Anybody who is an Intel fanatic must be knowing what Moore’s law is. Named after Gordon Moore, co-founder of Intel and Fairchild Semiconductors, his 1965 observation paper described it as a doubling up of components per integrated circuit every year and estimated that this rate of growth would continue for at least another decade.
Undoubtedly, Moore’s law ran successfully until now. The company has decided to move on with 9th Gen Core processors and ditch Moore’s law. The 9th Gen Core processors have been launched and it turns out to be that they are simply refreshed version of the last generation chips possessing the same 14nm process the company is been using since 2014.
Gauging these new babies based on their performance, Intel assures us that they are squeezing out higher speeds. As compared to the last year’s model, the new chips are 10% faster. Targeting the serious users Intel has added some nice features on the granular side which includes switching back to solder thermal interface material (STIM) over the paste.
But chiefly the new chips have the same components that last year’s chips had—more cores. Reason being Intel still hasn’t managed to take a leap from its 14nm manufacturing node to 10nm process.
The manufacturing node describing the size of the minimum feature on the wafer of silicon determines the processor manufacturing generation. Usually, the smaller the node, the more technology can fit in, and the better the overall performance.
Intel has now accelerated to its third gen products incorporating this old architecture: the 14nm node traces its roots to 2014’s Broadwell meaning that Intel hasn’t really progressed to an appreciably smaller transistor in nearly 50 years. This thing stands strikingly opposite to the long-flaunted Moore’s Law—Intel’s leading light for the past half-decade—that will discontinue its hold soon.
The added perks of increased transistor coming from shrinking die sizes—for instance, cheaper cost per chip, lower heat, decreased power demands, and improved yields—will be harder to achieve by going forward. Rather than spending time on the classic tick-tock structure, where Intel would fill in the mold of architecture size for one generation and enhance it in the next generation, we are spending time even more on the current step, and Intel must lookout for ways to step up its performance. The recently launched 9th Gen chips are just the upgraded version of the 2017 Coffee Lake architecture, which Intel fondly calls the 14nm++ node, which technically is the upgraded version of the initial 14nm node.
Exactly none of this is new. It’s been quite evident that Moore’s Law would be no longer applicable sooner or later. Transistors can be further minified, which Intel hasn’t achieved yet. Certainly, Intel is facing problems to make similar kind of jumps that it once did.
This points towards the adjustments that both company and customers must make regarding what to expect from future gen of chips as Intel is compelled to look for other ways to improve performance and add new features to boost customers to upgrade.
Whether the new technology can secure its place in the upcoming years, we will have to wait and see. But if this year’s batch is taken into consideration, it appears like Intel’s future will be more of gradual change in preference to a sprint towards the future.