Results

SambaNova and Intel outlined an inference system that will combine GPUs, SambaNova RDUs for decode and Intel Xeon 6 CPUs for agentic tools and system orchestration. The system will be available to enterprises and cloud platforms in the second half.

As part of the partnership, SambaNova said it will standardize on Xeon 6 to go with its RDUs on the inference system. The system is part of SambaNova's blueprint for inference.

SambaNova recently launched its SN50 AI inference chip that runs agentic AI 3x cheaper than GPUs and said SoftBank will be the first to deploy the chip in its AI data center in Japan. The company also said it raised $350 million in Series E funding.

SambaNova, Intel inference

Anthropic's development of its Claude frontier models has spooked the cybersecurity industry (or at least investors) much in the same way the LLM player has freaked out the SaaS industry. Now with the launch of Project Glasswing and the Claude Mythos Preview, Anthropic is securely entrenched in the cybersecurity value chain. As a result, its total addressable market just went up. So will Anthropic's $30 billion annual revenue run rate.

Coupa and AWS said they have signed a 5-year deal to partner on spend management data and apps. Under the partnership, the two companies will combine Coupa's business intelligence network with AWS AI tools. The goal is to deliver autonomous direct and indirect spend data across sourcing, payment to supply chain.

The two companies said AWS customers will be able to deploy Coupa Navi AI agents, which are built on Amazon Bedrock, to automate procurement workflows.

Today’s reading list has a common theme: Turmoil in the land of AI infrastructure. Here’s a trio of links:

The Information: OpenAI CEO and CFO Diverge on IPO Timing: The gist is this: OpenAI CEO Sam Altman wants to go public ASAP to fund his GPU promises. CFO Sarah Friar has argued that the company isn’t ready.

WSJ: An Inside Look at OpenAI and Anthropic’s Finances Ahead of Their IPOs: The money paragraph: The costs are set to be so mind-bogglingly high that both companies report two different measures of profitability—one that includes training costs, and one that leaves them out.

In other words, all you have to do is strip out the compute costs and these companies look pretty good. I just spit out my coffee writing that.

How Microsoft Vaporized a Trillion Dollars: A Substack from former Microsoft engineer on Azure’s architecture and scaling problems.

Here's an interesting post on org charts in the agentic AI age from Block's Jack Dorsey and Roelof Botha. A few observations after a lot of long talking:

  • Block sees three roles in the org chart. Individual contributors, directly responsible individuals on the hook for outcomes and player coaches.
  • "There is no need for a permanent middle management layer. Everything else the old hierarchy did, the system coordinates, and everyone is empowered, with a role that's much closer to the work and the customer."
  • "At Block, we're questioning the underlying assumption: that organizations have to be hierarchically organized with humans as the coordination mechanism. Instead, we intend to replace what the hierarchy does. Most companies using AI today are giving everyone a copilot, which makes the existing structure work slightly better without changing it. We're after something different: a company built as an intelligence (or mini-AGI)."

Amazon said it will add a 3.5% fuel and logistics-related surcharge across Fulfillment by Amazon (FBA) in the US and Canada as well as to Remote Fulfillment with FBA from the US into Canada, Mexico, and Brazil. The surcharge starts April 17.

Buy with Prime in the US and Multi-Channel Fulfillment (MCF) will wee the surcharge kick in on May 2. The surcharge applies to fulfillment fees not the sale price of items.

Cursor launched Cursor 3 and will more directly compete with Anthropic's Claude Code and OpenAI's Codex by giving developers the ability to manage multiple AI agents. In a blog post, Cursor said:

"The new Cursor interface brings clarity to the work agents produce, pulling you up to a higher level of abstraction, with the ability to dig deeper when you want. It's faster, cleaner, and more powerful, with a multi-repo layout, seamless handoff between local and cloud agents, and the option to switch back to the Cursor IDE at any time."

The interface of Cursor 3 provides access to multiple workspaces, runs agents in parallel and hands off between local and cloud environments.