


Born for AI: Samsung says it will put HBM4 video memory into production in 2025, competing for leadership in high-performance computing
The rapid increase in AI computing power in recent years has made computing cards a new sought-after target for major hardware manufacturers. In particular, computing cards launched by companies like NVIDIA are in short supply. In addition to NVIDIA's powerful GPUs, , including Samsung, Hynix and other storage manufacturers do not want to miss this AI feast, especially high-performance computing cards require high-performance graphics memory produced by them. Currently, a senior executive in the storage field of Samsung issued a document saying that Samsung plans to mass-produce in 2025. Produce the latest HBM4 video memory, thereby surpassing Hynix.
In 2016, Samsung officially began mass production of HBM video memory. Compared with GDDR video memory, HBM video memory has larger bandwidth, thereby achieving higher performance transmission. In the consumer market, graphics cards such as AMD's Radeon Fury use HBM video memory, but the sales volume of Fury graphics cards in the consumer market is not large and it has not caused much impact. However, in the field of high-performance computing, HBM video memory has performed well and is highly sought after by customers. According to reports, Hynix currently occupies 50% of the HBM video memory market, Samsung occupies 40%, and Micron occupies 10%
Currently, the latest generation of HBM video memory launched by Samsung is HBM3e, which has a maximum bandwidth of 9.8Gbps. Although this bandwidth far exceeds the needs of the consumer market, it is far from enough for supercomputing GPUs, especially for AI and other technologies. The rapid development of cloud computing has doubled customer demand for high-performance computing. According to previous news, Nvidia plans to launch the next generation computing card Blackwell-B100 in 2025, which will be better than the current H100 in terms of computing power, and will reach unprecedented heights in the field of AI. There is no doubt that Samsung hopes Use Nvidia's next-generation computing card to gain more market share with its HBM4 video memory.
As an old rival of Samsung, Hynix is expected to mass-produce HBM4 video memory in 2026, while Micron has no more news about HBM4 video memory for the time being. If Samsung can achieve mass production of HBM4 video memory one year earlier than its competitors, then there will be no Doubt will take the lead in the high-performance computing market. Of course, for us ordinary consumers, Samsung HBM4 video memory does not make much sense. After all, few consumer-grade graphics cards will use this video memory. Everyone still cares about GDDR7 When will the mass production of video memory be more reliable?
The above is the detailed content of Born for AI: Samsung says it will put HBM4 video memory into production in 2025, competing for leadership in high-performance computing. For more information, please follow other related articles on the PHP Chinese website!

Google is leading this shift. Its "AI Overviews" feature already serves more than one billion users, providing complete answers before anyone clicks a link.[^2] Other players are also gaining ground fast. ChatGPT, Microsoft Copilot, and Pe

In 2022, he founded social engineering defense startup Doppel to do just that. And as cybercriminals harness ever more advanced AI models to turbocharge their attacks, Doppel’s AI systems have helped businesses combat them at scale— more quickly and

Voila, via interacting with suitable world models, generative AI and LLMs can be substantively boosted. Let’s talk about it. This analysis of an innovative AI breakthrough is part of my ongoing Forbes column coverage on the latest in AI, including

Labor Day 2050. Parks across the nation fill with families enjoying traditional barbecues while nostalgic parades wind through city streets. Yet the celebration now carries a museum-like quality — historical reenactment rather than commemoration of c

To help address this urgent and unsettling trend, a peer-reviewed article in the February 2025 edition of TEM Journal provides one of the clearest, data-driven assessments as to where that technological deepfake face off currently stands. Researcher

From vastly decreasing the time it takes to formulate new drugs to creating greener energy, there will be huge opportunities for businesses to break new ground. There’s a big problem, though: there’s a severe shortage of people with the skills busi

Years ago, scientists found that certain kinds of bacteria appear to breathe by generating electricity, rather than taking in oxygen, but how they did so was a mystery. A new study published in the journal Cell identifies how this happens: the microb

At the RSAC 2025 conference this week, Snyk hosted a timely panel titled “The First 100 Days: How AI, Policy & Cybersecurity Collide,” featuring an all-star lineup: Jen Easterly, former CISA Director; Nicole Perlroth, former journalist and partne


Hot AI Tools

Undresser.AI Undress
AI-powered app for creating realistic nude photos

AI Clothes Remover
Online AI tool for removing clothes from photos.

Undress AI Tool
Undress images for free

Clothoff.io
AI clothes remover

Video Face Swap
Swap faces in any video effortlessly with our completely free AI face swap tool!

Hot Article

Hot Tools

WebStorm Mac version
Useful JavaScript development tools

MinGW - Minimalist GNU for Windows
This project is in the process of being migrated to osdn.net/projects/mingw, you can continue to follow us there. MinGW: A native Windows port of the GNU Compiler Collection (GCC), freely distributable import libraries and header files for building native Windows applications; includes extensions to the MSVC runtime to support C99 functionality. All MinGW software can run on 64-bit Windows platforms.

ZendStudio 13.5.1 Mac
Powerful PHP integrated development environment

Zend Studio 13.0.1
Powerful PHP integrated development environment

EditPlus Chinese cracked version
Small size, syntax highlighting, does not support code prompt function
