Micron Technology Defied the Odds

Show video

Nobody believed in Micron Technology. A startup starting ten years  late, dealing in commodity chips. Coming out of a city more known  for potatoes than silicon: Boise. Facing down the world's most  competitive semiconductor industry.

Somehow, they beat those odds. In this video,   we look at the memory underdog  from Idaho: Micron Technology. ## Beginnings I have never mentioned Boise before  on this channel - it is not exactly   near Asia - so let me introduce it to you. Boise the city began in 1835 as  a trading post called Fort Boise   set up by the Hudson Bay Company.  The name's origin is uncertain. It grew into prominence during the American  West's various gold rushes of the 1840s   and 1860s supplying materials to miners. In 1864,   it became the capital of the Idaho territory.  Idaho then became an American state in 1890.

Idaho is particularly known  for its potatoes. People   were farming potatoes there as early as 1837. But the industry really sprouted in the  1950s, after the newly-established Idaho   Potato Commission began marking their  potatoes as being "Grown in Idaho". The harsh lands of Idaho gave rise to some  of America's most extraordinary success   stories like the billionaire Jack R. Simplot,  founder of the massive JR Simplot Company.

Simplot started hustling for business at the age   of 14 and slowly built a processed food  business empire. That includes supplying   half of McDonald's French fries - or  chips if you want to call it that. These Boise lands will also give rise to  another extraordinary success story: Micron.

## The Parkinsons Identical twins Ward and Joe Parkinson grew  up in a small town in Idaho called Blackfoot. Their father was a hard-working  man who ran a 1-man insurance   agency. The Parkinson family emphasized education,   so they studied hard while working in potato  fields owned by the JR Simplot company. Both boys got good grades, won school  debates and science fairs so they were   accepted to Columbia. Joe Parkinson - the  more eloquent of the two - went into law,   and eventually returned to  Boise to open his own law firm. Ward, on the other hand, took a different  path. He dropped out of the Columbia pre-med  

program and did an EE degree at Utah State  while also working as a farmer. After that,   he joined the famed Fairchild Semiconductor  where he designed ECL logic and memory circuits. Fairchild was then going down the drain, but  Ward distinguished himself. So much so that   they offered to pay for his Masters  in computer architecture at Stanford.

After he graduated, he was poached by  his old boss at Fairchild to work at   a company called Reticon, producing  image sensors. Two years after that,   he jumped again to Motorola  where he designed memories. It was the late 1970s and it was becoming  clear to everyone in the industry that Metal   Oxide Semiconductors or MOS was the future.  Ward wanted to get more experience in that,   so he jumped to the leading  MOS memories company - Mostek.

## Mostek Mostek was founded in 1969 by a few  Texas Instruments folks led by LJ Sevin. Sevin and his co-founders were frustrated  with their managers' conservative insistence   on bipolar transistors over the high-potential  MOS as a challenger to magnetic core memories. The new Dallas company introduced a plethora  of memory breakthroughs like ion implantation   and address multiplexing. The latter  let Mostek make more complex Dynamic   RAM modules without adding more pins and  thus taking up valuable space on the board. Mostek quickly snagged the lead in world  Dynamic RAM market share. Particularly,  

a brilliant 4K product called the  MK4027, first introduced in 1976.   And the strength of their design was  what convinced Ward Parkinson to join. There, he joined a design team he greatly  respected. This included his manager,   Dr. Paul Schroeder, who helped design  the aforementioned MK4027. Dennis Wilson   and Doug Pitman in particular were  similarly great memory designers.

## Inmos & Micron But unexpectedly in 1976,  Mostek suffered a huge shock. One of its own cofounders Richard Petritz left to   start a competing semiconductor company  backed with British money called Inmos. Petritz hired over the aforementioned  Paul Schroeder by giving him the deputy   managing director title. Schroeder in turn  convinced his colleagues Parkinson, Wilson,   Pitman, and a fourth guy Dave Wooten  to leave Mostek and join them at Inmos. They were not the first to leave, but for  Mostek it was the last straw. Infuriated,  

LJ Sevin filed a lawsuit against Inmos, hoping  to get a permanent injunction against poaching. Doug Pitman and Ward Parkinson were personally  included on this suit. It bombastically accused   them of everything from stealing trade secrets to  violating antitrust law by hoarding design talent. Inmos later handily won the suit, but did  little to help the designers targeted by it.   Probably should have, considering  what happened next. In October 1978,   the three designers ditched Inmos to start their  own design consulting company, Micron Technology.

The new Micron team called LJ Sevin,  and this might sound strange but Sevin   quickly contracted them to design the next  generation of memory chips for Mostek. But   since the price was a bit low, Micron asked for  the option to manufacture any design they made.   Micron was then just three dudes without  even an office, so Sevin gave it to them. They worked out of Dallas for the first year,  but soon Pitman, who was from Idaho Falls,   wanted to go home to the Northwest and raise his  four kids. After looking at Montana and Utah,   Boise was chosen because it had an airport and  Joe Parkinson had many business contacts there. Whilst on a scouting trip, Pitman  scoured Boise for new space and found   a half-finished dentist's basement  for half the normal rental price.  

A donation to the Catholic Church  next door gave them free parking. ## Japanese Memory Despite receiving the option from  LJ Sevin, Micron never intended   to go into manufacturing - though  they sometimes fantasized about it. Then the market changed. In the  late 1970s, Japanese competition in   semiconductor memories intensified. Intel  shipped the first 1K Dynamic RAM in 1970. In the next generation, 4K, they and the  Americans took 90% market share while the   Japanese grappled with new innovations like  the aforementioned address multiplexing. Note that memory incremented  four times each generation.

But in the generation after that 16K, the  Japanese began gathering substantial market   share - thanks in part to memory module  pin standardizations that Mostek pioneered. Then in 1979, Hitachi, NEC and Fujitsu  shipped the 64K DRAM at the same time   as the Americans. And thanks to their  lower prices - stemming from superior   processes and government subsidies  - they took the market share lead. Mostek and the other American memory makers  began to experience declining, volatile profits.

It allowed the acquisitive United Technologies  - a conglomerate that made helicopters,   elevators and other equipment -  to buy them out in August 1979.   They promptly canceled the design  contract with Micron Technology. ## Hell No Micron now had no customers  and no source of revenue. But they still had the 64K DRAM  design, which they completed,   and a few others. So in what is perhaps a  fit of mania, the Micron team decided to   take matters into their own hands and go  make these chips right in Boise, Idaho.

It was legitimately a bit crazy. Leslie  Gill was Micron's first CTO. When Joe   Parkinson told her that they were going to  make chips, she thought "Potato or corn?" No surprise that traditional venture  capitalists refused to invest. Some   asked if they had to close the factory  during the harvest. Ward Parkinson   famously recalled that people didn't  just tell them "No", they said "Hell no". ## Yanke and Noble As fund-raising went on, Joe Parkinson  became increasingly important to the company.

He eventually joined Micron Technology  as its President and fourth co-founder,   though he still practiced  at his law firm for a while. It is through Joe that the Micron team met  Ron Yanke. Yanke took over his father's   machine shop and turned it into a success.  He parlayed that into successful mining,   timber, real estate, and aviation investments.

Yanke had a business partner - a potato  farmer named Allen Noble. Noble lived   one of those classic American success  stories. He ran away from home at the   age of 14, bought a few acres for farming  potatoes, and turned that into an empire. His flagship project was pulling water up 500  feet from the Snake river into the Dry Lake area,   transforming what before had been  a dry desert into a potato bonanza. A few months earlier, Noble and Yanke asked  Ward through Joe Parkinson to troubleshoot   an electronic controller for a broken  irrigation robot they were working on.

Ward warned them that this wasn't his  specialty but nevertheless worked for   several weeks on the side to fix it.  Noble grew to like the guy and later   said that he would have backed them  on anything they were going to do. So when the Micron team came calling,  he knew he would be in. Over lunch,   Noble and Yanke agreed to kick in  some seed money to get things rolling. ## The Odds Neither Noble or Yanke knew  semiconductors. Ward could not   take their money in good faith without  telling them what they were up against.

So he arranged for them to go to Texas to meet  with LJ Sevin. The August 1979 meeting was like   one of those "scared straight" therapies they do  for 12 year olds but for semiconductor investors.   The charismatic Sevin, himself a former  designer, laid out Micron's challenges. Micron was a design consultancy with  little experience in manufacturing.  

They were entering the market in 1979 - some ten   years behind established players  like Mostek, Intel, AMD or IBM. Sevin estimated it would cost about $100  million in startup money to get into the memory   market. And that market is extremely, sickeningly  volatile, estimated to lose $300 million in 1981. The Japanese were then going nuts  and pouring what was said to be   billions of state-backed dollars into  the industry. They were already seeding  

next-gen 256K and everyone expected  them to steamroll the Americans. Noble then asked Sevin, "Well  if the market is so bad,   why was Mostek still in it?  What hope did they have?" To this blunt question, Sevin gave a sad smile  and replied with a generic answer about the   importance of being nimble before adding  at the end, "But sometimes I wonder ..." On the way back to the airport, Ward asked  Noble and Yanke what they thought of what   they just learned - fully expecting the  worst. Noble reportedly said, "Let's do it". Yanke and Noble helped bring in a third investor,  Tom Nicholson, who owned one of the biggest sheep   ranches in the country. Nicholson provided the  initial land and water rights for the factory. For all that, they were still short. The Micron  team estimated that they needed $6 million in  

new loans and investments. Right then they had  a million, and another million in bank loans. Noble and Yanke were rich. But not  rich enough to single-handedly fund   a semiconductor fab. However, Noble  knew someone who was and that was   how Micron got a meeting with JR  Simplot, the potato billionaire.

"What'll take it to get her  rolling", he asked. "$4 million",   they said. Simplot later said that he saw  that these farm boys - referring to Yanke,   Noble, and the Parkinsons - had gone all in  on this venture. And for him, that was enough. "Alright, we'll go for it", he said. Simplot,  

his son Scott, and his companies  eventually bought 40% of Micron,   bankrolling chip R&D and factory construction  much to the consternation of his finance team. ## Building the Factory Micron's corporate strategy was  simple - a reliable product at   a low cost. They wanted to be the low-cost leader. Their first 64K chips were built in a  subcontracted National Semiconductor   fab in Utah while the Micron fab was  being built. That was completed in 1981   at the cost of $20 million, or about  $76 million today. Plenty of money,   but just a fifth of what a DRAM  factory cost even in those days. How?

First, land, labor, water, and energy  were far cheaper in Idaho than they   were in Silicon Valley and even Japan.  Hydropower in Idaho is quite common,   and rates were a third of those in California. But that is not enough, so we should  talk about equipment. One reason why a   TSMC fab is so expensive is because  they have a lot of expensive tools   inside. Micron generally does not use the  most advanced tools. For instance their   lithography tools were older models  from Perkin-Elmer, ASML, or Canon.

In fact, whenever new technology came  out, Micron would not adopt it to move   on to the next-generation density memory  like others did. Rather, they preferred   to use it to make a far cheaper version of  the current generation out in the market. Next, redundancy. Another reason why  a big TSMC fab is expensive is because   it has a lot of tools for redundancy.  So if one breaks down, the line does   not stop because you always can route the  work-in-progress to another, working machine.

When Micron was building their first factory,  they gave up on that option. As CFO Gill said: > "Long term, you may need to  back up all your equipment,   but to make the first chips, you don't.  You have one machine and hope it works" ## Design Cleverness Last and perhaps most importantly,   design and manufacturing were lock-step in  sync about being the low-cost leader. That   means pulling out all the stops on the design  side to achieve the lowest cost memory bit.

They were especially good with layouts.  Doug Pittman was a layout guy. If you   recall in the semiconductor design  process, layout is where you wire   together the various memory cells after  they have been chosen for the design. Companies doing logic chips usually do the layouts  with an electronic design automation software,   which does it relatively inefficiently  but cheaply. And that's fine for logic,   where there is relatively  available space on the silicon. But memory is different. A smart layout  can let Micron cram more memory cells  

into the same piece of real estate.  High density can mean a small die,   which in turn means more dies on a single wafer.  Which in turn means a cheaper per-die cost. Members of the product and process engineering  team also pitched in. For instance, by reducing   the number of masks. Masks represent a layer  in the final design, so the more you have,   the more complex the design is, and  thus the more expensive it is to make. Low mask counts - sometimes 30-40% lower - were  an underrated factor for Micron’s cheaper costs.  

The company regularly cites mask reductions in its  shareholder letters as an indication of cost cuts. Micron never had a shortage of cleverness. Early  on, Ward figured out an alternate use of their   defective DRAM chips - turning them into "optic  RAM" image sensors for early computer cameras. It sold quite well in the market for  several years as the MicronEye accessory.  

Though being sold in the early 1980s,  it was probably far ahead of its time. After a dodgy few months, the first  MT4264 DRAM chips started coming off   the line in late October 1981. They  let the first inventors - Simplot,   Yanke, Noble and others - look at them through a  microscope. JR remembers it being a wonderful day. ## Entering the Market Micron's chips first entered the market in  1982. That year, they sold a million chips. Customers noted how small the chips were:   33,000 square mils - 1 mil means 0.001  inch. This was 40% smaller than Motorola's,   15% smaller than Hitachi's. It was a  size that even TI struggled to achieve.

They were an ideal fit for the growing PC  market. Apple bought them for their Macintosh   and Lisa PCs. Jack Tramiel personally guaranteed  millions in sales from his company, Commodore. Despite this and improving yields over 30%,  the company was still losing money on each chip   sold. It became clear that something drastic was  needed. Fortunately, designers had deliberately  

made the chip bigger than it needed to be  in case of manufacturing process variations. But Micron's second fab had better  equipment and produced larger 5-inch   wafers. So the designers removed those  tolerances, shrinking the die to a mere   22,000 square mils. This and the bigger  wafers right-sized the chip economics.

Released in January 1983, the  new smaller 64K MT4264 chip   was an immediate hit. Sales surged from  half a million in 1982 to 10.9 million,   net losses narrowed from $6.8 million  to just $2.6 million. The backlog   reached $23 million so Micron immediately  started hiring and expanding the factory. Thanks to continued good DRAM sales  from the growing PC revolution,   Micron's sales the next year  1984 surged to $84 million,   producing $29 million in profits which  they funneled into 256K development.

That year, the company also went public, selling  2.1 million shares on the market. The stock   were offered at $14 and went to $40. JR  Simplot and his family and companies still   held almost 19% of Micron, and made  $150 million on his early investment. JR was such a believer that he told  his employees to buy the stock,   even on margin. And that he would make whole  any loss they might take on that borrowing.   He liked to say, "We are going to build  the memory of the world here in Boise".

## From Up to Down But LJ Sevin had not been wrong. Memory is a commodity and the  market can turn at the drop of a pin. In late 1984, Micron's salespeople  noticed that customers were starting   to ask for lower prices and concluded  that a market downturn was on its way. They decided that they should get ahead of  this decline. So in September, they shocked   the industry by cutting the price of their 64K  chips from $3.40 a chip to as low as $1.95.

It was a play for market share  that puzzled market observers   and infuriated the rest of the industry.  The stock declined 25% in just two days. Lane Mason of the research firm Dataquest said,  "A price drop of that magnitude was unwarranted,   unwise, and leaves a lot of money on the table". NEC Electronics' president Keiske Yawata said,   "I don't see marketing expertise  in what they are doing".

VC Pierre Lammond, who invested for Sequoia  and Khosla Ventures over a long career,   really laid in there with: "I think their  luck is going to run out. The large Japanese   companies and some of the US manufacturers will  eventually wipe them off the face of the map". ## 1985 Thanks to this price cut, Micron managed to pile  up orders and win favor with customers. Initially. Then came a disastrous 1985. The  year began with high hopes. The   Semiconductor Industry Association  predicted a nice 22% growth in sales. Then suddenly, it all fell apart thanks to  weakness in the American electronics industry.  

Prices on every piece of electronics  from CD players to gaming consoles to   PCs crashed. Several VC-funded PC  makers collapsed, hurting demand. American chip sales fell by 28%. National  Semiconductor in particular took a heavy   beating. Their sales fell 17% and  profits swung from plus 59 million   to negative 117 million. They shut down for  39 days of the year and laid off 2,200 people. The Japanese continued to produce,  even as the Americans cut back.  

Insulated from financial damage by  their sibling banks, electronics,   or real estate arms, they sold at  whatever it took to get the sale. After the price cut in September 1984,   prices for 64K memory fell from $2.20 per  chip to under $1.00. During the summer,   spot prices for 64K were hitting an unthinkable  35 or 40 cents. It felt like there was no bottom. In February 1985, Micron laid off  nearly 50% of its 1,250 employee base,   500 people. They also dialed down 64K production,  shifting more towards the next generation 256K.

Eleven American memory-makers exited  the market that year - most famously,   Intel. United Technologies sold Mostek to  Thomson of France. By the end of the year,   the only American companies still making Dynamic  RAM memory were Micron and Texas Instruments. ## The Petition Micron eventually decided that  greater measures were necessary. In June, they filed a petition with  the US International Trade Commission   and the US Department of Commerce  alleging a material injury due to   dumping practices by the Japanese  - essentially selling below cost.

They requested an investigation into this  and possible countermeasures. They also   filed a lawsuit against Japanese  firms themselves for dumping. Micron was the first to take  such steps and it demonstrated   their differences with the rest  of the American semiconductor   industry - represented by the Semiconductor  Industry Association or SIA. Joe Parkinson said: > It was very clear to me that they [SIA] had a   different agenda ... their strategy  was whatever the Japanese get into,  

let's get out. The people who are dominant  in SIA are not taking the Japanese on. Indeed, the SIA was deeply divided. Some  like Motorola and Charlie Sporck of National   Semiconductor wanted to fight. But there  was also a sizable faction that believed  

that the US should instead move on to more  valuable, design-rich products like Intel did. And several - including those in Japan -  thought that Micron was being self-serving.   Jack Beedle of In-Stat, a semiconductor  market research firm, said it best: > Everyone was cruising along at $3 until  they started the price war. Then they   turn around and sue Japan on prices ... any  company that only produces one kind of part   when that part is also manufactured by the  Japanese is going to have serious problems Not the most charitable view, but it  was reflective of the larger industry   in early 1985 - that Micron had made  its own bed and should sleep in it. The   Japanese makers cited Micron's actions  as exhibit number one in their defense.

## The Agreement I shall sidestep this era of semiconductor  trade discord between the US and Japan,   but as it became clear that something was  indeed going on, the industry joined in. In August 1985, the International  Trade Association made a preliminary   ruling that there was indeed some material  injury. With the help of Idaho's senators,   the issue escalated to the highest  levels of American government.

Eventually the Japanese came to the  belief that some accommodation was   necessary despite continued disagreement  on the dumping issue. Akio Morita of Sony   led informal talks with the Americans  on a resolution to bring back peace. Thus came the 1986 Semiconductor Agreement,  which put price floors on Japanese semiconductor   products and opened up new markets in  Japan. The guys in Idaho were happier,  

though skeptical that the Japanese  would actually follow through. All throughout this ordeal, Micron still  suffered $50 million of losses in 1986 and 1987. When the Japanese lagged in implementing  the 1986 agreement as the Idahoans feared,   Reagan imposed tariffs in early 1987, reducing  supply and production from the Japanese. When the market turned again in early 1988, it  caused a DRAM shortage. Memory prices shot up,   irritating American computer and electronics  makers, who screamed "Chip shortage!" But Micron - as one of the two last American  memory makers standing - did very well,   turning nearly $100 million in profit.  They had survived the Japanese wave.

## Conclusion In 1998, amidst another brutal memory downturn,  Micron bought Texas Instruments' DRAM business. It solidified their position as America's  final DRAM memory-maker - the last survivor   of one of the chip industry's most brutal spaces. Today, they are a $150 billion company.  Armed with $6 billion in CHIPS money,   they are setting out to build leading  edge memory fabs in Idaho and New York. The Micron Technology story is remarkable. You  can't make it up. The way these guys started out,  

persisted, and eventually won the market  through smart design, cost control,   and sheer determination is a demonstration that  semiconductor success is possible almost anywhere.

2024-09-06

Show video