惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

S
SegmentFault 最新的问题
Google DeepMind News
Google DeepMind News
G
Google Developers Blog
Martin Fowler
Martin Fowler
MongoDB | Blog
MongoDB | Blog
月光博客
月光博客
Jina AI
Jina AI
宝玉的分享
宝玉的分享
人人都是产品经理
人人都是产品经理
D
DataBreaches.Net
V
V2EX
WordPress大学
WordPress大学
T
The Blog of Author Tim Ferriss
Last Week in AI
Last Week in AI
B
Blog
博客园 - 叶小钗
小众软件
小众软件
Stack Overflow Blog
Stack Overflow Blog
P
Proofpoint News Feed
A
About on SuperTechFans
J
Java Code Geeks
Cyber Security Advisories - MS-ISAC
Cyber Security Advisories - MS-ISAC
Y
Y Combinator Blog
Microsoft Security Blog
Microsoft Security Blog

SiliconANGLE

Will agentic AI governance run amok? The lesson of Asimov’s Three Laws - SiliconANGLE AI + quantum, Amazon vs. Starlink and the wide-open US-China internet battle - SiliconANGLE Team Cymru launches Total Insights Feed to replace legacy threat intelligence lists - SiliconANGLE AI Mode in Chrome adds split-screen view to enhance the web search experience - SiliconANGLE Resolve AI raises $40M at $1.5B valuation to optimize production environments - SiliconANGLE How Zscaler and OpenAI turn zero-trust security into an AI accelerator - SiliconANGLE OpenAI ratchets up Codex's agentic capabilities to rival Claude Code - SiliconANGLE Anthropic launches Claude Opus 4.7 with coding, visual reasoning improvements - SiliconANGLE Slash raises $100M at a $1.4B valuation to expand AI-powered banking platform for online businesses - SiliconANGLE Canva unveils Canva AI 2.0, recasting its platform as an agentic system for work - SiliconANGLE Data center, consumer device chips boost TSMC’s revenue - SiliconANGLE Mission-critical security cannot be bolted on, says Oracle - SiliconANGLE Agentic infrastructure reshapes enterprise AI - SiliconANGLE Data quality, and data freedom, foundational for AI success - SiliconANGLE Data trust is a bedrock in successful, scalable AI outcomes - SiliconANGLE Google introduces new agentic AI-ready tools and resources for Android developers  - SiliconANGLE Agentic AI orchestration separates winners from laggards - SiliconANGLE Data-driven tools turning the tide against human trafficking - SiliconANGLE Achieving trusted AI development goes beyond 'vibes' - SiliconANGLE Impinj boosts edge computing power in updated R700 RAIN RFID reader - SiliconANGLE Certinia powers professional services with AI - SiliconANGLE Antioch prepares to accelerate simulated testing for autonomous robots after raising $8.5M - SiliconANGLE Developer tooling startup Expo nabs $45M investment - SiliconANGLE Solidroad lands $25M to bring AI to customer support interactions - SiliconANGLE DuploCloud lands compliance and AI governance certifications as enterprise buyers tighten scrutiny - SiliconANGLE Lua lands $5.8M to help businesses build and manage AI agent workforces - SiliconANGLE Best of frenemies: Oracle's and AWS' clouds unite with dedicated, private connectivity - SiliconANGLE NIST shifts National Vulnerability Database to risk-based triage as CVE submissions hit record levels - SiliconANGLE Cisco goes to the races with new Churchill Downs multiyear partnership - SiliconANGLE Susecon 2026 will tackle the future of open-source platforms - SiliconANGLE
Rambus targets agentic AI workloads with faster client me...
by Duncan Riley · 2026-05-27 · via SiliconANGLE

Rambus targets agentic AI workloads with faster client memory chipset

Rambus Inc. today announced a complete DDR5 9600 client memory module chipset designed to push PC memory speeds to 9,600 megatransfers per second, targeting the bandwidth and capacity demands of agentic artificial intelligence workloads on desktops and laptops.

The new offering from the chip and silicon intellectual property company combines a second-generation client clock driver, dubbed CKD02, with a PMIC5120 power management integrated circuit and a serial presence detect hub. Together, the three chips provide what Rambus describes as a complete solution for clocked DDR5 modules operating between 8,000 and 9,600 MT/s.

The chipset is built for the emerging CUDIMM and CQDIMM module formats used in desktops and the CSODIMM format used in laptops. Those clocked module designs are intended to overcome the signal integrity, clock jitter and timing instability problems that emerge when DDR5 memory is pushed beyond 6,400 MT/s. An on-module clock driver conditions and redistributes the clock signal across the dynamic random-access memory chips on the module, allowing higher data rates without the degradation that would otherwise occur.

Rambus argues the timing aligns with a shift in how PCs handle AI. Agentic systems, which plan, execute and adapt workflows in real time, require persistent context, concurrent processing and continuous data movement between the processor and system memory. That translates to sharp increases in both bandwidth and capacity at the module level, the company said.

“Agentic workloads are fundamentally more memory-hungry, driving the need for higher memory bandwidth, greater capacity and improved efficiency in AI-enabled PCs,” said Rami Sethi, senior vice president and general manager of memory interface chips at Rambus. “Our DDR5 9600 Client Chipset, featuring the Gen2 Client Clock Driver, delivers the performance foundation needed to enable this new era of intelligent, high-performance client systems for AI-driven productivity, next-generation gaming and professional content creation.”

Within the chipset, the Gen2 client clock driver retimes, conditions and distributes the clock signal sent from the processor to the DRAM devices on the module. The PMIC5120 steps down system voltage to the levels needed to power the DRAM and the other active chips on the module. The SPD hub handles module identification, configuration and telemetry communication.

Jeff Janukowicz, research vice president at International Data Corp., said memory is becoming a bottleneck as AI workloads spread across client devices. Clocked architectures like CUDIMM and CSODIMM are how the industry is dealing with the signal integrity and timing problems that show up at higher data rates, he added, and full chipset solutions that hold stable at those speeds will be a factor in how quickly next-generation AI PCs catch on.

The launch extends Rambus’ memory module chipset portfolio from server platforms into client systems, including next-generation AI PCs, notebooks and workstations.

Image: Rambus

A message from John Furrier, co-founder of SiliconANGLE:

Support our mission to keep content open and free by engaging with theCUBE community. Join theCUBE’s Alumni Trust Network, where technology leaders connect, share intelligence and create opportunities.

  • 15M+ viewers of theCUBE videos, powering conversations across AI, cloud, cybersecurity and more
  • 11.4k+ theCUBE alumni — Connect with more than 11,400 tech and business leaders shaping the future through a unique trusted-based network.

About SiliconANGLE Media

SiliconANGLE Media is a recognized leader in digital media innovation, uniting breakthrough technology, strategic insights and real-time audience engagement. As the parent company of SiliconANGLE, theCUBE Network, theCUBE Research, CUBE365, theCUBE AI and theCUBE SuperStudios — with flagship locations in Silicon Valley and the New York Stock Exchange — SiliconANGLE Media operates at the intersection of media, technology and AI.

Founded by tech visionaries John Furrier and Dave Vellante, SiliconANGLE Media has built a dynamic ecosystem of industry-leading digital media brands that reach 15+ million elite tech professionals. Our new proprietary theCUBE AI Video Cloud is breaking ground in audience interaction, leveraging theCUBEai.com neural network to help technology companies make data-driven decisions and stay at the forefront of industry conversations.