SAMSUNG FOUNDRY

Classical scaling hits the Tox scaling limit in the semiconductor industry about process evolution. From the begining of the 32LP process, continuous scale-down has been achieved by material innovation, and Samsung Foundry adopted HKMG as Foundry 1st. By extending HKMG to 28nm, it provides 28LP and 28LPP processes, which has been used in products of various customers.

By utilizing FDSOI device architecture, significant boost in device performance is achieved for both high performance and ultra low leakage.
28FDS has been in production already for Network, Consumer, MCU and IoT products and expanding into new applications such as mmWave.
Moreover industry's first eMRAM(embedded Magnetic Random Access Memory) with extremely small bitcell, endurance of 1E6 cycles and fast write speed ( ~1000x better than conventional eFlash) opens up new type of applications where power, speed and cost are important. 18FDS is the next generation node that has the best performance and power in 2x node (comparable to 14LPP FinFET), yet with advantage of body-bias control.

Musca-S1 is a versatile 28FDS eMRAM based IoT Reference Platform(Test Chip & Board),
Musca-S1 allows for Designing IoT Security System more Easily & accelerates your Time to Market through Chip to Cloud Design,
Musca-S1 is a platform developed by Samsung Foundry and SAFE™ Partners Arm, Cadence & Sondrel.
For more information, go to: Arm Musca-S1 promotion page

Samsung Foundry has taken a revolutionary footprint to the foundry industry by implementing the innovative 3D-structure FinFET technology which is the industry's the most important technology milestone into 14nm process transition for thesuperior performance/power/scaling benefit.

14nm is constantly evolving to provide even more benefits for long life time.
Also single platform technology covers various applications from high performance(AI, Network) to low power(Mobile, Crypto currency) with supporting 14LPP, 14LPC, 14LPU. 11LPP is being ready for the best class area stabilities and supporting Automotive application as well.

Samsung foundry has maximized utilization of multi-patterning technology to overcome the cost concern and 10nm will be another long-lived node like as 14nm. To support long life time, transistor performance has been boosted with optimization of fin structure and stress engineering. Moreover 8LPP evolving as the most optimized & economical node until EUV enabling. 8LPU shows extremely low power efficiency with similar speed to 7nm.

EUV lithography technology has competitive in PPA (power and performance, Scaling). Less number of mask and process steps with EUV adoption provide benefit of shorter TAT to support Time-to-Market and better defect density resulted in better yield.
Compared to ArF(i), EUV has better OPC model accuracy and fewer hot spots. HVM (High Volume Manufacturing) goal of 250W has been demonstrated. Moreover mask defect control is one of the key challenges, and Samsung Foundry has developed in-house mask defect inspection tool to enable stable mask making infra. This is unique benefit of Samsung Foundry EUV technology.
Dedicated EUV fab has been under construction to support customer demand, and mass production will start in 2019.

7nm with full EUV technology on track for 2019 production, as foundry's most scaled technology. Further more 5LPE defined for IP re-usability, yield learning leverage, and easy migration from 7LPP.