Ramp hits $60bn valuation in $1.85bn round, Bloomberg reports
THE SO WHAT
Ramp raising $1.85 billion at a ~$60 billion valuation tells you fintech infra that owns spend data and workflows is still getting premium multiples. If you sell into finance or ops, assume your customers are under pressure to consolidate tools around whoever controls the card and the ledger.
READ THE SOURCE
MORE FROM THE WIRE
Gen Alpha founders are already pitching VCs. Here's what investors want them to know.
If VCs are already pattern-matching on Gen Alpha founders, the bar for "young and technical" is about to rise again. For operators, this means more native-AI competitors who treat agents, social distribution, and global collaboration as defaults — assume faster copycats and shorter moats.
Startups & VentureSources detail how Firmus' IPO collapsed in 48 hours after US fund managers deemed its $30B valuation too rich for a company with just $51M in FY 2026 revenue
A 600x+ revenue multiple getting rejected in 48 hours is a clear line: even AI‑adjacent data center plays are being forced back toward fundamentals. If you’re raising, expect sharper scrutiny on revenue quality, not just “AI infra” narrative—clean up contracts, churn, and unit economics before you hit the road.
Startups & VentureSources: Nuvacore, a six-month-old Sequoia-backed chip startup that's designing a new central processor for data centers, is raising funds at a ~$2.5B valuation
A six-month-old data center CPU startup with no product raising hundreds of millions at a ~$2.5B valuation shows how much capital is hunting for alternatives to incumbent x86 and ARM in AI-era compute. If you’re planning long-lived infra, start tracking these new CPU roadmaps alongside GPUs—lock-in risk is rising on both sides of the socket.
Startups & VentureOxide Computer, which helps companies build their own clouds, raised a $445M Series D led by Eclipse at a $6B valuation, taking its total funding to ~$835M
$445M into Oxide at a $6B valuation is a clear bet that integrated on-prem clouds are a durable alternative to hyperscaler lock-in for AI-era workloads. If you’re planning heavy GPU or data residency spend, you now have to at least model a world where owning the rack — not just renting it — is financially rational.