Capgemini Engineering’s Post

2nm Process Node - Can be a game changer? The world’s leading semiconductor companies are racing to make so-called “2 nanometer” processor chips that will power the next generation of smartphones, data centers, and artificial intelligence. TSMC remains the analysts’ favorite to maintain its global supremacy in the sector, but Samsung Electronics and Intel have identified the industry’s next leap forward as a chance to close the gap. For decades, chipmakers have sought to make ever more compact products. The smaller the transistors on a chip, the lower the energy consumption and the higher their speed. Today, terms such as “2 nanometer” and “3 nanometer” are widely used as shorthand for each new generation of chip, rather than a semiconductor’s actual physical dimensions. [[[ 2nm : the "2 nm process" is the next MOSFET die shrink after the "3 nm" process node. The term "2 nanometer" or alternatively "20 angstrom" (a term used by Intel) has no relation to any actual physical feature (such as gate length, metal pitch or gate pitch) of the transistors. "2.1 nm node range label" is expected to have a contacted gate pitch of 45 nanometers and a tightest metal pitch of 20 nanometers. ]]] Samsung sees 2 nanometer as a game-changer. “But people are still doubtful it can execute the migration better than TSMC.” Meanwhile, former market leader Intel is promoting its next generation “18A” node at technology conferences and offering free test production to chip design firms. The US company says it is set to begin production of 18A in late 2024, potentially making it the first chipmaker to migrate to the next generation. https://lnkd.in/g4q7hWEZ

1nm! This is an incredible achievement! You may not understand what 1nm means, but in the tech world, it means you can now fit additional smaller transistors on a smaller CPU at a lower cost. It also means increased performance, improved power efficiency, and less cooling required. The performance increase comes from having more transistors per square nm, and they are also faster because smaller transistors mean less distance an electrical signal has to travel between them making the performance of a CPU faster. For example, Apple's A14 chip uses a 5nm fabrication process and has 11.8 billion transistors, as opposed to Apple's A12 chip which uses a 7nm fabrication process and has 6.9 billion transistors. Moving from 7nm to 5nm allowed Apple to improve the A14's performance by 40%, increase efficiency by 30% and extend battery life per day in handsets by 5 Hours by consuming 40% less power than an A12 chip. Moving to 1nm is no joke. Consider this, human hair is roughly 80,000nm to 100,000nm in diameter, and a human DNA strand is roughly 2.5nm in diameter. This means that the wires connecting the transistors inside a CPU fabricated using a 1nm process are more than twice as small as a human DNA strand. This is going to open the door to so many advances in computing that it's hard to explain. Performance, power efficiency, and OPS all go up, while size, cooling, and power consumption all go down. Think about it, the Samsung S23 Ultra phone sold in the U.S. uses the latest Qualcomm Snapdragon 8 Gen 2 chip, uses a 4nm process, and has 16 billion transistors. The Apple iPhone 14 Pro uses the latest A16 chip uses a 4nm process, and has 16 billion transistors, now compare the last phone you had to the latest phones offered by either Samsung or Apple. As TSMC opens its doors to its Global Research and Development Center in Hsinchu, Taiwan, to drive the next generations of semiconductor technology. I can only wish we were doing the same thing here in the U.S. With global tensions in Europe and Asia, having a manufacturing fab right now with that technology and capability here in the U.S. would give us a huge advantage as we try to reduce our manufacturing dependence outside the U.S. The new center in Taiwan will house 7000 researchers working on TSMC's 1nm and below process technologies as well as working on novel materials and transistor structures. We need more investment in technology development and manufacturing here in the U.S. like Micron's $100 Billion investment announcement in Oct. 2022 to build a mega fab in upstate New York. TSMC is investing $12B to build a second fab in Arizona by 2024, it open its first fab in Washington in 1996. Samsung announced in Nov. 2021 that it plans to invest $17B to build a fab in Texas by 2025. Unfortunately, when these fabs come online they'll only offer a 5nm or 4nm manufacturing process until they get additional investment and training to move to 1nm, but by then TSMC Taiwan will be working on pm processes.

There is no digital society without chips! 📱 At Eindhoven University of Technology, we are going to push the limits of the machines that manufacture chips to enable a sustainable digital society. While the number of connected devices will keep growing exponentially, we will make them in a more sustainable way, taming (or even reducing) the total energy consumption. 🌍 #futurechipsweek #SustainableTech #ChipManufacturing #DigitalSociety

It's hard to imagine a world without chips in smartphones, cars, and computers. 🖥️ As we place more demands on these technologies, they are consuming more and more energy. If we don't find ways to make our computing technologies more sustainable, then the footprint of our digital world will become unsustainably large. 🌍 At the High Tech Systems Center (HTSC) on our campus, our 400-strong team of researchers are advancing fabrication approaches for the next generation of chips with energy-efficiency very much in mind. Learn more in this article as part of Future Chips Week at TU/e 👉 https://lnkd.in/epKWSvmi #futurechipsweek #semicon #semicontechnology #chips

https://lnkd.in/gGUGS5DW Some chips anyone ⁉️ The future is here!! Chipping away at competition With The 2nm Chip The transition to 2nm represents a quantum leap in chip design and fabrication. This node will enable transistors to be packed even more densely, boosting performance and reducing power consumption, essential for the next wave of technological innovations. Estimates suggest a 45% increase in processing power, 75% reduction in power consumption, and more features on a single chip. Smartphones with even more powerful cameras, laptops with advanced AI capabilities for multitasking, and faster internet connectivity due to more efficient processing. Further, AWS and other cloud providers could significantly reduce their energy costs and carbon footprints while increasing their computational power, enabling faster and more efficient cloud services. TSMC currently reigns supreme in chip manufacturing, holding the trust of major tech companies like Apple and AMD. It has plans for mass production of 2nm chips in 2025. But the US said checkmate with it's Chips Act. Intel's audacious strategy involves countering with a 1.8nm chip,! Gee, hope the lab has a prototype in place! But don't ignore that the US has thrown its weight behind the company by backing its proposed project in Ohio. Samsung owing to its vertical integration and innovation could be a good candidate for semiconductor supremacy.

In the relentless pursuit of smaller, faster, and more efficient computing power, the semiconductor industry has reached a groundbreaking milestone: the advent of 2nm chips.

2nm Process Node - Can be a game changer? The world’s leading semiconductor companies are racing to make so-called “2 nanometer” processor chips that will power the next generation of smartphones, data centers, and artificial intelligence. TSMC remains the analysts’ favorite to maintain its global supremacy in the sector, but Samsung Electronics and Intel have identified the industry’s next leap forward as a chance to close the gap. For decades, chipmakers have sought to make ever more compact products. The smaller the transistors on a chip, the lower the energy consumption and the higher their speed. Today, terms such as “2 nanometer” and “3 nanometer” are widely used as shorthand for each new generation of chip, rather than a semiconductor’s actual physical dimensions. [[[ 2nm : the "2 nm process" is the next MOSFET die shrink after the "3 nm" process node. The term "2 nanometer" or alternatively "20 angstrom" (a term used by Intel) has no relation to any actual physical feature (such as gate length, metal pitch or gate pitch) of the transistors. "2.1 nm node range label" is expected to have a contacted gate pitch of 45 nanometers and a tightest metal pitch of 20 nanometers. ]]] Samsung sees 2 nanometer as a game-changer. “But people are still doubtful it can execute the migration better than TSMC.” Meanwhile, former market leader Intel is promoting its next generation “18A” node at technology conferences and offering free test production to chip design firms. The US company says it is set to begin production of 18A in late 2024, potentially making it the first chipmaker to migrate to the next generation. https://lnkd.in/g4q7hWEZ

The Inevitable Death of CMOS: A Journey Through Time CMOS has been the backbone of electronics, powering everything from supercomputers to smartphones. Despite predictions of its demise, CMOS technology continues to thrive and evolve. Our CEO shares insights on the resilience of CMOS and the future of semiconductor innovation. Read the full article on our blog: https://lnkd.in/g4iyRzZx #CMOS #Semiconductors #Innovation #TechFuture #QuantumComputing #AI #Electronics #TechTrends #Technology #FutureOfTech #Engineering #Science #TechNews #ElectronicDevices #TechInnovation

Racing to make “2 nm” processor chips Semiconductor giants TSMC, Samsung & Intel are racing to make next generation of “2 nm” processor chips that will power the next generation of smartphones, data centers and artificial intelligence. The smaller the transistors on a chip, the lower the energy consumption and the higher their speed. Taiwan Semiconductor Manufacturing Company remains the analysts’ favorite to maintain its global supremacy in the sector, but Samsung Electronics and Intel have identified the industry’s next leap forward as a chance to close the gap. Any company that opens up a technological lead in the next generation of advanced semiconductors will be well placed to dominate an industry that pulled in well over $500bn in global chip sales last year. That is projected to grow further due to a surge in demand for the data center chips that power generative AI services. TSMC, which dominates the global market in processors, has already shown the process test results for its “N2” — or 2 nm — prototypes to some of its biggest customers, including Apple and Nvidia. It seems Samsung said the Korean chipmaker was offering cut-price versions of its latest 2 nm prototypes in an effort to attract the interest of big-name customers including Nvidia. Former market leader Intel has also made bold claims about producing its next generation of chips by the end of next year #semiconductor #semiconductormanufacturing #semiconductorindustry #semiconductortrends

🚀 AI models are rapidly evolving, outpacing hardware capabilities, and Arm is seizing the opportunity to innovate across the compute stack. Recently, Arm unveiled new chip blueprints and software tools aimed at enhancing smartphones' AI efficiency. But that's not all – they've also revolutionized blueprint delivery, potentially accelerating adoption. Introducing Arm Compute Subsystems (CSS) for Client, the latest compute solution for AI applications in smartphones and PCs, promising over 30% increased compute and graphics performance, and a remarkable 59% faster AI inference. Arm is not only powering the smartphone revolution but also gaining traction in PCs and data centers. They've launched new CPU designs optimized for AI workloads, new GPUs, and software tools to streamline AI app development on Arm chips. The real game-changer is delivering physical chip blueprints ready for manufacturing, saving immense time. Partnerships with Samsung and TSMC have resulted in cutting-edge solutions meeting the soaring demand for generative AI in mobile devices. This collaboration empowers designers to push semiconductor innovation's boundaries for unmatched AI performance and efficiency. Arm’s goal is to enable faster time-to-market by providing optimized designs for neural processors, delivering cutting-edge AI performance. Together, we're accelerating AI innovation using the most advanced technologies and design solutions. 🌟🔍💡 #AI #Innovation #TechRevolution #ArmTech #Semiconductors #MachineLearning #Smartphones #PCs #DataCenters #AITools #OpenAI #GenerativeAI #TechInnovation