At the heart of your phone, tablet and computer lies the microprocessor, a tiny chip home to billions of transistors capable of processing an immense amount of information.
Without the microprocessor, modern technology could not exist, which is why in this post we’ll be looking at the company that started it all, Intel.
December 23, 1947.
After two years of restless labor at Bell Laboratories, these three men stood in awe of the transistor, their greatest invention.
The man in the middle was William Shockley, an entrepreneurial fellow who realized what a fortune he could make from this new technology.
In 1956 he moved to the west coast, establishing the first silicon device company in what came to be known as Silicon Valley.
He couldn’t convince any of his former colleagues at Bell Labs to leave with him and so he resorted to hiring fresh university graduates.
In an ironic twist of fate, just one year later eight of his brightest employees got together and left the company in the same way that he had left Bell Laboratories.
Under the patronage of industrialist Sherman Fairchild, the “Traitorous Eight”, as they were called, founded Fairchild Semiconductor.
Much to Shockley’s dismay, Fairchild became one of the leaders of the industry while his own venture failed.
In 1959 one of the original “Traitorous Eight”, Robert Noyce, created the first integrated circuit.
Like the transistor before, the integrated circuit was a technology with huge potential, and he knew that.
In 1968 he left Fairchild to start his own company and he was joined by his colleague and fellow ‘traitor’ Gordon Moore, who had famously postulated Moore’s law.
To fund their venture they went to Arthur Rock, the acclaimed investor who had arranged their original deal with Sherman Fairchild a decade earlier.
With $3 million of initial capital and the creative portmanteau of integrated electronics, Noyce and Moore founded Intel on July 18, 1968.
Behind their venture was the ambitious plan to build large-scale integrated semiconductor memories.
Back then, they were ten times more expensive than standard magnetic core memories, which were much slower and less efficient.
Nine months after its creation, Intel had developed its first product: the 3101 Schottky bipolar memory.
It was the world’s first solid state memory device and it was capable of storing a whopping 64 bits.
One year later, Intel became pioneers in dynamic random access memory, or DRAM, by creating the first commercially available DRAM chip, the 1103.
Its success marked the beginning of the end for magnetic memory and established DRAM as the primary storage medium of modern computers.
Intel’s reputation grew quickly, and not just in the United States.
A Japanese calculator company called Busicom had reached out to Intel in 1969 with a request to build integrated circuits for their calculators.
While working on this project, Intel engineer Ted Hoff figured out a way to build a central processing unit onto a single chip.
By cramming 2,300 transistors onto a one-eighth- by one-sixth-inch chip, Hoff’s invention had the same power as the ENIAC computer that was the size of a room.
Intel had unwittingly stumbled upon the foundation of modern computing, the microprocessor.
They called it the 4004 and started selling it in 1971.
A year later, Intel unveiled the 8008, an 8-bit microprocessor.
Intel’s first general-purpose microprocessor, the 8080, came in 1974 and it essentially became the industry standard, finding its way into almost every cash register, calculator and traffic light of its day.
Interestingly enough, the 8080 was designed for almost everything except computers.
At the time, computers were manufactured entirely in-house, with a single company building its own terminals, compilers, and operating systems.
The 8080, however, became so popular that the manufacturers, starting with Hewlett Packard, eventually began designing their systems around it.
In 1978 Intel released the 8086, a 16-bit processor that would eventually become Intel’s saving grace.
Up until that point Intel’s revenues were coming almost entirely from their DRAM division, but Japan’s rising semiconductor industry was quickly eating away at their profits.
Intel’s only way forward was microprocessors, and they went all in by partnering up with IBM.
We’ve already covered IBM in a previous post, but just to recap, in the early 1980s IBM were struggling to catch up with the rise of the personal computer.
At first, IBM didn’t think PCs would be worth it to the average person, but once that started happening anyway, IBM’s bureaucracy made developing their own PC a nightmare.
They ended up partnering with Intel for their processor and with Microsoft for their operating system, which allowed them to develop their IBM PC in just under a year.
It was released in 1981 and it became the dominant personal computer of its time, establishing Intel as the chief supplier of processors.
The IBM PC used a modified 8086 processor, and although IBM eventually lost the personal computer market to cheap compatible copycats, Intel remained at the heart of every personal computer made over the next decade.
The legacy of the 8086 remains to this day, as the vast majority of modern computers are based on its derivative x86 architecture.
During the 1980s Intel emerged as the most profitable hardware supplier to the rising PC industry.
They reached $1 billion of revenue in 1983, and the same amount as net income just nine years later.
In 1993 Intel released the Pentium line, their fifth generation of processors.
For this generation Intel started building dedicated motherboards alongside its processors, a move that kept them ahead of their competition and doubled their net income that year to $2.3 billion.
Throughout the 90s Intel continued to develop more powerful processors, more or less in accordance with Moore’s law.
In 1998 Intel branched out into the value-PC market by releasing the cheap, low-performance Celeron line.
The new millennium, however, would be a much more difficult time for Intel.
The dot-com crash and fierce competition from AMD saw Intel fall below 80% market share for the first time in decades.
The situation became so bad that in 2001 Intel’s profits had slumped by a stunning 87%.
By that point it became clear that racing to build faster and faster processors wasn’t the way to go, especially when most people were using their computers just to read their email or browse the web.
Intel shifted their focus accordingly, building a more efficient, less power-hungry line called Centrino.
Released in 2003, the Centrino wasn’t actually a processor but a fully functional platform, complete with a chipset and wireless network.
It worked extremely well on portable computers just around the time when laptops were finally starting to take off, lifting Intel back to the top of the industry.
In line with their new philosophy, Intel began developing multi-core processors, releasing their first dual-core in 2005.
In general, the past few generations have been split into three main categories based on processing power: i3, i5, and i7.
Up until last year, Intel were operating on a “Tick-Tock” model, where they either shrink the size of the current microarchitecture to make it more efficient or release an entirely new one every 18 months.
The performance of the last two generations hasn’t improved that much though, and Intel have also attracted a lot of antitrust litigation.
In 2009 the European Union fined Intel more than one and a half billion dollars for bribing computer manufacturers to use their processors.
Similar accusations have sprung up in the US, Japan and South Korea.
Despite the lawsuits, Intel’s business has been going great, and they’ve been able to branch out into various other tech markets, usually through acquisitions.
Among other things they’re working on solid-state drives, machine learning and autonomous vehicles.
Some of these projects are more successful than others, but it’s unlikely that they’ll be replacing Intel’s main microprocessor business any time soon.
0 comments:
Post a Comment