BIG TECH

Nvidia bets it has the CPU for the AI agent era

It's official. Nvidia is no longer just a GPU company. Its a full-stack AI company.

The chipmaker may have backed into its plum role as the chief supplier and kingmaker of the AI revolution, but no one can argue that it isn't seizing the moment. 

On Monday, Nvidia used the annual Computex event in Taiwan to unleash a flurry of announcements that ranged across the AI stack from CPUs to data centers to open models. Here's our quick analysis of the most important news: 

• Vera CPU now designed for agents: Nvidia declared that its new Vera CPU is specifically designed for agentic workloads, including tool use, writing code, and processing data. It claimed that Vera is a new type of CPU that can complete these tasks 1.8x faster than traditional x86 CPUs. It has signed up OpenAI, Anthropic, SpaceXAI, CoreWeave, Lambda, Dell, HPE, Lenovo, and others to use the chips.

• New platform for AI factories: The company launched the DSX software platform for designing, simulating, building and operating AI factories. That includes DSX MaxLPS, which optimizes power and cooling to run 40% more GPUs on the same power budget, and DSX OS, which the company is pitching to become the open-source operating system for AI factories. 

• AI supercomputer for Windows desktops: The new DGX Station for Windows can run AI models up to 1 trillion parameters locally for better performance, security, and cost savings. This is aimed at engineers, researchers, developers, and enterprises.

• New models for robots and robotaxis: The company announced Cosmos 3 (covered by The Deep View's Nat Rubio-Licht), a new world model that can work across text, images, video, sound, and actions to power physical AI and robotics. It also announced Alpamayo 2 Super, a new open reasoning model for robotaxis and H2 Plus, an open reference design for humanoid robots that combines hardware, software, and onboard compute.

• Vera Rubin hits full production: The next-gen AI infrastructure platform that combines GPUs, CPUs, networking, storage, and security and was unveiled at CES is ramping into full production across 350 factories in 30 countries. It delivers 10x higher agent throughout than the previous generation Grace Blackwell platform, Nvidia claims. Expect this to become the system that will power the world's largest and most powerful AI factories in the years ahead.

Pay special attention to the CPU news. Nvidia CEO Jensen Huang said, in a statement, “AI agents will be the largest users of computing. Vera is the first CPU designed for that future — built to run agentic AI at hyperscale with extraordinary performance, efficiency and programmability."

All the announcements have substance, but the Vera CPU announcement was arguable the most significant, as the GPU company pushes even further into the primary tech that has powered the biggest advances of the past 50 years. Thinking of Nvidia as a CPU competitor to Intel and ARM chipmakers such as Apple and Qualcomm still takes some getting used to. It's likely even more helpful to think of Nvidia as a full-stack AI company. It's pushing hard into other areas of AI because its platform advantage with CUDA, which put it in the plum role of AI kingmaker, is likely to evaporate in the next several years. Expect the Nvidia we're seeing at Computex 2026 to be the Nvidia we see from now on — a company reaching into lots of new areas to try to establish a foothold.

TOGETHER WITH SNYK

OWASP MCP Top 10: Securing the New Trust Boundary for Agentic AI

On June 4, Vandana Verma Sehgal will break down the OWASP MCP Top 10 using practical examples and real-world attack paths.

As AI agents gain access to more systems and workflows, security teams are struggling to govern what those agents can see, access, and execute.

Topics include:

• Prompt injection through tools and resources

• Unsafe tool execution and excessive agency

• Governance gaps and insecure integrations

Attend to earn 1 CPE credit and learn how to secure the new trust boundary for agentic AI.

PRODUCTS

Why ChatGPT won't stay a chatbot forever

In this episode of The Deep View Conversations, senior reporter Nat Rubio-Licht sits down with Ian Silber, Head of Product Design at OpenAI, to explore how ChatGPT is evolving for the future.

Silber explains why designing for AI requires a different mindset than designing traditional apps. Instead of treating the model as something behind the interface, he says designers now have to think of the model itself as part of the material they work with. That shift changes everything from product decisions and user experience to ethics, safety, and human judgment.

The conversation also covers Silber’s experience at Instagram, how that shaped his approach to building fast-growing consumer products, and how OpenAI’s design team is thinking about the next phase of ChatGPT, including more proactive and agentic experiences. Silber also shares how he personally uses AI in his work, from brainstorming design principles to prototyping ideas with Codex.

Topics covered:

• How OpenAI approaches product design for ChatGPT

• Why AI changes the traditional design process

• What designers can learn from fast-growing consumer products like Instagram

• How ethics, safety, and responsibility show up in AI design

• Why human judgment will become more important as AI tools improve

• How Codex and image generation are changing prototyping workflows

• What young designers should know as they enter the AI era

• Why curiosity and adaptability matter in a fast-changing industry

If you’re a designer, product leader, builder, technologist, or anyone trying to understand how AI is changing creative work, product development, and human decision-making, you don’t want to miss this episode.

📺 Watch on YouTube
🎧 Listen in your favorite podcast player

Subscribe to Deep View Conversations for interviews with the leaders shaping the future of AI, business, and technology.

TOGETHER WITH KORE

Build with AI: ABL is the only compilable agent language on the market.

Every new agent built on any other platform is a guess until it runs in production.

ABL — Agent Behavior Language — generates compiler-validated agents from intent. Structurally sound before they ever run. No other platform builds this way.

Stop re-engineering the foundation. Start shipping agents.

Watch the demo.

CULTURE

Exclusive: TwelveLabs’ bid to stop AI Slopageddon

Generative video models have led to a proliferation of AI slop on the internet. It’s why TwelveLabs is automating everything but the video. 

On Tuesday, the AI lab unveiled Rodeo, its first application-layer product that allows creatives and consumers to take advantage of its AI video editing capabilities. The tool allows users to find, edit and assemble footage with simple natural language prompts. 

Rodeo is built on TwelveLabs proprietary models, Marengo and Pegasus, which work in tandem to contextually understand a user’s footage, down to the emotion, pacing, dialogue, narrative structure and visual sequences.

• To start, users can import their footage, and Rodeo’s agents do the work of sifting and labeling for you. Then, users can simply describe the clips they’d like to retrieve, and Rodeo surfaces them.

• Once the clips are retrieved, users can drag and drop them into the platform’s video editor. Each clip also has an AI-suggested alternative, and a transcript if it contains dialog. 

• If the clips that Rodeo surfaced aren’t a fit, you can go back and refine the prompt to make it more specific. 

• Once you’re done, files can be exported and edited in post-production software formats, streamlining the hours it takes for video editors to sift through footage. 

In a demo, Ryan Khurana, lead of agents and applications at TwelveLabs, showed The Deep View the process of editing down footage with Rodeo, cutting down 40 hours of NASA footage into a 33-second sizzle reel in minutes. 

“I really believe in the thesis that people want to use AI not to just conjure videos from scratch, but as a way to take advantage of things they already have,” Khurana said.

The company called Rodeo a “natural progression” of its existing products, which are built directly into enterprise production pipelines, typically for unscripted TV production or documentary companies, and can’t be accessed by individual creatives. Rodeo, instead, targets nontechnical creators who require video-editing tools, such as YouTubers or influencers, a segment which Khurana said presents a “way bigger opportunity.” 

AI slop is quickly flooding the internet, with some estimates claiming that up to 50% of content is now AI-generated. Though bringing AI into the entertainment industry is something that artists and creators tend to balk at, Khurana said the goal is to reduce the amount of AI-generated slop that’s clogging our feeds, not increase it. “We would never want to build a system that took away creative judgment. That’s not what AI is good at.” 

His thesis is that by reducing the time required to collate and edit human-authored content, more authentic content will prevail. Instead of forcing creatives to spend hours weeding through footage, Khurana said, the goal is for Rodeo to supercharge their ability to produce more by casting off the time constraint. 

“When the demand for content is much higher than the supply of new content, obviously that gap is filled by a lower tier,” said Khurana. “If we can make the supply of real content increase by unblocking creatives, that provides a more level playing field for good content to get to more eyeballs.”

TwelveLabs’ thesis runs counter to the recent explosion in low-quality AI-generated slop in social media videos and ads. It’s even popping up in professional content, with an AI-generated film called “Hell Grind” debuting at Cannes. The prospect has Hollywood so wound up that the institutions behind the Oscars and the Golden Globes had to clarify their rules on how the tech can and can’t be used in order to be considered for nomination. But TwelveLabs' Rodeo is a clear example of how AI could actually make sense in entertainment. It has the potential to lower the barrier of entry for new creators and filmmakers and speed up the production process for professionals. Along with not seeking to substitute for human creativity, as Khurana said, fighting the rising tide of AI-generated content with higher-quality storytelling benefits the people consuming it, especially as it becomes increasingly difficult to differentiate real from synthetic. 

LINKS

• Amazon shuts down internal leaderboard to track employee AI usage  

• NYC AI startup will clean homes for free to use footage for robot training

• Asana acquires workflow automation firm StackAI for $75 million  

• Dell stock soars, bolstered by AI server revenue 

• ByteDance, InnoStar Semiconductor partner to develop new AI chips

• Windows PCs powered by Nvidia chips could launch this week

• Microsoft is looking to declare AI independence this week at Build 2026

• Microsoft Copilot: Redesigned app and how it shows up across Microsoft 365

• ElevenLabs: Introduced Dubbing v2, a ‘revolutionary new dubbing model’

• Lovable: Supports the new Claude Opus 4.8 model

• Gemini App: Updates were rolled out to make quotas stretch further

• Broadridge: VP, AI Quality Software Engineering

• EXL: Data/AI Client Engagement Lead - Insurance - Senior AVP

• SpaceX: AI Engineer, Platform Infrastructure, Special Programs

GAMES

POLL RESULTS

Is your company investing in AI agent deployments?

Yes (34%)
Somewhat (32%)
No (22%)
Other (12%)