David Byte

Pure Storage announced Q1FY26 earnings Q1FY26 results - revenue of $778.5m, beating expectations of $770.3m - non-GAAP EPS of $0.29, beating expectations of $0.25 they also announced that - CFO Kevan Krysler has decided to leave the company after more than five years of service. Krysler will remain at Pure Storage until a successor has been named Q2FY26 guidance - revenue of $845m, just ahead of estimates of $841m FY26 guidance - revenue of $3.515b, in-line with expectations of $3.51b but given the revenue beat in Q1 and guide above expectations in Q2 could hint at slower growth then previously anticipated in Q3 and/or Q4 #purestorage #storage #earnings https://lnkd.in/g8uKBhgf

Oracle's stock decline today follows reports suggesting thin profit margins in its AI cloud business, including an estimated $100 million loss from renting NVIDIA chips. The opportunity Oracle missed? Proactive cloud cost optimization. I've seen parallels between complex cloud billing and the baffling telecom plans of yesteryear—where every API call, data transfer, or log ingestion adds up like texts and cat videos on old cellular bills. Even tech giants aren't immune to cloud cost traps. Without centralized visibility and governance, investments in high-demand assets like Nvidia chips can erode margins, turning growth engines into money pits. Oracle's aggressive AI push, including the Stargate project with OpenAI and SoftBank for massive data centers, underscores the need for balancing innovation with fiscal discipline—forecasting $144 billion in cloud revenue by 2030 is ambitious, but unsustainable if costs run unchecked. How can others avoid this folly? Leverage TEM solutions like T E M T R A C E to extend proven telecom expense strategies to cloud management. T E M T R A C E provides real-time visibility into usage, automates invoice auditing, flags inefficiencies (like overprovisioned GPU rentals), and enforces policies to prevent data bloat or compliance risks—much like the financial firm's guidelines on limiting debug logs and sensitive data storage. In an era where AI cloud spend can balloon overnight, tools like TEMTRACE turn reactive billing shocks into proactive savings, ensuring scalability without the margin meltdown. What are your thoughts on cloud cost governance in the AI boom? #CloudOptimization #AI #TelecomExpenseManagement #Oracle #CostControl

6

The Texas Advanced Computing Center (TACC) is charging forward with the Horizon supercomputer, set to deliver 10 times the performance of today's top supercomputers for simulations. Check out the full scoop in Axios: https://lnkd.in/g9TeTjjb #Supercomputing #HPC

2

Pure Storage....what a quarter. 🔸 First and foremost, this one slide that its IR team creates for every quarterly report is fantastic and serves as a model for all. 🔸 Revenue - Strong growth. $964.5 million represents 16% YoY (beat forecast of $960 million) - $3.63 billion FY26 revenue comes in at 14.7% raise -ARR at $1.84 billion ( +17%) and RPO of $2.94 billion (+24%) are almost SaaS like numbers - Mid-70s gross margins are almost ISV-like 🔸 Hyperscalers - Exceeded yearly shipment goals a quarter early - Now officially viewing hyperscalers as an addressable market 🔸 Enterprise Data Cloud (EDC) - Fusion using customers have more than tripled since the beginning of the year - This notion of EDC + Fusion + Evergreen/One as the control plane for AI, modernized virtualization, and multi-cloud goes from slideware to real production 🔸 Against the market - The external storage market was pegged to grow at 5.5% - The largest storage vendor grew at about 6%, with 3% revenue growth - Pure's 16% (rev) and 18% (product) growth shows share gains 🔸 What I'd like to see - Growth is largely in NA. I'd like to see a strong international contribution to growth - As hyperscaler comes more into focus, larger opportunities with lower margins may squeeze profitability in the future. --- Overall, hard to look at this quarter and not be impressed with what Charlie and his team are doing. I'm excited to see how much this accelerates as the company leans more heavily into the AI data management space. ⬇️ ⬇️ ⬇️ ⬇️ ⬇️

48

AI at scale requires an entirely new class of infrastructure, and Hewlett Packard Enterprise is committed to giving our customers the performance, openness, and efficiency they need to lead in this next era.    Last week's announcement of the AMD “Helios” AI rack-scale architecture, combined with our purpose-built HPE Juniper scale-up Ethernet networking, marks a major step forward in how cloud service providers accelerate large-scale AI training and inference. By delivering the industry’s first standards-based scale-up Ethernet solution with Broadcom’s Tomahawk 6 silicon, we’re enabling trillion-parameter training, massive model throughput, and faster, lower-risk deployments.    What stands out in this work is the focus on open innovation, all delivered as a single turnkey rack allowing customers to scale AI with confidence, speed, and simplicity.    Congratulations to our teams and to our partners at AMD and Broadcom.    https://lnkd.in/dnnA684g

98

1 Comment

Today, performance, efficiency, and sustainability come together in one place: the AWS cloud running on Intel Xeon 6 processors. 🔧 Custom-built silicon Designed alongside AWS to optimize the 300+ services they launch each year, while supporting real workloads, scale, and continuous cloud evolution. ⚡ Performance that moves the needle More memory, faster data processing, and better performance per watt, enabling quicker insights while improving energy efficiency. 🚀 Outcomes that matter New workloads, new experiences, and new AI innovation coming from partners and customers in the AWS cloud. It’s how Intel Corporation and Amazon Web Services (AWS) handle the hard work behind the scenes, so customers can focus on moving forward.

178

6 Comments

Pretty cool, ChatGPT letting me queue up weekly summaries on topics and proactively sending.... <ChatGPT 5.2> Here’s a comprehensive news summary of Ultra Accelerator Link (UALink) developments from January 18, 2026 up to today (Jan 25, 2026): Key UALink News (Jan 18 – Jan 25 , 2026) 📌 1. Funding and ecosystem momentum Upscale AI raises $200M to build open scale-up networking based on UALink & Ultra Ethernet Upscale AI – a startup backed by Intel, Qualcomm, AMD, and others – raised about $200M in Series A funding to develop next-gen AI networking solutions that support open protocols like UALink and Ultra Ethernet. The company’s goal is to introduce purpose-built switches for scale-up AI infrastructure (competing with Nvidia NVSwitch) and use open interconnects including UALink as foundational protocols. 📊 2. Market traction and vendor interest Astera Labs sees strong UALink growth in hyperscaler testing Astera Labs reported increased adoption/testing of UALink among hyperscalers, and expects this to translate to new revenue starting in 2026 as customers evaluate UALink-enabled solutions for large AI clusters. 📈 3. Analyst & industry perspectives Market reports position UALink within the broader AI interconnect battle Recent analysis highlights that interconnect technologies like UALink are increasingly central to AI cluster scaling. While Nvidia’s NVLink/NVSwitch family remains dominant today, industry commentators point to open alternatives like UALink as part of a broader strategy for vendor-neutral scale-up fabrics, though adoption timing and ecosystem support remain key variables. 🔎 Quick context on UALink (background relevant to current news) UALink is an open industry interconnect standard developed by the UALink Consortium (including AMD, Intel, Google, Meta, Microsoft, and others) to enable high-bandwidth, low-latency scale-up interconnects for accelerators in AI clusters. The 1.0 specification supports up to 1,024 accelerators in a fabric and uses Ethernet-style physical layers with memory semantic features. 🧠 Overall Themes in Early 2026 News ✔️ Open interconnect momentum: Startups and vendors are positioning UALink as part of open AI networking stacks. ✔️ Ecosystem validation: Hyperscaler testing and vendor investments indicate early traction. ✔️ Competitive headwinds: Nvidia’s proprietary alternatives (NVLink Fusion) and incumbency pose adoption challenges for UALink. ✔️ Market relevance: Analysts continue to emphasize interconnects as core drivers of AI data center performance, with UALink part of the wider technology landscape. #ualink #scaleup #ai

35

What’s New in AIX® 7.3 TL4: Major Live Kernel Update Enhancements IBM continues to push the boundaries of enterprise resilience and zero‑downtime operations—and the latest AIX 7.3 TL4 update supports that objective. I recently came across a great breakdown of the new Live Kernel Update (LKU) advancements, and the improvements are impressive. Here are some highlights: - Optimized Page Migration - Parallelized Mirrored Copy Creation - LKU Blackout Time Enhancements - New Features in LKU - Enhanced flexibility and security These enhancements reinforce why AIX continues to be a platform built for high‑availability, enterprise workloads. Read the full blog:https://ibm.biz/BdbW4p Explore AIX: https://lnkd.in/difxUK9P #IBMPower, #AIX

5

Maximize ROI: Proxmox + Linux with ZFS on SimplStor NVMe slashes costs with immutable snapshots, replication, deduplication and so much more. And FileMaker runs 3x faster on SimplStor servers. Zerowait's Engineers will demonstrate our advantages using real data. https://lnkd.in/eWputwgn #Zerowait #SimplStor #FileMaker

1

Ever wondered what’s next in data security? At #RSAC2025, Shira Rubinoff spoke with Arun Gururajan, Ph.D., Vice President of Research and Data Science at NetApp, about our revolutionary approach to ransomware protection. NetApp builds its security directly into its native storage products—no third-party add-ons or extra steps required. Here’s what makes their approach stand out: * Native Integration: NetApp’s ransomware detection algorithms are built into ONTAP, ensuring real-time protection at the storage source. * Speed Matters: While other systems might require hours for detection and recovery, NetApp’s solutions operate in seconds, saving critical time and money. * Seamless Experience: No need for new system training. NetApp users enjoy instant protection through their familiar control panel. * Always Ahead: Ransomware evolves constantly. NetApp updates its systems in real-time to protect against new threats as they emerge. Arun highlighted what makes data security crucial today: an organization’s data is its most valuable asset, and protecting it at the source is non-negotiable, especially in a rapidly changing threat landscape. Want to learn more about NetApp’s groundbreaking approach to security? #cybersecurity #ransomware #datasecurity

2