), decimal digits, bit fields, triple-word (96 bits), and quad-word (128 bits). The Numerics architecture adds floating-point reals (32, 64, or 80 bit) while the Protected architecture adds byte strings with decidedly CISC-like instructions to act on them.28

When you get to the Extended architecture with objects, process management, and interprocess communication instructions, the large instruction set seems obviously CISC.29 (The instruction set is essentially the same as 432 and the 432 is an extremely CISC processor.) You could argue that the i960 Core architecture is RISC and the Extended architecture is CISC, but the problem is that they are identical chips.

Of course, it doesn't really matter if the i960 is considered RISC, CISC, or CISC instructions running on a RISC core. But the i960 shows that RISC and CISC aren't as straightforward as they might seem.

Summary

The i960 chips can be confusing since there are four architectures, along with scalar vs. superscalar, and multiple families over time. I've made the table below to summarize the i960 family and the approximate dates. The upper entries are the scalar families while the lower entries are superscalar. The columns indicate the four architectural variants; although the i960 started with four variants, eventually Intel focused on only the Core. Note that each "x" family represents multiple chips.

CoreNumericsProtectedExtended
KAKBMCXAOriginal (1988)
SASB  Entry level, 16-bit data bus (1991)
Jx   Midrange (1993-1998)
Rx,VH   I/O interface (1995-2001)
CA,CF MMMXSuperscalar (1989-1992)
Hx   Superscalar, higher performance (1994)

Although the i960 is now mostly forgotten, it was an innovative processor for the time. The first generation was Intel's first RISC chip, but pushed the boundary of RISC with many CISC-like features. The i960 XA literally set the standard for military computing, selected by the JIAWG as the military's architecture. The i960 CA provided a performance breakthrough with its superscalar architecture. But Moore's Law means that competitors can rapidly overtake a chip, and the i960 ended up as history.

Thanks to Glen Myers, Kevin Kahn, Steven McGeady, and others from Intel for answering my questions about the i960. Thanks to Prof. Paul Lubeck for obtaining documentation for me. I plan to write more, so follow me on Twitter @kenshirriff or RSS for updates. I've also started experimenting with Mastodon recently as @[email protected] and Bluesky as @righto.com so you can follow me there too.

Notes and references

  1. The 432 used two chips for the processor and a third chip for I/O. At the time, these were said to be "three of the largest integrated circuits in history." The first processor chip contained more than 100,000 devices, making it "one of the densest VLSI circuits to have been fabricated so far." The article also says that the 432 project "was the largest investment in a single program that Intel has ever made." See Ada determines architecture of 32-bit microprocessor, Electronics, Feb 24, 1981, pages 119-126, a very detailed article on the 432 written by the lead engineer and the team's manager. 

  2. The performance problems of the iAPX 432 were revealed by a student project at Berkeley, A performance evaluation of the Intel iAPX 432, which compared its performance with the VAX-11/780, Motorola 68000, and Intel 8086. Instead of providing mainframe performance, the 432 had a fraction of the performance of the competing systems. Another interesting paper is Performance effects of architectural complexity in the Intel 432, which examines in detail what the 432 did wrong. It concludes that the 432 could have been significantly faster, but would still have been slower than its contemporaries. An author of the paper was Robert Colwell, who was later hired by Intel and designed the highly-successful Pentium Pro architecture. 

  3. You might expect the 8086, 186, and 286 processors to form a nice progression, but it's a bit more complicated. The 186 and 286 processors were released at the same time. The 186 essentially took the 8086 and several support chips and integrated them onto a single die. The 286, on the other hand, extended the 8086 with memory management. However, its segment-based memory management was a bad design, using ideas from the Zilog MMU, and wasn't popular. The 286 also had a protected mode, so multiple processes could be isolated from each other. Unfortunately, protected mode had some serious problems. Bill Gates famously called the 286 "brain-damaged" echoing PC Magazine editor Bill Machrone and writer Jerry Pournelle, who both wanted credit for originating the phrase.

    By 1984, however, the 286 was Intel's star due to growing sales of IBM PCs and compatibles that used the chip. Intel's 1984 annual report featured "The Story of the 286", a glowing 14-page tribute to the 286. 

  4. Given IBM's success with IBM PC, Intel was puzzled that IBM wasn't interested in the 386 processor. It turned out that IBM had a plan to regain control of the PC so they could block out competitors that were manufacturing IBM PC compatibles. IBM planned to reverse-engineer Intel's 286 processor and build its own version. The computers would run the OS/2 operating system instead of Windows and use the proprietary Micro Channel architecture. However, the reverse-engineering project failed and IBM eventually moved to the Intel 386 processor. The IBM PS/2 line of computers, released in 1987, followed the rest of the plan. However, the PS/2 line was largely unsuccessful; rather than regaining control over the PC, IBM ended up losing control to companies such as Compaq and Dell. (For more, see Creating the Digital Future, page 131.) 

  5. The 386 team created an oral history that describes the development of the 386 in detail. Pages 5, 6, and 19 are most relevant to this post. 

  6. You might wonder why the processor was codenamed P4, since logically P4 should indicate the 486. Confusingly, Intel's processor codenames were not always sequential and they sometimes reused numbers. The numbers apparently started with P0, the codename for the Optimal Terminal Kit, a processor that didn't get beyond early planning. P5 was used for the 432, P4 for the planned follow-on, P7 for the i960, P10 for the i960 CA, and P12 for the i960 MX. (Apparently they thought that x86 wouldn't ever get to P4.)

    For the x86 processors, P1 through P6 indicated the 186, 286, 386, 486, 586, Pentium, and Pentium Pro as you'd expect. (The Pentium used a variety of codes for various versions, such as P54C, P24T, and P55C; I don't understand the pattern behind these.) For some reason, the i386SX was the P9 and the i486SX was the P23 and the i486DX2 was the P24. The Pentium 4 Willamette was the first new microarchitecture (NetBurst) since P6 so it was going to be P7, but Itanium took the P7 name codename so Willamette became P68. After that, processors were named after geographic features, avoiding the issues with numeric codenames.

    Other types of chips used different letter prefixes. The 387 numeric coprocessor was the N3. The i860 RISC processor was originally the N10, a numeric co-processor. The follow-on i860 XP was the N11. Support chips for the 486 included the C4 cache chip and the unreleased I4 interrupt controller. 

  7. At the time, Intel had a family of 16-bit embedded microcontrollers called MCS-96 featuring the 8096. The 80960 name was presumably chosen to imply continuity with the 8096 16-bit microcontrollers (MCS-96), even though the 8096 and the 80960 are completely different. (I haven't been able to confirm this, though.) Intel started calling the chip the "i960" around 1989. (Intel's chip branding is inconsistent: from 1987 to 1991, Intel's annual reports called the 386 processor the 80386, the 386, the Intel386, and the i386. I suspect their trademark lawyers were dealing with the problem that numbers couldn't be trademarked, which was the motivation for the "Pentium" name rather than 586.)

    Note that the i860 processor is completely unrelated to the i960 despite the similar numbers. They are both 32-bit RISC processors, but are architecturally unrelated. The i860 was targeted at high-performance workstations, while the i960 was targeted at embedded applications. For details on the i860, see The first million-transistor chip

  8. The Intel 80387 floating-point coprocessor chip used the same floating-point unit as the i960. The diagram below shows the 80387; compare the floating-point unit in the lower right corner with the matching floating-point unit in the i960 KA or SB die photo.

    The 80837 floating-point coprocessor with the main functional blocks labeled. Die photo courtesy of Antoine Bercovici. 80387 floor plan from The 80387 and its applications.

    The 80837 floating-point coprocessor with the main functional blocks labeled. Die photo courtesy of Antoine Bercovici. 80387 floor plan from The 80387 and its applications.

     

  9. I compared the instruction sets of the 432 and i960 and the i960 Extended instruction set seems about as close as you could get to the 432 while drastically changing the underlying architecture. If you dig into the details of the object models, there are some differences. Some instructions also have different names but the same function. 

  10. The first i960 chips were described in detail in the 1988 book The 80960 microprocessor architecture by Glenford Myers (who was responsible for the 80960 architecture at Intel) and David Budde (who managed the VLSI development of the 80960 components). This book discussed three levels of architecture (Core, Numerics, and Protected). The book referred to the fourth level, the Extended architecture (XA), calling it "a proprietary higher level of the architecture developed for use by Intel in system products" and did not discuss it further. These "system products" were the systems being developed at BiiN. 

  11. I could find very little documentation on the Extended architecture. The 80960XA datasheet provides a summary of the instruction set. The i960 MX datasheet provides a similar summary; it is in the Intel Military and Special Products databook, which I found after much difficulty. The best description I could find is in the 400-page BiiN CPU architecture reference manual. Intel has other documents that I haven't been able to find anywhere: i960 MM/MX Processor Hardware Reference Manual, and Military i960 MM/MX Superscalar Processor. (If you have one lying around, let me know.)

    The 80960MX Specification Update mentions a few things about the MX processor. My favorite is that if you take the arctan of a value greater than 32768, the processor may lock up and require a hardware reset. Oops. The update also says that the product is sold in die and wafer form only, i.e. without packaging. Curiously, earlier documentation said the chip was packaged in a 348-pin ceramic PGA package (with 213 signal pins and 122 power/ground pins). I guess Intel ended up only supporting the bare die, as in the Hughes HAC-32 module. 

  12. According to people who worked on the project, there were not even any bond wire changes or blown fuses to distinguish the chips for the four different architectures. It's possible that Intel used binning, selling dies as a lower architecture if, for example, the floating point unit failed testing. Moreover, the military chips presumably had much more extensive testing, checking the military temperature range for instance. 

  13. The original i960 chips (KA/KB/MC/XA) have a large number of pins that are not connected (marked NC on the datasheet). This has led to suspicious theorizing, even on Wikipedia, that these pins were left unconnected to control access to various features. This is false for two reasons. First, checking the datasheets shows that all four chips have the same pinout; there are no pins connected only in the more advanced versions. Second, looking at the packaged chip (below) explains why so many pins are unconnected: much of the chip has no bond pads, so there is nothing to connect the pins to. In particular, the right half of the die has only four bond pads for power. This is an unusual chip layout, but presumably the chip's internal buses made it easier to put all the connections at the left. The downside is that the package is more expensive due to the wasted pins, but I expect that BiiN wasn't concerned about a few extra dollars for the package.

    The i960 MC die, bonded in its package. Photo courtesy of Antoine Bercovici.

    The i960 MC die, bonded in its package. Photo courtesy of Antoine Bercovici.

    But you might wonder: the simple KA uses 32 bits and the complex XA uses 33 bits, so surely there must be another pin for the 33rd bit. It turns out that pin F3 is called CACHE on the KA and CACHE/TAG on the XA. The pin indicates if an access is cacheable, but the XA uses the pin during a different clock cycle to indicate whether the 32-bit word is data or an access descriptor (unforgeable pointer).

    So how does the processor know if it should use the 33-bit object mode or plain 32-bit mode? There's a processor control word called Processor Controls, that includes a Tag Enable bit. If this bit is set, the processor uses the 33rd bit (the tag bit) to distinguish Access Descriptors from data. If the bit is clear, the distinction is disabled and the processor runs in 32-bit mode. (See BiiN CPU Architecture Reference Manual section 16.1 for details.) 

  14. The 432 and the i960 both had unforgeable object references, the Access Descriptor. However, the two processors implemented Access Descriptors in completely different ways, which is kind of interesting. The i960 used a 33rd bit as a Tag bit to distinguish an Access Descriptor from a regular data value. Since the user didn't have access to the Tag bit, the user couldn't create or modify Access Descriptors. The 432, on the other hand, used standard 32-bit words. To protect Access Descriptors, each object was divided into two parts, each protected by a length field. One part held regular data, while the other part held Access Descriptors. The 432 had separate instructions to access the two parts of the object, ensuring that regular instructions could not tamper with Access Descriptors. 

  15. The name "BiiN" was developed by Lippincott & Margulies, a top design firm. The name was designed for a strong logo, as well as referencing binary code (so it was pronounced as "bine"). Despite this pedigree, "BiiN" was called one of the worst-sounding names in the computer industry, see Losing the Name Game

  16. Some sources say that BiiN used the i960 MX, not the XA, but they are confused. A paper from BiiN states that BiiN used the 80960 XA. (Sadly, BiiN was so short-lived that the papers introducing the BiiN system also include its demise.) Moreover, BiiN shut down in 1989 while the i960 MX was introduced in 1990, so the timeline doesn't work. 

  17. The superscalar i960 architecture is described in detail in The i960CA SuperScalar implementation of the 80960 architecture and Inside Intel's i960CA superscalar processor while the military MM version is described in Performance enhancements in the superscalar i960MM embedded microprocessor

  18. I don't have a die photo of the i960 MM, so I'm not certain of the relationship between the MM and the MX. The published MM die size is approximately the same as the MX. The MM block diagram matches the MX, except using 32 bits instead of 33. Thus, I think the MM uses the MX die, ignoring the Extended features, but I can't confirm this. 

  19. The military's Ada mandate remained in place for a decade until it was eliminated in 1997. Ada continues to be used by the military and other applications that require high reliability, but by now C++ has mostly replaced it. 

  20. The military standard was decided by the Joint Integrated Avionics Working Group, known as JIAWG. Earlier, in 1980, the military formed a 16-bit computing standard, MIL-STD-1750A. The 1750A standard created a new architecture, and numerous companies implemented 1750A-compatible processors. Many systems used 1750A processors and overall it was more successful than the JIAWG standard. 

  21. Chip designer and curmudgeon Nick Tredennick described the market for Intel's 960MX processor: "Intel invested considerable money and effort in the design of the 80960MX processor, for which, at the time of implementation, the only known application was the YF-22 aircraft. When the only prototype of the YF-22 crashed, the application volume for the 906MX actually went to zero; but even if the program had been successful, Intel could not have expected to sell more than a few thousand processors for that application." 

  22. In the early 1970s, chip designs were created by cutting large sheets of Rubylith film with X-Acto knives. Of course, that technology was long gone by the time of the i960.

    Intel photo of two women cutting Rubylith.

    Intel photo of two women cutting Rubylith.

     

  23. The Rx I/O processor chips combined a Jx processor core with a PCI bus interface and other hardware. The RM and RN versions were introduced in 2000 with a hardware XOR engine for RAID disk array parity calculations. The i960 VH (1998) was similar to Rx, but had only one PCI bus, no APIC bus, and was based on the JT core. The 80303 (2000) was the end of the i960 I/O processors. The 80303 was given a numeric name instead of an i960 name because Intel was transitioning from i960 to XScale at the time. The numeric name makes it look like a smooth transition from the 80303 (i960) I/O processor to the XScale-based I/O processors such as the 80333. The 803xx chips were also called IOP3xx (I/O Processor); some were chipsets with a separate XScale processor chip and an I/O companion chip.  

  24. Although the technical side of RISC vs. CISC is interesting, what I find most intriguing is the "social history" of RISC: how did a computer architecture issue from the 1980s become a topic that people still vigorously argue over 40 years later? I see several factors that keep the topic interesting:

    • RISC vs. CISC has a large impact on not only computer architects but also developers and users.
    • The topic is simple enough that everyone can have an opinion. It's also vague enough that nobody agrees on definitions, so there's lots to argue about.
    • There are winners and losers, but no resolution. RISC sort of won in the sense that almost all new instruction set architectures have been RISC. But CISC has won commercially with the victory of x86 over SPARC, PowerPC, Alpha, and other RISC contenders. But ARM dominates mobile and is moving into personal computers through Apple's new processors. If RISC had taken over in the 1980s as expected, there wouldn't be anything to debate. But x86 has prospered despite the efforts of everyone (including Intel) to move beyond it.
    • RISC vs. CISC takes on a "personal identity" aspect. For instance, if you're an "Apple" person, you're probably going to be cheering for ARM and RISC. But nobody cares about branch prediction strategies or caching.

    My personal opinion is that it is a mistake to consider RISC and CISC as objective, binary categories. (Arguing over whether ARM or the 6502 is really RISC or CISC is like arguing over whether a hotdog is a sandwich RISC is more of a social construct, a design philosophy/ideology that leads to a general kind of instruction set architecture that leads to various implementation techniques.

    Moreover, I view RISC vs. CISC as mostly irrelevant since the 1990s due to convergence between RISC and CISC architectures. In particular, the Pentium Pro (1995) decoded CISC instructions into "RISC-like" micro-operations that are executed by a superscalar core, surprising people by achieving RISC-like performance from a CISC processor. This has been viewed as a victory for CISC, a victory for RISC, nothing to do with RISC, or an indication that RISC and CISC have converged. 

  25. The quote is from Microprocessor Report April 1988, "Intel unveils radical new CPU family", reprinted in "Understanding RISC Microprocessors". 

  26. John Mashey of MIPS wrote an interesting article "CISCs are Not RISCs, and Not Converging Either" in the March 1992 issue of Microprocessor Report, extending a Usenet thread. It looks at multiple quantitative factors of various processors and finds a sharp line between CISC processors and most RISC processors. The i960, Intergraph Clipper, and (probably) ARM, however, were "truly on the RISC/CISC border, and, in fact, are often described that way." 

  27. The i960 datasheet lists an extensive set of addressing modes, more than typical for a RISC chip:

    • 12-bit offset
    • 32-bit offset
    • Register-indirect
    • Register + 12-bit offset
    • Register + 32-bit offset
    • Register + index-register×scale-factor
    • Register×scale-factor + 32-bit displacement
    • Register + index-register×scale-factor + 32-bit displacement
    where the scale-factor is 1, 2, 4, 8, or 16.

    See the 80960KA embedded 32-bit microprocessor datasheet for more information. 

  28. The i960 MC has string instructions that move, scan, or fill a string of bytes with a specified length. These are similar to the x86 string operations, but these are very unusual for a RISC processor. 

  29. The iAPX 432 instruction set is described in detail in chapter 10 of the iAPX 432 General Data Processor Architecture Reference Manual; the instructions are called "operators". The i960 Protected instruction set is listed in the 80960MC Programmer's Reference Manual while the i960 Extended instruction set is described in the BiiN CPU architecture reference manual.

    The table below shows the instruction set for the Extended architecture, the full set of object-oriented instructions. The instruction set includes typical RISC instructions (data movement, arithmetic, logical, comparison, etc), floating point instructions (for the Numerics architecture), process management instructions (for the Protected architecture), and the Extended object instructions (Access Descriptor operations). The "Mixed" instructions handle 33-bit values that can be either a tag (object pointer) or regular data. Note that many of these instructions have separate opcodes for different datatypes, so the complete instruction set is larger than this list, with about 240 opcodes.

    The Extended instruction set, from the i960 XA datasheet. Click for a larger version.

    The Extended instruction set, from the i960 XA datasheet. Click for a larger version.

The Group Decode ROM: The 8086 processor's first step of instruction decoding

A key component of any processor is instruction decoding: analyzing a numeric opcode and figuring out what actions need to be taken. The Intel 8086 processor (1978) has a complex instruction set, making instruction decoding a challenge. The first step in decoding an 8086 instruction is something called the Group Decode ROM, which categorizes instructions into about 35 types that control how the instruction is decoded and executed. For instance, the Group Decode ROM determines if an instruction is executed in hardware or in microcode. It also indicates how the instruction is structured: if the instruction has a bit specifying a byte or word operation, if the instruction has a byte that specifies the addressing mode, and so forth.

The 8086 die under a microscope, with main functional blocks labeled. This photo shows the chip with the metal and polysilicon removed, revealing the silicon underneath. Click on this image (or any other) for a larger version.

The 8086 die under a microscope, with main functional blocks labeled. This photo shows the chip with the metal and polysilicon removed, revealing the silicon underneath. Click on this image (or any other) for a larger version.

The diagram above shows the position of the Group Decode ROM on the silicon die, as well as other key functional blocks. The 8086 chip is partitioned into a Bus Interface Unit that communicates with external components such as memory, and the Execution Unit that executes instructions. Machine instructions are fetched from memory by the Bus Interface Unit and stored in the prefetch queue registers, which hold 6 bytes of instructions. To execute an instruction, the queue bus transfers an instruction byte from the prefetch queue to the instruction register, under control of a state machine called the Loader. Next, the Group Decode ROM categorizes the instruction according to its structure. In most cases, the machine instruction is implemented in low-level microcode. The instruction byte is transferred to the Microcode Address Register, where the Microcode Address Decoder selects the appropriate microcode routine that implements the instruction. The microcode provides the micro-instructions that control the Arithmetic/Logic Unit (ALU), registers, and other components to execute the instruction.

In this blog post, I will focus on a small part of this process: how the Group Decode ROM decodes instructions. Be warned that this post gets down into the weeds, so you might want to start with one of my higher-level posts, such as how the 8086's microcode engine works.

Microcode

Most instructions in the 8086 are implemented in microcode. Most people think of machine instructions as the basic steps that a computer performs. However, many processors have another layer of software underneath: microcode. With microcode, instead of building the CPU's control circuitry from complex logic gates, the control logic is largely replaced with code. To execute a machine instruction, the computer internally executes several simpler micro-instructions, specified by the microcode.

Microcode is only used if the Group Decode ROM indicates that the instruction is implemented in microcode. In that case, the microcode address register is loaded with the instruction and the address decoder selects the appropriate microcode routine. However, there's a complication. If the second byte of the instruction is a Mod R/M byte, the Group Decode ROM indicates this and causes a memory addressing micro-subroutine to be called.

Some simple instructions are implemented entirely in hardware and don't use microcode. These are known as 1-byte logic instructions (1BL) and are also indicated by the Group Decode ROM.

The Group Decode ROM's structure

The Group Decode ROM takes an 8-bit instruction as input, along with an interrupt signal. It produces 15 outputs that control how the instruction is handled. In this section I'll discuss the physical implementation of the Group Decode ROM; the various outputs are discussed in a later section.

Although the Group Decode ROM is called a ROM, its implementation is really a PLA (Programmable Logic Array), two levels of highly-structured logic gates.1 The idea of a PLA is to create two levels of NOR gates, each in a grid. This structure has the advantages that it implements the logic densely and is easy to modify. Although physically two levels of NOR gates, a PLA can be thought of as an AND layer followed by an OR layer. The AND layer matches particular bit patterns and then the OR layer combines multiple values from the first layer to produce arbitrary outputs.

The Group Decode ROM. This photo shows the metal layer on top of the die.

The Group Decode ROM. This photo shows the metal layer on top of the die.

Since the output values are highly structured, a PLA implementation is considerably more efficient than a ROM, since in a sense it combines multiple entries. In the case of the Group Decode ROM, using a ROM structure would require 256 columns (one for each 8-bit instruction pattern), while the PLA implementation requires just 36 columns, about 1/7 the size.

The diagram below shows how one column of the Group Decode ROM is wired in the "AND" plane. In this die photo, I removed the metal layer with acid to reveal the polysilicon and silicon underneath. The vertical lines show where the metal line for ground and the column output had been. The basic idea is that each column implements a NOR gate, with a subset of the input lines selected as inputs to the gate. The pull-up resistor at the top pulls the column line high by default. But if any of the selected inputs are high, the corresponding transistor turns on, connecting the column line to ground and pulling it low. Thus, this implements a NOR gate. However, it is more useful to think of it as an AND of the complemented inputs (via De Morgan's Law): if all the inputs are "correct", the output is high. In this way, each column matches a particular bit pattern.

Closeup of a column in the Group Decode ROM.

Closeup of a column in the Group Decode ROM.

The structure of the ROM is implemented through the silicon doping pattern, which is visible above. A transistor is formed where a polysilicon wire crosses a doped silicon region: the polysilicon forms the gate, turning the transistor on or off. At each intersection point, a transistor can be created or not, depending on the doping pattern. If a particular transistor is created, then the corresponding input must be 0 to produce a high output.

At the top of the diagram above, the column outputs are switched from the metal layer to polysilicon wires and become the inputs to the upper "OR" plane. This plane is implemented in a similar fashion as a grid of NOR gates. The plane is rotated 90 degrees, with the inputs vertical and each row forming an output.

Intermediate decoding in the Group Decode ROM

The first plane of the Group Decode ROM categorizes instructions into 36 types based on the instruction bit pattern.2 The table below shows the 256 instruction values, colored according to their categorization.3 For instance, the first blue block consists of the 32 ALU instructions corresponding to the bit pattern 00XXX0XX, where X indicates that the bit can be 0 or 1. These instructions are all decoded and executed in a similar way. Almost all instructions have a single category, that is, they activate a single column line in the Group Decode ROM. However, a few instructions activate two lines and have two colors below.

Grid of 8086 instructions, colored according to the first level of the Group Decode Rom.

Grid of 8086 instructions, colored according to the first level of the Group Decode Rom.

Note that the instructions do not have arbitrary numeric opcodes, but are assigned in a way that makes decoding simpler. Because these blocks correspond to bit patterns, there is little flexibility. One of the challenges of instruction set design for early microprocessors was to assign numeric values to the opcodes in a way that made decoding straightforward. It's a bit like a jigsaw puzzle, fitting the instructions into the 256 available values, while making them easy to decode.

Outputs from the Group Decode ROM

The Group Decode ROM has 15 outputs, one for each row of the upper half. In this section, I'll briefly discuss these outputs and their roles in the 8086. For an interactive exploration of these signals, see this page, which shows the outputs that are triggered by each instruction.

Out 0 indicates an IN or OUT instruction. This signal controls the M/IO (S2) status line, which distinguishes between a memory read/write and an I/O read/write. Apart from this, memory and I/O accesses are basically the same.

Out 1 indicates (inverted) that the instruction has a Mod R/M byte and performs a read/modify/write on its argument. This signal is used by the Translation ROM when dispatching an address handler (details). (This signal distinguishes between, say, ADD [AX],BX and MOV [AX],BX. The former both reads and writes [AX], while the latter only writes to it.)

Out 2 indicates a "group 3/4/5" opcode, an instruction where the second byte specifies the particular instruction, and thus decoding needs to wait for the second byte. This controls the loading of the microcode address register.

Out 3 indicates an instruction prefix (segment, LOCK, or REP). This causes the next byte to be decoded as a new instruction, while blocking interrupt handling.

Out 4 indicates (inverted) a two-byte ROM instruction (2BR), i.e. an instruction is handled by the microcode ROM, but requires the second byte for decoding. This is an instruction with a Mod R/M byte. This signal controls the loader indicating that it needs to fetch the second byte. This signal is almost the same as output 1 with a few differences.

Out 5 specifies the top bit for an ALU operation. The 8086 uses a 5-bit field to specify an ALU operation. If not specified explicitly by the microcode, the field uses bits 5 through 3 of the opcode. (These bits distinguish, say, an ADD instruction from AND or SUB.) This control line sets the top bit of the ALU field for instructions such as DAA, DAS, AAA, AAS, INC, and DE that fall into a different set from the "regular" ALU instructions.

Out 6 indicates an instruction that sets or clears a condition code directly: CLC, STC, CLI, STI, CLD, or STD (but not CMC). This signal is used by the flag circuitry to update the condition code.

Out 7 indicates an instruction that uses the AL or AX register, depending on the instruction's size bit. (For instance MOVSB vs MOVSW.) This signal is used by the register selection circuitry, the M register specifically.

Out 8 indicates a MOV instruction that uses a segment register. This signal is used by the register selection circuitry, the N register specifically.

Out 9 indicates the instruction has a d bit, where bit 1 of the instruction swaps the source and destination. This signal is used by the register selection circuitry, swapping the roles of the M and N registers according to the d bit.

Out 10 indicates a one-byte logic (1BL) instruction, a one-byte instruction that is implemented in logic, not microcode. These instructions are the prefixes, HLT, and the condition-code instructions. This signal controls the loader, causing it to move to the next instruction.

Out 11 indicates instructions where bit 0 is the byte/word indicator. This signal controls the register handling and the ALU functionality.

Out 12 indicates an instruction that operates only on a byte: DAA, DAS, AAA, AAS, AAM, AAD, and XLAT. This signal operates in conjunction with the previous output to select a byte versus word.

Out 13 forces the instruction to use a byte argument if instruction bit 1 is set, overriding the regular byte/word pattern. Specifically, it forces the L8 (length 8 bits) condition for the JMP direct-within-segment and the ALU instructions that are immediate with sign extension (details).

Out 14 allows a carry update. This prevents the carry from being updated by the INC and DEC operations. This signal is used by the flag circuitry.

Columns

Most of the Group Decode ROM's column signals are used to derive the outputs listed above. However, some column outputs are also used as control signals directly. These are listed below.

Column 10 indicates an immediate MOV instruction. These instructions use instruction bit 3 (rather than bit 1) to select byte versus word, because the three low bits specify the register. This signal affects the L8 condition described earlier and also causes the M register selection to be converted from a word register to a byte register if necessary.

Column 12 indicates an instruction with bits 5-3 specifying the ALU instruction. This signal causes the X register to be loaded with the bits in the instruction that specify the ALU operation. (To be precise, this signal prevents the X register from being reloaded from the second instruction byte.)

Column 13 indicates the CMC (Complement Carry) instruction. This signal is used by the flags circuitry to complement the carry flag (details).

Column 14 indicates the HLT (Halt) instruction. This signal stops instruction processing by blocking the instruction queue.

Column 31 indicates a REP prefix. This signal causes the REPZ/NZ latch to be loaded with instruction bit 0 to indicate if the prefix is REPNZ or REPZ. It also sets the REP latch.

Column 32 indicates a segment prefix. This signal loads the segment latches with the desired segment type.

Column 33 indicates a LOCK prefix. It sets the LOCK latch, locking the bus.

Column 34 indicates a CLI instruction. This signal immediately blocks interrupt handling to avoid an interrupt between the CLI instruction and when the interrupt flag bit is cleared.

Timing

One important aspect of the Group Decode ROM is that its outputs are not instantaneous. It takes a clock cycle to get the outputs from the Group Decode ROM. In particular, when instruction decoding starts, the timing signal FC (First Clock) is activated to indicate the first clock cycle. However, the Group Decode ROM's outputs are not available until the Second Clock SC.

One consequence of this is that even the simplest instruction (such as a flag operation) takes two clock cycles, as does a prefix. The problem is that even though the instruction could be performed in one clock cycle, it takes two clock cycles for the Group Decode ROM to determine that the instruction only needs one cycle. This illustrates how a complex instruction format impacts performance.

The FC and SC timing signals are generated by a state machine called the Loader. These signals may seem trivial, but there are a few complications. First, the prefetch queue may run empty, in which case the FC and/or SC signal is delayed until the prefetch queue has a byte available. Second, to increase performance, the 8086 can start decoding an instruction during the last clock cycle of the previous instruction. Thus, if the microcode indicates that there is one cycle left, the Loader can proceed with the next instruction. Likewise, for a one-byte instruction implemented in hardware (one-byte logic or 1BL), the loader proceeds as soon as possible.

The diagram below shows the timing of an ADD instruction. Each line is half of a clock cycle. Execution is pipelined: the instruction is fetched during the first clock cycle (First Clock). During Second Clock, the Group Decode ROM produces its output. The microcode address register also generates the micro-address for the instruction's microcode. The microcode ROM supplies a micro-instruction during the third clock cycle and execution of the micro-instruction takes place during the fourth clock cycle.

This diagram shows the execution of an ADD instruction and what is happening in various parts of the 8086. The arrows show the flow from step to step. The character µ is short for "micro".

This diagram shows the execution of an ADD instruction and what is happening in various parts of the 8086. The arrows show the flow from step to step. The character µ is short for "micro".

The Group Decode ROM's outputs during Second Clock control the decoding. Most importantly, the ADD imm instruction used microcode; it is not a one-byte logic instruction (1BL). Moreover, it does not have a Mod R/M byte, so it does not need two bytes for decoding (2BR). For a 1BL instruction, microcode execution would be blocked and the next instruction would be immediately fetched. On the other hand, for a 2BR instruction, the loader would tell the prefetch queue that it was done with the second byte during the second half of Second Clock. Microcode execution would be blocked during the third cycle and the fourth cycle would execute a microcode subroutine to determine the memory address.

For more details, see my article on the 8086 pipeline.

Interrupts

The Group Decode ROM takes the 8 bits of the instruction as inputs, but it has an additional input indicating that an interrupt is being handled. This signal blocks most of the Group Decode ROM outputs. This prevents the current instruction's outputs from interfering with interrupt handling. I wrote about the 8086's interrupt handling in detail here, so I won't go into more detail in this post.

Conclusions

The Group Decode ROM indicates one of the key differences between CISC processors (Complex Instruction Set Computer) such as the 8086 and the RISC processors (Reduced Instruction Set Computer) that became popular a few years later. A RISC instruction set is designed to make instruction decoding very easy, with a small number of uniform instruction forms. On the other hand, the 8086's CISC instruction set was designed for compactness and high code density. As a result, instructions are squeezed into the available opcode space. Although there is a lot of structure to the 8086 opcodes, this structure is full of special cases and any patterns only apply to a subset of the instructions. The Group Decode ROM brings some order to this chaotic jumble of instructions, and the number of outputs from the Group Decode ROM is a measure of the instruction set's complexity.

The 8086's instruction set was extended over the decades to become the x86 instruction set in use today. During that time, more layers of complexity were added to the instruction set. Now, an x86 instruction can be up to 15 bytes long with multiple prefixes. Some prefixes change the register encoding or indicate a completely different instruction set such as VEX (Vector Extensions) or SSE (Streaming SIMD Extensions). Thus, x86 instruction decoding is very difficult, especially when trying to decode multiple instructions in parallel. This has an impact in modern systems, where x86 processors typically have 4 complex instruction decoders while Apple's ARM processors have 8 simpler decoders; this is said to give Apple a performance benefit. Thus, architectural decisions from 45 years ago are still impacting the performance of modern processors.

I've written numerous posts on the 8086 so far and plan to continue reverse-engineering the 8086 die so follow me on Twitter @kenshirriff or RSS for updates. I've also started experimenting with Mastodon recently as @[email protected]. Thanks to Arjan Holscher for suggesting this topic.

Notes and references

  1. You might wonder what the difference is between a ROM and a PLA. Both of them produce arbitrary outputs for a set of inputs. Moreover, you can replace a PLA with a ROM or vice versa. Typically a ROM has all the input combinations decoded, so it has a separate row for each input value, i.e. 2N rows. So you can think of a ROM as a fully-decoded PLA.

    Some ROMs are partially decoded, allowing identical rows to be combined and reducing the size of the ROM. This technique is used in the 8086 microcode, for instance. A partially-decoded ROM is fairly similar to a PLA, but the technical distinction is that a ROM has only one output row active at a time, while a PLA can have multiple output rows active and the results are OR'd together. (This definition is from The Architecture of Microprocessors p117.)

    The Group Decode ROM, however, has a few cases where multiple rows are active at the same time (for instance the segment register POP instructions). Thus, the Group Decode ROM is technically a PLA and not a ROM. This distinction isn't particularly important, but you might find it interesting. 

  2. The Group Decode ROM has 38 columns, but two columns (11 and 35) are unused. Presumably, these were provided as spares in case a bug fix or modification required additional decoding. 

  3. Like the 8008 and 8080, the 8086's instruction set was designed around a 3-bit octal structure. Thus, the 8086 instruction set makes much more sense if viewed in octal instead of hexadecimal. The table below shows the instructions with an octal organization. Each 8×8 block uses the two low octal digits, while the four large blocks are positioned according to the top octal digit (labeled). As you can see, the instruction set has a lot of structure that is obscured in the usual hexadecimal table.

    The 8086 instruction set, put in a table according to the octal opcode value.

    The 8086 instruction set, put in a table according to the octal opcode value.

    For details on the octal structure of the 8086 instruction set, see The 80x86 is an Octal Machine