Bobby Miller
Writer Website
Is Moore's Law Meant to be Broken?
1,800 words
Nearly fifty years ago, in a paper for Electronics Magazine, a man named Gordon Moore (b. 1929) made an interesting statement. He claimed that the number of transistors on a microchip will double approximately every two years. According to The Economist, Moore later said that it was a somewhat offhand comment at the time. However, this statement has since then become known as “Moore’s Law,” and it’s the guiding principle behind Intel, a semiconductor chip company he co-founded in 1968.
Gordon Moore, co-founder of Intel (Credit: Slate.com)
Basically, Moore’s Law says that computer chips are improving at a faster rate as time goes on. We can squeeze more transistors onto a microchip in order to allow for faster data transfer without using more space or money.
A transistor is responsible for guiding the flow of electricity while amplifying its signals, as TechTarget.com explains. So, more transistors in a microchip means that more data can flow through it. This in turn allows us to transfer high amounts of data at a faster rate without more microchips taking up more physical space.
This graph demonstrates how Intel’s major processors have gained more transistors for three decades. (Credit: Associate Professor Rodney Van Meter of Keio University)
Moore’s Law is the reason why computing technology has become faster and more powerful while also becoming smaller and more affordable. For example, a smartphone today might be a thousand times more powerful than a computer from forty years ago, according to Time magazine. As Alan Seabaugh of IEEE Spectrum magazine also notes, the transistors are “less power hungry than their predecessors,” meaning they’re more energy efficient.
Transistor prices are dropping as well, which is why powerful computers are affordable to most consumers. CNET.com explains that the average transistor cost $5.52 in 1954. But in 2004, the average transistor cost only 191 nanodollars, which is 191 billionths of a dollar. This is remarkable considering how many product prices actually rise over time thanks to inflation.
Clearly, Moore’s Law affects more than just microchips. It revolutionizes all technology based on it. “Advances in process technology and reductions in cost make computing devices accessible to an ever-increasing number of people worldwide,” according to Intel’s official website. It empowers “innovations across the computing continuum—from the smallest handheld devices to the largest cloud-based servers.”
Moore’s Law often applies to technology other than transistors in microchips. In a talk with Scientific American, Mark Kryder, PhD, noted that hard drive space is increasing at a faster rate over time while dropping in price as well.
Here’s a good example of that. In the 1980s, a 15 MB hard drive was advertised for $2,495 (without the installation costs). Nowadays, you can get a 2 TB [that is, 2,000,000 MB] external hard drive for $100. That’s about 133,333 times the storage for 1/25 of the price. More transistors make it perfectly reasonable to manage that amount of data at a quick pace in a small space.
(Credit: BoingBoing.net and Amazon.com)
It’s important to note that Moore’s Law isn’t simply claiming that technology becomes more powerful and more compact over time, though. It’s improving at an exponential rate. In other words, each year, computing technology sees a larger improvement than it did the previous year.
Intel’s Plans for Smaller Transistors
Obviously, making transistors smaller than they already are requires innovative techniques in producing them. To live up to its co-founder’s vision today, Intel is moving from 2-D (planar) transistors to 3-D ones. The “22nm 3-D tri-gate transistor technology assures us that the promise of Moore’s Law will continue to be fulfilled,” the company explains on its website.
22nm (22 nanometers) pretty darn tiny considering that “a human blood cell is about 7,000nm while a strand of DNA is 2.5nm,” as Kyle Russell of BusinessInsider.com points out.
After this transition, the company hopes that extreme ultraviolet lithography (EUV) will keep Moore’s Law alive even longer. This budding technique might let Intel map even more intricate circuits onto microchips. “EUV should allow Intel to continue to make its chips smaller, faster, and more efficient,” as Russell explains. “But first Intel has to figure out how to get it to work on a commercial scale.”
The company started experimenting with EUV as early as 2001, according to CNET.com. The process begins with a “mask,” a large representation of the circuits that will eventually appear on the chip. A diffused laser light bounces off this mask to capture the image, and then the laser bounces off a series of mirrors that make the mask’s image small enough to fit on a microchip.
(Credit: CNET.com)
There is hope for this technique. In 2009, Dutch firm ASML showed it could “reliably produce wavelengths of light appropriate” for EUV, as Russell of BusinessInsider.com reports. Yet, there has not been much progress in the field since then, so Intel invested $4.1 billion into ASML during the summer of 2013. It also made an “undisclosed” investment into one of ASML’s competitors so that it will have multiple suppliers if EUV becomes marketable.
If these prove to be worthy investments, EUV will probably be the key to Intel’s transistors after its most recent 22nm model. The company hopes to develop 10nm transistors by 2015 and 7nm transistors by 2017.
Why The Law Might Soon Be Broken
However, recent reports suggest that Moore’s Law might be coming to an end soon no matter what Intel does. At one point, transistors can’t become even smaller to fit more on a microchip due to the laws of physics. If Moore’s vision does prove unsustainable, it will hurt not just Intel’s profits but also the advancement of technology as a whole.
Moore himself confessed to TechWorld.com that he doesn’t believe the so-called law can last forever. “In terms of [transistor] size you can see that we’re approaching the size of atoms which is a fundamental barrier,” he said. He did add, though, that “we have another 10 to 20 years before we reach a fundamental limit.” He made this comment back in 2005, so that point might be sneaking up on us.
The problem is that transistors are made of silicon. Physicist Michio Kaku explained to Geek.com that once chip production reaches 5nm, silicon won’t work for us anymore. Smaller processors would just overheat.
This member of the periodic table of elements is deeply ingrained in the tech world, hence the name “Silicon Valley,” but it might betray Moore’s Law soon. (Credit: Chemistry_King of Zazzle.com)
We have quantum mechanics to thank for this dilemma. You see, a transistor should keep electrons flowing smoothly, like water in a riverbed. But as Seabaugh of IEEE Spectrum magazine explains, transistors can’t do this if they become too small. “The electron has a pesky ability to penetrate barriers—a phenomenon known as quantum tunneling,” he said. “As chipmakers have squeezed ever more transistors onto a chip, transistors have gotten smaller, and the distances between different transistor regions have decreased. So today, electronic barriers that were once thick enough to block current are now so thin that electrons can barrel right through them.”
If Moore’s Law is broken, that doesn’t mean technology will stop improving. It simply means that it will stop improving at a faster rate over time. Because of this, it will take longer for prices on the latest computing technology to drop. This will give people more time to adapt to new tech items, which could be helpful for those who aren’t exactly electronic experts. We may also become more responsible with our current gadgets since they won’t be made obsolete as quickly. This in turn will reduce e-waste.
However, consumer electronics wouldn’t be the only field hindered by the end of Moore’s Law. We would also see slower progress in medical technology and other essential fields of research. And with microchip costs no longer dropping as quickly, getting computers and Internet connectivity to poor areas of the world for their education and advancement would take significantly longer.
The Future of Moore’s Law
With Moore’s Law in a precarious state, there are two main ways that tech thinkers can deal with the issue.
First, they might just try to find ways to speed up their products in ways other than adding more advanced processing cores from Intel. As Russell of BusinessInsider.com explains, companies can try having “other processors work in conjunction with one another” to help the most central cores of the product.
For example, in a Sony PlayStation 4 gaming console, the central processors are devoted to the video game being played. Meanwhile, other tasks, such as downloading files or recording the on-screen action for future viewing, are relegated to other processors. This allows the most powerful cores to do the most important jobs without being slowed down.
Companies could pull off similar effects by assigning certain tasks to a cloud if the device is connected to the Internet. This also places more transistors at a device’s disposal, but without them actually being inside the device.
Second, there are some researchers and companies who don’t believe it’s necessary to try weaseling around the alleged demise of Moore’s Law. Rather, they believe the law can continue indefinitely if we use different tricks at our disposal.
Michael Kanellos of CNET.com is completely skeptical of doomsayers when it comes to Moore’s Law. He notes that this famous computing principle’s “demise has been predicted repeatedly over the last few decades,” but it has stood the test of time for fifty years.
To work around the issue of electrons flowing outside their designated zone when chips become smaller than 5nm, we could try using materials other than silicon to make the microchips. Seabaugh of IEEE Spectrum magazine speculates that the next generation of microchip materials might come from a mixture of elements contained in columns III and V of the periodic table of elements. These elements might allow electrons to flow through barriers safely if transistors are redesigned.
Physicist Michio Kaku believes that “molecular, optical, and quantum computers” might act as potential successors to silicon chips, but even these would pose “some very difficult problems to overcome if they succeed.”
Despite the issues scientists today face, Futurist Ray Kurzweil believes Moore’s Law is far from over. In addition to the solutions optical and quantum computing might provide, we can look into DNA computing as well. On his official website, he asserts that Moore’s Law will continue to the point that, in the next few decades, we will reach the Singularity, a new era when major technological progress constantly occurs.
Even if we don’t look that far into the future, the survival of Moore’s Law has huge implications for future generations. Its demise would not only slow down how quickly we get cool new gadgets but also how quickly major fields of science progress. The advancement of technology itself has occurred exponentially for decades now, yet that period of rapid progress could become a blip in humanity’s history.
Bobby Miller originally wrote this article for the digital magazine Nvate in January of 2014. Due to its age, some of the information or links in the article may be outdated.