Samsung Semiconductor

🚀 From learning about Agentic AI to exploring what it can unlock for silicon design. #SamsungSemiconductorIndia Day 2 of IEEE CEDA SWEAT 2026 is here, and as the AI Strategic Partner, Samsung is bringing perspectives, ideas and hands-on experiences to the conversation around the future of semiconductor engineering. 🎤 Hear from our leaders: Manish Goel, VP & Head of System LSI, will deliver the keynote: “AI Agents in Silicon Design: From Intelligence to Autonomy” Amit Jain, VP & Head of SoC, will join the panel: “The Agentic Semiconductor Revolution: Road to 2030 Chip Era” And that’s not all. 👇 🔹 Drop by the Samsung booth to explore our demos 🔹 Experience how AI and semiconductor innovation are coming together 🔹 Interact with our teams and exchange ideas on what’s next If you're at IEEE CEDA SWEAT 2026 today, come meet us, hear from our people, and experience the innovation firsthand. 📍 DoubleTree by Hilton Bengaluru Whitefield 📅 26-27 September 2026 #IEEECEDA #SWEAT2026 #Samsung #AgenticAI #Semiconductors #PreSilicon #SystemLSI #SoC #SiliconDesign #Innovation

🔗 𝗥𝗘𝗚𝗜𝗦𝗧𝗘𝗥 𝗡𝗢𝗪: https://bit.ly/SWEAT2026 🎙️ 𝗔 𝗸𝗲𝘆𝗻𝗼𝘁𝗲 𝗳𝗿𝗼𝗺 𝘁𝗵𝗲 𝘄𝗼𝗿𝗹𝗱 𝗼𝗳 𝗦𝘆𝘀𝘁𝗲𝗺 𝗟𝗦𝗜 𝗮𝗻𝗱 𝗮𝗱𝘃𝗮𝗻𝗰𝗲𝗱 𝗰𝗵𝗶𝗽 𝗱𝗲𝘀𝗶𝗴𝗻 We are honoured to welcome Manish Goel, 𝗩𝗶𝗰𝗲 𝗣𝗿𝗲𝘀𝗶𝗱𝗲𝗻𝘁 𝗮𝗻𝗱 𝗛𝗲𝗮𝗱 𝗼𝗳 𝗦𝘆𝘀𝘁𝗲𝗺 𝗟𝗦𝗜, 𝗦𝗮𝗺𝘀𝘂𝗻𝗴 𝗦𝗲𝗺𝗶𝗰𝗼𝗻𝗱𝘂𝗰𝘁𝗼𝗿 𝗜𝗻𝗱𝗶𝗮 𝗥𝗲𝘀𝗲𝗮𝗿𝗰𝗵, as the 𝗞𝗲𝘆𝗻𝗼𝘁𝗲 𝗦𝗽𝗲𝗮𝗸𝗲𝗿 at 𝗜𝗘𝗘𝗘 𝗖𝗘𝗗𝗔 𝗦𝗪𝗘𝗔𝗧 𝟮𝟬𝟮𝟲. 💡 𝘞𝘪𝘵𝘩 𝘢 𝘭𝘦𝘢𝘥𝘦𝘳 𝘧𝘳𝘰𝘮 𝘚𝘢𝘮𝘴𝘶𝘯𝘨 𝘚𝘦𝘮𝘪𝘤𝘰𝘯𝘥𝘶𝘤𝘵𝘰𝘳 𝘐𝘯𝘥𝘪𝘢 𝘙𝘦𝘴𝘦𝘢𝘳𝘤𝘩, 𝘔𝘢𝘯𝘪𝘴𝘩 𝘎𝘰𝘦𝘭 𝘣𝘳𝘪𝘯𝘨𝘴 𝘱𝘦𝘳𝘴𝘱𝘦𝘤𝘵𝘪𝘷𝘦 𝘧𝘳𝘰𝘮 𝘰𝘯𝘦 𝘰𝘧 𝘵𝘩𝘦 𝘸𝘰𝘳𝘭𝘥'𝘴 𝘭𝘦𝘢𝘥𝘪𝘯𝘨 𝘴𝘦𝘮𝘪𝘤𝘰𝘯𝘥𝘶𝘤𝘵𝘰𝘳 𝘤𝘰𝘮𝘱𝘢𝘯𝘪𝘦𝘴 𝘵𝘰 𝘚𝘞𝘌𝘈𝘛 𝟮𝟬𝟮𝟲. ⚡ 𝗜𝗘𝗘𝗘 𝗖𝗘𝗗𝗔 𝗦𝗪𝗘𝗔𝗧 𝟮𝟬𝟮𝟲 𝘛𝘩𝘦 𝘈𝘨𝘦𝘯𝘵𝘪𝘤 𝘚𝘦𝘮𝘪𝘤𝘰𝘯𝘥𝘶𝘤𝘵𝘰𝘳 𝘙𝘦𝘷𝘰𝘭𝘶𝘵𝘪𝘰𝘯 📅 𝟮𝟳 𝗦𝗲𝗽𝘁𝗲𝗺𝗯𝗲𝗿 𝟮𝟬𝟮𝟲 | 𝗦𝘂𝗻𝗱𝗮𝘆 📍 𝗗𝗼𝘂𝗯𝗹𝗲𝗧𝗿𝗲𝗲 𝗦𝘂𝗶𝘁𝗲𝘀 𝗯𝘆 𝗛𝗶𝗹𝘁𝗼𝗻, 𝗪𝗵𝗶𝘁𝗲𝗳𝗶𝗲𝗹𝗱, 𝗕𝗲𝗻𝗴𝗮𝗹𝘂𝗿𝘂 🕘 𝟵:𝟬𝟬 𝗔𝗠 – 𝟱:𝟬𝟬 𝗣𝗠 🌐 𝗛𝗲𝗮𝗿 𝗳𝗿𝗼𝗺 𝗶𝗻𝗱𝘂𝘀𝘁𝗿𝘆. 𝗘𝘅𝗽𝗹𝗼𝗿𝗲 𝗻𝗲𝘄 𝗶𝗱𝗲𝗮𝘀. 𝗘𝗻𝗴𝗮𝗴𝗲 𝘄𝗶𝘁𝗵 𝘁𝗵𝗲 𝗳𝘂𝘁𝘂𝗿𝗲 𝗼𝗳 𝘀𝗲𝗺𝗶𝗰𝗼𝗻𝗱𝘂𝗰𝘁𝗼𝗿 𝗶𝗻��𝗼𝘃𝗮𝘁𝗶𝗼𝗻. #IEEECEDA #SWEAT2026 #SamsungSemiconductor #SystemLSI #Semiconductors IEEE Council on Electronic Design Automation IEEE CEDA Bangalore Chapter IEEE BANGALORE SECTION Samsung Semiconductor Sajal Mittal | Dr Jalaja S | Girish Aramanekoppa Subbarao | Vamsi Krishna Addepalli | Dr. Vimala P | Apratim Chatterjee | Lakshmanan Balasubramanian | Dr. Venugopal.N PESU N | Dr.Mohammad Ziaullah | Saptarshi Pani | Dr. Jyothi A P | Dr.Mohammed Zakir B | Dr. Nagaraja Rao | Rajeela Deshpande | Dr. Arun Balodi | Dr. Karunakara Rai | Garima Srivastava | Dr. Renuka V Tali | Aloke Das | Dr.Parameshachari B D | Vishal Shah | Ian O'Connor | Cristiana Bolchini | Iris Hui-Ru Jiang | Prasant Misra, Ph.D. | Gaurav Pathak

Under this year’s harvest moon, our hearts grow as full as the moon itself. 🌕 Chuseok is a time to return home, gather around the table with family, share delicious food, and give thanks for all that has carried us here. As we honor traditions passed down through generations, we celebrate the joy of being together and the abundance of the season. What's one thing you'd love to do with your family this Chuseok? Share it in the comments. 🤝 #SamsungSemiconductor #Chuseok

AI Infra Summit is officially in the books! From tackling the critical scaling challenges of next-gen AI to breaking through the notorious memory bottleneck, this year’s sessions sparked incredible discussions. And our Samsung experts were proud to be key contributors to the conversations happening onsite.   On Day 1, Vasanthi Jagatha teamed up with Alexander Nails of SGLang to present "Composable AI Memory Node for Inference at Scale," showcasing Samsung’s innovative AI Memory Node™ solution that, in a production deployment, showed 5-7X improvement in key token metrics and KV utilization with half the GPU nodes. This performance and cost optimization was achieved by intelligently composing Samsung’s cutting-edge memory technologies like DRAM and CXL through an open memory orchestration layer called Samsung Cognos. Indong Kim took the stage on Day 2 to present "Driving the Wave of AI Revolution: 3D Innovations in Memory & Storage Architecture." His session outlined accelerated industry timelines and power and thermal challenges for HBM while discussing how our 3D innovations in memory (#zHBM) and storage (#zNAND-O) are being co-designed to support the next generation of AI workloads, closing with a powerful reminder that Samsung is uniquely positioned as a complete IDM solution provider.   Then Hokyoon Lee joined industry leaders from OpenAI and Intel on a high-impact panel hosted by Rambus. The discussion underscored a central truth: While compute power continues to climb, memory bandwidth remains the primary bottleneck for AI accelerators, particularly for low-latency inference. Samsung continues to explore how memory and storage can advance AI systems, while developing future solutions optimized for each layer of the memory hierarchy.   #AIInfraSummit #SamsungMemory #ArchitectingAISystems #SamsungSemiUS

Industry-First LPDDR6 Validation for Next-Generation Mobile AI 🏆 Samsung Semiconductor has successfully validated its 16GB LPDDR6, operating at I/O speeds of up to 10.7Gbps, on Qualcomm Technologies' Snapdragon® 8 Elite Extreme Gen 6, supporting increasingly sophisticated on-device AI with higher bandwidth, improved power efficiency, and enhanced reliability. ✅ Up to 114GB/s memory bandwidth Enhanced I/O speeds and architectural improvements help accelerate data movement between memory and the processor for AI workloads. ✅ Up to 21% improved power efficiency Dynamic Voltage and Frequency Scaling (DVFS) and Dynamic Efficiency Mode (DEM) optimize power consumption as workload demands change. ✅ Enhanced memory reliability Per-Row Activation Counting (PRAC) monitors repeated activation patterns to help mitigate potential data interference. The validation marks the industry's first LPDDR6 validation on Qualcomm Technologies' next-generation mobile platform, reflecting close collaboration between Samsung Semiconductor and Qualcomm Technologies. 🔗 Learn more about the LPDDR6 innovations behind this industry-first validation. #SamsungSemiconductor #LPDDR6 #OnDeviceAI #AI #Memory

✨ Stacking Higher, Storing More: V10 BV-NAND As AI systems continue to generate more data, storage density is becoming increasingly important. V10 BV-NAND (Bonding V-NAND) is Samsung Semiconductor's 10th-generation V-NAND architecture that applies wafer bonding to enable an ultra-high-layer structure with more than 400 layers. Unveiled at FMS 2026, V10 BV-NAND introduces a new approach to scaling storage density, performance, and efficiency for the AI era. ✅ More than 400 layers enabled through a new wafer-bonding architecture ✅ 58% higher bit density than V9 ✅ 45% faster program time and 33% faster I/O speed than V9 ✅ Greater storage capacity within the same footprint for thinner devices and more efficient data centers At FMS 2026, V10 BV-NAND marks the next step in Samsung’s V-NAND evolution, 13 years after introducing the industry’s first V-NAND technology. By enabling more storage within the same footprint, V10 BV-NAND reflects Samsung Semiconductor’s vision for the next generation of storage. #SamsungSemiconductor #V10BVNAND #VNAND #NAND #FMS2026

✨ High-Performance Storage for On-Device AI: zNAND-O As AI workloads expand from cloud servers to smartphones, PCs, and other edge devices, on-device AI requires high-performance storage within limited space and power constraints. zNAND-O is Samsung Semiconductor's high-performance NAND solution designed for on-device AI. Using TSV-based 3D packaging, zNAND-O vertically stacks multiple V-NAND chips to enable high storage density, high data-loading bandwidth. At FMS 2026, Samsung Semiconductor introduced zNAND-O as an industry-first concept model for high-performance storage. ✅ Shorter data paths for faster, more power-efficient processing ✅ TSV-based 3D packaging with vertically stacked V-NAND chips ✅ <3 μs read latency, 200-400 GB/s bandwidth, and 1 TB/s+ through multi-chip stacking As AI moves closer to the device, zNAND-O reflects Samsung Semiconductor's vision for bringing high-performance data access closer to where AI is processed. 🔍 Discover more memory innovations unveiled by Samsung Semiconductor at FMS 2026. #SamsungSemiconductor #zNAND_O #OnDeviceAI #NAND #FMS2026

🔗 𝗥𝗘𝗚𝗜𝗦𝗧𝗘𝗥 𝗡𝗢𝗪: https://bit.ly/SWEAT2026 🎤 𝗪𝗵𝗲𝗻 𝗔𝗜 𝗶𝘀 𝗿𝗲𝗱𝗲𝗳𝗶𝗻𝗶𝗻𝗴 𝘀𝗲𝗺𝗶𝗰𝗼𝗻𝗱𝘂𝗰𝘁𝗼𝗿 𝗱𝗲𝘀𝗶𝗴𝗻, 𝘁𝗵𝗲 𝗿𝗶𝗴𝗵𝘁 𝗰𝗼𝗻𝘃𝗲𝗿𝘀𝗮𝘁𝗶𝗼𝗻 𝗺𝗮𝘁𝘁𝗲𝗿𝘀. We are pleased to have Amit Jain, 𝗩𝗶𝗰𝗲-𝗣𝗿𝗲𝘀𝗶𝗱𝗲𝗻𝘁, 𝗦𝗮𝗺𝘀𝘂𝗻𝗴 𝗦𝗲𝗺𝗶𝗰𝗼𝗻𝗱𝘂𝗰𝘁𝗼𝗿 𝗜𝗻𝗱𝗶𝗮 𝗥𝗲𝘀𝗲𝗮𝗿𝗰𝗵, join 𝗜𝗘𝗘𝗘 𝗖𝗘𝗗𝗔 𝗦𝗪𝗘𝗔𝗧 𝟮𝟬𝟮𝟲 as a 𝗣𝗮𝗻𝗲𝗹𝗶𝘀𝘁. 💬 𝘐𝘯 𝘵𝘩𝘪𝘴 𝘱𝘢𝘯𝘦𝘭, 𝘪𝘯𝘥𝘶𝘴𝘵𝘳𝘺 𝘱𝘦𝘳𝘴𝘱𝘦𝘤𝘵𝘪𝘷𝘦𝘴 𝘮𝘦𝘦𝘵 𝘦𝘮𝘦𝘳𝘨𝘪𝘯𝘨 𝘪𝘥𝘦𝘢𝘴 𝘢𝘳𝘰𝘶𝘯𝘥 𝘈𝘐 𝘢𝘯𝘥 𝘵𝘩𝘦 𝘧𝘶𝘵𝘶𝘳𝘦 𝘰𝘧 𝘴𝘦𝘮𝘪𝘤𝘰𝘯𝘥𝘶𝘤𝘵𝘰𝘳 𝘪𝘯𝘯𝘰𝘷𝘢𝘵𝘪𝘰𝘯. ⚡ 𝗜𝗘𝗘𝗘 𝗖𝗘𝗗𝗔 𝗦𝗪𝗘𝗔𝗧 𝟮𝟬𝟮𝟲 𝘛𝘩𝘦 𝘈𝘨𝘦𝘯𝘵𝘪𝘤 𝘚𝘦𝘮𝘪𝘤𝘰𝘯𝘥𝘶𝘤𝘵𝘰𝘳 𝘙𝘦𝘷𝘰𝘭𝘶𝘵𝘪𝘰𝘯 📅 𝟮𝟳 𝗦𝗲𝗽𝘁𝗲𝗺𝗯𝗲𝗿 𝟮𝟬𝟮𝟲 | 𝗦𝘂𝗻𝗱𝗮𝘆 📍 𝗗𝗼𝘂𝗯𝗹𝗲𝗧𝗿𝗲𝗲 𝗦𝘂𝗶𝘁𝗲𝘀 𝗯𝘆 𝗛𝗶𝗹𝘁𝗼𝗻, 𝗪𝗵𝗶𝘁𝗲𝗳𝗶𝗲𝗹𝗱, 𝗕𝗲𝗻𝗴𝗮𝗹𝘂𝗿𝘂 🕘 𝟵:𝟬𝟬 𝗔𝗠 – 𝟱:𝟬𝟬 𝗣𝗠 🚀 𝗕𝗿𝗶𝗻𝗴 𝘆𝗼𝘂𝗿 𝗾𝘂𝗲𝘀𝘁𝗶𝗼𝗻𝘀. 𝗝𝗼𝗶𝗻 𝘁𝗵𝗲 𝗰𝗼𝗻𝘃𝗲𝗿𝘀𝗮𝘁𝗶𝗼𝗻. #IEEECEDA #SWEAT2026 #AgenticAI #Semiconductors #ChipDesign IEEE Council on Electronic Design Automation IEEE CEDA Bangalore Chapter IEEE BANGALORE SECTION Samsung Semiconductor Sajal Mittal | Dr Jalaja S | Girish Aramanekoppa Subbarao | Vamsi Krishna Addepalli | Dr. Vimala P | Apratim Chatterjee | Lakshmanan Balasubramanian | Dr. Venugopal.N PESU N | Dr.Mohammad Ziaullah | Saptarshi Pani | Dr. Jyothi A P | Dr.Mohammed Zakir B | Dr. Nagaraja Rao | Rajeela Deshpande | Dr. Arun Balodi | Dr. Karunakara Rai | Garima Srivastava | Dr. Renuka V Tali | Aloke Das | Dr.Parameshachari B D | Vishal Shah | Ian O'Connor | Cristiana Bolchini | Iris Hui-Ru Jiang | Prasant Misra, Ph.D. | Gaurav Pathak

💡What if AI could help make semiconductor materials research faster and more scalable? Developing new semiconductor materials requires extensive time and resources. As AI-driven approaches gain traction in materials research and simulation, the ability to efficiently explore more materials and complex atomic structures is becoming increasingly important. EquFlashV2, developed by Samsung, is an E(3)-equivariant Machine Learning Force Field (MLFF) designed to accelerate atomic-level materials simulation. By making large-scale atomistic simulations more practical, EquFlashV2 opens up new possibilities for exploring and evaluating semiconductor materials—and for advancing AI-driven approaches to semiconductor R&D. 👇Read the full article below to learn more about EquFlashV2 and the possibilities of AI-driven materials simulation. #SamsungSemiconductor #AI #AtomisticSimulation #SemiconductorMaterials #AIforMaterials

✨ Bringing Memory Closer to Compute: zHBM As AI models continue to grow, moving data between memory and AI accelerators is becoming a critical challenge for next-generation AI infrastructure. zHBM is Samsung Semiconductor's next-generation 3D memory architecture that vertically stacks HBM directly above the AI accelerator, reducing the distance between memory and compute. At FMS 2026, Samsung Semiconductor unveiled the industry's first zHBM concept model, introducing a new approach to scaling AI performance, efficiency, and system integration. ✅ Up to 8× higher performance than HBM5 through zHBM's next-generation interface system ✅ More than 10× the memory density of HBM5 enabled by wafer-bonded zHBM architecture ✅ 3× higher energy efficiency and over 50% lower thermal resistance than HBM5 ✅ Customized IP integration between memory and AI accelerators for customer-specific architectures By bringing memory closer to compute, zHBM represents Samsung Semiconductor's vision for the next generation of AI memory architecture. 🚀 Explore Samsung Semiconductor's latest innovations unveiled at FMS 2026. #SamsungSemiconductor #zHBM #HBM #AIInfrastructure #FMS2026