December 19th, 2017 ~ by admin

Chip of the Day: TRW MPY-16AJ – Making Multiplication Manageable

TRW MPY16AJ – 1978

In Primary School students are tasked with memorizing their multiplication tables.  Taking the time to manually calculate 6×5 is much slower than simply committing the result to memory.  This allows more complex math to be processed quicker as the students skills develop.  Typically this is limited to numbers up to 12×12, resulting in 144 results to ‘store.’  In computing the same can be done.  A ROM can be used as a lookup table for multiplication.  The problem is it does not scale well.  Handling 4×4-bit multiplication requires a 256×8 ROM (2m+n addresses and m+n outputs). This could be handled by a many ROMs available in the 1970’s.  Anything more than 4-bits though was simply not possible.  This gave rise to the need for multipliers to calculate the result.

TRW MPY16AJ – Large Heatsink affixed to package to dissipate its 5W

This was a problem TRW set out to rectify in 1976.  TRW LSI Products was formed in the 1960’s to commercialize the transistor products that had been developed by Pacific Semiconductors, a division of TRW.  It was James Buie who invented the TTL logic gate in 1961 while working for TRW.  TTL went on to become the logic standard throughout the 1970’s and 80’s.   TRW was involved in aerospace, helping design planes, satellites, and missiles, fields that required processing of signals data, what became known as Digital Signal Processing (DSP).  In the 1980’s processors were designed to handle this, such as the TI TMS320 series, but in the 1970’s it had to be done with discrete components.  DSP systems had several needed blocks, Fast ADCs, ALUs, and multipliers.  TRW invented fast ADCs to handle the inputs, and ALUs were available such as AMDs Am2901 or even the TTL series 74181s.  Multipliers however were not widely available, especially for large bit-widths.

MPY-16 die

TRW’s first multiplier was a custom device to work with their own avionics processing system.  It was made on a Bipolar process, and multiplexed the entire product, using around 40 pins total (the entire product was multiplexed with the operands).  It could handle a multiply in 330ns worst case.  Interestingly yields of the device were considered ‘excellent’ at 3 working devices per wafer (out of 19 per wafer (most likely a 2″ wafer)).  Today, yields like that would be completely unacceptable.

TRW designed the MPY-16AJ as a brute-force 16×16 multiplier.  It was designed on a Bipolar process with around 3600 gates.  It implements a series of AND gates and CARRY-SAVE-ADDERS to implement the multiplication.  There are faster methods, but they come at the cost of complexity and power draw.  As designed the the MPY-16AJ dissipates 5 Watts while handling a signed (2’s complement) multiplication in a worse case 230ns).  They MPY16 was packaged in a large 64-pin package to limit the # of pins that had to be multiplexed.  The lower 16-bits of the product are multiplexed with one of the operands.  This is acceptable as in many applications the upper 16-bits of the product are sufficient accuracy.  The 64-pin package allowed for not less multiplexing, but also a much larger surface for heat dissipation.  A heatsink was also affixed to the package as well.

Micron (Russia) 1802VR5 – MPY16HJ Clone made in 1992

Later versions of the MPY-16 added support for unsigned multiplication as well (the MPY16H) and became the standard for 16-bit multipliers.  Compatible multipliers were made by Analog Devices (ADSP1016, 40-50ns at 150mW) and LOGIC LMU16/216) in CMOS, by Weitek (WTL1516/A/B, 50-100ns at 0.9-1.8W) in NMOS, by Synertek (SY66016 100ns at 1.5W) in HMOS, by AMD (Am29516 38ns at 4W) in ECL, as well as many others.  These were implemented internally with different processes, and different multiplier algorithms but externally they all mimicked the standard TRW MPY16J and served as the basis of many signal processing and high end math computers.  As a testament to their usefulness, the MPY16 was also copied by the USSR as the 1802VR5.  The TRW MPY16 was last made in the mid-1980’s but its clones continued to be made into the 1990’s.  Today its functions can be handled by any DSP, CPU or even coded into a FPGA, but for a time, the MPY16 multiplied the efficiency of many processing systems.

March 17th, 2011 ~ by admin

Chips of the day: TI TMS320E The tale of the two dies

Its fairly common for a manufacturer to make several devices out of a single actual die.  Just disable part of the die, whether because its faulty, or not needed, or simply do not connect the pins to that feature.  Intel did this a lot with the Celeron, and PIII line, disable some L2 cache on a PIII and you get a Celeron.  Today it is done with multi-core processors.

TI TMS320E17JDL

Using a common wafer for several products saves a large amount of money, no need for a second mask set, and testing systems.  Here we have a Texas Instruments TMS320E17JDL.  The TMS320 is the industry standard in DSPs (Digital Signal Processors). The E17  from 1990 runs at 20.5MHz has a 4K EPROM, 256 bytes of RAM, and a pair of serial ports.  You can see the large sections of the die devoted to the ROM, RAM, and MAC (Multiply and Accumulate).

TI TMS320E15JDL

This is the TI TMS320E15JDL.  It is the same basic DSP core as the E17, it includes the same 4K EPROM, the same 256 bytes of RAM and the same MAC unit.  It has some I/O ports tasked with doing different things, but thats a relatively minor difference.  The big difference is the E15 lacks the 2 serial ports of the E17.  You can see on the die where that hardware does not exist, its a large black spot, void of any circuitry.  A very interesting and unusual occurrence.

TI either used a completely different mask for the E15, or they simply chose to not expose that small part of the mask.

Posted in:
CPU of the Day

November 16th, 2010 ~ by admin

The History of the SMS300 and Signetics 8X300 Processors

On November 20th, 1969 a small company was formed in Mountain View, CA called Scientific Micro Systems Inc (SMS). They would join the dozens of post-Fairchild semiconductor startups in Silicon Valley.  Many of these we remember and know well, Intel, AMD, Zilog, MOS all are familiar and have designed processors that left a story, if not a legacy, in history. SMS has became a forgotten player in the roaring 70’s but they did introduce a few important things to the market. First 4096-bit bipolar Schottky ROM? SMS. First 256-bit bipolar Schottky RAM? Again SMS.

Signetics N8X300I – Early 1978

In January 1975 SMS announced prototypes of their own 8-bit microcontroller.  The SMS300 was a non-traditional design.  It focused on manipulation of signals.  It had 16-bit instructions, but operated on data 8 bits at a time.  It had very limited ways of accessing external memory (and no real way to access data memory).  It was designed as perhaps the first DSP.  It was fabricated in bipolar Schottky transistor technology.  This allowed it to be incredibly fast (albeit very power hungry) for its time.  Initial clock speeds were 6.66MHz and quickly ramped to 8MHz in 1976.  The SMS300 was initially not available for sale as a single chip.  It was sold as a single board computer called the SMS330 (as well as the SMS331 and SMS332) which contained everything needed to run the SMS300.  This was packaged like a oversized 64pin DIP (similar to how some of the BASIC Stamp microntrollers are today). These systems started at $370 and topped out at $1460.  In August of 1975 SMS ‘unbundled’ the SMS300 and began selling it (and its support chips) separately to those who wanted them.

Soviet Electronika KM1818VM01A 8X300 Clone

Soviet Electronika KM1818VM01A 8X300 Clone

SMS did not make the SMS300 themselves, they contracted another Silicon Valley company to fabricate them.  Signetics, founded in 1961 by ex-Fairchild workers, was the first company founded to solely manufacture ICs rather then discrete transistors.  In 1975 Signetics was purchased by Philips but continued to operate under the Signetics trademark until 1993. Signetics made all of the SMS300 devices for SMS, and in 1976 Signetics became a second-source, and could sell the SMS300 under the Signetics brand.  By 1978 Signetics had purchased the rights to the SMS300 and renamed it the 8X300.  1977 or 1978 is generally when people think the 8X300 was develped. This is, unfortunately, due to forgotten history as by the time Signetcs bought the design, it had been on the market and in use for over 2 years.  Signetics continued to make the 8X300 into the early 1990s where it found wide use in disk controllers, telecommunications and other DSP like environments.  The N8X300 was also second sourced by AMD though I have yet to see one.  The 8X300 was also *second sourced* by the Soviets in the 80’s and early 90’s by the Electronika state electronics company in what is now Voronezh, Russia.  These of course were not licensed copies but they are however, still of interest.

Signetics N8X305N Early 1988

In around 1982 Signetics released the N8X305, the successor to the 8X300.  It was functionally compatible but increased the general purpose registers to 13 from 8 among some other instruction improvements that greatly improved upon the data handling deficiencies of the 8X300. Processor speed was also boosted to 10MHz. (200ns instruction time vs 250ns for the 8X300) The N8X305 also used the same pin-out as the 8X300 and the same 50pin DIP package (as well as adding a 68pin PLCC).  AMD second-sourced the 305 as the AM29X305.  The N8X305 continued to be made into the 1990s and saw use in many military applications as well.  Because of this Signetics/Philips could not simply terminate production.  They sold production rights to Lansdale Semiconductor Inc., who still offers the N8X305 to this day.

Signetics N8X401I – 1988

In 1986 Signetics again revised the design and released the N8X401.  This processor now had a full 32 instruction (including a RETURN instruction allowing the use of subroutines) The N8X401 also added an 8-bit data bus making data handling somewhat simpler but also increasing the package size to a 64 pin DIP.  Internal usable registers was increased to 16 and the instruction width was increased to 20-bits. Clock speed remained at 10MHz but it was now fabricated in ECL (Emitter Coupled Logic).  This among other improvements, helped to result in a 35% speed boost over the N8X305.

AMD AM29X305DC – N8X305 Second Source – 1985

The N8X401 was not a great success, primarily because of the competition at the time.  By 1986 8-bits, even for a DSP, was rapidly becoming out of date, especially with such purpose built DSPs such as the Motorola DSP56000 and the industry standard setting TI TMS320 series.

Thus is the story of one of the most odd processors to come out of the chip boom of the 1970’s (EA 9002, MMI 6701, AMI S2000 et al).  There are several other forgotten processors of that era which are also deserving of some remembrance.  In time we’ll try to document their history here at the CPU Shack as well.

Posted in:
Research