Intel Redefines the Numbers
The following is from our J.Gold Associates, LLC. July 26, 2021 Technology Currents research note that analyzes the recent Intel announcements. To obtain a copy or to sign up for our mailing list, send an email to: info (at) jgoldassociates (dot) com
Intel is announcing its plans for its IDM 2.0 strategy, and one of its major themes includes the renaming of its process nodes to be more in line with others in the industry. Intel has been labeled as a chip manufacturing laggard, even as it dominated process technology only a few years ago. Indeed, Intel has had a number of manufacturing “hiccups” in the past couple of years, and is working hard to retake the lead in process technology while others like TSMC, Samsung, etc. surpassed Intel manufacturing capabilities. But while many claim a major and longer term advantage with a one to two process technology node advantage, is the process technology gap really as great as it seems?
What’s in a Name (nm)?
Unfortunately the marketing hype around chip size is more about “specsmanship” than about a standardized way to measure size. In fact, there is no consistent way that the industry measures the process node size, and many liberties are taken in stated sizing. Some measure it by transistor or gate size, (but fail to measure the density of transistors arrayed on the chip), or pick a specific array of transistors for their density (e.g., in a uniform memory array). This inconsistency leads to a wide array of “nanometer claims” that in reality are often apples to oranges comparisons.
Intel believes that its naming convention for nodes is more realistic than some of its competitors, and that its current 10nm process is at least as compact as many 7 nm processes available in the market. The perception of this “process node competition” has placed Intel at a disadvantage in the market as it appears to be at least one and possibly two generations behind in process technology, which many chip vendors have used for marketing advantage (e.g., AMD, Qualcomm, Nvidia). This has had major consumer and financial consequences.?
No longer willing to play this game of “specsmanship” with its competitors, Intel has decided to rename its processes to be more in line with the market requirements. It believes that the critical determination of value for many customers is performance per watt which requires a blending of power, area and performance, and not just an optimization of process density. As a result, it will be dropping direct references to its process node size and instead implementing a node generation scheme. Its renaming its now-in-production enhanced super FinFet technology to Intel 7, it’s previously referred to 7nm technology as Intel 4 technology with chips like Meteor Lake coming in the?second half of 2021, an Intel 3 process it plans for the second half of 2023, and its planning an advance that Intel calls 20 Angstrom technology with a totally new type of transistor (RibbonFET) that it will produce in first half of 2024.
While no doubt many competitors will see this as just “Intel marketing spin”, we believe that it’s an important step for Intel in a few ways. First, it shows the market that Intel is not nearly as far behind in process technology as some perceive and that other things are also critical to performance. Second it gives Intel employees an additional boost in making sure they can provide improved process steps going forward (it’s always better for a team to be close in score to the competition rather than way behind). Finally, Intel is setting a stake in the ground that says to its competitors you can take liberties with your measurements, but we won’t play that game.
It’s all in the package
There is also a component of Intel’s strategy that it does excel at over the competition, and that is separate from the process node specsmanship battle. Future complex chips will be made of many “chiplets” tied together on a silicon substrate, much like a PC board. By creating a new 3D packaging technology (Foveros) that is ahead of the competition, Intel can create very complex chips using a variety of its own (or even third party) chiplets. Single monolithic processors are simply becoming too complex and the more complex (and larger) they become the more costly to process and build without defects they become as well. So having a leading packaging capability is critical to all semiconductor operations and this is an area with a renewed emphasis by many third party producers that also see this as critical technology.
Intel’s next generation Foveros technology (Foveros Omni with power and signal integrity boost, and Foveros Direct with direct copper to copper chip connections and through holes (PowerVIA) will give it a distinct advantage in producing the complex higher end System on a Chip (SoC) products needed in future data centers, AI, Machine Learning Edge Computing, and what it calls XPU designs. Intel’s competitors, while not sitting still, will have to play catch-up to Intel’s innovations in packaging. While this is less of an issue for smaller less complex chips that can be monolithically produced, future high performance and complex SoC designs will require advanced packaging techniques and not just advanced processing nodes.
Bottom line: While some will look at the Intel announcements as simply a marketing exercise, we believe there is real “meat” in these announcements that will put Intel back in a highly competitive, and potentially leadership position over the next 1-3 years. There is no doubt Intel will face fierce competition from other chip vendors in both process and packaging, but it’s important to recognize that Intel is not as far behind as many believe it to be, and it has the tools and processes to remain a leader in the markets it serves.
About J.Gold Associates, LLC.
J.Gold Associates provides advisory services, syndicated research, strategic consulting and in-context analysis to help its clients make important technology choices and to enable improved product deployment decisions and go to market strategies. We provide our clients with insightful, meaningful and actionable analysis of trends in the computer and technology industries. We have acquired a broad based knowledge of the technology landscape and business deployment requirements, and bring that expertise to bear in our work. We cover the needs of business users in enterprise and SMB markets, plus focus on emerging consumer technologies that will quickly be re-purposed to business use.?
?