engineering sample samples qualification cpu processor prozessor information mhz pictures core frequency chip packaging info ic x86 museum collection amd cyrix harris ibm idt iit intel motorola nec sgs sgs-thomson siemens ST signetics mhs ti texas instruments ulsi hp umc weitek zilog 4004 4040 8008 808x 8085 8088 8086 80188 80186 80286 286 80386 386 i386 Am386 386sx 386dx 486 i486 586 486sx 486dx overdrive 80187 80287 387 487 pentium 586 5x86 386dlc 386slc 486dx2 mmx ppro pentium-pro pro athlon duron z80 sparc alpha dec dirk oppelt
News Archive 2004

News Archive 2006
News Archive 2005

December 2004
November 2004
October 2004
August - September 2004
April - July 2004
March 2004
February 2004
January 2004

News Archive 2003

December 2004 updates and changes top of page
* 2004-12-24
In October I added a nice service from HitMaps that collects hits to a web page by including a small image into the page. The IP address of the page requester is translated into geographical coordinates and stored in a database. Once a day, the HitMaps map images are updated according to the database. Now, about 2 months later, I am happy to present the HitMap that shows my site visitors' locations all over the world:

Locations of visitors to this page
(click on the map thumbnail to enlarge)

I did not expect such a broad audience from so many different locations - thank you all very much for your interest! My best wishes for the holidays go to all my readers around the world.
* 19 new additions to the collection today:

  • Math Coprocessors

    - for 8086/88 Systems
    Intel D8087-1
    The 8087 was the first coprocessor that Intel made available for the 80x86 family. It was introduced in 1980 and therefore does not have full compatibility with the IEEE-754 standard for floating-point arithmetic (which was finally released in 1985). It complements the 8088 and 8086 CPUs and can also be interfaced to the 80188 and 80186 processors.
    The 8087 is implemented using NMOS. It comes in a 40-pin CERDIP and is available in 5 MHz, 8 MHz, and 10 MHz versions. Power consumption is rated at max. 2400 mW.
    A neat trick to enhance the processing power of the 8087 for computations that use only the basic arithmetic operations (+,-,*,/) and do not require high precision is to set the precision control to single-precision. This gives one a performance increase of up to 20%.
    With the help of an additional chip, the 8087 can in theory be interfaced to an 80186 CPU. The 80186 was used in some PCs (e.g. from Philips, Siemens) in 1982/1983.

    - for 80286 Systems
    IIT 2C87-8
    IIT 2C87-10
    IIT 2C87-20
    The IIT 2C87 was the first Intel 287 clone. It was introduced in 1989 and has about the same speed as the Intel 287XL. The 2C87 implements the full 387 instruction set. It is available for speeds of up to 20 MHz. The 2C87 implements the full 80387 instruction set but is not fully compatible with the IEEE-754 standard for floating-point arithmetic.

    Cyrix FasMath CX-82S87-NP-SV
    This 80287-compatible chip was developed from the Cyrix 83D87 and has been available since 1991. It complies completely with the IEEE-754 standard for floating-point arithmetic and features nearly total compatibility with Intel's coprocessors, including implementation of the full Intel 80387 instruction set. It implements the transcendental functions with the same degree of accuracy and the superior speed of the Cyrix 83D87. This makes the Cyrix 82S87 the fastest and most accurate 287 compatible coprocessor available.
    Documentation by Cyrix rates the 82S87 at 730 kWhets/sec for a 12.5 MHz system, while the Intel 287XL performs only 552 kWhets/sec. 82S87 chips manufactured after 1991 use the internals of the Cyrix 387+, which succeeds the original 83D87.

    - for 386 Systems
    Intel A80387-16B
    Intel A80387-20
    This chip was the first generation of coprocessors designed specifically for the Intel 80386 CPU. It was introduced in 1986, about one year after the 80386 was brought to market. Early 386 system were therefore equipped with both a 80287 and a 80387 socket. The 80386 does work with an 80287, but the numerical performance is hardly adequate for such a system.
    The 80387 has itself since been superseded by the Intel 387DX introduced by a quiet change in 1989. The old 80387 is about 20% slower than the newer 387DX.

    Cyrix FasMath CX-83D87-25-GP-KN
    Cyrix FasMath CX-83D87-25-GP-XA
    Cyrix FasMath CX-83D87-33-GP-KN
    The CX-83D87 was introduced in 1989. It is the fastest 387-compatible coprocessor and provides up to 50% more performance than the Intel 387DX. The 83D87 also offers the most accurate transcendental functions of all coprocessors. It is the 387 clone with the highest degree of compatibility to the Intel 387DX. Unlike the Intel 387DX, the 83D87 (and all other 387-compatible chips as well) does not support asynchronous operation of CPU and coprocessor. To reduce power consumption the 83D87 features advanced power saving features. Those portions of the coprocessor that are not needed are automatically shut down. If no coprocessor instructions are being executed, all parts except the bus interface unit are shut down.

    Cyrix FasMath CX-83S87-16-JP
    Cyrix FasMath CX-83S87-16-KP
    Cyrix FasMath CX-83S87-20-JP
    Cyrix FasMath CX-83S87-25-JP
    Cyrix FasMath CX-83S87-33-JP
    The 83S87 is the SX version of the Cyrix 83D87. Just as the 83D87 is the fastest 387-compatible coprocessor, the Cyrix 83S87 is the fastest of the 387SX compatible coprocessors, as well as providing the most accurate transcendental functions. 83S87 chips manufactured after 1991 use the internals of the Cyrix 387+, the successor to the original 83D87. The Cyrix 83S87 is ideally suited to be used with the Cyrix Cx486SLC CPU, a 486SX compatible CPU which is a replacement chip for the Intel 386SX CPU.
    The 83S87 is packaged in a 68-pin PLCC and is available in 16, 20, 25, and 33 MHz versions. Due to the advanced power saving features of the Cyrix coprocessor, the typical power consumption of the 20 MHz version is only about 350 mW, while maximum power dissipation is 1.6 W.

    - for 486 Systems
    IIT XC87DLC-33
    IIT XC87SLC-33
    IIT 4C87DLC-33
    The IIT DLC/SLC are the coporocessors for 386/486 hybrid processors like the Cyrix Cx486 DLC/SLC or Texas Instruments TX486DLC. They can also be used interchangeably with the IIT 3C87 as normal 386 FPUs.

    - for NexGen Systems
    NexGen Nx587
    A math coprocessor for the NexGen Nx586 processor, announced when the part first shipped, but never officially introduced. Due to various issues the two-socket solution envisioned turned out to be impractical, so finally NexGen shipped the Nx586 without a math unit (and therefore poor floating point performance) followed by the Nx586FP. The Nx586FP is a multi-chip module containing two dies, the Nx586 CPU and the FPU.

    Most of the information provided here is from Norbert Juffa's excellent FPU reference "Everything you always wanted to know about math coprocessors", please refer to 'copro16a.txt' for more information.


  • November 2004 updates and changes top of page
    * 2004-11-28
  • 14 processors for sale on eBay is offering 14 processors on eBay:
    AMD A80486DXL2-66
    AMD5k86-P75 SSA/5-75ABR
    AMI S6820, old (1977)
    Cyrix 5x86-100GP
    Cyrix Cx486DX2-80
    Cyrix Cx486S-40GP FasCache
    Cyrix M II-300GP, 66MHz Bus
    IIT FPU 4C87DLC-40
    Intel FPU A80387DX 16-33
    Intel i486 Overdrive ODPR486DX-33
    Intel Pentium 166 MMX BP80503166 Boxed Version
    SGS-Thomson It's ST486 DX2-66
    SGS-Thomson It's ST486 DX2-66GS
    Texas Instruments TI486DX2-G80-GA
    The processors will be shipped worldwide. Auctions end Sunday, December 5th. Here's a list of all items at eBay.

    19 new additions to the collection today:

  • MIPS Rx000 Processors

    MIPS (Microprocessor without interlocked pipeline stages) is a RISC microprocessor architecture developed by MIPS Computer Systems Inc in the early eighties. MIPS processors were used in high-end servers from Siemens and DEC and especially in SGI's computer product line, and have found broad application in embedded systems, Windows CE devices, and Cisco routers. The Nintendo 64 and Sony PlayStation 2 consoles also use MIPS processors. The architecture was very successful, about one third of all RISC chips produced in mid 1990s were MIPS based designs. In 1999 the ARM/StrongARM architecture took over rather decisively in thanks to cell phone and PDA usage.
    The MIPS design is licensed to many 3rd party vendors, among them IDT, Siemens, NEC, HP and Toshiba. It is one of the most popular licensed architectures. This is not because the MIPS architecture is particularly good in a specific task. It isn't. MIPS is just a convenient, clean, and easily scaled architecture around which special-purpose network processors or protocol engines can be added. It is one of the cleanest and most generic processor designs around, finely tuned for absolutely nothing.
    References:  MIPS architecture at Wikipedia  Number one RISC architecture (1997)  MIPS architecture documentation  MIPS R series tech docs  Comprehensive SGI information at SGIstuff

    New chips in the collection:

    R3000 series
    Siemens R3000A-33-AE
    IDT 79R3000A-12G
    The first commercial MIPS CPU model, the R2000, was announced in 1985 as a 32-bit implementation. It was followed by the R3000, the first successful MIPS design in the marketplace with more than 1 million processors made. The R3000 was used in high-end UNIX computers by Siemens and DEC and in the Silicon Graphics SGI Personal IRIS 4D/20 graphic workstations. On this machines 3D sequences for movies like The Abyss, Jurrasic Park or Terminator 2 were rendered.
    IDT Rx000 FPU 79R3010A-12QJ
    The R3010 is the FPU for the R3000.
    Reference:  R3000 at SGIstuff

    R4400 series
    IDT 79R4400MC-50G447
    NEC D30412RJ-50 VR4400MC
    NEC D30412RJ-75 VR4400MC
    NEC D30412LRJ-200 VR4400MC
    NEC D30412LRJ-250 VR4400MC
    Toshiba TC86R4400MC-200
    The R4000 series, released in 1991, extended the MIPS instruction set to a full 64-bit architecture and moved the FPU onto the main die to create a single-chip system. The design was so important to SGI, at the time MIPS' major customer, that SGI bought the company in 1992. A number of improved versions soon followed, including the R4400.
    The R4x00 processors have been available and used in different versions. PC (as in R4400PC) denotes primary cache only and SC denotes secondary cache. The MC versions contain special support for cache architectures in multiprocessor systems.
    R4400 processors where used in many Silicon Graphics computers, e.g. the Indigo2 workstations, Onyx supercomputers and Challenge servers.
    References:  MIPS R4400 Product Information  R4400 at SGIstuff

    R4300 series
    NEC D30200GD-100 VR4300
    The R4300, introduced in 1995, is a low-cost but very powerful processor with its external bus reduced to 32 bit. It was used in the Nintendo64 game console, the NeoGeo Hyper 64 arcade machine system and many other devices like set-top boxes, video entertainment systems, arcade games, x-terminals, network hardware and laser printers.
    References:  Nintendo64 Tech Info

    R4600 series
    IDT 79R4600-100G
    IDT 79R4600-133G
    IDT 79R4600-133G 'Orion' Labeled
    The R4600 'Orion' was designed by Quantum Effect Devices (QED), a company started by former MIPS employees. R4600 processors where used in Silicon Graphics Indy workstations and are still used in Cisco series 4000 routers.
    References:  MIPS R4600 Product Information  IDT 79R4700 Documents  R4600 at SGIstuff

    R5000 series
    IDT 79RV5000-150G
    The R5000, also a QED design, replaced the R4600 in 1996. The R5000 FPU had more flexible single precision floating-point scheduling than the R4x00, and as a result, R5000-based SGI Indys had much better graphics performance than similarly clocked R4400 Indys with the same graphics hardware. SGI gave the old graphics board a new name (XL8 -> XGE8) when it was combined with R5000 in order to emphasize the improvement.
    References:  MIPS R5000 introduction  MIPS R5000: Fast, Affordable 3-D  IDT 79RV5000 Documents  R5000 at SGIstuff

    R10000 series
    NEC D30700LRS-180 VR10000
    NEC D30700LRS-250 VR10000
    Toshiba TC86R10000-195
    The R10000 'T5' is the first single-chip 4-issue superscalar processor by MIPS, introduced in 1995. It gave a 70 to 100% performance increase over the R4400 at 250 MHz, previously the most powerful MIPS processor. The R10000 CPU was used in NEC supercomputers and various SGI machines like the Indigo2, O2, Octane, Octane 2, Onyx, Onyx 2, Origin and Challenge.
    References:  R10000 User Guide  R10000 at SGIstuff  NEC VR10000 announcement  T5: Brute Force (

    R12000 series
    NEC D30710RS-300 VR12000
    The R12000, released in 1998, is an enhanced R10000 and used an improved process to shrink the chip and run it at higher clock rates. It was used in NEC supercomputers and several SGI machines like the O2, Octane 2, Onyx 2 and later Origin.
    References:  RISC Fights Back with the MIPS R12000  R12000 at SGIstuff  NEC VR12000 launch

    Sony Emotion Engine
    Sony CXD9615GB 'Emotion Engine'
    The Emotion Engine is the CPU for Sony's PlayStation 2, developed in a Sony - Toshiba cooperation and introduced in 1999. Data bus, cache memory as well as all registers are implemented in 128 bit technology, integrated on a single 0.18 micron process technology chip, making it the first commercial 128 bit CPU. The Emotion Engine, based on the MIPS R5900, is sort of a combination CPU and DSP processor, whose main function is simulating 3D worlds. It integrated all necessary units on the die: The MIPS III CPU core, 2 vector units, FPU, image processing unit (basically an MPEG2 decoder with some other capabilities), 10-channel DMA controller, graphics interface unit, RDRAM and I/O interfaces, all connected via a shared 128-bit internal bus.
    References:  Sony announcement  Ars Technica Technical Overview  Die shots
  • * 2004-11-07
  • 15 processors for sale on eBay is offering 15 processors on eBay:
    AMD A80486SX2-66
    AMD N80L286-12/S
    IBM 486 DX4 (486-4V3100GC)
    IBM 6x86MX PR300, black
    IBM Blue Lightning DX2, 80MHz (486-V580GA)
    Intel DX4ODPR100 Overdrive V1.1
    Intel ODPR486DX-25 Overdrive V3.0
    Intel Pentium P5 A80501-60 (old version, no goldcap)
    Intel Pentium P5 A80501-66
    Intel R80286-10
    Motorola PowerPC 604e, XPC604ERX166PD
    Motorola PowerPC 750 (G3), XPC750ARX233PE
    Motorola XC68040RC33M
    Texas Instruments TI486DX2-G66-GA
    ULSI FPU Math-Co DX, 40MHz
    The processors will be shipped worldwide. Auctions end Sunday, November 14th. Here's a list of all items at eBay.

    31 new additions to the collection today:

  • Intel 80186/80188 Processors

    The 80186 microprocessor was developed by Intel in 1982. It is an improved 8086 with several common support functions built in: clock generator, system controller, interrupt controller, DMA controller, and timer/counter. It also added 8 new instructions and executes instructions faster than the 8086. As with the 8086, it has a 16-bit external bus and is also available as the 80188, with an 8-bit external data bus. The initial clock rate of the 80186 and 80188 was 6 MHz. They were, and still are, generally used as embedded processors but also as the CPU of few personal computers:
    • The Mindset graphics computer, a very advanced computer for the time. It had proprietary chips that enhanced and sped up the graphics.
    • The original Gateway Handbook, a small subnotebook computer.
    • The Telenova Compis, a Swedish school computer.
    • The Tandy 2000, a somewhat PC-compatible workstation featuring particularly advanced graphics for its time.
    In 1987 Intel announced the second generation of the 80186 family: the 80C186/C188. The 80186 was redesigned as a static, stand-alone module known as the 80C186 Modular Core and is pin compatible with the 80186 family, while adding an enhanced feature set. The high-performance CHMOS III process allowed the 80C186 to run at twice the clock rate of the NMOS 80186, while consuming less than one-fourth the power.
    In 1991 the 80C186 Modular Core family was again extended with the introduction of the 80C186XL. The 80C186XL/C188XL is a higher performance, lower power replacement for the 80C186/C188.

    New chips in the collection:

    Intel 80186
    Ceramic LCC package:
    Intel C80186-6, 6 MHz
    Intel C80186-3, 8 MHz
    Intel C80186, 8 MHz
    Intel R80186, 10 MHz
    Intel R80C186-16, 16 MHz
    Ceramic PGA package:
    Intel A80C186-16, 16 MHz
    Plastic PLCC package:
    Intel N80C186XL10, 10 MHz
    Intel N80C186XL12, 12 MHz
    Intel TN80C186-12, 12 MHz
    Intel TN80C186XL12, 12 MHz

    Second source versions by AMD and Siemens:
    Ceramic LCC package:
    AMD C80186-3, 8 MHz
    AMD R80186-10, 10 MHz
    Siemens SAB 80186-R, 8 MHz
    Plastic PLCC package:
    AMD N80C186, 10 MHz
    AMD N80C186-16, 16 MHz

    The Intel 80188 is a version of the 80186 with an 8 bit external data bus, instead of 16 bit. This makes it less expensive to connect to peripherals.

    Intel 80188
    Ceramic LCC package:
    Intel R80188, 10 MHz
    Plastic PLCC package:
    Intel N80C188, 8 MHz
    Intel TN80C188, 8 MHz
    Intel N80C188XL12, 12 MHz
    Intel TN80C188EB13, 13 MHz
    Intel N80C188XL16, 16 MHz
    Intel S80C188XL20, 20 MHz

    Second source versions by AMD and Siemens:
    Ceramic LCC package:
    AMD R80188, 10 MHz
    AMD R80188B, 10 MHz
    Siemens SAB 80188-R, 8 MHz
    Plastic PLCC package:
    AMD N80C188, 8 MHz
    AMD N80C188-12, 12 MHz
    AMD IN80C188-12, 12 MHz
    AMD N80C188-16, 16 MHz
    AMD N80C188-20, 20 MHz
    AMD Am188ES-20KC, 20 MHz

    References:  Manuals at  Intel 80186 overview  Intel 80186 documentation

  • October 2004 updates and changes top of page

  • 16 processors for sale on eBay is offering 16 processors on eBay:

    AMD 5k86-P90 SSA/5-90ABQ
    Cyrix Cx486S-40 + Cyrix FPU Cx487S-40QP
    Cyrix 5x86-100GP w/ Cooler
    Intel FPU A80387DX 16-33
    Intel A80486DX2-66 w/ blue Cooler
    Intel Pentium 166 MMX BP80503166, Boxed Version
    Intel Pentium 200 MMX BP80503200, Boxed Version
    Intel i486 OverDrive DX2ODPR66, V3.0
    Intel i486 OverDrive ODPR486DX-33, V3.0
    Intel Pentium P5, 66MHz A80501-66
    Intel PODP5V83 Pentium Overdrive
    It's ST 486DX2-80GS
    Texas Instruments TI486DX2-G80-GA
    ULSI FPU Math-Co US83C87, 33MHz
    UMC Green CPU U5SX 486-33

    The processors will be shipped worldwide. Auctions end Sunday, October 31st. Here's a list of all items at eBay.
  • 2004-10-05

    5 new additions to the collection today:

  • Intel Pentium II/III Xeon Processors

    Pentium II Xeon
    Intel Pentium II Xeon 400MHz / 512KB Cache
    Intel Pentium II Xeon 450MHz / 1MB Cache
    The Pentium II Xeon was introduced in June 1998 as Intel's new line of server and workstation processors. The PII Xeon core, manufactured in a .25 micron process, is not much different from the Pentium IIs core. It added multiprocessing support for quad CPU systems and even 8 CPUs in one system and a few enhancements, such as support for more than 4GB of memory. The biggest difference in both performance and architecture comes from the PII Xeons L2-cache, at 512KB, 1 or even 2 MB and running at full clockspeed, while the Pentium II was only offered with a 512KB L2 cache running at half CPU clockspeed.
    The PII Xeon also introduced Intel's biggest processor module so far (about twice the size of a Pentium II cartridge) to include the CPU core and up to four 512KB L2 cache chips.
    References:  Intel Technical Documents  Tom's Hardware Guide  Xeon at Wikipedia  Pictures and Info at

    Pentium III Xeon
    Intel Pentium III Xeon 80525KY500512 Engineering Sample
    This is an Engineering Sample of a Pentium III Xeon 500MHz in a Pentium II Xeon cartridge. It was manufactured in October 1998, almost half a year before its official introduction in March 1999.
    Intel Pentium III Xeon 667/133/256
    Intel Pentium III Xeon 80526KZ800256 Engineering Sample
    The Pentium III Xeon was introduced in March 1999. Early versions (Code named Tanner) where built in .25 micron technology and featured a 100MHz front side bus, 512KB to 2MB L2 off-die cache sizes and multiprocessor support for up to 8 CPUs. In October 1999 Intel introduced the 2nd generation PIII Xeon Cascades, using the same core as the .18 micron Pentium III Coppermine processor. As a result the Cascades has its 256KB on-die L2 cache running at full clock speed and all the other Coppermine features like Advanced Transfer Cache, Advanced System Buffering, support for 133 MHz front speed bus, SSE (Streaming SIMD Extensions), etc.
    However, the introduction of Cascades had some severe drawbacks: The small 256KB L2 cache took back the Xeon's performance advance over the standard Pentium III and the 133MHz bus protocol did not support more than 2 CPUs in a system. In May 2000 Intel finally introduced a "large cache" version with up to 2MB (on-die) L2 cache and 100MHz front side bus, again supporting multiprocessing with up to 8 CPUs.
    References:  Intel Technical Documents  Pictures and Info at

    There is one Pentium III Xeon 667/133/256 available for trade, including its original package.
  • 2004-10-03

  • 12 processors for sale on eBay is offering 12 processors on eBay:

    AMD Am486DX2-50
    AMD Am486SX2-66
    Cyrix 6x86MX-PR200, 66MHz Bus
    Cyrix M II-300GP, 75MHz Bus
    IBM 486DX, rare OEM-Version
    IDT 79R4600-100G
    Intel FPU A80387DX-33
    Intel DX4ODPR100 Overdrive
    Intel Pentium P5 A80501-60, 'Processor'-Marking
    Intel PODP5V83 Pentium Overdrive
    Motorola PowerPC 750 (G3), XPC750ARX233PE
    NexGen Nx586-P90

    The processors will be shipped worldwide. Auctions end Sunday, 10th. Here's a list of all items at eBay.

  • August - September 2004 updates and changes top of page

    21 new additions to the collection today:

  • PowerPC Processors

    PowerPC is a RISC microprocessor architecture created by the 1991 Apple-IBM-Motorola alliance (AIM).
    'Power' (Performance Optimization With Enhanced RISC) and was adopted from IBM's POWER architecture from their RS/6000 series.
    The PowerPC is designed along RISC principles, and allows for a superscalar implementation. Versions of the design exist in both 32-bit and 64-bit implementations. Starting with the basic POWER specification, the PowerPC added:
    • big or little-endian modes (requiring a reboot)
    • single-precision floating point in addition to double-precision
    • additional floating point instructions at the behest of Apple
    • a complete 64-bit specification, which is backward compatible with the 32-bit mode
    • removal of some of the more esoteric POWER instructions, which are emulated in microcode
    The first single-chip implementation of the design was the MCP601 and released in Apple's PowerMac in March 1994. For a list of Apple computers using a specific PowerPC CPU please refer to
    PowerPC on Wikipedia

    PowerPC 601
    IBM PowerPC 601, PPC601FD-080-2, 80MHz
    The 601 microprocessor is the first member of a family of devices which IBM and Motorola jointly developed in Austin, Texas. It is considered the first-generation PowerPC because it was a hybrid chip based on IBM's POWER architecture and Motorola's 88110.
    References:  Motorola MPC 601 User's Manual  Motorola MPC 601 Doc Archive  IBM PPC 601 White Paper

    PowerPC 603
    Motorola PowerPC 603, XPC603PRX166LC, 166MHz
    Motorola PowerPC 603, XPC603RRX200LC, 200MHz
    Motorola PowerPC 603, XPC603RRX250LC, 250MHz
    IBM PowerPC 603ev, PPC603evBB200r, 200MHz
    IBM PowerPC 603ev, PPC603evFB200r, 200MHz
    The PowerPC 603 was designed to run in low cost systems and portable machines. The first PowerPC 603 was slower than the PowerPC 601 at the same clock speed - partially due to the 16KB L1 cache as opposed to the 32KB L1 cache in the PowerPC 601. Because of its size and low power usage, it was quickly ramped up to higher clock speeds.
    The PowerPC 603 was followed by the 603e, adding 16KB of L1 cache to some models, and the 603ev which shrunk the manufacturing process from .5 microns to .35 microns on some models, allowing it to be clocked even higher.
    References:  Motorola MPC 603/603e User's Manual  Motorola MPC 603e Docs  Motorola MPC 603 Doc Archive  IBM PPC 603 White Paper  IBM PPC 603e Technical Library

    PowerPC 604
    Motorola PowerPC 604e, XPC604ERX166PD, 166MHz
    Motorola PowerPC 604e, XPC604RRX300QB, 300MHz
    IBM PowerPC 604, PPCA604FC1202PQ, 120MHz
    IBM PowerPC 604, PPCA604BE133aCPQ, 133MHz
    IBM PowerPC 604e, XPPC604eBC166aC relabeled from 180 to 166MHz
    IBM PowerPC 604e, PPC604e2BE233eE, 233MHz
    The PowerPC 604 was the high end PowerPC when released, performing at 1.5 times the speed of a PowerPC 601 at equal clock speeds. The chip was later updated by doubling the L1 cache to 64KB and shrinking the processor, now dubbed the 604e.
    The fastest versions of the 604e version were nicknamed "Mach 5" and ran at 250-350MHz.
    References:  Motorola MPC 604 User's Manual  Motorola MPC 604 Hardware Specs  Motorola MPC 604e Hardware Specs  Motorola MPC 604 Doc Archive  IBM PPC 604e Technical Library

    PowerPC 740/750 (G3)
    Motorola PowerPC 740 (G3), XPC740PRX300LE, 300MHz
    Motorola PowerPC 750 (G3), XPC750ARX233PE, 233MHz
    Motorola PowerPC 750 (G3), XPC750ARX266PE, 266MHz
    Motorola PowerPC 750 (G3), XPC750PRX300PB, 300MHz
    Motorola PowerPC 750 (G3), XPC750PRX300RB, 300MHz
    IBM PowerPC 750 (G3), PPC750--DB0M266, 266MHz
    IBM PowerPC 750 (G3), PPC750--EB5M300, 30MHz
    Codenamed "Arthur", the PowerPC G3 name comes from the third generation of PowerPC microprocessor. It was used in Apple Macintosh computers such as the PowerBook G3, the iMacs, iBooks and several desktops, including the Power Macintosh G3s.
    The G3 was introduced in two different versions, derived from the PPC 603 series of microprocessors: the PPC 740 and PPC 750 microprocessors. The PPC 740 slightly outperformed Pentium IIs while consuming less than 20% of the amount of power and size. Derived from the PPC 740, the PPC 750 had a faster way to access L2 cache, which allowed higher performance.
    The earlier versions, made by Motorola, used an aluminium process for fabrication, and were limited to 400 MHz speeds. Later versions, manufactured by IBM with a "silicon-on-insulator" fabrication process, achieved speeds of 500 MHz and beyond. All G3 versions did not completely implement a standard for symmetric multiprocessing computers, which made design and manufacture of a SMP computer comparatively difficult. The PowerPC G4 corrected this deficiency.
    With its combination of small size and low power requirements, the G3 proved an ideal laptop microprocessor in its era. Apple ceased using the G3 on October 22, 2003.
    References:  Motorola MPC 740 Docs  Motorola MPC 750 Docs  IBM PPC 750 Technical Library

    PowerPC 7400 (G4)
    Motorola PowerPC 7400 (G4), XPC7400 RX400TK, 400MHz
    The fourth generation of PowerPC processors, the PPC G4, is used in Apple Macintosh computers such as the G4 PowerBook, the 2nd generation "Flat Panel" iMac, the eMac, the 3rd generation iBook, and the desktop G4 Power Mac.
    Most of the G4 design was done by Motorola in close cooperation with Apple. IBM, the third member of the AIM alliance, chose not to participate in the design of the G4 in part owing to microprocessor design disagreements concerning a Vector Processing Unit on the chip. Ultimately, the G4 architecture design contained a 128-bit vector processing unit called AltiVec.
    With the AltiVec unit, the G4 microprocessor can do four-way single precision floating point math, or 16-way byte math in a single cycle. Furthermore, the vector processing unit on the G4 is superscalar, and can do two vector operations at the same time. Compared to Intel's x86 microprocessors at the time, this feature offered a substantial performance boost, if the application was coded to take advantage of the AltiVec unit.
    References:  Motorola MPC 7400 Hardware Specs  Motorola MPC 7400 Doc Archive

    PowerPC 860
    Motorola PowerPC 860, XPC860MHZP50C1, 40MHz
    PowerQUICC Integrated Communications Processor
    The MPC860 PowerQUICC is a versatile one-chip integrated microprocessor and peripheral combination that can be used in a variety of controller applications, excelling particularly in communications and networking products.
    There is one PPC 860 available for trade.
    References:  Motorola MPC 860 Fact Sheet  Motorola MPC 860 Docs

  • April - July 2004 updates and changes top of page

    ...::: now has more than 500 chips online :::...

    11 new additions to the collection today:

  • several math coprocessors from ULSI and Weitek

    ULSI Advanced Math Coprocessor
    ULSI Math-Co DX (w/o speed markings)
    ULSI Math-Co US83C87, 25MHz
    ULSI Math-Co US83C87, 25MHz, USA (different markings)
    ULSI Math-Co US83C87, 33MHz
    ULSI Math-Co DX, 33MHz
    ULSI Math-Co DX, 40MHz, USA (different markings)
    The ULSI 83C87 or Math-Co DX is an 80387-compatible coprocessor first introduced in early 1991, well after the IIT 3C87 and Cyrix 83D87 appeared. Like other 387 clones, it is somewhat faster than the Intel 387DX, particularly in its basic arithmetic functions. The transcendental functions, however, show only a slight speed improvement over the Intel 387DX.
    Like other non-Intel 387 compatibles, the 83C87 does not support asynchronous operation of the CPU and the coprocessor. This means that the 83C87 always runs at the full speed of the CPU. It is available in 20, 25, 33, and 40 MHz versions. The ULSI is produced in low power CMOS; power consumption at 20 MHz is max. 800 mW (400 mW typical), at 25 MHz it is max. 1000 mW (500 mW typical), at 33 MHz it is max. 1250 mW (625 mW), and at 40 MHz it is max. 1500 mW (750 mW typical). The 83C87 is packaged in a 68-pin ceramic PGA.
    The Math-Co DX w/o speed markings is a donation by Jorge Segovia. Thanks a lot!

    ULSI US83C87 DX/DLC, 40MHz
    ULSI US83C87 DX/DLC, 40MHz (different markings)
    The DX/DLC is a slightly enhanced version of the original ULSI 83C87 with increased standards compatibility, but still not 100% IEEE conform. Some transcendental functions have also been sped up.

    Weitek Abacus
    Weitek Abacus 3167, 20MHz
    Weitek Abacus 3167, 25MHz
    Weitek Abacus 3167, 33MHz
    The 3167 was introduced in 1989 to provide the fastest floating point performance possible on a 386 based system at that time. The Weitek Abacus is not a real coprocessor, strictly speaking, but rather a memory mapped peripheral device. It was optimized for speed wherever possible. Besides using the faster memory mapped interface to the CPU (the 80x87 uses IO-ports), it does not support many of the features of the 80x87 coprocessors, allowing all of the chip's resources to be concentrated on the fast execution of the basic arithmetic operations.
    The Weitek Abacus provided up to 2.5 times the performance of an Intel 387DX coprocessor, but it's performance would drop to about half the stated rate for double precision, while the value for the Intel 387DX would not change much. Anyhow, before the advent of the Intel RapidCAD, the Weitek 3167 usually beat all 387 compatible coprocessors even for double precision operations. For typical applications the advantage of the Weitek 3167 over the 387 clones is much smaller.
    The Abacus 3167 is packaged in a 121-pin PGA that fits into an EMC socket provided by most 386 based systems. It does not fit into the normal coprocessor socket designed to hold a 387 compatible coprocessor in a 68-pin PGA. To get the best of both worlds, one might want to use a Weitek 3167 and a 387 compatible coprocessor in the same system. These coprocessors can coexist in the same system just fine. Only problem is that most 386 based systems contain only one coprocessor socket, usually of the EMC (extended math coprocessor) type. Thus, you can install either a 387 coprocessor or a Weitek 3167, but not both. There are little daughter boards available though that fit into the EMC socket and provide two sockets, an EMC and a standard coprocessor socket.
    While support for 80x87 coprocessors is very common in application programs, the Weitek Abacus coprocessors do not enjoy such wide spread support. Due to their high price, only a few high-end PCs have been equipped with Weitek coprocessors. Therefore most of the programs that support these coprocessors are also high-end products like AutoCAD and Versacad-386.

    Most of the information provided here is from Norbert Juffa's excellent FPU reference "Everything you always wanted to know about math coprocessors", please refer to 'copro16a.txt' for more information.
    Weitek at Wikipedia

  • 2004-04-02
    17 new additions to the collection today:

  • 80386 Processors

    The Intel 80386, introduced in October 1985, was the successor to the 80286 processor and the first Intel processor with 32-bit data and address busses. It allows multiple application programs to run at the same time (when running under 386-specific operating systems) using Protected Mode. The 386 can address four gigabytes (2^32 bytes) of memory using Enhanced Mode, a 32-bit extension of Protected Mode. As it was in the 80286, segment registers were used to index inside a segment table that described the division of memory. Unlike the 286, however, inside each segment one could use 32-bit offsets, which allowed every application to access the 4GB of memory. However, 16 megabytes was a typical maximum in IBM PCs.
    In addition, Enhanced Mode supported paging, a mechanism which made it possible to use virtual memory.
    The first IBM compatible to use the 386 was the Compaq 386, before IBM used it in high-end models of their PS/2 series.
    With the 386, Intel introduced the 'DX' - 'SX' naming system. DX stands for Double-word eXternal, SX for Single-word eXternal. The SX versions therefore are lower-speed version of the 386(DX), introduced in 1989. They use a 16-bit instead of a 32-bit data bus.
    Some 386DX 16MHz Intel processors had a small bug which appeared as a software problem. The bug involved occurred when running true 32-bit code in a program such as within OS/2 2.x, UNIX/386, or Windows in Enhanced mode. The bug would cause the system to lock up and is a difficult issue to determine without having Intel actually look at the chip. Chips that passed the test, and all subsequent chips which were bug-free, were marked with a ΣΣ symbol. 386DX chips that are not marked with these symbols may have not been tested by Intel and may be defective. Chips having an 'IV' marking also don't have this bug.

    Intel i386 DX
    Intel A80386-16 ΣΣ
    Intel A80386DX-16 ΣΣ
    The 16 MHz versions with original and 'new' DX markings. One A80386-16 is for trade.

    Intel A80386-20 ΣΣ
    Intel A80386DX-20 ΣΣ
    Two 20 MHz versions, one without the DX and one early DX version without i386 logo, which was introduced in 1989.

    Intel A80386DX-25 IV
    Intel A80386DX-25 IV, new i386 DX logo, '85
    25 MHz versions with i386 logo and the even newer i386 DX logo, introduced in late 1991. The first of the two is a donation by Jorge Segovia. Thanks a lot!

    Intel A80386DX-33 IV
    Intel A80386DX-33 IV, new i386 DX logo, '85
    The fastest i386 DX with 33 MHz, one with old, one with new style logo. There are some old style logo versions for trade.

    Intel i386 SX
    Intel NG80386SX-16, w/o Intel logo
    Intel NG80386SX-16
    An early i386 SX, without Intel logo and a newer version. Another donation by Jorge Segovia, thanks again!
    One NG80386SX-16 is also new in my tradelist.

    Intel NG80386SX-25
    Intel NG80386SX33
    The 25 and 33 MHz versions of the i386 SX.

    References:  i386 Manuals and Datasheets  Wikipedia  Intel

    Intel KU80386EX25
    The 80386 EX microprocessor is designed for embedded applications that require high integration and low power. Key features include power management, low-voltage operation, and on-chip integration of numerous common peripherals such as interrupt controllers, chip selects, counters and timers.
    References:  Datasheet  User's Manual

    Intel KU80386SLBA-25 Engineering Sample
    In October 1990, Intel released the 80386SL, which was basically an 855,000 transistor version of the 386SX processor, with cache, bus, and memory controllers, ISA compatibility and power management circuitry. It added a special system management mode (SMM), in which the BIOS could more easily perform power management and other functions without requiring OS support. The 386SL was the first chip specifically made for portable computers.
    This item here is an engineering sample.

    AMD Am386SX/SXL
    In 1991 AMD introduced advanced versions of the 386SX processor - not a second source production of the Intel chip, but a reverse engineered pin compatible version. In fact, it was AMD's entry in the x86 market other than as a second source for Intel. AMD 386SX processors were available at faster clock speeds at the time they were introduced and still cheaper than the Intel 386SX.
    Produced in 0.8 m technology and using a static core their clock speed could be dropped down to 0 MHz, consuming just some mWatts. Power consumption was up to 35% lower than with Intel's design and even lower than the 386SL's, making the AMD 386SX the ideal chip for both desktop and mobile computers. The SXL versions featured advanced power management functions and used even less power.
    AMD Am386SX-20
    AMD Am386SX/SXL-25
    AMD Am386SX/SXL-33
    There is one AMD Am386SX-20 in my tradelist.

    References:  Am386SX/SXL/SXLV Datasheet

  • March 2004 updates and changes top of page

    17 new additions to the collection today:

  • Motorola MC68000 processors

    The MC68000 was the first member of Motorola's very successful family of 16- and 32-bit processors. Introduced in 1979 it was actually a 32 bit architecture internally, but had a 16 bit data bus and 24 bit address bus to fit in a 64 pin package. It was used in a wide variety of computers at that time, including the Commodore Amiga, Apple Macintosh, Atari ST and the original Sun and SGI UNIX machines, but also several game consoles like Sega Genesis/MegaDrive, NeoGeo and many arcade machines were based on the MC68000. It also had a great success as a controller, chosen by many medical manufacturers and printer manufacturers like HP, Printronix and Adobe because of its low cost, convenience, and good stability.
    Variants of the 68000 include the 68HC000 (a low-power HCMOS implementation, 0.13-0.26 W compared to 1.2 W NMOS) and the 68EC000 with selectable 8 or 16 bit data bus.

    Plastic DIP-64 versions (68000P):
    Motorola MC68HC000P8, 8 MHz, low power (HC)
    Motorola MC68000P12, 12.5 MHz
    Motorola MC68HC000P12F, 16.67 MHz, low power (HC)
    Additionally, there is an MC68000P8 for trade.

    Plastic LCC-48 versions (68000FN):
    Motorola MC68000FN8, 8 MHz
    Motorola MC68000FN10, 10 MHz
    Motorola MC68HC000FN12, 12.5 MHz, low power (HC)
    Motorola MC68HC000FN16, 16 MHz, low power (HC)
    Motorola MC68EC000FN16, 16 MHz, selectable data bus width (EC)

    Ceramic PGA-68, gold lead version (68000RC):
    Motorola MC68HC000RC10, 10 MHz, low power (HC)

    Second source versions:
    Rockwell R68000C8E, 8 MHz ceramic DIP-64
    Signetics SCN68000C8I64, 8 MHz ceramic DIP-64
    Signetics SCN68000C8N64, 8 MHz plastic DIP-64
    Thomson TS68000CP8, 8 MHz plastic DIP-64
    Thomson EF68000CM10, 10 MHz ceramic DIP-64
    The Signetics SCN68000C8N64 is a donation by Jorge Segovia. Thanks a lot!

  • Motorola MC68008 processors

    The MC68008 is an 8-bit bus version of the MC68000 and can address 1MB or 4MB of memory (using a 48-pin or 52-pin package). The original 48-pin package version was used in the Sinclair QL, to make maximum use of inexpensive 8-bit technology on the motherboard while computing internally with the relatively fast 16/32-bit processing engine.

    Motorola MC68008P8, 8 MHz plastic DIP-48
    Motorola MC68008L10, 8 MHz side-brazed ceramic DIP-48
    The MC68008P8 is another donation by Jorge Segovia. Thanks very much!

  • Motorola MC68010 processor

    The 68010 is a 68000 with the addition of virtual machine and virtual memory capabilities and a "loop mode" which acts like a 3 word instruction cache. It is largely similar to the Motorola 68000 CPU with the exception of the addition of several instructions for breakpoint and register control, as well as the ability to save all of the processor state on an interrupt. This made it far easier to use for virtual memory applications, for which the 68000 was unsuited. A 68010 will plug into a 68000 socket and work in most systems.

    Motorola MC68010R8, 8 MHz ceramic PGA-64

    References: M68K FAQ  Wikipedia  Great Microprocessors of the Past and Present  Motorola identification at CPU World

  • February 2004 updates and changes top of page

    7 new additions to the collection today:

  • Intel OverDrive Processors

    Intel OverDrive processors were a category of various Intel CPUs that were produced to upgrade computers. ODP is the acronym for Over Drive Processor, however most OverDrives are ODPR (Over Drive Processor Replacement) chips: To upgrade a computer to a new processor you just replace it with the new one. Their interfaces appear to the motherboard like an older chip, so you don't need to worry about whether or not the motherboard supports the higher speeds. ODP chips (like the DX2ODP66 or the ODP486SX-25) in contrast were made for additional sockets on the mainboard, the Performance Upgrade Socket.
    On older versions Intel labeled the OverDrive chips with the external local bus speed they supported, like the ODP486SX-25 and ODPR486DX-33. Since all these OverDrive chips have the clock doubling feature, it was assumed by Intel that everyone knows that the CPU runs at twice the bus speed internally.

    Intel DX2ODPR66, V3.0 Overdrive
    Intel DX2ODPR66, V4.0 Overdrive
    Upgrade 33MHz Intel 486DX and SX based systems to 66MHz.
    There is not much difference between an DX2ODPR66 and a standard 486DX2-66. Intel was marketing their speed doubling technology in different products. DX2 chips were sold to Intel OEMs and OverDrive chips were sold directly to end-users. The DX2 chips typically have no built-in heatsink and have a distinctive DX2 label. The OverDrive chips have a distinctive OverDrive label and may have a built-in heatsink.
    The 'V3.0' and 'V4.0' marks the third and fourth stepping of the processor since introduction.
    References:  Intel

    Intel DX40DPR75 Overdrive
    Upgrades 25MHz Intel 486DX & 16/20/25MHz 486SX based systems to 75MHz.
    Intel DX40DPR100 Overdrive
    Upgrades 33MHz Intel 486DX & SX based systems to 100MHz.
    DX4 OverDrives typically have qualities different from standard 486s of the same speed. These include built-in voltage regulators, different pin-outs, write-back cache instead of write-through cache, built-in heatsinks, and fanless operation.

    Intel PODP5V83 Pentium Overdrive
    Upgrades Intel 486 based systems to Pentium technology at 83MHz.
    The Pentium OverDrive is an Intel Pentium processor for 486 Socket 3 and Socket 2 motherboards. It was provided as a means to give a Pentium performance-level upgrade option for owners of 486 computer systems. It was however criticised for being more expensive and slower than competing CPU-upgrade options such as the AMD Am5x86 and Cyrix 5x86, and being too late to the market. Mainboard compatibility was also a problem, it turned out that many boards didn't support this new chip.
    To perform properly, the Pentium OverDrive was dependent on a high amount of secondary cache ram being present on the motherboard; without it the chip was only a trivial amount faster than an DX4.
    References:  Datasheet  Intel

    Intel PODPMT66X200 Pentium MMX Overdrive
    Upgrades original Pentium 100/133/166MHz to Pentium MMX 200MHz, 90/120/150MHz to Pentium MMX 180MHz and 75MHz to Pentium MMX 150MHz.
    Most older Pentium boards cannot handle the Pentium with MMX because of its requirement for a 2.8V core. The Pentium MMX OverDrive includes a converter that lets it run in Socket 5 and Socket 7 motherboards that do not have 2.8V support. Otherwise the chip is identical to a standard Pentium with MMX.
    References:  Datasheet

    Intel PODP66X333 Pentium II Overdrive
    Upgrades 150 and 180MHz Pentium Pro based systems to Pentium II 300MHz, 166 and 200MHz Pentium Pro based systems to Pentium II 333MHz.
    The Pentium II OverDrive is a processor upgrade designed for Pentium Pro (Socket 8) based systems. By using the 0.25 micron Deschutes core of the Pentium II, Intel managed to put a quick end to the 16-bit weakness of the Pentium Pro, and in doing so, they also managed to double the amount of L1 cache on the processor to 32KB. The PII OverDrive is nothing more than a Pentium II Xeon on a smaller scale, using the Pentium II's design to place the L2 cache externally off of the CPU, but on a card that would allow it to operate at the same clock speed as the CPU itself. The chip has a built in clock multiplier of 5.0x, and it derives its clock speed based on the motherboard's set FSB frequency, so on a 60Mhz board (150/180MHz Pentium Pro systems) the processor operates at 300MHz, on 66MHz systems (166/200MHz Pentium Pro) at 333MHz.
    References:  Intel  Introduction Press Release

  • January 2004 updates and changes top of page

    9 new additions to the collection today:

  • math coprocessors for the i386SX

    Intel N80387SX-16
    Intel N80387SX-20
    Intel N80387SX (16-25 MHz)
    The 387SX is the coprocessor for the Intel 386SX. It has all the features the Intel 80387 offers, including the ability for asynchronous operation of CPU and coprocessor. Due to the 16 bit data path between the CPU and the coprocessor, the 387SX is a bit slower than a 80387 operating at the same frequency. In addition, the 387SX is based on the core of the original 80387, which executes instructions slower than the second generation 387DX. The 387SX comes in a 68-pin PLCC package and is available in 16 MHz and 20 MHz versions.

    Intel N80387SL
    The 80387SL is a coprocessor designed for use in systems based on an Intel 386SL CPU. The 386SL is derived from the 386SX. It is a static CHMOS IV design with very low power requirements that is intended to be used in notebook and laptop computers.

    The 387SL, introduced in early 1992, has been designed to accompany the 386SL in machines with low power consumption and substitute the 387SX for this purpose. It features advanced power saving mechanisms. It is based on the 387DX core, rather than on the older and slower 80387 core (which is used by the 387SX).

    IIT 3C87SX-16
    IIT 3C87SX-20
    IIT 3C87SX-25
    IIT 3C87SX-33
    The 3C87SX is the version of the IIT 3C87 that is intended for use with an Intel or AMD 386SX CPU. It is functionally equivalent to the IIT 3C87. Due to the 16-bit data path between the CPU and the coprocessor in a 386SX based system, coprocessor instructions will execute somewhat slower than on the 3C87. It is the only 387SX coprocessor that is offered at speeds of 16, 20, 25, and 33 MHz. The 3C87SX is packed in a 68-pin PLCC.

    ULSI Math-Co SX, 20MHz
    The Math-Co SX is the SX version of the ULSI 83C87/Math-Co DX for operation with an Intel 387SX or an AMD Am387SX. It is functionally equivalent to the 83C87. To aid low power laptop designs, the ULSI 83S87 features an advanced power saving design with a sleep mode and a standby mode with only minimal power requirements. The ULSI 83S87 is packaged in a 68-pin PLCC.