In coin collecting often times an example is valued not because of its perfection, but because of its imperfections.  An off-center print, the obverse being printed upside down, or the double strike, where a coin doesn’t get cleared form the die and gets hit twice.

Such appears to be the case with this Intel A80386DX-33.  It clearly went through the engraver twice. A similar example (from the same exact lot) is fine, so clearly this one, made in early 1992, was a mistake that was not caught.  I have seen mis-aligned prints, off center etc, but this is the first example i have seen that was engraved twice.  It is interesting that even within the same lot, the spacing of the markings varied somewhat.  Notice that on the right side of the chips the sets of markings line up but they diverge towards the left.  It appears the stepper motors moving the tooling or the chips were a bit sloppy or out of calibration.

Have you seen any other double engraved comments? Let us know in the comments.

NRAO Radio Telescope

There are many greats of the CPU industry, some, such as Federico Faggin (designer of the 4004 and worked on the 8008, then founded Zilog) are fairly well known.  Others include Gelsinger and Meyer (of x86 fame) perhaps even Gordon Moore, of which a  ‘law’ is named.  Chuck Peddle and Bill Mensch designed the ubiquitous 6502 processor, but there were more, many more. Engineers whose names have been oft forgotten, but whose work has not.  The 1970’s and 80’s were the fast and the furious of processor designs.  Some designs were developed, sold, or canceled in weeks, months; years were not a period of time that was available to these designers, for in a year, a new technology would dictate a new design.

One of these designers is Charles H. Moore. (aka Chuck Moore).  Chuck is perhaps best known for inventing the FORTH programming language in 1968, originally to control telescopes.  It was a stack based language, and lended itself well to small microcomputers and microcontrollers.  Some microcontrollers even embedded a FORTH kernel in ROM.  It was also designed to be able to be ported to different architectures easily.  FORTH continues to be used today for a variety of applications.  However Chuck did not just invent a 1970’s programming language.

Read More »

Usually we find vintage processors here at the CPU Shack Museum, however, from time to time, we get our hands on something very new, and usually significant.  If by significant one means the processor from a Top500 supercomputer then yes, it is significant.

IBM 51Y7638 – Produced Early 2012 – Blue Gene/Q 1.6GHz 18 Core PowerPC-A2

This is a Compute card from an IBM Blue Gene/Q (specifically the 6 rack BG/Q running at England’s Science & Technology Facilities Council Daresbury Lab in Cheshire).  A Blue Gene/Q system is made up of these cards, 32 per ‘Node Card’, and 1024 per rack. This doesn’t count the I/O board which use a similar design and contains 8 Compute cards per rack.

BlueGeneQ ASIC die shot

Each of the Compute cards contains a large ASIC (the large chip in the middle).  This ASIC contains 18 PowerPC-A2 processor cores running at 1.6GHz.  16 of them are ‘User’ cores, 1 is for system management (handles interrupts  message passing, etc) and the 18th is a spare, for increased fault tolerance. The ASIC also contains 32MB of shared L2 cache and a dual 1.3GHz memory controller for the 16GB of DDR3 memory on the card.   All said this 45nm chip contains 1.47 Billion transistors, but only dissipates 55Watts, granted, that adds up when you have thousands of them.

A ‘basic’ system contains 4 racks, so 4096 compute cards (4128 if you count the the I/O boards). Together this is 65,536 user cores and consumes upwards of 85kW of power (this actually makes it one of the most efficient super computers available).

So how do these cards become available?  Simply put when you have so many in a system, statistically you are going to have failures, and somewhat frequently.  IBMs target failure rate, based on a 96 rack system (which is massive) is 70 hours.  That’s one failure  every 3 days.  At this point the common reaction is to express shock at the dismal reliability of such a system, however, lets put it another way, that’s one failure out of 98,000+ Compute cards (yes there are other failure points but for the sake of argument we’re using just the compute cards).  If you run an IT department that services nearly 100,000 computers and you only have to fix something twice a week, there is a good chance you should get a raise.

Mask ROM

Users of the Python programming language often say it can do anything, and that may just be true.  Microcontrollers through out their history have had a variety of ways to store the programs they run.  Unlike a microprocessor, a microcontroller typically has a fixed, or somewhat fixed, program that it runs.  This program is often referred to as firmware (its not software, as its cant easily be changed, and its not hardware as it isn’t discrete chips, thus firmware).

There are several common ways to store firmware:

• UV-EPROM: The microcontroller has a UV-EPROM as part of its die (or in some cases separate but on the same package).  This can be programmed using higher voltages, and erased/updated, albeit not in the field.  This was popular in the 80’s for prototyping work.
• Flash (or EEPROM):  This replaced UV-EPROM program storage as it was update-able in circuit, allowing for such things as user BIOS upgrades, updating firmware on CD/Hard Drives etc.  This has become fairly standard for any firmware that is likely to need to be upgraded in the future.
• 

  Intel B2616 – Unless you clean the paint off

  OTP: One Time Programmable Read Only Memory is useful for medium to large scale production runs.  This allows the code to be ‘burnt’ onto a chip prior to shipping. Often all these were were UV-EPROM chips in a plastic package.  Early Intel’s even used UV-EPROM chips, and simply painted over the window.  a 2708 UV-EPROM became a 2608 PROM with the simple application of some nail polish.  There has been some experimentation and success in erasing/reusing these with the use of X-Rays. (they can penetrate the plastic package).

• Mask ROM: A Mask ROM is just as it sounds, the program code is actually added to the actual mask itself when making the microcontroller die.  This is the most economical  and reliable for very large production runs.  Obviously one wants to make very sure the code is correct before cutting a mask with it, masks are expensive, and manufacturers are not keen on giving do overs.   In 1978 Intel charged $1000 Mask fee, and a minimum order of 100 units for an 8k ROM (~$10).   By 1986 that Mask fee had risen to $3000 and min units to 1000.

So what happens when 20+ years later you need to figure out whats ON a mask ROM?  The paper tape, 8″ floppy or punch card the program original was stored on is long since gone.  Being that its a mask ROM one can actually SEE the connections, so its possible to decap a device, and visually determine the code, albeit with a lot of tedious work.  Adam Laurie of Aperture Labs developed a Python script to automate some of it, and wrote an article explaining it, which covers some every interesting Mask ROM info.  Not to mention some very nice pictures, so check it out.

Click for much larger

Every now and then I will get a chip in that I cannot ID.  This is a particularly perplexing one.  It looks like it should be something fairly well known, but I cannot determine what.  By the dates its a 2005 vintage IBM, MCM, on a fairly large ceramic package with 1077 lands.  It contains a pair of Infineon HYB39S256160DT-7 256Mbit (4Mbitx16bit) DRAMs which are 7ns 143MHz max, commonly used on PC133 SDRAM.   That works out to 64MB.  Also on the package is a IBM0436A8ACLAB 8Mbit (256Kx36) 4.5ns (222MHz) 1Mbyte SRAM.

IBM MCM die

Markings on the die are:
0FE45000L3
AKESXEX0
1 10-10
09K2262

If you have any ideas what it is, or what it may be, post a comment.  I may just give you one.  These came in with a lot of HP PA-RISC processors, so perhaps related?

UPDATE (10/20/2016): Mystery solved. These are processors from a Cadence Palladium emulator system. Read more about them here

PIC17C766/CL ES

If there is anything a Microchip PIC is known for typically ‘large’ is not what comes to mind.  PICs were originally developed in the 1970s as a peripheral controller and ended up finding uses in products of every sort, for 35 years and counting.  The PIC17 series extended the original 12 bit architecture to 16 bits (16 bit instructions, ALU and registers are still 8-bit).  It added many new instructions (58 total) and an 8×8 hardware multiplier. Max clock speed was 33MHz. It was considered the ‘high end’ of the PIC line but now has been replaced by the PIC18 line.  Most of the PIC17s produced were in the 40-68 pin range. Many designers considered the PIC17 to be a less then great processor and in 2000 Microchip replaced it with the much better PIC18 line.

This PIC17C766/CL was one of only 7 variants in the line (17C42,43,44, 752,756,762 and 766) compared to the many dozens in the PIC16 or 18 lines.  Produced in an 84 pin CLCC with UV EPROM window (for its 16k EPROM) the 17C766 provided 66 I/O lines, more then enough for any project.  It was used in some PIC development systems and emulators which were some of the few systems that really needed a PIC with 84 pins.  This particular example also happens to be an Engineering Sample and was produced in mid-2001, AFTER the introduction of the PIC18, it seems there was still at least some demand and use for the PIC17 a decade after its introduction.

Intel 486 With new logo – 2007

Intel introduced the 486 in 1989, 24 years ago so it may be a bit surprising to see a 486 with the more modern Intel non-‘dropped e’ logo.  Intel began using the current logo in 2006, well after the height of the 486 market.  However Intel continued to make several 486 processors clear up through 2007.  These went to supply many embedded applications (medical, industrial, etc) that had originally been designed with a 486 and remained in production.  These types of devices either did not need a better processor or due to regulatory reasons, could not use one.  This is actually a common issue with medical devices, once designed, they are certified by the FDA (in the US, other agencies in the EU, Canada etc).  This certification is very specific to that exact hardware.  Often you cannot even change the S-spec/revision of the processor.  If it was built/certified with an SX911 A80486DX2-66 that’s what it must continue to use.  Re-certification is very expensive and time consuming.  It may seem overkill but these type of applications are often life sustaining, a failure could mean more then a reboot and lost game of Doom.  This is why Intel (and other companies) will choose a few products out of each line to be long-term production, this helps engineers select a product to design into their device, that has a guaranteed production life.  Currently Intel’s long term products guarantee a 7-year production.

Original DX2 Late 1994

This particular chip was made in April 2007. Intel announced it would be discontinued by the end of 2007 so this is truly one of the last (Intel) 486s ever made.  It is a SX911 &E5V2X version which is an SL Enhance (&E) 5 Volt Mobile (2X) processor.  The 2X means that the processor requires 2 external clocks, typically one would be used for when a laptop was plugged in (full speed, 33MHz in this case) and a second, which could be anything lower, for battery operation.  Obviously these can be implemented in any application, not just a laptop.  Registers contents, interrupts  etc are all preserved when the processor switches clock frequency so this is done on the fly, just like today’s processors.  The SX911 spec, which first was released in 1994, came in both 1X (desktop) and 2X (mobile) versions.  Throughout its 13 years of production it remained on a 150mm (6-inch) wafer process at 0.8u.

Cypress CY7C601-25GC – First package with heatspreader – Omitted on later versions

In Mid-1987 Sun Microsystems (now owned by Oracle) released the SPARC (Scalable Processor ARChitecture)  processor architecture to be used in their computers (replacing the 68k based systems they had previously used).  The SPARC was designed from the outset to be an open architecture, allowing manufactures to license and built processors that implemented it using whatever technology they wished.  The goal of this was to 1) build a large SPARC ecosystem and 2) keep prices in check by fostering competition among manufacturers.  The SPARC is still used today by Oracle, Fujitsu, the European Space Agency and others, owing largely to its design as an open architecture from the very beginning.

The first version was made by Fujitsu on a 20,000 gate array at 1.2 micron and ran at 16.6MHz.  In July 1988 Cypress  (later to be spun off as Ross and make the famous HyperSPARC line) announced the CY7C601.  This was the fastest implementation of the SPARC at the time.  It was made on 0.8u CMOS process and contained 165,000 transistors, dissipating around 3.3Watts.  As was typical of many processor designs of the time, it was an integer only processor, requiring a separate chip (the CY7C602) for floating point work.  In September of 1988, Cypress cross licensed the ‘601 to Texas Intruments in exchange for rights to the 8847 floating point processor.  This was mainly to appease one of Cypress main customers who demanded that a second source for the ‘601 chips be available, a demand more common in the 1970s then in 1988 but Cypress obliged.  Cyrpress also gained the rights to make the next generation SPARC processor that TI was developing.  TI would go on to make many SPARC processors, and continued to be the primary fab for Sun up through the SPARC T2 Plus in 2008.  Oracle now used TSMC to fab the T3 and T4 SPARC processors.

Original Intel 6MHz 80186 Made in 1984

2012 marked the 30th anniversary of the introduction of the Intel 80186 and 80188 microprocessors.  These were the first, and arguably only, x86 processors designed from the beginning as embedded processors.  It included many on-chip peripherals such as a DMA channels, timers and other features previously handled by external chips.  Initially released at 6MHz, clock for clock many instructions were faster then the 8086 it was based on, due to hardware improvements.

In 1987 Intel move the 186 to a CMOS process and added more enhancement including math co-processor support, power down modes and a DRAM refresh controller.  Speeds were increased up to 25MHz (from the 10MHz max of the NMOS version).  Through the years Intel continued to developed new versions of the 186 with added features, lower voltages, and different packages.  It was not until 2007 that Intel finally stopped production of the 186 series.  It continued to be made by others under license including AMD, who made versions running up to 50MHz.  Fujitsu and Siemens also produced the 186 series. Like the 8051 the 186 gained significant support, being embedded in millions of devices.  The instruction set was familiar, debugging and development systems were (and are) plentiful so the 186 core continues to be in wide use.

As IC complexity and transistor counts increased the need for a processor core that could not just be embedded into a system, but be embedded into a custom ASIC or SoC became apparent. IC’s were being designed to handle things like DVD playback, set-top boxes, flat panel control and more.  These applications still required some sort of processor to handle them but having to have a separate IC for it was not economical.

pixelworks PW166B – 67MHz Tubro186 based Flatpanel Controller made in 2004

VAutomation (founded in 1994) designed Verilog and VHDL synthesizable cores (meaning they could be ‘dropped’ into an IC design or FPGA).  In November 1996 VAutomation licensed the 8086/8, 80186/8 and the CMOS versions from Intel.   This gave them them ability to design their own compatible models of these processors without fear of litigation.  More importantly it allowed them to sub-license these designs to others.  In 1997 VAutomation demo’d their first 186, the V186 core.  This was a Intel 80186 compatible core that could be synthesized into a customers design.  It was ‘technology independent  which means it was not restricted to a certain process or even technology.  It could be used in CMOS, ECL, 0.35u, 1 micron, whatever the client needed.  On a 0.35u CMOS process it was capable of speeds in excess of 60MHz, and did so with less then 28,000 gates. One of the first licensees was Pixelworks, which made controllers for monitors.  Typical licensing was a $25,000 fee up front and royalties on a per device basis usually split into a high volume (over 500,000 units) and low volume.  Typical price per chip was $0.25-$2.00, which was cheaper then the $15 price Intel was charging for a discrete 80C186.

Read More »

A very interesting post by Jack Ganssle over at Embedded.com about his adventures and lessons learned from a trip to the Bahamas and a chat with the engineer charged with maintaining an island Intel 8085 based telecom system.  An apt discussion especially with the CES (Consumer Electronics Show) going on, an event showcasing the latest and greatest of what mostly is ‘disposable’ technology.

Many devices, however, must still be built for areas, industries, and regions where replacement is simply not a viable option, whether its because of remoteness or lack of parts.  In the early days of computing surface mount devices were few and far between.  The DIP rained supreme, and the end user was often very good at debugging with simple tools.  Sockets made repairing boards much easier and even discrete resistors could be changed out without much trouble.  Upgrading your computer often involved replacing the processor with a faster one, then changing the clock crystal to support it, and switching out individual RAM chips to ones with a faster access time to support the now faster processor.  You then used the left over parts to either repair another system, or build another project.

From the article:

“Here in Staniel Cay there’s no FedEx; no UPS; no postal system. The mail boat arrives once every two weeks, more or less, if the weather is good. Occasionally it sinks.”

JITS (Just In Time Supply) is simply not an option there. And its not just the Bahamas where repair is the only option. It’s not limited to non hi tech areas.  When something breaks in space, perhaps the most hi tech environment of all, you can’t go to town and get new parts, you have to make do, that’s real engineering.  Apollo 13 showed how things could be repaired with very unusual articles, and similar ‘fixes’ are common on the International Space Station.

Give it a read, its as entertaining as it is educational.