Welcome to this edition of our Weekly Analyst Insights roundup, which features key insights that our analysts have developed based on the past week’s events.
Last week our datacenter analyst Matt Kimball published both a long research piece and a shorter article assessing how Pure Storage has evolved over the years — including into its brand-new identity as Everpure. Matt is exactly the right person to analyze this, given his long background in datacenters and storage, not to mention his years of experience covering this particular company. We were happy that the folks at Pure/Everpure briefed us ahead of time so that Matt’s thoroughly informed research piece could appear on the same day as the announcement of the new company name.
Patrick Moorhead and his broadcasting partner Daniel Newman (right) interview Everpure CEO Charlie Giancarlo (left)
at the Pure//Accelerate conference in 2024 (Credit: Six Five Media)
As the 2024 image above suggests, I’ve known Everpure CEO Charlie Giancarlo for years, and I’ve followed the company’s successes pretty closely myself. I’ll be eager to see how Everpure continues to keep pace — and, usually, set the pace — for its market in this era of pervasive AI.
This week, Anshel and I are in Barcelona for Mobile World Congress. The rest of the team is deep in advisory, research, and preparing for a busy season of client and vendor events. If you’ll be attending any of the same events, or if you see that we’ll be in your city, please reach out.
Last week, MI&S analysts appeared in top-tier business and technology coverage in The New York Times, CNBC, Yahoo Finance, Computerworld, CIO, DataCenter Knowledge, TechTarget, The Indian Express, DT Next, AI Daily Shot, and RTech Round. Our commentary spanned NVIDIA and AMD AI datacenter chips, Meta’s chips-for-stock deal, AWS Kiro AI coding reliability and review gaps, Google search ranking changes driven by European regulation, Samsung Galaxy S26 multi‑agent AI and privacy display features, Salesforce’s AWU productivity metric and Q4 earnings, ServiceNow’s autonomous workforce plans for L1 service desk roles, and whether quantum computing represents the next datacenter revolution or another tech hype cycle.
Our MI&S team also published 11 deliverables — 1 Forbes article, 2 Research Papers, 4 Research Notes, 3 Analyst Insights, and 1 Podcast.
Check out this week’s Analyst Insights roundup for more from the MI&S team, including what’s top of mind for each analyst, our thoughts on vendor announcements, press quotes, and more.
Have a great week!
Patrick Moorhead
ServiceNow made multiple announcements last week suggesting a substantial move towards autonomous agents to supplement support teams. While many role-based agents are planned, the first to come to market is an L1 support agent that can take on basic IT customer service tasks. While this specific capability has been around for a little while, the supporting management and observability infrastructure has been upgraded in the wake of the integration of the Moveworks acquisition and updates to AI Control Tower capabilities. This strategy ties in well with my notion that service and software companies will start to become harder to define in the agentic era. The ability to spin up role-based resources trained in an enterprise’s specific ITOM processes on demand is not only a software advance but a competitive maneuver to compete with outsourcing companies. I explore this thought more deeply in an article I published last week on LinkedIn.
Last week I was also briefed by a startup that is making some headway with an end-to-end agentic development framework. While most of the attention so far has been given to the likes of AWS Bedrock, Microsoft Foundry, and Google Vertex, I liked what I saw from UnifyApps. For starters, UnifyApps is a hybrid platform enabling easier access across clouds, models, and on-premises data. But I was also impressed by how the team has a very good line of sight to the customer challenge with respect to scaling agents. The services that UnifyApps provides are necessary in terms of data normalization, application integration, governance, and tooling. It’s a bit of a Goldilocks approach, where some of the bigger cloud solutions are more complex and may be too much for an enterprise — but also a lot more robust than what would typically be provided by an agentic development framework. The fact that UnifyApps also partners heavily with cloud players helps customers integrate with their existing cloud investments. If you are considering a new framework to help scale your agents better, it could be worth a look.
One of the cooler trends in agentic technology is the crop of startups using agents — and AI more broadly — to deliver a new class of vertical business solutions. This is not your typical horizontal technology play, but rather actual companies solving specific business problems. On that front, H2L Marketing recently released its Ellipse Platform, which was built to help large rental organizations. This is a space that is interesting because of both its size and challenges. About 50% of inbound calls to large residential rental complexes never get replies. That’s a lot of lost revenue due to overstretched staffing. Ellipse is a voice-driven agent solution (which also works with text) that helps manage inbound inquiries and can automate the entire rental process, including scheduling showings and providing leases. H2L has not fully closed the loop to the point that a renter can complete the entire process yet, but what it has implemented is already gaining traction and pipeline for its customers. My thought is that companies like H2L will be in the next tranche of unicorns, given how AI can be used to get products to market and ultimately to scale faster.
IBM took a pretty significant hit last week after Anthropic announced its new COBOL migration capabilities. I am guessing that this is going to be a sign of the times, where jittery investors are looking for any signs of an AI disruption (or bubble). Like the downturn in pure-play SaaS players earlier in February, the IBM news is overblown. But while the SaaS events had been brewing for a while, the IBM news came as a bit of a surprise.
While many have weighed in on this topic, I have a particular perspective on this since I have led product and go-to-market efforts for legacy products and programming languages. Here’s my rationale for saying that this Anthropic and IBM story is overblown — my quick thoughts as a former VP of product management who worked at a legacy company for a decade.
Recalling my own experience as a product executive, most customers that walked away from our legacy solution ended up paying more than the legacy system was costing them. And in many cases, they paid that to us anyway.
Finally, I wanted to once again mention that Matt Kimball and I launched our IT Talk Podcast on YouTube last month. Episode 3 will stream on Friday March 6. You can find episodes 1 and 2 here.
ServiceNow made multiple announcements last week suggesting a substantial move towards autonomous agents to supplement support teams. While many role-based agents are planned, the first to come to market is an L1 support agent that can take on basic IT customer service tasks. While this specific capability has been around for a little while, the supporting management and observability infrastructure has been upgraded in the wake of the integration of the Moveworks acquisition and updates to AI Control Tower capabilities. This strategy ties in well with my notion that service and software companies will start to become harder to define in the agentic era. The ability to spin up role-based resources trained in an enterprise’s specific ITOM processes on demand is not only a software advance but a competitive maneuver to compete with outsourcing companies. I explore this thought more deeply in an article I published last week on LinkedIn.
Last week I was also briefed by a startup that is making some headway with an end-to-end agentic development framework. While most of the attention so far has been given to the likes of AWS Bedrock, Microsoft Foundry, and Google Vertex, I liked what I saw from UnifyApps. For starters, UnifyApps is a hybrid platform enabling easier access across clouds, models, and on-premises data. But I was also impressed by how the team has a very good line of sight to the customer challenge with respect to scaling agents. The services that UnifyApps provides are necessary in terms of data normalization, application integration, governance, and tooling. It’s a bit of a Goldilocks approach, where some of the bigger cloud solutions are more complex and may be too much for an enterprise — but also a lot more robust than what would typically be provided by an agentic development framework. The fact that UnifyApps also partners heavily with cloud players helps customers integrate with their existing cloud investments. If you are considering a new framework to help scale your agents better, it could be worth a look.
One of the cooler trends in agentic technology is the crop of startups using agents — and AI more broadly — to deliver a new class of vertical business solutions. This is not your typical horizontal technology play, but rather actual companies solving specific business problems. On that front, H2L Marketing recently released its Ellipse Platform, which was built to help large rental organizations. This is a space that is interesting because of both its size and challenges. About 50% of inbound calls to large residential rental complexes never get replies. That’s a lot of lost revenue due to overstretched staffing. Ellipse is a voice-driven agent solution (which also works with text) that helps manage inbound inquiries and can automate the entire rental process, including scheduling showings and providing leases. H2L has not fully closed the loop to the point that a renter can complete the entire process yet, but what it has implemented is already gaining traction and pipeline for its customers. My thought is that companies like H2L will be in the next tranche of unicorns, given how AI can be used to get products to market and ultimately to scale faster.
IBM took a pretty significant hit last week after Anthropic announced its new COBOL migration capabilities. I am guessing that this is going to be a sign of the times, where jittery investors are looking for any signs of an AI disruption (or bubble). Like the downturn in pure-play SaaS players earlier in February, the IBM news is overblown. But while the SaaS events had been brewing for a while, the IBM news came as a bit of a surprise.
While many have weighed in on this topic, I have a particular perspective on this since I have led product and go-to-market efforts for legacy products and programming languages. Here’s my rationale for saying that this Anthropic and IBM story is overblown — my quick thoughts as a former VP of product management who worked at a legacy company for a decade.
Recalling my own experience as a product executive, most customers that walked away from our legacy solution ended up paying more than the legacy system was costing them. And in many cases, they paid that to us anyway.
Finally, I wanted to once again mention that Matt Kimball and I launched our IT Talk Podcast on YouTube last month. Episode 3 will stream on Friday March 6. You can find episodes 1 and 2 here.
Google has released Gemini Nano Banana 2, refining one of the best image-generation models out there to address one of my biggest problems with the first version: the need for higher-resolution images. The original model supported images only up to 1024 x 1024 and didn’t really enable those images to be used for real applications. The new version brings in a much more work-friendly model refined to meet enterprise demands. I also enjoy using Nano Banana as a photo editor, and working at higher resolution means not having to reduce the quality of my images to get them edited. Google says this model will be embedded into the Gemini app and will be available virtually everywhere.
The drama around Anthropic and the U.S. Department of Defense only got bigger as OpenAI swooped in to capture the contract — and raised $110 billion in the same week. While it’s unclear who is in the wrong between Anthropic and the DoD regarding the limitations Anthropic put on its models, what does seem clear is that Anthropic technology was used to enable the DoD to eliminate Iran’s supreme leader, Ayatollah Khamenei. This surely won’t be the last we hear of Anthropic and the DoD, as the company has already said it will go to court regarding the disagreement.
On February 27, 2026 — last Friday — Anthropic responded to U.S. Secretary of War Pete Hegseth’s stated intent to designate Anthropic as a supply chain risk. This escalation was apparently the result of Anthropic’s refusal to allow its Claude AI to be used for mass domestic surveillance or fully autonomous weapons. Anthropic CEO Dario Amodei has said that he believes today’s models are too unreliable for lethal autonomy and that mass surveillance violates fundamental rights.
Anthropic regards Secretary Hegseth’s intended designation as unprecedented and legally unsound. The company also points out that it has supported U.S. classified networks since 2024 without hindering government missions. It also contends that Hegseth lacks the statutory authority to prevent companies that do business with the U.S. military from doing business with Anthropic, and says that the company “will challenge any supply chain risk designation in court.”
One important wrinkle: Since China has no restrictions on what it can use AI for, limiting its uses could put the United States at a disadvantage.
AMD and Nutanix Partner to Bring AI to the Commercial Enterprise
As if on cue, enterprise adoption came into sharp focus as AMD announced a strategic partnership with Nutanix that includes a $150 million equity investment and up to $100 million in joint engineering and go-to-market efforts. The companies intend to co-develop an integrated platform designed to simplify deployment and operation of enterprise AI workloads across hybrid and multicloud environments. The partnership focuses on aligning AMD’s silicon roadmap with Nutanix’s hybrid infrastructure software stack, with the goal of enabling enterprises to deploy AI in environments they already operate rather than requiring purpose-built infrastructure.
Market Impact
This reflects a broader shift in the AI infrastructure market from component-level competition toward platform alignment. The next phase of enterprise AI adoption will not be defined solely by accelerator performance, but by how easily infrastructure can be deployed, operated, and integrated into existing environments. Silicon vendors are increasingly aligning with infrastructure software providers to create validated platforms that reduce deployment friction. This approach also allows chip vendors to attach themselves higher in the stack, where long-term architectural standardization occurs.
Enterprise IT Impact
Enterprises — specifically those that don’t have operating practices similar to hyperscalers — are not building AI infrastructure from scratch. They are extending existing operational models. Nutanix already serves as the operational control plane for hybrid infrastructure in many organizations. Integrating AMD acceleration into that environment lowers adoption barriers by aligning with existing lifecycle management, governance, and operational workflows. This reduces integration complexity and allows enterprises to deploy AI capabilities without introducing a separate infrastructure model.
My Take
I may be out of the mainstream, but I see this partnership as perhaps more impactful than AMD’s Meta partnership. (See the “Datacenter — Silicon” section of this page for more details on that.) The AMD investment itself is certainly notable, but the strategic alignment is far more important. AMD is positioning itself not just as a silicon supplier, but as part of the enterprise AI platform layer. If one believes that AI will be embedded in every workload, every application, every workflow, then enablement has to include enablement at the lowest levels. Because of this, enterprise adoption will ultimately favor platforms that integrate cleanly into existing infrastructure. This partnership aligns with that reality and strengthens AMD’s ability to participate in enterprise AI deployments that will undoubtedly prioritize operational consistency over raw performance metrics.
VAST Forward Highlights Expansion Toward Operational AI Platform Role
At its inaugural Forward event, VAST Data made several announcements aimed at reducing the friction of deploying, optimizing, and managing AI data environments. The company introduced Polaris, a global control plane designed to orchestrate distributed VAST deployments across hybrid and multicloud settings. The company also expanded its collaboration with NVIDIA, introduced GPU-accelerated infrastructure configurations, and announced new services focused on governance and lifecycle management of AI systems, including PolicyEngine and TuningEngine.
What This All Means
I think this reflects a broader shift in AI infrastructure toward operational platforms rather than standalone systems. As AI deployments scale, infrastructure coordination, lifecycle management, and governance become critical operational requirements. The introduction of a global control plane reflects increasing recognition that AI infrastructure must be managed as a distributed system rather than isolated deployments.
Trying to build AI infrastructure by assembling so many different elements is difficult for the most sophisticated datacenter teams and IT organizations. And for those that aren’t the largest of the large or the most advanced in terms of capabilities, it is impossible. But this is what VAST solves for.
What This Means for Enterprise IT
The enterprise IT organization is exactly who I referenced above. Enterprise AI deployments increasingly span multiple environments. Coordinating infrastructure across hybrid and multicloud environments introduces operational complexity. Control plane capabilities that unify lifecycle management and governance reduce operational friction and improve infrastructure consistency. This becomes particularly important as AI systems transition from experimental deployments to persistent operational workloads.
My Take
VAST is positioning itself to participate in the operational layer of AI infrastructure, not just the performance layer. Long-term infrastructure value increasingly resides in operational coordination and lifecycle management. Platforms that simplify operational complexity will be better positioned as enterprise AI deployments scale.
The question — or tension — is this: While enterprise IT organizations can see the intrinsic value in the VAST AI OS, they will also bristle a bit at the thought of possible vendor lock-in for what will effectively be the enterprise. If one believes that AI will power every facet of the business, it is logical to see how this lock-in could happen. VAST must stress its openness and partnerships (beyond NVIDIA and neoclouds) to gain traction in enterprise environments.
Dell Earnings Highlight Record Infrastructure Growth Driven by AI Demand
Dell reported record financial results for fiscal 2026, driven primarily by strength in its Infrastructure Solutions Group (ISG). The group’s revenue reached a record $19.6 billion in the fourth quarter, up a whopping 73% year over year, reflecting unprecedented demand for AI infrastructure. AI-optimized server revenue alone reached approximately $9 billion in the quarter, up a staggering 342% YoY, and totaling $24.7 billion for the full fiscal year, with cumulative AI server orders passing $64 billion (with a backlog of $43 billion exiting the year). Dell also guided toward continued expansion, with expectations that AI server revenue could approach $50 billion in fiscal 2027, reflecting sustained demand.
AI Is the Gasoline on the Fire
These results confirm that enterprise and hyperscale AI infrastructure is now driving server market growth. AI infrastructure is beginning to gain momentum, shifting from pilots to sustained procurement. OEMs like Dell are benefiting from their ability to integrate accelerators into complete systems, including power, cooling, networking, and lifecycle management. This shifts value from accelerators to complete infrastructure platforms capable of operating AI workloads reliably at scale.
What This Means for Enterprise IT
Enterprises are increasingly evaluating and buying AI infrastructure as integrated solutions, rather than building from pieces. Dell’s growth reflects demand for proven infrastructure solutions that take the complexity out of deployment and operational integration. This reduces integration complexity and accelerates enterprise adoption. The scale of Dell’s backlog also shows that enterprise and hyperscale customers are planning infrastructure deployments over several years — not just one-time projects.
My Take
Dell’s infrastructure results point to a structural shift in the server market. The company is no longer primarily selling general-purpose infrastructure — it is increasingly serving as a system integrator for accelerated computing environments. Long-term differentiation in this market will depend less on server hardware itself and more on the ability to deliver complete, operationally ready infrastructure platforms that integrate accelerators, networking, storage, and lifecycle management. Dell’s scale, supply chain, and enterprise reach position it well to participate in this transition.
NVIDIA Earnings — Wow!
NVIDIA reported record revenue of $68.1 billion for Q4 FY2026, up 20% sequentially and 73% year over year, and $215.9 billion for the full fiscal year, representing 65% annual growth. Datacenter revenue reached a record $62.3 billion in the quarter, up 22% sequentially and 75% YoY, and $193.7 billion for the full year, up 68% YoY. This confirms that the datacenter business now represents the overwhelming majority (almost 92%) of NVIDIA’s revenue and is the primary driver of its financial performance. Gross margins remained strong at approximately 75%, reflecting continued pricing power. The company also set revenue guidance of approximately $78 billion for the next quarter, demonstrating incredible growth — even at this scale.
What This Means for the Market as a Whole
These numbers should tell us that accelerated computing is now the structural layer of global infrastructure. The scale of datacenter revenue (approaching $200 billion annually) shows sustained hyperscale and large enterprise investment in AI. This also shows that demand is not limited to training infrastructure but increasingly includes inference infrastructure, which requires sustained deployment across production environments.
What This Means for Enterprise IT
For enterprise IT organizations, the signals should be clear. AI infrastructure is no longer experimental or limited to hyperscale environments. It is becoming standard infrastructure for modern application platforms. As mentioned in my discussion of AMD’s moves, the scale of deployment at hyperscale drives ecosystem maturity, including software tooling, operational models, and validated architectures. These improvements make accelerated infrastructure easier to adopt within enterprise environments. At the same time, continued hyperscale demand will influence supply dynamics and platform availability, reinforcing the need for more thoughtful, considered infrastructure planning.
My Take
NVIDIA is no longer benefiting from early-cycle deployment alone; it is operating at a sustained infrastructure scale. The market’s muted reaction, while seemingly irrational, likely reflects elevated expectations rather than underlying weakness. When companies operate at this scale, continued growth — even at record levels — has to be evaluated against future growth assumptions rather than past performance. From an infrastructure perspective, the trajectory remains clear: accelerated computing is becoming a permanent and foundational layer of enterprise and hyperscale infrastructure.
AMD and Meta Go Big
If the Nutanix partnership (discussed under “Datacenter — AI”) is about the volume of the market, AMD and its Helios platform are all about hyperscale enablement. In the company’s announced partnership with Meta, we will see an expanded strategic partnership under which Meta plans to deploy up to 6 gigawatts of AMD Instinct GPUs over time. The agreement includes milestone-based incentives and performance-based warrants tied to deployment thresholds. This represents one of the largest publicly disclosed infrastructure commitments involving AMD accelerators and reflects deeper alignment between the companies’ infrastructure roadmaps.
This Is About Scale
This reinforces a structural shift toward multi-vendor AI infrastructure ecosystems. Hyperscale operators are aligning closely with silicon providers to ensure supply continuity, influence roadmap development, and reduce architectural dependency on any single vendor. These relationships increasingly resemble strategic infrastructure partnerships rather than traditional supplier agreements. The scale of this deployment also reflects the growing importance of inference infrastructure, which requires sustained deployment of accelerators optimized for operational efficiency rather than peak training performance.
This is not a shift away from NVIDIA. This is an acknowledgement that inference at scale is an ecosystem play that does not suggest reliance on a single supplier.
What This Means for Enterprise IT
As touched on above, hyperscale adoption accelerates ecosystem maturity. Large deployments drive software optimization, operational tooling improvements, and broader ecosystem support. These benefits eventually propagate into enterprise infrastructure offerings. Enterprises evaluating alternative accelerator platforms will increasingly find more mature software stacks, validated deployment models, and clearer operational guidance as hyperscale deployments scale.
My Take
This agreement confirms that AMD is now a strategic infrastructure supplier at hyperscale scale. More importantly, it accelerates ecosystem development around AMD’s platform. Hyperscale deployments establish operational credibility and drive ecosystem readiness. Enterprise adoption typically follows once platforms demonstrate operational stability and integration maturity at scale.
I’ve written about this several times over the past few years when CSPs such as Azure and OCI deployed AMD’s MI-Series GPUs at scale. The optimizations made and fed back into the ecosystem enable an enterprise IT organization (and developer community) to adopt and deploy with confidence. Further, even though hyperscale wins grab headlines, enterprise adoption will determine long-term success.
ServiceNow has launched Autonomous Workforce and EmployeeWorks, extending its AI platform from assisting with work to fully executing it across IT service desk and employee-support workflows. The first “AI specialist” targets Level 1 service desk roles and shifts the focus from drafting suggested responses for human agents to resolving common issues end-to-end. EmployeeWorks combines Moveworks’ conversational interface and enterprise search with ServiceNow workflows so employees can ask for help in natural language across Teams, Slack, or the browser and have the platform orchestrate multi-system actions behind the scenes.
The strategic signal in this launch is ServiceNow’s emphasis on governance and cross-system execution as the center of its AI story. Its AI Control Tower is positioned as a hub to manage policies, access, and auditability across heterogeneous AI systems, including third-party models, which aligns directly with CIO concerns that many AI projects stall because of issues with data quality, risk, and change management rather than model capability.
For customers already standardized on the company’s Now Platform, the potential upsides are faster resolution times, fewer tickets reaching human agents, and a more intuitive employee entry point into services. The constraint is that these outcomes depend on clean data, mature governance, and meaningful integration work, especially in fragmented or less-standardized environments. In parallel, ServiceNow’s experimentation with consumption- and outcome-oriented pricing for AI agents is an important signal as investors debate whether agentic AI compresses traditional seat-based SaaS in the emerging “SaaSpocalypse” narrative. (For more on this, and the SaaSpocalypse, see Jason Andersen’s article on LinkedIn here.)
The way enterprises think about data platforms has changed. It used to be about selecting a primary platform and standardizing around it. That approach does not hold up anymore.
Why? No single platform can do everything. What matters now is how the ecosystem works together across ingestion, governance, transformation, analytics, and AI. Performance is expected. The harder problem is keeping identity, metadata, and data quality consistent across multi-cloud and on-premises environments without creating fragmentation.
This new reality shifts the role of IT leadership. You are not just choosing a platform. You are designing and managing an ecosystem. The real questions are practical. How quickly can you onboard new data? Are business definitions enforced consistently across different tools? Do insights flow back into operational systems in a way that drives measurable outcomes?
AI also changes the equation. Once models start influencing customer, operational, and financial decisions, governance cannot be an afterthought. A weak semantic layer becomes an operational and regulatory risk. The vendors that build interoperability and governance into their core architectures are the ones that will win. The ecosystem is what you are really buying now.
For more of my analysis on this shift, check out my latest Forbes article: “Beyond The Enterprise Data Platform: Why Ecosystems Win.”
I was in Atlanta this past week for SAP’s Strategic Analyst Council. The specifics of those conversations are under NDA until the Sapphire conference in May, so I’ll keep this at a high level.
SAP’s theme is the shift toward the autonomous enterprise. The company’s vision is to move ERP beyond being a system of record and toward becoming a system that can guide, execute, and serve as a system of action. I speak and write a lot about event-driven ERP, which is the primary objective. Think in terms of embedded intelligence across finance, supply chain, procurement, workforce, and customer processes, with AI layered into core workflows rather than sitting off to the side.
Thinking back about last week’s event, I liked how SAP discussed wanting to have a pace of innovation for their customers. SAP, like a number of other ERP vendors, is accelerating around AI and adding more agentic capabilities. The bigger question is how quickly customers can realistically absorb that change. Many enterprises are still stabilizing cloud migrations, cleaning up master data, and standardizing global processes. Governance, trust, and change management will matter just as much as the technology itself.
SAP’s vision makes sense, and the opportunity is significant. The real variable is innovation aptitude on the customer side — especially in terms of how ready organizations are to adopt, operationalize, and trust more autonomous systems inside mission-critical processes.
I was recently a guest on the Acumatica ERP Podcast with Lauren O’Hara, and we had a great discussion about where ERP modernization is really headed and what it actually takes to make it work.
One thing I keep saying is that modernization is not just about buying new software. Most organizations are using only a portion of the ERP capabilities they already have. Moving to a new platform or upgrading does not automatically create results. It takes clear process improvement, disciplined data management, and alignment around what the business is truly trying to fix or improve. Implementation and change management do not get enough attention, but they are usually what separate small improvements from real operational impact.
We also spent time on data issues and the bigger shift happening in ERP. This will be a familiar refrain if you regularly read my work, or even if you only read my entries on this page: ERP is no longer just a back-office system of record. It is becoming the operational core that feeds analytics, automation, and AI. As more event-driven and agentic approaches emerge, ERP will increasingly help guide what happens next in a workflow rather than simply recording what already happened. That can only work if the data is clean, governance is strong, and the organization is ready to evolve with it.
If you are thinking about ERP modernization and want a practical perspective on what actually drives value, my conversation with O’Hara is a good one to check out. You can see a short clip on LinkedIn, or watch the whole episode on YouTube.
Workday and Insperity have announced the general availability of HRScale, a co-branded solution that integrates Workday’s HCM software with Insperity’s professional employer organization (PEO) shared HR solutions services. This offering is specifically designed for small and mid-sized businesses (SMBs) that require enterprise-grade human resources technology but lack the internal infrastructure to manage complex administrative tasks like payroll, compliance, and benefits in-house. By pre-configuring Workday’s suite to integrate with Insperity’s service model, the partnership aims to lower the barrier to entry for high-end HR tech, which has historically been cost-prohibitive or too complex for the SMB segment.
This move signals a strategic shift for Workday as it seeks new growth channels outside of its traditional large-enterprise stronghold. By leveraging Insperity’s existing sales force and service expertise, Workday wants to scale into the mid-market without significantly increasing its own customer support and implementation overhead. For Insperity, the integration provides a technical competitive advantage over other PEOs that often rely on fragmented or legacy software stacks. From an industry perspective, this partnership intensifies competition for other players, including ADP and Paycheque, that have long dominated the SMB space with bundled service-and-software models. The success of this initiative will likely depend on how effectively the two companies manage the handoff between software updates and manual service delivery, but it establishes a clear roadmap for how enterprise SaaS providers can penetrate down-market through strategic service alliances.
Today’s entry is about Aliro — a new industry standard for residential and commercial access control. Think of it as “Matter for locks.” It has the potential to do for smart locks what Matter does for the smart home, but across a much broader market. The technologies, silicon components, and early ecosystem support are already in place, so I expect Aliro-enabled products in the market later this year, with a rapid ramp in 2027.
On February 26, the Connectivity Standards Alliance (CSA — the home of Matter) published the Aliro 1.0 specification, which is a communication protocol and credential standard for digital access control. Essentially, it’s an evolution of Apple’s own Home Key and the CCC Digital Key (which Apple helped write). Aliro aims to replace the current fragmented jumble of proprietary apps and key cards with mainstream ecosystems that already live on your phone. The standard isn’t just for residences. It also targets commercial buildings — offices, universities, hospitals, hotels, apartments, parking garages, entry gates — and potentially anything else you want to lock up.
How It Works
Aliro defines how smartphones and wearables authenticate with door readers using public-key infrastructure. Credentials live in your mobile wallet — e.g., Apple, Google, or Samsung — with no dedicated app or cloud dependency. The spec supports NFC for tap-to-access, Bluetooth LE for longer-range access, and BLE plus UWB for hands-free proximity unlock.
A two-phase protocol handles both familiar and first-time encounters. When a reader recognizes your device, authentication completes instantly. When it doesn’t (for example because the lock is offline or it’s your first visit), the protocol transmits a digitally signed access document that provides the reader with enough information to make an access decision without a network connection.
Who Built It
The CSA led the development effort with more than 220 member companies. Apple, Google, and Samsung anchor the platform side, and integration with existing wallets gives the standard an immediate path to billions of devices. ASSA ABLOY, Allegion, Infineon, NXP, STMicroelectronics, and Nordic Semiconductor contributed to the spec and are among the first to pursue certification.
Aliro’s protocol architecture draws from Apple’s Home Key and the Car Connectivity Consortium’s Digital Key standard, with different cryptographic primitives and expanded capabilities for enterprise-scale deployment. The pattern echoes Matter: proprietary approaches converging into an open specification. The unifying principle is to remove undifferentiated friction from the marketplace.
Why It Matters
Matter targets the smart home, but the CSA positions Aliro for corporate, educational, hospitality, and multi-family environments — markets where access control is a serious operational and security requirement, not just a convenience feature.
I think the enterprise value proposition is compelling. Legacy badge systems lock organizations into a single vendor’s readers, software, and credentialing stack. By contrast, Aliro acts as a universal interoperability layer. A certified reader from one manufacturer works with credentials from any compliant wallet. This reduces integration costs, simplifies maintenance, and enables system owners to mix and match hardware. Aliro devices can also work without network connections in garages, elevators, and other dead zones where connected systems commonly fail. Aliro locks can also let you into your home when power or internet access is down.
What to Watch
According to the press release, the CSA expects Apple, Allegion, Aqara, Google, HID, Kastle, Kwikset, Last Lock, Nordic Semiconductor, Nuki Home Solutions, NXP Semiconductors, Qorvo, Samsung Electronics, and STMicroelectronics to be the first suppliers to achieve Aliro 1.0 certification. The certification program is open, and the first certified devices should arrive later this year. Three factors will determine how quickly Aliro reaches critical mass.
It’s About Time!
Today’s smart locks and access control systems often seem like artifacts from the 2000s with system-specific access cards, multiple proprietary phone apps, confusing setup procedures, and the high cost and persistent friction of walled-garden ecosystems. My car unlocks automatically when I walk up to it, so why can’t my home do the same? We finally have line-of-sight to standards-based locks that use your phone’s native ecosystem to store keys for your home, entry gate, offices, rental properties, gym locker, or just about anything with a door. And UWB is a game-changer for proximity features. The open questions are about adoption velocity, not architecture or technology, so I expect rapid uptake.
Samsung has taken the wraps off the Samsung Galaxy S26 family of smartphones and the Buds4 family of earbuds. During the launch event, the company unsurprisingly spent a lot of time talking about AI, detailing updates to Galaxy AI in Bixby, Gemini, and Perplexity and continuing to lean into agentic experiences. The Galaxy S26 Ultra saw the most updates compared to the last generation, for example by improving charging to 60W wired and 25W with Qi2 wireless. Samsung also released a new wireless charger to enable this, although it didn’t include the charging magnets in the phone, therefore requiring a Qi2 case that includes those magnets.
The biggest upgrade Samsung included on the S26 Ultra came in the form of a new per-pixel privacy display, allowing the user to decide what someone beside them might be able to see. This feature also allows the user to block only certain parts of the display like notifications and passwords, or to enable it only when using certain apps. The security implications of this feature are huge; it could prevent countless breaches just by keeping passwords safe from prying eyes. Excitement around this feature has already been so strong that it’s likely that Samsung will waterfall the capability into lower-cost devices in the next generation. Considering that it’s such a new feature, it makes sense to release it initially into a lower-volume, higher-cost device like the S26 Ultra to make sure that people really want it. This could also represent a hardware moat for Samsung — an area where it might have exclusive capabilities for quite some time.
Superconducting quantum computers are one of the leading platforms for next-generation quantum computing. They are the fastest modality, but they face a major traffic jam problem. To work, they need millions of qubits, but each qubit needs its own connection to the fridge — superconduction requires low temperatures — from its location in the room-temp electronics. Managing that many wires is a challenge.
A new paper in Nature Microsystems & Nanoengineering presents a fix for the problem using MEMS switches. These are very small components acting as multiplexers which can handle many signals with fewer wires. The researchers tested MEMS switches at temperatures below -441°F. At these extremely cold temps, the switches actually worked better, using less power and performing with more stability.
The team even solved a “bouncing” problem where the tiny switch would flutter in the vacuum of its packaging, ensuring it could flip more than 100 million times without breaking. That said, the developers also acknowledged that there are still dielectric challenges at high frequencies. Nonetheless, by successfully implementing basic logic operations (like NAND and NOR gates) in the cold, the researchers showed that these tiny switches could be a key to building the massive quantum computers of the future.
Oracle and Red Bull Racing have formalized a multi-year extension of their title partnership, a strategic alignment designed to navigate the 2026 Formula 1 technical overhaul. This transition represents the most significant shift in modern F1 history, mandating a move to 100% sustainable fuels and a redesigned hybrid power unit that increases electrical output from 120kW to 350kW, creating a near 50/50 split between combustion and electric power. The agreement centers on deploying Oracle Cloud Infrastructure and specialized AI stacks to support Red Bull Ford Powertrains, a new entity tasked with developing an independent engine to compete with established manufacturers such as Ferrari and Mercedes-Benz.
This integration is critical because the 2026 regulations also introduce “active aerodynamics,” which are moveable front and rear wings that switch between high-downforce and low-drag modes. This adds thousands of new variables to race strategy. By using OCI for high-performance computing, Red Bull can run granular simulations that model energy deployment and thermal management in a virtual environment. This computational approach is a necessary equalizer for a team without a legacy engine manufacturing infrastructure, and it should allow them to iterate on designs and strategy agents far faster than traditional physical prototyping would permit under strict cost-cap constraints.
By deepening this technical dependency, Red Bull is moving toward a vertical technology stack similar to the partnerships seen between Mercedes and TeamViewer or Ferrari and AWS, where cloud providers supply core architectural components rather than merely being sponsors. However, Red Bull’s specific focus on a trackside AI “strategy agent” to automate real-time telemetry ingestion suggests an attempt to shift from reactive data visualization to proactive predictive modeling. This mirrors a broader enterprise trend in time-critical industries, where organizations are moving toward bespoke, edge-capable AI environments to maintain a marginal-gains advantage. The success of this partnership will serve as a primary indicator of whether advanced cloud-scale compute can effectively substitute for decades of specialized automotive engineering experience.
Last week, I attended RingCentral’s 2026 analyst summit — and came away from the event seeing a company leaning hard into agentic AI, backed by disciplined, cash‑generative execution. The sessions were under NDA, but what I can tell you is that the roadmap and architecture presented solid backing for my view that RingCentral is moving from being a traditional UCaaS provider to an AI‑driven communications platform. Given RingCentral’s depth in telephony, its strength in voice technology should also serve it well as voice increasingly becomes a primary UI for agentic AI.
Product discussions centered on agentic, system‑level AI that coordinates work across applications rather than isolated features. RingCentral highlighted voice‑centric AI and offerings such as AVA, AIR, and ACE as orchestration layers on top of its telephony and contact center stack, aimed at automating workflows, summarizing and routing interactions, and connecting front‑office and back‑office systems with concrete customer use cases. This technology is backed by a financial approach focused on profitable growth, with AI already contributing meaningful ARR and expected to become a larger share of the business over time.
From my vantage point, the leadership team is one of RingCentral’s strongest indicators of health. The executive bench includes a notable number of boomerang employees who have returned to the company, which suggests there is something compelling about the current strategy, culture, or opportunity drawing experienced talent back. I am encouraged by the conviction and clarity they showed for both the technology roadmap and the financial trajectory, and that tells me a lot about how they see the durability of the company’s shift. In my experience, leadership quality often tells you a good portion of what you need to know about a company, and in this case the mix of returning leaders, experienced operators, and visible enthusiasm for the AI‑driven future was the most important signal from the event.
AI Chips / Matt Kimball / DataCenter Knowledge
Inference Becomes the Next AI Chip Battleground
AMD & Meta Chips, Stock Deal / Matt Kimball / DataCenter Knowledge
AMD, Meta Strike $100B, 6 GW Chip Deal as AI Race Heats Up
AMD & Meta Chips, Stock Deal / Patrick Moorhead / NY Times
Racing to Catch Up With Nvidia, AMD Signs Chips-for-Stock Deal With Meta
AMD & Meta Chips, Stock Deal / Patrick Moorhead / DT Next
Meta, AMD in major chips-for-stock deal
AWS / Kiro, AI Coding / Jason Andersen / Tech Target
AWS Kiro ‘user error’ reflects common AI coding review gap
Google / Search Engine Results / Anshel Sag / Computer World / Europe forces a search reset: Google experiments with fairer rankings
NVIDIA / Earnings / Patrick Moorhead / CNBC
Tech Nvidia’s stock wrapping up tough week as Wall Street focuses more on competition than growth
NVIDIA / Earnings / Matt Kimball / DataCenter Knowledge
Wall Street’s Sour Response to Nvidia’s Record Earnings
NVIDIA / Galaxy AI Chips / Patrick Moorhead / AI Daily Shot
Nvidia Galaxy AI Chips Announced: What the Game-Changer Means for 2026
Quantum Computing / Patrick Moorhead / RTech Round
Quantum Computing: Is It The Next Data Centre Revolution, Or Just Another Tech Hype Cycle?
Samsung / Multi-Agent AI / Anshel Sag / The Indian Express
Samsung banks on new ‘Privacy Display’ tech, agentic AI with flagship Galaxy S26 series
Salesforce / AWU / Robert Kramer / CIO
AWU by Salesforce: A shiny new metric that tells CIOs little of value
ServiceNow / Autonomous Workforce / Melody Brue / Computer World
ServiceNow plans automation of L1 Service Desk roles, promises more AI ‘specialists’ to come
Unless otherwise noted, our analysts will be attending the following events in person.
July events coming soon.
August events coming soon.
September events coming soon.
November events coming soon.
December events coming soon.
Founder, CEO and Chief Analyst | + posts
Patrick Moorhead is the founder, CEO, and chief analyst of Moor Insights & Strategy. His big-picture view of technology is grounded in more than 20 years as an executive leading strategy, product management, product marketing, and corporate marketing functions at NCR, AT&T, Compaq, and AMD. He has shared his expertise in areas from silicon to infrastructure to enterprise SaaS and everything in-between in thousands of national broadcast appearances (CNBC, Yahoo Finance), articles (Forbes, CIO), research-based analyses, and podcast episodes. Today, he has 100+ CXO-level advisory clients and is often ranked the #1 technology industry analyst by ARInsights.
VP & Principal Analyst at Moor Insights & Strategy | + posts
Robert Kramer is vice president and principal analyst covering enterprise data, including data management, databases, data lakes, data observability, data analytics, and data protection. Robert has over 30 years of proven experience with startups, IT companies, global marketing, detailed strategies, business modeling, and planning, working with enterprise companies, GTM assets, management, and execution.
Matt Kimball is a Moor Insights & Strategy senior datacenter analyst covering servers and storage. Matt’s 25 plus years of real-world experience in high tech spans from hardware to software as a product manager, product marketer, engineer and enterprise IT practitioner. This experience has led to a firm conviction that the success of an offering lies, of course, in a profitable, unique and targeted offering, but most importantly in the ability to position and communicate it effectively to the target audience.
Senior Analyst-in-Residence | + posts
Bill Curtis is the Moor Insights & Strategy Analyst in Residence for large-scale Internet of Things systems. Bill helps enterprises design distributed solutions that integrate the full end-to-end IoT stack from real-world devices to analytics.
Jason Andersen is vice president and principal analyst covering application development platforms, technologies, and services. Jason brings over 25 years of experience in product management, product marketing, corporate strategy, sales, and business development at Red Hat, IBM, and Stratus to his work for MI&S and its advisory clients. Working both in the field and in the headquarters of some of the most innovative technology companies, Jason has a wealth of experience in building great products and driving their adoption across a broad spectrum of industries and use cases.
Paul Smith-Goodson is the Moor Insights & Strategy Vice President and Principal Analyst for quantum computing and artificial intelligence. His early interest in quantum began while working on a joint AT&T and Bell Labs project and, during 360 overviews of Murray Hill advanced projects, Peter Shor provided an overview of his ground-breaking research in quantum error correction.
Anshel Sag is Moor Insights & Strategy’s in-house millennial with over 18 years of experience in the IT industry. Anshel has had extensive experience working with consumers and enterprises while interfacing with both B2B and B2C relationships, gaining empathy and understanding of what users really want. Some of his earliest experience goes back as far as his childhood when he started PC gaming at the ripe of old age of 5, building his first PC at 11, and learning his first programming languages at 13.
Mel Brue is vice president and principal analyst covering modern work and financial services. Mel has more than 25 years of real tech industry experience in marketing, business development, and communications across various disciplines, both in-house and at agencies, with companies ranging from start-ups to global brands. She has built a unique specialty working in technology and highly regulated spaces, such as mobile payments and finance, gaming, automotive, wine and spirits, and mobile content, ensuring initiatives address the needs of customers, employees, lobbyists and legislators, as well as shareholders.
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。