Nima Badiey

Watching Jensen Huang’s keynote at GTC Paris this morning, one thing is clear: AI infrastructure is now economic infrastructure. And agentic AI is its next great workload. At KX, we’re proud to be building for this new era. Today we launched the AI Banker Agent—our first production-grade Agentic AI Blueprint, built in collaboration with NVIDIA. It brings together: ⚡ NVIDIA NeMo, Nemotron & NIM microservices 📊 KX’s high-performance time series & vector database 🧠 Domain-specific AI workflows for global markets trading This is real-time, applied AI—engineered for the speed, complexity, and compliance rigor of capital markets. 📣 Read the full press release: https://lnkd.in/e_JSnsxJ 🎥 Watch Jensen’s keynote from GTC Paris: https://lnkd.in/ekH_N4CV This is how we make the AI Factory real. #AgenticAI #AIInfrastructure #KX #NVIDIA #GTCParis #AIinFinance #KDBAI #RealTimeAI #FinancialServices #TradingAI #AppliedAI #AITransformation

38

1 Comment

Thrilled to be partnering with Amazon Web Services (AWS) to bring Cerebras speed to all AWS developers. Watch David Brown's video to learn more about our combined solution and get ready to build faster ;)

70

EdgeMode CEO Charlie Faulkner discusses demand for AI Data center infratsructure capacity and how EdgeMode is solving the power supply problem. https://lnkd.in/eDdKN9vj

8

AMPED about this strategic partnership with Intel Corporation and a big thanks to Kevork Kechichian and the whole team. 🙏 More to come as we get past our SN50 launch today! https://lnkd.in/efyYnttY

10

Thrilled to share that a Digital Twin Framework (ExaDigiT), developed in partnership with Hewlett Packard Enterprise, is a first-of-its-kind digital twin for supercomputer data centers like Oak Ridge National Laboratory’s Frontier. ExaDigiT combines workload, power, and cooling modeling with interactive dashboards, enabling optimization, what-if analysis, and virtual prototyping tailored to each system. Already in use on leadership-class systems worldwide — this innovation is transforming how we model, manage, and optimize supercomputing infrastructure. Huge congratulations to the incredible teams at ORNL and HPE who made this milestone possible! https://lnkd.in/gcA_idVa

251

4 Comments

Ronnie Vasishta, SVP of Telecom at NVIDIA, testified on the critical need for the United States to lead with AI-RAN to build a stronger and more connected nation. House Committee on Energy and Commerce. https://lnkd.in/gqxVFCa3 The NVIDIA AI-RAN is a driving catalyst for the national buildout of INTELLIGENT INFRASTRUCTURE. AI-enabled orchestration to support demanding applications like autonomous systems, spatial computing, intelligent transportation, and enhanced public safety and response. Jensen Huang - NVIDIA and Masayoshi Son - SoftBank Group Corp. - AITRAS - It will free our Spectrum with AI-RAN. ERA of AUTONOMY, advancing reconfigurability & baseband processing. AI-GRID: https://lnkd.in/d765bSFk Intelligent Infrastructure Economic Zones activate Digital Twin-aaS, Drone-as-a-Service, AI Inference aaS, AIRE M2M Platform, Maintenance-aaS, Mobility-aaS, Delivery-aaS, Freight-aaS, etc. Jensen Huang, CEO of NVIDIA. “AI meets $100 trillion of the world's industry.” https://lnkd.in/exXzcJSt INTELLIGENT INFRASTRUCTURE bridges the Civic and market Economies to enable Society 5.0. It merges the Digital and Physical Worlds and enables the AI economy. Building Stronger and More Connected Communities. Autonomy Institute INTELLIGENT INFRASTRUCTURE is enabling AI on a sidewalk near you.  https://lnkd.in/gK4-J8qF

41

10 Comments

Memory modules are just the starting point, Memory Pools represent the solution to accelerate real AI platforms. AI workloads are rapidly scaling, particularly with trillion-parameter large language models (LLMs) and real-time inference. Traditional architectures are facing significant memory bottlenecks. While GPUs and accelerators are essential for AI compute, limitations in memory, latency, and siloed resources are hindering performance. At UnifabriX, we see CXL as more than just an interconnect, it is the cornerstone for the next generation of AI infrastructure. Our advanced Memory Pooling technology harnesses the full potential of AI workloads by eliminating memory silos and enabling dynamic, scalable memory access across diverse compute environments.

26

Nice summary from the Synergy Research Group on 4Q26 hyperscaler performance “New data from Synergy Research Group shows that Q4 enterprise spending on cloud infrastructure services jumped by almost $12 billion from the previous quarter and by $29 billion from Q4 of 2024. The scale of these increments far surpasses anything previously seen in this market. The 2025 full-year market reached $419 billion. Backing out the impact of some major currency fluctuations over the period, this represents growth of 30% from the fourth quarter of last year. This marks the ninth consecutive quarter of accelerating year-on-year growth and the highest growth rate the market has seen in more than three years. GenAI is clearly the primary driver of these changing market dynamics. Among the major cloud providers, Amazon maintains a strong lead in the market, though Microsoft and Google continue to achieve substantially higher growth rates. Their Q4 worldwide market shares were 28%, 21%, and 14% respectively.” #cloud #ai #google #microsoft #amazon #hyperscaler #marketshare #aws #azure #googlecloud #genai https://lnkd.in/gCJqq3W6

2

It's been incredible work with metrobloks on Project Kati-2. Take a listen to their founder and CEO Ernest Popescu. He shares incredible insights on how the data center landscape has changed since his days at Amazon Web Services (AWS), AI at Meta, and Iron Mountain.

12