←
Matt Marshall1,758 posts
Matt Marshall avatar

Matt Marshall

@mmarshall

Founder, @VentureBeat

◷ Joined April 2007▧ 45 Photos and videos
Matt Marshall avatar
Matt Marshall@mmarshall·Oct 23

Even among companies running their own stack, 56% say the GPUs they use sit idle at least half the time. One reason: When an agent waits on a tool call or a user, GPU compute drops to zero while the session’s memory cache stays locked. Agents run out of memory before compute.

Matt Marshall avatar
Matt Marshall@mmarshall·Oct 23

Why? Companies are paying for two kinds of AI. They lined up GPU capacity out of fear of shortages, while their people ran up token spending on frontier models from OpenAI and Anthropic that can’t run on those GPUs. The full @VentureBeat analysis: venturebeat.com/infrastructu…

Matt Marshall avatar
Matt Marshall@mmarshall·Oct 23

The AI bubble isn’t where Wall Street is looking. It’s inside the enterprise. 79% of companies running AI say the GPUs they use sit idle at least half the time. Yet 70% of those companies plan new AI infrastructure, according to three months of VentureBeat Intelligence surveys.

Post attachment
Matt Marshall avatar
Matt Marshall@mmarshall·Oct 16

Reflection’s Ioannis Antonoglou also made the case against running Chinese models on your own servers. You can’t be sure a model wasn’t “trained to just activate certain behaviors in certain scenarios,” he told me. Then: “This is not to say that Chinese models are bad.”

Matt Marshall avatar
Matt Marshall@mmarshall·Oct 16

Reflection’s bet: Buyers will care more about the cost of each finished task than the price of each token. Nobody has published cost per task for these models yet, so we’re measuring it. My interview with Reflection’s technology chief: venturebeat.com/technology/r…

Matt Marshall avatar
Matt Marshall@mmarshall·Oct 16

America answered China’s lead in open AI models twice in the last 24 hours, with opposite bets. Meta’s Muse Spark Open is built for power. Reflection’s Beam, out today, was trained to use fewer tokens. On one coding test, it used about a third of the tokens of China’s GLM-5.2.

1:17
Matt Marshall avatar
Matt Marshall@mmarshall·Oct 13

Model switching is costly: 59% of companies that swapped the model in a live app had to rework the app or revert. Home Depot CIO Angie Brown: “One of the things is the context training that has to be done anytime there’s a switch, so there’s a stickiness level..."

Matt Marshall avatar
Matt Marshall@mmarshall·Oct 13

Companies pay for Claude. They run their agents on OpenAI. Ramp reports more businesses now pay Anthropic than OpenAI. Yet in VentureBeat’s September survey, twice as many companies run AI agents on OpenAI’s platform as on Anthropic’s: 43% to 22%.

Post attachment
Matt Marshall avatar
Matt Marshall@mmarshall·Sep 18

Caveats, because they matter. 169 respondents, all at companies with 100+ employees. Self-selected panel, not a probability sample. This wave skewed mid-market. 16% were at companies over 10,000 employees, against 51% the month before. August snapshot, not a trend.

Matt Marshall avatar
Matt Marshall@mmarshall·Sep 18

Primary platform, full August ranking: OpenAI 33% Google 24% Microsoft 13% Anthropic 11% Enterprise apps (Salesforce SAP, Workforce, etc) 11% Amazon 3% Anthropic is fourth. Behind Google and Microsoft. (@VentureBeat data)

Post attachment
Matt Marshall avatar
Matt Marshall@mmarshall·Sep 18

Anthropic leads on the forward-looking measure. Of the 121 enterprises shopping for an agent platform in the next year, 36 named Anthropic. (45 use it today). No other platform draws that much new interest relative to its current base. Most evaluated. Least committed to.

Matt Marshall avatar
Matt Marshall@mmarshall·Sep 18

Ramp and OpenRouter published data showing that OpenAI passed Anthropic this month. But that data tracked spend and tokens. We asked a different question. Which platform have you standardized on? Not what you're trying; what you're committed to. venturebeat.com/orchestratio…

Matt Marshall avatar
Matt Marshall@mmarshall·Sep 18

Enterprises run several AI agent platforms at once. Only one gets named the primary. Of the 75 companies running OpenAI's agent platform, 52 named it the primary. That's 69%. Of the 45 running Anthropic's, 17 did. That's 38%. New VentureBeat Research. 169 enterprises.

Post attachment