48 Years Micro Chip and Beyond
Four decades ago today — November 15, 1971 — Intel placed an advertisement for the first single-chip CPU, the Intel 4004, in Electronic News. Designed by the fantastically-forenamed Federico Faggin, Ted Hoff, and Stanley Mazor, the 4004 was a 4-bit, 16-pin microprocessor that operated at a mighty 740KHz — and at roughly eight clock cycles per instruction cycle (fetch, decode, execute), that means the chip was capable of executing up to 92,600 instructions per second. We can’t find the original list price, but one source indicates that it cost around $5 to manufacture, or $26 in today’s money.
The 4004 used state-of-the-art Silicon Gate Technology (SGT) PMOS logic — a technique that Faggin perfected at Fairchild Semiconductor in 1968 — the world’s first metal-oxide-silicon (MOS) process. This breakthrough allowed the 4004 to have no less than 2,300 transistors and a feature size of 10 micron.
By comparison, there are half a billion transistors in a Sandy Bridge chip, and each one is just 0.032 micron. Considering a human hair is around 100 micron, the 4004 was still rather impressive — but irrespective of feature size or transistor count, the fact that it was carved from a single piece of silicon is what made the 4004 truly spectacular. Faggin was so proud of his creation that he even signed the chip "FF”, which you can see in the top right of the image below.
In real-world use, the 4004′s 92,600 instructions per second equated to the addition of two eight-digit numbers in 850 microseconds, or around 1,200 calculations per second. It’s perhaps not surprising that the first use of the 4004 was in the Japanese Busicom 141-PF calculator — and in fact, it was Busicom who originally asked Intel to create the 4004, as its in-house engineers needed 12 integrated circuits to make the 141 calculator work.
Intel C4004 vs. Pentium 4 vs. XEON 5500
Intel C4004 has 2300 transistors / Pentium 4 with 188 million transistors
For example, compared to the Intel 4004, today’s second-generation Intel Core processors are more than 350,000 times the performance and each transistor uses about 5,000 times less energy. In this same time period, the price of a transistor has dropped by a factor of about 50,000.
Intel XEON 731 millionen transistors 300,000 x C4004
The introduction of the 5600 marks the next step in the transition to the Intel 32-nanometer processors, formerly code-named Westmere. The 32nm logic technology uses Intel’s second generation high-k metal gate transistors to increase speed and decrease energy consumption. Processors within the Xeon 5600 family range from a four core L5609 at 1.8GHz all the way up to a six core X5680 running a 3.33GHz. All chips have 12MB of L3 cache regardless of core count.
Intel XEON 5680 Core 6 Full-size image 3,2 MB
Google quantum computer leaves old-school supercomputers in the dust
A Google quantum computer has far outpaced ordinary computing technology, an achievement called quantum supremacy, that's an important milestone for a revolutionary way of processing data. Google disclosed the results in the journal Nature on Wednesday. The achievement came after more than a decade of work at Google, including the use of its own quantum computing chip, called Sycamore.
"Our machine performed the target computation in 200 seconds, and from measurements in our experiment we determined that it would take the world's fastest supercomputer 10,000 years to produce a similar output," Google researchers said in a blog post about the work.
The news, which leaked into the limelight in September with a premature paper publication, offers evidence that quantum computers could break out of research labs and head toward mainstream computing. They could perform important work like creating new materials, designing new drugs at the molecular level, optimizing financial investments and speeding up delivery of packages. And the quantum computing achievement comes as progress with classical computers, as measured by the speed of general-purpose processors and charted by Moore's Law, has sputtered.
IBM challenges Google's quantum results
IBM is a major quantum computing fan, but it questioned Google's prematurely released results in a blog post Monday.
"We argue that an ideal simulation of the same task can be performed on a classical system in 2.5 days and with far greater fidelity," the IBM researchers wrote. They suggested different algorithms and a different classical computer design in a preprint paper of their own.
Google said it welcomes improvements to quantum computer simulation techniques but said its overall result is "prohibitively hard for even the world's fastest supercomputer, with more double exponential growth to come. We've already peeled away from classical computers, onto a totally different trajectory."
And you can try for yourself if you like. Google released its quantum computer's raw output to encourage others to see if they can do better at simulating a quantum computer. "We expect that lower simulation costs than reported here will eventually be achieved, but we also expect that they will be consistently outpaced by hardware improvements on larger quantum processors," the Google researchers said. Intel didn't offer an opinion on Google's results, but did say quantum supremacy is "a strategic benchmark."
"We are committed to moving quantum from the lab to commercialization," said Jim Clarke, Intel Labs' director of quantum hardware, in a statement.