HBM3 Memory: Break Through to Greater Bandwidth


Delivering unrivaled memory bandwidth in a compact, high-capacity footprint, has made HBM the memory of choice for AI/ML and other high-performance computing workloads. HBM3 as the latest generation of the standard raises data rates to 6.4 Gb/s and promises to scale even higher. The Rambus HBM3 controller provides industry-leading support of the extended roadmap for HBM3 with performance to 9.6... » read more

DRAM Choices Are Suddenly Much More Complicated


Chipmakers are beginning to incorporate multiple types and flavors of DRAM in the same advanced package, setting the stage for increasingly distributed memory but significantly more complex designs. Despite years of predictions that DRAM would be replaced by other types of memory, it remains an essential component in nearly all computing. Rather than fading away, its footprint is increasing,... » read more

The Power Of HBM3 Memory For AI Training Hardware


AI training data sets are constantly growing, driving the need for hardware accelerators capable of handling terabyte-scale bandwidth. Among the array of memory technologies available, High Bandwidth Memory (HBM) has emerged as the memory of choice for AI training hardware, with the most recent generation, HBM3, delivering unrivaled memory bandwidth. Let’s take a closer look at this important... » read more

Understanding Memory Protection Units


Let’s talk about the various security measures we take to protect our homes. We use locks, install security cameras to detect any suspicious activities, etc., all to safeguard our loved ones and valuable possessions. In similar sense, the tiny microcontrollers inside our devices have protection units (PU) that safeguard the precious resource: memory. One of the most important types of prote... » read more

Closing The Performance Gap Between DRAM And AI Processors


As the workhorse of semiconductor memory, DRAM holds a unique place in the industry thanks to its large storage capacity and ability to feed data and program code to the host processor quickly. Lately, this unsung hero of the circuit board has been taking a backseat to its logic counterparts, as a wave of high-performance FPGAs, CPUs, GPUs, TPUs and custom accelerator ASICs emerges to meet t... » read more

The Limits Of AI-Generated Models


In several recent stories, the subject of models has come up, and one recurrent theme is that AI may be able to help us generate models of a required abstraction. While this may be true in some cases, it is very dangerous in others. If we generalize, AI should be good for any model where the results are predominantly continuous, but discontinuities create problems. Unless those are found and... » read more

Causes Of Memory Unsafety


Memory unsafety is a characteristic of many of today’s systems. The root cause of buffer bounds vulnerabilities such as buffer overflows and over-reads is unsafe programming. Major software vendors consistently report memory unsafety problems. For example, the Chromium open-source browser project has stated that 69% of CVEs (Common Vulnerabilities and Exposures) reported relate to memory... » read more

The Impact Of Channel Hole Profiles On Advanced 3D NAND Structures


In a two-tier 3D NAND structure, the upper and lower channel hole profile can be different, and this combination of different profiles leads to different top-down visible areas. The visible area is the key metric to determine whether the bottom SONO layer can be punched through and ensure that the bit cells connect to the common source line. Performing channel hole profile splits on a silicon w... » read more

Memory And High-Speed Digital Design


As DRAM gets faster, timing constraints, jitter, and signal integrity become harder to control. The real challenge is to understand what can go wrong early in the design process, and that becomes more complex with each new version of memory and higher signal speeds. Stephen Slater, product manager for EDA products at Keysight, talks about how simulation can be applied to these issues, what to t... » read more

Memory Technologies Key To Advancing AI Applications


Memory is an integral component in every computer system, from the smartphones in our pockets to the giant data centers powering the world’s leading-edge AI applications. As AI continues to rise in reach and complexity, the demand for more memory from data center to endpoints is reshaping the industry’s requirements and traditional approaches to memory architectures. According to OpenAI,... » read more

← Older posts Newer posts →