The Everspin blog is our way of communicating to customers, partners, and the industry in general exactly what MRAM is and how it can be used in the development of solutions that range from industrial to enterprise to transportation and beyond. If you have a request for us to address a particular topic, you can use our information request form and suggest a topic in the Describe Your Request section.
Data Persistence with MRAM Over CXL®
MRAM offers unique attributes that span a wide range of use cases. Access time, as well as endurance, can be optimized for four main use cases. For persistent SRAM use cases, MRAM can be optimized for high endurance and fast access. MRAM can also address NOR flash use cases, enabling reliable code execution in computing where MRAM has higher endurance and offers significatnly faster writes than NOR Flash. For configuration or distributed persistent memory use cases, MRAM enables fast access with a limited number of updates. Lastly, MRAM can also address OTP use cases, providing a superior solution to other OTP technologies. All these use cases typically assume a direct attach leveraging one of the legacy interfaces such as parallel or serial memory and flash interfaces. This limits the degrees of freedom and value proposition that MRAM natively offers.
The emerging CXL interface is about to change this as it offers a non-deterministic access paradigm optimal for MRAM as well as protocols enabling computer memory as well as storage solutions all the way to accelerators that can be combined to address the most demanding compute applications, especially in the Artificial Intelligence Domain. Additionally, the CXL interface allows for the MRAM to be disaggregated and shared with a large number of compute nodes on the CXL bus in e.g. HPC use cases. Furthermore, CXL attached MRAM offers remote persistent memory functionality and can be used to log data, as a write cache or temporarily protect data in flight, eliminating complex power fail mechanisms. In the configuration as a CXL offload accelerator, MRAM may provide persistent memory that can also be used to store weights for AI inferencing applications. Clearly, CXL also has the benefit of seamlessly connecting MRAM into applications over generations, reducing the interface R&D efforts for the device and host. CXL finally offers an interface option that empowers users to deploy MRAM in a wide variety of use cases, enhancing system performance, power efficiency, and reliability.
A potential implementation of CXL persistent memory is described in the following whitepaper, authored by Bill Gervasi of Wolley Inc. with contributions from Joe O’Hare. The concept of a wide IO MRAM can be utilized in many design approaches. Everspin will continue to explore ways to provide a true persistent memory over CXL.
For the complete Article CLICK HERE
Traditional Memory vs. MRAM: Revolutionizing Non-Volatile Memory
In spearheading the semiconductor industry’s resurgence in 2024, the memory market took the lead, surpassing micro, logic, and then analog ICs. Despite lower demand in other sectors, memory products, particularly MRAM, exhibited consistent and positive sales. Industry projections indicate a significant uptick in MRAM adoption, aligning seamlessly with the surge in artificial intelligence (AI) and underscoring the pivotal role memory plays in evolving AI technology. With its remarkable blend of non-volatility, efficiency, and endurance, MRAM is poised to become the universal memory solution, increasingly favored over traditional counterparts for applications in automotive, industrial, aerospace, and data center storage...
For the complete article CLICK HERE
A New Era in Everspin's Persistent Memory Solutions
MRAM has proven itself as the highest performing persistent memory available in the commercial memory industry, combining the best aspects of random access memory with non-volatility. In a world in which protection of mission critical data is paramount, MRAM has stepped in and filled a void that conventional memory types such as SRAM, DRAM and Flash have left open. Everspin’s broad portfolio of MRAM fulfills the need for persistence, endurance, low latency.
Driven by pervasive applications such as data analytics, cloud computing, both terrestrial and extraterrestrial, artificial intelligence (AI), Edge AI including Industrial IoT, the market for persistent memory is projected to grow at a CAGR of 27.5% between 2020 and 2030[1]. This growth calls for a cohesive product strategy to address the market requirements.
Since Everspin introduced our first family of magnetoresistive random-access memory (MRAM) persistent memories, we have been a pioneering force, pushing the boundaries of what’s possible in data storage and processing. Today, we mark the next chapter in our journey with the launch of PERSYST: the new brand for our top-tier persistent memory families.
Everspin’s Road to Embedded World 2024
It’s that time of the year again – Everspin is heading back to Germany for Embedded World on April 9-11! This conference has always been an important gathering for the embedded community, bringing together experts and suppliers from across the globe together. It’s a fun three days, with many presentations, classes and networking opportunities that foster many new relationships and innovative ideas.
Last year, the conference attracted a total of 950 exhibitors and 27,000 visitors. Everspin also exhibited, and we had some very memorable moments. Our team had the opportunity to speak with various industry media outlets, including TechInsights, Embedded Computing Design and EE Times.
This year, we are set to unveil more updates to our persistent memory products, and we’re particularly excited to announce a new brand name on the first day of Embedded World! Not only that, but we have planned some live and static demos of EMxxLX products that you can experience at our booth.
Visit Everspin at Embedded World 2024 (booth 548, Hall 3)!
The Advantages of using MRAM with FPGA’s, a perspective from Sanjeev Aggarwal, President and CEO of Everspin Technologies
Field-programmable gate arrays (FGPAs) are finding their way into many different applications as 5G technology, and other LPWAN connectivity solutions roll out. Automotive, IoT operations, and 5G edge applications are implementing FPGA chips due to their fast time-to-market and their ability to be reprogrammed. In many cases, the FPGA is installed in applications where frequent system updates are needed. Often these bitstream updates are made over the air (OTA. An FPGA with a magnetoresistive random-access memory (MRAM) can simplify the software associated with the management process of updating new bitstreams into memory. System designers can take advantage of MRAM’s faster write capabilities which can lead to lower energy usage. New MRAM developments are underway to combine the unique benefits of MRAM and FPGA.
For the complete article CLICK HERE
Sanjeev Aggarwal Describes the New EMxxLX STT-MRAM Family
The EMxxLX family is the highest performing discrete persistent memory in the industry with up to 400MB/s of read and write bandwidth. We have taken steps to make this product easy to design-in for our customers by using the standard xSPI octal protocol, standard packages, and eliminating any need for endurance wear-leveling or external error correction. It represents the decisive move to make our STT-MRAM technology suitable for industrial markets with enhanced temperature, endurance and data retention as compared to earlier STT-MRAM products. Production availability now means that our customers can take full advantage of the performance and reliability of this new offering in their industrial, aerospace, storage, networking, transportation, IoT and edge, AI, metering, and infrastructure systems that all demand the highest of data integrity over the life of the systems in harsh environments. Additionally, we have matched the command protocol for serial NOR Flash memories which gives customers the option to use MRAM in FPGA based system for very fast configurations and performance advantages with the extremely fast writing capability, 1000x times faster than worst case NOR programming speeds. Announcing production availability of the EMxxLX is a big milestone for Everspin, but more importantly for our customers that will take advantage of the performance, backed by the strong reputation for reliability for our products.
Joe O'Hare Interviewed in Storage Unpacked
This week, Chris and Martin dig deeper into storage media and look at magneto-resistive RAM, commonly known as MRAM in this conversation with Joe O’Hare, Director of Distribution and Product Marketing at Everspin Technologies.
Catching Up on the Latest Developments in MRAM
Electronic Design Editor Bill Wong interviewed Everspin CEO Kevin Conley about the state of MRAM and its track to becoming a universal memory. In this Q&A, Kevin discusses MRAM pricing, the potential of Everspin’s 1 Gb STT-MRAM solution, embedded MRAM, Toggle MRAM and more. To read the full Q&A interview, visit Electronic Design.
Peeling Back the Covers of STT-MRAM
STT-MRAM is a persistent memory solution that enables higher density with the ability to scale while consuming less power. With fast read/write speed similar to DRAM and the ability to handle memory workloads, STT-MRAM is considered the “leading contender for universal memory,” the “world's fastest write speed for embedded memory,” or as we like to say, “the industry’s most durable, highest performance persistent memory.” There’s no question STT-MRAM is highly regarded for its versatility, reliability, speed and endurance, especially for applications where data persistence is paramount, such as enterprise storage, industrial and transportation systems. For example, IBM has adopted Everspin STT-MRAM for use as a persistent write buffer in the flash modules used to build their All Flash Arrays (AFA). Read more...
Choices: Toggle or STT-MRAM?
There are clear benefits to using Toggle MRAM and STT-MRAM which is why Everspin continues to invest in both product lines. In June 2019, we announced production phase for our 1 Gb STT-MRAM product. During the announcement, many media and analysts viewed this as a move away from our Toggle MRAM product line. That couldn’t be further from the truth, evidenced by our recent expansion of our Toggle MRAM solutions. We deliver choices to our customers to meet their broad application requirements. We’re now offering a wide variety of Toggle MRAM densities, with new 2Mb, 8MB and 32Mb Toggle MRAM solutions to accompany our 16MB and other low-density solutions, adding to memory and price point variety. Read more...