Chris DePuy | 650 Group

@nokia's Aurelis for Data Centers reflects a broader shift toward simpler, greener, fiber-based architecture for data center operations. Using PON for OOBM is a smart approach to reducing space, power and cost as AI and cloud workloads expand. Glad to contribute my perspective. #DataCenters #CloudInfrastructure #Fiber #PON #Nokia

AI is forcing a rethink of enterprise security infrastructure. In my latest blog, I look at how @CatoNetworks is leveraging @nvidia technology to enhance its security services by embedding GPU compute across its global SASE PoPs. The bigger takeaway: security platforms are starting to evolve their underlying architectures to both secure AI and use AI to strengthen security operations. We look forward to discussing this more with the Cato team at RSA. #SASE #AISecurity #Cybersecurity #NVIDIA #RSAConference

Ready to deploy AI agents? NVIDIA NemoClaw simplifies running @openclaw always-on assistants with a single command. 🦞 Deploy claws more safely ✨ Run any coding agent 🌍 Deploy anywhere Try now with a free NVIDIA Brev Launchable 🔗 nvidia.com/nemoclaw

@cdepuy @nvidia @openclaw @steipete @650Group @AlanWeckel bro tagged himself in the tweet 🤦‍♂️

Jensen @nvidia says every company needs an @OpenClaw strategy and thanks Peter Steinberger (@steipete) for joining at #gtc26. And it introduced nemoclaw!! @cdepuy @650group @alanweckel

Jensen @nvidia at #gtc26 shows the roadmap chart. Pursuing both copper and CPO scale-up @650Group @cdepuy @AlanWeckel

At #gtc26 Jensen of @nvidia just announced that: -Groq 3rd Generation is in full production Spectrum X CPO in full production NV Link rack with 256 nodes using Ethernet @cdepuy @AlanWeckel @650Group

Jensen Huang of @nvidia kicks off the #gtc26 keynote starting by thanking VIPs for attending.  Jensen then joked that he will spend the next two hours on the next set of slides about structured data and unstructured data.  @alanweckel @cdepuy @650group

We are at @nvidia #GTC26 in San Jose, CA awaiting Jensen’s keynote at SAP Center. We enjoyed the very optimistic presentation by Peter Steinberger (@steipete), developer of OpenClaw, who is now with @OpenAI. 🦞 @alanweckel @cdepuy @650Group

The AI-RAN chip wars were one of the clearest themes at #MWC26. @nokia and @SamsungNetworks are betting on merchant silicon. @ericsson, @Huawei and @ZTEPress are sticking with ASICs. At stake: flexibility vs optimization, shared infrastructure vs single-purpose design, and ultimately the future economics of the RAN. In my latest blog, I break down where each vendor stands and why this matters for operators as AI moves deeper into network strategy. 650group.com/blog/mwc-2026-… #AIRAN #RAN #Telecom #vRAN #AIInfrastructure #650Group

I was at @Fortinet Accelerate 2026 in Las Vegas this week, where one theme came through clearly: AI readiness is reshaping how vendors think about security and networking together. In my latest blog, I break down the key announcements, what stood out at the event and what it says about Fortinet’s broader strategy. Read it here: 650group.com/blog/fortinet-… #Fortinet #Cybersecurity #Networking #AI

Fortinet keynote opens at its #accelerate26 conference in Las Vegas. Russ Schafer, EVP Marketing, shared the big product announcements while Founder and CEO Ken Xie discussed how AI is integral to its products and how Edge will eat the Cloud. @Fortinet @650Group

At Mobile World Congress today, Nokia and Amazon Web Services presented today about how both companies are working together to deliver Telecom Core functions both on cloud and on premises #MWC26 #Nokia #NokiaCore #NokialAR @AWS @nokia @650Group

At Nokia’s MWC26 press event, its management team pitched the entire portfolio (optical, routing, fixed networks, RAN and Core) with AI tying it together. @justinhotard @650group @nokia #mwc26 #AIRAN

@SpaceX acquiring @xai is a big AI capex signal. The key isn’t “data centers in space” headlines — it’s funding durability. This combo makes it more likely xAI keeps building data centers aggressively, backed by SpaceX/Starlink scale and IPO optionality. 3 implications we’re tracking: ➡️Broadband/LEO: constellation expansion pressures rural terrestrial ROI ➡️RAN: stronger direct-to-cell path tightens the rural tower economics ➡️DC infra: another leg under the $1T capex cycle + new compute/networking innovation vectors Blog here: 650group.com/blog/spacex-ac… #AIInfrastructure #DataCenters #Capex #Starlink #LEO #Telecom #Networking

As we head into 2026, it’s clear the industry is moving from discussion to execution. Real-world constraints, not theory, are shaping infrastructure decisions and market direction. I’m grateful for the thoughtful conversations and engagement throughout the past year and look forward to continuing the dialogue as these markets evolve. Wishing everyone a strong start to the year ahead.

As the year comes to a close, I want to thank the industry for the thoughtful discussions and engagement throughout 2025. Wishing you a well-deserved holiday break and a productive year ahead!

Every wave of compute has rewritten the networking roadmap — but AI is rewriting it at record speed. In my @EthernetAllianc TEF 2025 session, I shared why Ethernet is entering a $200B growth cycle as AI fabrics push bandwidth demand from 30–40% CAGR to 70–80% and accelerate the move toward one billion data-center ports. That expansion is driving massive opportunity across LPO, LRO, ACC, AEC and CPO technologies as Ethernet scales up, out and across the AI stack. Appreciate @ConvergeDigest and Converge! Network Digest for capturing the moment and helping amplify the discussion. 📹 Video here: nextgeninfra.io/2025-tef/?spot… #TEF2025 #EthernetForAI #AIInfrastructure #DataCenters #Networking #650Group

AI is accelerating demand for deep observability. @650Group's Deep Observability Quarterly Market and Long-Term Forecast Report shows 25% YoY growth as LLMs drive the need for trusted, network-derived telemetry. @gigamon led with 50% share, with @AristaNetworks, @kentikinc, @Keysight and @NETSCOUT also shaping the market. Release: gigamon.com/company/news-a…

CPO is moving to the center of AI networking — and @MarvellTech just made its intentions clear. At its Industry Analyst Meeting, Marvell outlined a roadmap that puts CPO sampling in 2027, backed by the Celestial AI acquisition and a push to redefine how Ethernet switches scale for AI workloads. Other key signals: 👉Scale-Up: ESUN & UALink sampling in late ’26–’27, creating an alternative path to NVLink while still supporting custom NVLink solutions. 👉Scale-Out: Switch architectures shifting toward tighter optical/electrical integration. 👉Scale-Across (DCI): 1.6T ZR/ZR+ “coming soon,” targeting 2nm, QAM16 & 1000km+ reach. Marvell is gearing up for a share grab in what we expect to be a $200B/yr DC networking market as hyperscalers architect the next wave of AI factories. Full analysis here ⬇️ 650group.com/blog/marvells-… #CPO #AIInfrastructure #Ethernet #Optics #Networking #DataCenter #650Group