Leading the industry in GAA-FET, Samsung Foundry’s 3nm is the first commercially available technology to enable voltage supply level scaling for traditional FinFET technology. Reducing power by nearly half (45% lower) compared to 5nm FinFET and improving performance by 23% while reducing area by 16%, Samsung leverages Nanosheet GAA transistors and the ability to adjust nanosheet width to deliver PPA optimization for more adaptable designs.

New leaps in low power and high performance make 3nm ideal for the speed and battery demands of flagship 5G smartphones, while also powering breakthroughs in AI and HPC. Second generation 3nm GAA takes you even further, dropping power consumption by 50% from 5nm levels while improving performance by 30% inside a 35% smaller area.

Samsung Foundry’s 4nm represents the ultimate FinFET process technology. It enables the world’s smallest 4nm standard cell, with a cell height of 200nm, a new MOL architecture, and scaled Mx pitch. The 4nm family covers a range of performance driven processes, with SF4, SF4E, SF4P, SF4X, and SF4A.

SF4, SF4E, and SF4P are tailor made to power mobile products, beginning mass production in 2021. SF4E has been developed further into SF4 and SF4P, increasing device performance. And SF4X brings the highest performance in the 4nm family by adding higher drivability transistors and improved BEOL processes for reduced RC delay—bringing better support to HPCs.

Bringing breakthroughs in power and efficiency, 5nm enables HPC systems to do more than ever with less power, shrinking energy usages while improving speeds to break barriers in performance. With a 25% density improvement and a 10% performance uplift over 7nm, it cuts power consumption by 20%.

In the automotive world, Samsung Foundry’s 5nm process is powering the new wave of smarter, more responsive vehicles. With a world-class design infrastructure and a long list of IPs tailored to Automotive, IoT, and HPC, Samsung Foundry is ready to jumpstart 5nm designs and power tomorrow’s automotive and HPC innovations.

The 7nm process created new ways to improve both scalability and efficiency, revolutionizing not just how chips work, but how they’re made. 7nm introduced extreme ultraviolet (EUV) lithography to the world for the first time, putting silicon wafers through an exposure process with finer detail than conventional immersion ArF. Requiring only a single mask layer as opposed to the multiple masks used in previous process nodes, EUV technology shortens turnaround time while also improving yield rates.

All this adds up to a process that cuts area by 40% while doubling power savings. And with 20% fewer masks used during production, it brings cost-saving scalability to a range of products. Samsung began mass producing 7nm in 2019, with further expansions to its EUV capabilities in the following years.

Samsung Foundry’s 8nm lauded as the best and most balanced process before using EUV, provides a combination of high performance, scalability and cost in the traditional FinFET technology. It is an ideal process for a broad range of applications and offers customized extensions such as RF, eMRAM, and automotive to address specialty markets.

8nm RF process was designed to take 5G communications further, delivering a special architecture for a special chip. Samsung overcame scaling challenges for radio frequencies (RF) by mobilizing its RFeFET™ architecture—a technology exclusive to 8nm—to achieve a design that increased battery efficiency by 35% while also slashing chip area by 35%.

Covering markets from sub-6 GHz to mmWave applications, Samsung’s 8nm RF is a one-chip solution ideal for the rapid information sharing of 5G, delivering long-lasting power and crystal-clear signal quality from compact mobile devices. It represented yet another step in Samsung’s RF market leadership.

FD-SOI (Fully Depleted – Silicon on Insulator) is a process that overcomes some of the limitations of Bulk CMOS technology with improved performance, power and area. Utilizing FD-SOI device architecture boosts device performance and brings ultra-low leakage advantages. 28FDS has been in production for Network, Consumer, MCU, and IoT products and has expanded into new applications such as mmWave. Moreover, the industry's first eMRAM (embedded Magnetic Random Access Memory) with extremely small bitcell, endurance to 1E6 cycles and fast write speed (~1000× better than conventional eFlash) opens up new types of applications where power, speed, and cost are important.

Samsung Foundry’s integration of 3D FinFET technology into its 14nm process represented a light-speed jump in the foundry industry, raising performance, boosting power efficiency, and widening scaling capabilities. This triangulation of Performance, Power, and Area (PPA) was possible thanks to a design philosophy that zeroed in on 3D Fin structure. The result was a smaller chip size—with the industry’s smallest SRAM bit cell in the 14nm-class—and a process with an elegant simplicity despite the complexity at work.

14nm FinFET’s product options—14LPE, 14LPP, 14LPC, and 14LPU—are specialized to suit the fast-paced conditions of the mobile market, meeting the high-performance expectations of today’s mobile customers while delivering fast turnaround times and improved specs all around.

And to bring even more capabilities to our customers, Samsung Foundry is launching an eMRAM solution under the 14nm LPU FinFET-based NVM. That flash-type eMRAM will support automotive grade 1 demands with potential to power better performance in 5G, IoT, electric vehicles, and AI.

Samsung’s 28nm process became an instant classic, with high demand and continued industry staying power. With a low-power High-K Metal Gate Process, 28nm created a remarkably simple migration path that made it a long-lived node powering multiple generations of products. With a growing need for different variations of the process, Samsung Foundry broadened its portfolio to include RF and eNVM variations.

The 28nm Radio Frequency (RF) process lets chip designers enhance connectivity-focused applications with advanced RF functions, 28RF process design kits (PDK), and verification methods.

28nm eNVM (embedded non-volatile memory) technology, meanwhile, focuses on making new computing systems faster and more scalable. Within the eNVM category, 28nm eMRAM (embedded magnetic random access memory), uses a lower voltage than eFlash memory with writing speeds up to 1000 times faster. Lower power and high scalability allow customers to save cost and replace conventional eFlash, making 28nm eMRAM ideal for advanced microcontroller unit (MCU) applications within the automotive and IoT fields.

Pushing the lower limits of 8-inch technology, Samsung’s 70nm process sets a brisk pace for innovation. With improved speed and a low power demand over 90nm, it increases yield rates to deliver a higher number of good dies per wafer.

This has made Samsung’s 70nm process a good fit for large panels such as high resolution 8K TVs. It also supports modules specialized to use W-damascene processes for small panel products; its reduced chip size and high-density memory have made it perfect for micro displays.

The 90nm system-on-chip design marked Samsung’s first leap into nanotechnology. Reducing die size by more than 50% from 130nm, it provides a low power option with reduced dynamic power. Along with a 30% increase in speed, the 90nm process pushes boundaries as a solution for high-density non-volatile memory (NVM) and mobile applications. With more power at half the size, it solidified Samsung as a major player in a semiconductor industry undergoing major change.

Samsung Foundry is now extending the 90nm process with a specialized process for low-end MCU with clocks below 200MHz. The 90nm process powers a range of cutting-edge devices, including Fingerprint on Display (FOD) systems that read fingerprints optically. It uses Split Photo Diode to deliver higher dynamic range in vehicle camera sensors, while allowing a higher capacity for global shutter functions in today’s cameras. It also supports the addition of Single-Photon Avalanche Diode (SPAD) pixels for Direct Time of Flight applications.

Optimized for ULL (Ultra low leakage), the 130nm process is a go-to option for wearable and IoT devices with power management to keep devices performing their best on less power. As a process designed for analog performance, 130nm nodes excel as eFlash processes for MCU application, BCD processes for power management IC, and DDI processes for large panel displays—whether it’s a TV, PC, or tablet.

From its outset, the 130nm process allowed Samsung to pack a world of new advances into a single chip, triggering new waves of performance
for device makers. Using the world’s latest lithography techniques to make the speed of devices even more competitive, the 130nm
process powered a myriad of chips and amplified progress across the industry.

Samsung Foundry’s 180nm process offers exceptional accuracy and reliability with ideal support for analog applications. Stacked with reliable device characteristics, it’s known as a cost-effective solution catered to the demands of sectors like IoT, analog solutions, and automotive devices, bringing more digital parts of everyday life.

With its low power technology and high reliability, Samsung’s 180nm set the stage for more powerful mobile and smart devices that reshaped the world in a decade.