Developing Chip Innovations Faster and Cheaper with AI Study from Lam Research

Semiconductor Review | Saturday, August 05, 2023

'AI-human approach' reduces chip development costs by 50 percent, according to a Lam Research study in Nature.

FREMONT, CA: "New approaches in innovation are needed to enable the industry to scale fast enough to meet the data-driven world's evolving demand for next-generation chips," said Tim Archer, president, and chief executive officer at Lam Research. "The opportunity for greater collaboration between talented engineers and machines in process engineering highlighted in Lam's study in Nature is a potential game-changer for our customers and our industry at large. This research is a testament to Lam's more than 40-year heritage of industry leadership and semiconductor manufacturing innovation. I congratulate the Lam team on this exciting work."

In a recent study, Lam Research investigated the potential for artificial intelligence (AI) to be used in chip manufacturing process development, which is currently a human-driven process that is necessary for the mass manufacture of every new advanced semiconductor in the world. The study, recently published in the journal Nature, identifies an opportunity to address two major challenges facing the industry: lowering development costs and quickening the pace of innovation to meet the rising demand for next-generation chips. Experts predict the semiconductor market will reach $1 trillion in annual revenue by 20301. According to the study, a "human first, computer last" strategy can achieve process engineering goals significantly faster and for half the price of the current strategy.

Stay ahead of the industry with exclusive feature stories on the top companies, expert insights and the latest news delivered straight to your inbox. Subscribe today.

The complexity of upcoming chips continues to make process development more difficult and expensive. Researchers at Lam pitted skilled process engineers against AI-enabled computer algorithms to find a more effective strategy.

Experienced and talented engineers must first develop a customized recipe that details the precise parameters and permutations required for each process step to construct each chip or transistor that has been built. The construction of these nanometer-sized devices on a silicon wafer involves hundreds of steps. On several occasions, small layers of materials are routinely deposited onto silicon wafers, and the surplus material is then precisely removed via atomic-scale etching. Currently, human engineers carry out this crucial stage of semiconductor development, mostly relying on their intuition and a "trial and error" method. Process development can be time-consuming, expensive, and labor-intensive, with every recipe being specific to the chip design and more than 100 trillion alternatives to consider. As a result, it takes longer and longer to develop new technologies.

A targeted process development recipe was created at the lowest cost based on variables related to test batches, metrology, and overhead costs in the Lam study. In the study, researchers found that humans still excel at dealing with difficult, unconventional challenges, even when using a hybrid human-first, computer-last approach.

"Although critical to the creation of each and every chip produced, the plasma physics of process engineering has been for decades rooted in the same scientific approach that Thomas Edison used: trial and error," said Rick Gottscho, executive vice president and strategic advisor to the CEO – Innovation Ecosystem at Lam Research and co-author of the study. "Our research showed that while engineering talent remains essential to innovation, process engineering costs can be reduced by 50 percent by integrating AI at the right stage and with the right data. The study provides a prescriptive approach for bringing together the best of human-led engineering and what data science and machines offer to create a combination that performs better than one alone. If realized, this hybrid approach can lead to significant savings in both dollars and engineering time for the industry."

Lam is currently incorporating the study's primary findings into its development processes. The Lam study offers preliminary recommendations on effectively combining human expertise, talent, and experience with AI's capacity to quickly evaluate a wide range of potential combinations in process engineering.

"By complementing engineering expertise with AI using the human first, computer last approach, the tedious and laborious aspects of design are alleviated for engineers, freeing them up to focus on the creative areas of development and explore innovations that may have been out of reach either due to bandwidth or cost," said Keren Kanarik, technical managing director of Lam Research, lead author of the research paper and a former process engineer. "While the application of AI in process engineering is still in its infancy and human expertise and domain knowledge is essential for the foreseeable future, the results point us to a path to foundationally change the way processes are developed for manufacturing chips."

More in News

PCB design software is now becoming increasingly inseparable from workforce issues for electronics companies. This challenge goes beyond having access to a good platform for engineers; many organizations find themselves facing problems when trying to adopt software due to their inability to recruit experienced PCB experts or provide proper training programs. This challenge tends to appear most strongly in companies scaling up electronics capabilities with limited hardware design teams in place. PCB designs continue requiring highly specialized skills, particularly when it comes to routing, signal integrity and manufacturing processes that might not be immediately obvious at earlier stages of the process. The increased complexity of software plays a big role here as well. Some PCB design environments incorporate extensive feature sets that have been gradually built up over years in the industry, meaning that veteran engineers can easily feel at home in them while relatively new professionals find difficulty adjusting to the workflow of such systems. This challenge creates an interesting disconnect between software features and engineering teams' ability to adopt the platform effectively. Managers who invest money in state-of-the-art software often find out that the time required for onboarding is longer than anticipated or that a few senior PCB designers end up being involved in almost all key layout-related decisions. These issues have a significant impact on how electronics organizations are preparing for product development in terms of staff planning. Companies are competing more than ever to recruit engineers who are comfortable with moving from schematic designs to layout reviews and further to manufacturing without undergoing extensive training programs. PCB software skills are now becoming an important topic in recruitment interviews rather than just preferences of engineers. Training issues themselves create further challenges for electronics organizations in their efforts to scale up their capabilities. Mentorship has always been a major aspect of developing PCB skills, yet shorter deadlines make it more difficult to engage experienced workers in the process of onboarding new recruits who can take over their jobs in a year or two. Collaboration through remote channels has also made a difference, since it enables electronics companies to become more flexible when recruiting talent, but reduces interactions through which newer employees could gain insight into layout-related processes or manufacturing issues that emerge during project implementation. Some electronics organizations respond to such challenges by limiting the complexity of some parts of the workflow or restricting software usage to essential features available for broad implementation across teams. Others continue prioritizing highly specialized design environments despite their longer onboarding process needs. Software vendors might find themselves under greater pressure related to user interface, features and training as a consequence of current workforce-related concerns among buyers. Those companies will start focusing more on how long it takes to turn new recruits into productive members of the design team independent of experienced engineers. It seems that these workforce challenges cannot go away anytime soon. The PCB designing process remains highly technical and tightly connected with manufacturing requirements, meaning that electronics organizations have to be careful not to simplify everything too much. Still, the discourse surrounding PCB software adoption is beginning to extend into a discussion of workforce issues. ...Read more
PCB software purchasing processes have become more focused on preparation for manufacturing, at least among businesses working in a tight development schedule and producing fewer prototypes per cycle. Its an indication of growing concern regarding a practical situation within the industry. Completion of a design no longer means it will be manufactured efficiently. Engineering teams spend more time checking whether design outputs and documentation provide a clean transfer without requiring further clarification from manufacturers. It has been an ongoing problem. The recent state of production, however, seems to draw attention to the consequences. Component swapping, procurement complications, and fabrication schedule changes can reveal gaps in the documentation process that would otherwise go unnoticed under normal production conditions. Purchasing conversations related to PCB software have begun addressing the issue. Design tools that allow to update the bill of materials, provide consistent fabrication outputs, and communicate design rules effectively across teams are being prioritized. Software is judged based on the amount of ambiguity it leaves after design and before production. It creates even greater pressure for small-volume PCB manufacturing. Independent electronics developers and specialized device designers frequently work with external fabrication vendors that produce devices for multiple clients at once. In such a setting, poor documentation can lead to a rapid drop in priority. The issue is further complicated by emerging problems with component procurement. Routing and board redesign due to sourcing requirements may impact the size, thermal properties, and other factors that need to be considered throughout the rest of the development process. Software environments lacking proper revision tracking may introduce complications with understanding which designs and outputs need to be applied. Such discussions have expanded beyond individual engineers' concerns to cover overall team experience. A successful completion of the layout work does not automatically guarantee smooth production. Many cases have seen manufacturing teams identify preventable errors during their reviews. In certain settings, such incidents can be attributed to software workflow problems. There are also financial ramifications. A delay in the prototype creation may interfere with testing schedules and customer demonstrations. In such a case, a design system capable of minimizing production-related clarifications may be considered more valuable than one providing additional features that engineers never use. The competitive environment in the industry is likely to change as a result. Feature enhancement is traditionally the cornerstone of vendor strategy when selling a PCB solution. Some potential customers, however, place greater value on consistent preparation for manufacturing than in increasingly complex design tools. It does not necessarily mean the advanced features lose importance. There remains a high demand for comprehensive simulation and layout management for complex electronics. Still, many users care about their consistency rather than capabilities. Overall, there is an emerging shift in value assessment criteria. Instead of evaluating performance in the engineering department, the discussion begins addressing productivity issues during production, communication with the supplier, and revision tracking. ...Read more
Shrinking timelines for hardware engineering have arrived in areas where many teams developing printed circuit boards do not feel prepared. Revisions of the product that took longer times to make and move to manufacturing review are expected to be completed far faster, particularly in industries with electronics associated with connected devices, industrial equipment or compact consumer devices. These conditions are changing expectations of PCB design solutions. Customers are looking carefully at the speed of moving design modifications across stages of layout reviews, simulations and fabrication preparation without producing new verification bottlenecks further down the road. This does not relate to the need to incorporate additional capabilities into design platforms; the goal here is the reduction of bottlenecks between stages already stuffed with dependencies. Increasingly dense circuitry designs, higher density of component placement and greater interplay between electrical and mechanical engineering teams cause delays when it comes to handling revisions. Engineering teams used to thinking of PCB design platforms as tools for drafting the layout now pay more attention to the software as a layer of coordination between decisions about the layout and the schedule for production. This affects software purchasing conversations in practical terms. Workflows for importing designs, managing revisions and matching manufacturing needs now play a more significant role, as redesign iterations are becoming costly. Small hardware companies seem particularly susceptible to this trend. Big corporations can cover additional verification rounds with engineering capacity, while smaller operations are forced to rely on a few experts who will perform layout validation, comply with requirements and arrange fabrication in parallel. Any delay in handing off the design will cause the postponement of procurement or prototyping by days or even weeks. PCB solution vendors find themselves dealing with more divided audiences as well. While some customers seek simulation capabilities, others will look for consistent output or a strong library management system. These differences have produced a fragmentation in the industry because each engineering challenge has its own definition of what a streamlined process looks like. A new problem has appeared regarding onboarding. Experienced engineers are still hard to replace, although some software requires high levels of familiarity with the workflow. Training time has become a factor in evaluating solutions for PCB design, particularly for electronics engineering teams that include embedded software developers or people responsible for production. Distributed engineering also complicates matters. Reviews of the printed circuit board layout could involve employees operating from different locations or independent contractors working on a limited basis. Software platforms that confuse the user on versioning or the owner of a particular design can introduce issues in fabrication reviews, which are detected much later down the line. None of this means that there is an impending revolution coming to the PCB design solutions segment. Many teams prefer well-established design platforms and are reluctant to change their existing workflow. The work with the layout is too close to fabrication for electronics engineering teams to replace existing solutions easily. Nevertheless, conversations about PCB design software are starting to go beyond the questions of interfaces or feature sets. Engineering deadlines vary considerably across electronic markets, which prompts discussions of whether there might be too much room for delays in design processes. ...Read more
Ultrashort laser pulse solutions are redefining possibilities across high-precision industries and scientific fields. Known for their exceptional control and minimal heat impact, these lasers enable clean, accurate processing of delicate materials and biological tissues. Their applications span from advanced micromachining to groundbreaking medical procedures and ultrafast scientific experiments. By offering unmatched precision and versatility, ultrashort laser pulses are becoming essential tools in driving innovation, efficiency, and discovery in modern technology and research. Precision in Micromachining and Material Processing Ultrashort laser pulse solutions are transforming the field of micromachining and advanced material processing. These lasers emit pulses measured in femtoseconds or picoseconds, allowing them to interact with materials in an extremely controlled and localized manner. Because the energy is delivered in such short bursts, there is minimal heat transfer to the surrounding area. This results in high-precision cutting, drilling, or structuring without causing thermal damage or creating unwanted deformations on sensitive substrates. This level of precision is particularly beneficial for industries working with delicate or composite materials, including medical devices, semiconductors, and aerospace components. Renesas Electronics Corporation supports advanced semiconductor development through integrated hardware, software, and system design capabilities. For instance, in electronic manufacturing, ultrashort laser pulses can create micro-holes and intricate patterns in thin films and multilayered boards without impacting adjacent structures. This capability supports ongoing miniaturization efforts while preserving the integrity and performance of critical components. These laser solutions are ideal for processing transparent or brittle materials like glass, sapphire, and ceramics. Traditional methods often struggle with these materials due to cracking or chipping, but ultrashort pulses can process them cleanly and efficiently. The non-thermal interaction mechanism allows for superior surface quality and edge sharpness, making this technology a key enabler of next-generation precision manufacturing. Miller Sales Engineering supports precision manufacturing through engineered solutions that enhance component performance, reliability, and process efficiency. Advances in Medical and Scientific Applications Ultrashort laser pulse technology is also making significant contributions in medical procedures and scientific research. In the biomedical field, these lasers are used in procedures requiring extreme accuracy, such as eye surgeries and microscale tissue ablation. The ultrashort pulse duration promotes faster healing, minimizes collateral damage to surrounding tissues, and reduces the risk of complications. Surgeons and researchers benefit from the high level of control, which is essential when working in sensitive biological environments. In scientific research, ultrashort laser pulses are enabling breakthroughs in imaging, spectroscopy, and diagnostics. They are used to generate high-resolution images at the cellular or even molecular level, helping scientists understand biological processes in unprecedented detail. Their ability to produce high peak intensities also makes them useful for nonlinear optical experiments, where multiple photons are absorbed simultaneously to reveal information that is not accessible through conventional light sources. These laser systems are used in time-resolved studies to observe ultrafast phenomena, such as electron movement and chemical reactions. Researchers can use this data to develop more efficient materials, study complex biological systems, or enhance drug development. By delivering extremely short, high-intensity light pulses, ultrashort laser technology opens new frontiers in precision and discovery across disciplines. ...Read more