Majestic Labs’ $100M Push to Smash the AI Memory Limit

Harry Eaton
Majestic Labs Raises $100M to Tear Down AI's Memory Wall

Three ex-Google and Meta chip executives just raised over $100 million to solve one of AI's biggest headaches: the memory bottleneck choking data centers worldwide.

Majestic Labs emerged from stealth in November 2025 with a bold promise to pack 1,000 times more memory into a single server than today's top GPUs can handle.

The Team Rewriting AI Infrastructure

Majestic Labs was founded in late 2023 by Ofer Shacham (CEO), Masumi Reynders (COO), and Sha Rabii (President) — a powerhouse trio who built Meta's FAST (Facebook Agile Silicon Team) and led Google's GChips division.

Together, they hold over 120 patents and have shipped hundreds of millions of custom silicon units, including the world's first mobile AI processors and Meta's augmented reality compute platform.

After Meta's 2023 layoffs forced Shacham to shut down parts of FAST, the three founders saw an opportunity to tackle AI's most expensive problem: memory capacity.

Majestic Labs Funding Overview

RoundAmountLead InvestorYearValuation Estimate
Seed$10MLux Capital2024Undisclosed
Series A$90MBow Wave Capital2025$300-400M
Total$100M+$300-400M

Why This $100M Bet Changes the Game?

The funding breaks down as a $10 million seed round led by Lux Capital in 2024, followed by a $90 million Series A in September 2025 led by Bow Wave Capital.

Joining them are SBI, Upfront, Grove Ventures, Hetz Ventures, QP Ventures, Aidenlair Global, and TAL Ventures. Industry sources estimate Majestic's valuation at $300-400 million.

AI models now demand memory measured in terabytes, but even Nvidia's best GPUs max out at 192 gigabytes.

  • Majestic's servers flip that equation: each unit delivers 128 TB of high-bandwidth memory and can replace ten conventional server racks.

The result? Over 50x performance gains, drastically lower power consumption, and a data center footprint slashed by orders of magnitude.

This matters now because hyperscalers like Google, Meta, Microsoft, and Amazon are on track to spend $380 billion on data center infrastructure this year alone.

Key Investment Details

ParameterDetails
Funding StageSeries A
Total Raised$100M+
Valuation$300-400M (est.)
Target MarketHyperscalers, financial institutions, pharma
Key Innovation1,000x memory capacity vs. standard GPUs
AICurator Rating4.5/5 ⭐

AICurator Rating based on: founder pedigree (5/5), market timing (4.5/5), technical innovation (5/5), funding strength (4/5)

What's Next for Majestic Labs?

The fresh capital will scale Majestic's team of AI scientists, silicon engineers, and systems architects while pushing the product toward general availability.

Pilot deployments with hyperscalers and enterprise clients in finance and pharma are already underway to validate real-world performance.

The company faces stiff competition from Nvidia's GPU dominance and Google's in-house TPUs, but Majestic's singular focus on the memory-compute imbalance positions it to capture the massive market created by AI's explosive growth.

As large language models explode in size and complexity, Majestic's architecture could unlock AI capabilities that are currently impossible or too expensive to deploy at scale.​

Connect With Majestic Labs

PlatformLink
Websitemajestic-labs.ai
LinkedInlinkedin.com/company/majestic-labs-ai
LocationSan Francisco, CA & Tel Aviv, Israel
Share This Article
Leave a review