Each week NAND Research puts out a newsletter for our industry customers taking a look at what’s driving the week, and what happened last week that caught our attention. Below is a excerpt from this week’s, May 12, 2025.
Driving the Week
There are no big industry events this week, but tech earnings continue. Rapid7 announces on Monday, while CyberArk lands Tuesday. Wednesday is a big day with Cisco, Dynatrace, DXC, and CoreWeave (with its first earnings since going public), all announcing. Alibaba rounds out the week, releasing Thursday.
Enterprise AI
What’s driving enterprise AI spending? If you believe the results of IBM’s recent CEO Survey then it’s the fear of missing out, with about two-thirds of respondents saying that they’re adopting AI before fully understanding what they’re going to do with it. This makes it less suprising to see that only about half of respondents are seeing a clear path to ROI.
Fielding more than just surveys, it was all about enabling enterprise AI at last week’s IBM Think, with Big Blue launching its new Orchestrate tool to help enterprises deploy agentic AI. IBM also announced expanded partnerships with Oracle and AWS, both focused on enabling agentic AI.
Running faster than OpenAI, open-source AI enabler, and social media giant, Meta showed off the power of its new partnership with start-up accelerator company Cerebras. The companies demonstrated Meta’s llama API running about 18 times faster than OpenAI – processing 2,600 tokens/second versus the 130 tokens/second than ChatGPT delivers.
Despite the performance gains, it’s still difficult to get your hands on Cerebras parts. It does show, however, that Nvidia’s biggest exposure is in the inference market where it’s missing the software lock-in it enjoys in training.
Maybe they should call Cerebras: in its recent earnings, Microsoft warned that its customers might see “service disruptions” as demand outpaces supply. This comes on the back of recent reports that the cloud giant is scaling back long-term data center spending. When confronted with this, Microsoft CFO Amy Hood pointed out that near-term and long-term demand aren’t the same thing.
Infrastructure
The one growing mainframe segment is IBM’s LinuxONE, which shares a base platform with IBM’s z-series mainframes. Following last month’s launch of the latest generation mainframe, the z17, IBM last week launched its latest LinuxONE platform. The new platform supports IBM’s Tellumand Spyre AI processors.
Nutanix closes a big VMware feature gap with its new support for external storage, which the company announced at its 2024 NutanixNext. This year, Nutanix delivered on the promise with Pure Storage emerging as its first dedicated external storage solution. We asked Nutanix CEO Rajiv Ramaswami why they chose Pure, to which he responded: “Pure is growing in the market segments we care most about.”
This follows last week’s announcement by Dell that its providing vxRAIL-like software-defined storage capabilities with Nutanix.
Deal Watch
In its fourth acquisition of the year, observability leader Datadog plans to acquire feature-management company Eppo. Observability has become a critical component in managing enterprise infrastructure and workloads. Datadog has been especially aggressive in consolidating functionality into its platform through targeted acquisitions.
There’s a widespread rumor that Databricks is close to a deal to acquire commercial PostgreSQL supplier Neon for more than a billion dollars. It’d be a big acquisition for Datadog, but one that makes sense. We’ll be watching to see if this one comes to fruition.
What We’re Reading
Who doesn’t love a good detective story? The closest thing we have to detectives in the cyber world are our big-time threat hunters. Bloomberg takes a fascinating look at one such group: The Elite Microsoft Unit Constantly Working to Thwart Hackers.
It doesn’t take a detective to guess where New York Magazine is going with its piece, Everybody is Cheating Their Way Through College.
Finally, there’s more to your work wardrobe than Casual Friday, as Reconteur finds as it looks at the history of work-wear by asking the question: When Wore it Best?
Companies mentioned: Alibaba, AWS, Broadcom, Cerebras, Cisco, CoreWeave, Databricks, Datadog, Dell, DXC, Dynatrace, Eppo, IBM, Meta, Microsoft, Nutanix, OpenAI, Pure Storage, Rapid7, VMware
