Envisioned by the leading engineers and scholars from Samsung and Harvard University, the insight was published as a Perspective paper, titled ‘Neuromorphic electronics based on copying and pasting the brain’, by Nature Electronics. Donhee Ham, Fellow of Samsung Advanced Institute of Technology (SAIT) and Professor of Harvard University, Professor Hongkun Park of Harvard University, Sungwoo Hwang, President and CEO of Samsung SDS and former Head of SAIT, and Kinam Kim, Vice Chairman and CEO of Samsung Electronics are the co-corresponding authors.

Image of rat neurons on CNEA (CMOS nanoelectrode array).

The essence of the vision put forward by the authors is best summed up by the two words, ‘copy’ and ‘paste’. The paper suggests a way to copy the brain’s neuronal connection map using a breakthrough nanoelectrode array developed by Dr. Ham and Dr. Park, and to paste this map onto a high-density three-dimensional network of solid-state memories, the technology for which Samsung has been a world leader.

Through this copy and paste approach, the authors envision to create a memory chip that approximates the unique computing traits of the brain – low power, facile learning, adaptation to environment, and even autonomy and cognition – that have been beyond the reach of current technology.

The brain is made up of a large number of neurons, and their wiring map is responsible for the brain’s functions. Thus the knowledge of the map is the key to reverse engineering the brain.

While the original goal of neuromorphic engineering, launched in the 1980s, was to mimic such structure and function of the neuronal networks on a silicon chip, it proved difficult because, even until now, little is known of how the large number of neurons are wired together to create the brain’s higher functions. Thus, the goal of neuromorphic engineering has been eased to designing a chip ‘inspired’ by the brain rather than rigorously mimicking it.

This paper suggests a way to return to the original neuromorphic goal of the brain reverse engineering. The nanoelectrode array can effectively enter a large number of neurons so it can record their electrical signals with high sensitivity. These massively parallel intracellular recordings inform the neuronal wiring map, indicating where neurons connect with one another and how strong these connections are. Hence from these telltale recordings, the neuronal wiring map can be extracted, or ‘copied’.

The copied neuronal map can then be ‘pasted’ to a network of non-volatile memories – such as commercial flash memories that are used in our everyday life in solid-state drives (SSD), or ‘new’ memories such as resistive random access memories (RRAM) – by programming each memory so that its conductance represents the strength of each neuronal connection in the copied map.

(From the left) Donhee Ham, Fellow of Samsung Advanced Institute of Technology (SAIT) and Professor of Harvard University, Hongkun Park, Professor of Harvard University, Sungwoo Hwang, President and CEO of Samsung SDS (former Head of SAIT) and Kinam Kim, Vice Chairman and CEO of Samsung Electronics, the co-corresponding authors.

The paper takes a step further and suggests a strategy to rapidly paste the neuronal wiring map onto a memory network. A network of specially-engineered non-volatile memories can learn and express the neuronal connection map, when directly driven by the intracellularly recorded signals. This is a scheme that directly downloads the brain’s neuronal connection map onto the memory chip.

Since the human brain has an estimated 100 billion or so neurons, and a thousand or so times more synaptic connections, the ultimate neuromorphic chip will require 100 trillion or so memories. Integrating such vast number of memories on a single chip would be made possible by 3D integration of memories, the technology led by Samsung that opened up a new era for memory industry.

Leveraging its leading experience in chip manufacturing, Samsung is planning to continue its research into neuromorphic engineering, in order to extend Samsung’s leadership in the field of next generation AI semiconductors.

“The vision we present is highly ambitious,” said Dr. Ham. “But working toward such a heroic goal will push the boundaries of machine intelligence, neuroscience, and semiconductor technology.”

First Integration of HBM-PIM Into an AI Accelerator

In February, Samsung introduced the industry’s first HBM-PIM (Aquabolt-XL), which incorporates the AI processing function into Samsung’s HBM2 Aquabolt, to enhance high-speed data processing in supercomputers and AI applications. The HBM-PIM has since been tested in the Xilinx Virtex Ultrascale+ (Alveo) AI accelerator, where it delivered an almost 2.5X system performance gain as well as more than a 60% cut in energy consumption.

“HBM-PIM is the industry’s first AI-tailored memory solution being tested in customer AI-accelerator systems, demonstrating tremendous commercial potential,” said Nam Sung Kim, senior vice president of DRAM Product & Technology at Samsung Electronics. “Through standardization of the technology, applications will become numerous, expanding into HBM3 for next-generation supercomputers and AI applications, and even into mobile memory for on-device AI as well as for memory modules used in data centers.”

“Xilinx has been collaborating with Samsung Electronics to enable high-performance solutions for data center, networking and real-time signal processing applications starting with the Virtex UltraScale+ HBM family, and recently introduced our new and exciting Versal HBM series products,” said Arun Varadarajan Rajagopal, senior director, Product Planning at Xilinx, Inc. “We are delighted to continue this collaboration with Samsung as we help to evaluate HBM-PIM systems for their potential to achieve major performance and energy-efficiency gains in AI applications.”

DRAM Modules Powered by PIM

The Acceleration DIMM (AXDIMM) brings processing to the DRAM module itself, minimizing large data movement between the CPU and DRAM to boost the energy efficiency of AI accelerator systems. With an AI engine built inside the buffer chip, the AXDIMM can perform parallel processing of multiple memory ranks (sets of DRAM chips) instead of accessing just one rank at a time, greatly enhancing system performance and efficiency. Since the module can retain its traditional DIMM form factor, the AXDIMM facilitates drop-in replacement without requiring system modifications. Currently being tested on customer servers, the AXDIMM can offer approximately twice the performance in AI-based recommendation applications and a 40% decrease in system-wide energy usage.

“SAP has been continuously collaborating with Samsung on their new and emerging memory technologies to deliver optimal performance on SAP HANA and help database acceleration,” said Oliver Rebholz, head of HANA core research & innovation at SAP. “Based on performance projections and potential integration scenarios, we expect significant performance improvements for in-memory database management system (IMDBMS) and higher energy efficiency via disaggregated computing on AXDIMM. SAP is looking to continue its collaboration with Samsung in this area.”

Mobile Memory That Brings AI From Data Center to Device

Samsung’s LPDDR5-PIM mobile memory technology can provide independent AI capabilities without data center connectivity. Simulation tests have shown that the LPDDR5-PIM can more than double performance while reducing energy usage by over 60% when used in applications such as voice recognition, translation and chatbot.

Energizing the Ecosystem

Samsung plans to expand its AI memory portfolio by working with other industry leaders to complete standardization of the PIM platform in the first half of 2022. The company will also continue to foster a highly robust PIM ecosystem in assuring wide applicability across the memory market.

In a keynote speech at the 2021 Global Semiconductor Alliance (GSA) Memory+ Conference, Jinman Han, Executive Vice President and Head of Memory Global Sales and Marketing at Samsung Electronics, proclaimed that Samsung was ready to work with the entire industry to develop next-generation memory solutions that will be needed in the near future. The GSA Memory+ Conference is the Global Semiconductor Alliance’s premier event, which brings the global memory, logic and system design communities together to discuss advancing memory and system architectures.

The memory industry has long focused its efforts on satisfying demand for larger capacities and faster speeds, as well as higher bandwidth. However, in order to keep up with growing demand, the industry will need to look beyond individual innovations and assess the bigger picture. Samsung is committed to collaborating with the industry in this area and intends to lead the development of next-generation solutions.

The massive and ongoing rise in data has exposed a need for revolutionary change in memory computation and subsystems. In response, Samsung has been leading the development of technologies that can be used to give existing system architectures a much-needed update, including HBM-PIM, AXDIMM, Smart SSDs and CXL-based DRAM.

Having been the first in the memory industry to utilize EUV and HKMG process technology, Samsung is uniquely positioned to introduce in a new era of powerful memory innovations. The company has been optimizing thermal performance by advancing innovations like Hybrid Copper Bonding (HCB), while its V-NAND technologies continue to redefine how memory is stacked.

“I believe our industry will rise to this challenge not by innovating independently, but by finding solutions together,” said Han. “Going forward, we’ll develop a new IT ecosystem together that is much more powerful and sustainable, one which will be needed to architect the digital future.”

Click on the video below to watch the keynote.

Samsung Electronics, the world leader in advanced memory technology, today announced that it has begun mass producing its latest smartphone memory solution, the LPDDR5 UFS-based multichip package (uMCP). Samsung’s uMCP integrates the fastest LPDDR5 DRAM with the latest UFS 3.1 NAND flash, delivering flagship-level performance to a much broader range of smartphone users.

“Samsung’s new LPDDR5 uMCP is built upon our rich legacy of memory advancements and packaging know-how, enabling consumers to enjoy uninterrupted streaming, gaming and mixed reality experiences even in lower-tier devices,” said Young-soo Sohn, vice president of the Memory Product Planning Team at Samsung Electronics. “As 5G-compatible devices become more mainstream, we anticipate that our latest multichip package innovation will accelerate the market transition to 5G and beyond, and help to bring the metaverse into our everyday lives a lot faster.”

Based on the latest mobile DRAM and NAND interfaces, Samsung’s uMCP can deliver lightning-fast speed and high storage capacity at very low power. This combination will allow more consumers to immerse themselves in numerous 5G applications that were previously only available on premium flagship models, including advanced photography, graphics-intensive gaming and augmented reality (AR). Such flagship-level capabilities are made possible by a nearly 50% improvement in DRAM performance, from 17 gigabytes per second (GB/s) to 25GB/s, and a doubling of NAND flash performance, from 1.5GB/s to 3GB/s, over the previous LPDDR4X-based UFS 2.2 solution.

The new uMCP also helps to maximize space efficiency within a smartphone by integrating DRAM and NAND storage into a single compact package that measures only 11.5mm x 13mm, allowing more space for other features. With DRAM capacities ranging from 6 gigabytes (GB) to 12GB and storage options from 128GB to 512GB, the Samsung uMCP can be easily customized to accommodate the diverse needs of 5G smartphones throughout the mid- and high-end segments.

Samsung has successfully completed compatibility testing of the LPDDR5 uMCP with several global smartphone manufacturers, and expects its uMCP-equipped devices to hit mainstream markets starting this month.

At the heart of how this data remains saved for us to access anytime is Samsung Electronics’ market-leading semiconductor solutions, and so in order to raise awareness of just how crucial a role these chips play in our day-to-day lives, Samsung is launching its all-new SAVE campaign.

The word ‘save’ has many meanings: ‘to store’, ‘to gather’, ‘to cherish’, ‘to protect’, ‘to recover’ and ‘to liberate’ being just a few. Accordingly, the SAVE campaign’s message is ‘Samsung Memory Saves All the Progress in the World.’ This slogan summarizes Samsung’s commitment to contributing to society in ways that benefit everyone and bring about a better world for all through technology solutions that preserve the multifarious values and experiences of societies from all corners in the world.

In order to raise awareness of the role of the company’s memory solutions in everyday life, Samsung has collaborated with world-renowned K-pop artist NCT 127 to release a powerful single and music video, ‘SAVE.’

Furthermore, Samsung has released a dedicated campaign film showcasing the long-term vision and values that drive Samsung Memory.

• SAVE Campaign interview
• SAVE Music video

Samsung has also created a new, modern design for the SAVE campaign based on the iconic ‘save’ symbol in order to highlight the importance of saving and put forward the idea of saving for a new generation.

Combining elements of the original floppy disc shape, today’s SSD solutions, modern SD card design and the ‘download’ arrow symbol, this new design brings together saving solutions old and new to operate as a universal symbol for memory – and the concept of saving.

Stay tuned for more information on the SAVE campaign and to learn more about how Samsung is re-defining the word ‘save’ to bring new memory solutions to more people.

For further information about SAVE campaign, visit the page below.

www.samsung.com/semiconductor/sustainability/save-all-the-progress/

“As the leading provider of SAS storage for a decade, Samsung has been offering the most advanced and reliable enterprise solutions in full support of the critical workloads of global server OEMs, governments and financial institutions. Samsung enterprise solutions are also accredited by the U.S. National Institute of Standards and Technology for the most powerful data security,” said Kwangil Park, senior vice president of the Memory Product Planning Team at Samsung Electronics. “Like we have done with the PM1653, Samsung will continue to collaborate with our customers to accommodate the ever-growing demand of the enterprise server market for the most uncompromising offerings available.”

In order to build a robust 24G SAS ecosystem across the enterprise market, host bus adapter (HBA) system readiness is also essential. Samsung has been closely working with Broadcom, a leading HBA provider, to ensure a seamless and timely transition to this new SAS milestone.

“Broadcom values the innovation that 24G SAS products, such as the Samsung SSD PM1653, will bring to the enterprise server market,” said Jas Tremblay, vice president and general manager of the Data Center Solutions Group at Broadcom. “The combination of the PM1653 SSD and Broadcom’s next-generation SAS RAID products delivers up to 5X RAID 5 performance; which will be critical to meet the strong demands of the 24G ecosystem.”

The PM1653 has been optimized for high-performance enterprise servers, offering the industry’s highest random read speed – a key metric for server storage performance – of up to 800K IOPS. Its sequential read speed can reach 4,300MB/s, which is the maximum available speed for the 24G SAS interface and twice the speed of the previous-generation PM1643a drive.

Leveraging its dual-port system, the new SSD will give enterprise server OEMs the flexibility of using one or both ports depending on their system environment. Even if one port was to experience a failure during operation, data can be transferred and accessed through the other port with superior enterprise-grade reliability. The PM1653 is capable of supporting 24G and the legacy 12G SAS-3 platform as well.

Samsung has begun sampling the PM1653 SSD with select customers, and plans to begin mass production in the second half of this year.