Former Google and Meta Executives Launch Majestic Labs with $100M to Revolutionize AI Server Architecture

Former Google, Meta executives raise $100 million for high-capacity AI servers startup

Majestic develops its entire chip system, much like Nvidia, but each of the startup's servers may replace up to 10 of today's leading racks, the co-founders said. Three former Meta and Google silicon executives on Monday announced they've raised a total of $100 million to build technology they say will reduce cloud companies' spend on data center buildouts. Called Majestic Labs, the startup's co-founders are Ofer Shacham, Sha Rabii and Masumi Reynders, all of whom spent years working together leading silicon products at Meta and Google. Majestic's patent-pending silicon design architecture includes 1,000 times the memory of a typical enterprise-grade server, the co-founders told CNBC. Majestic develops its entire chip system, much like Nvidia, but each of the startup's servers may replace up to 10 of today's leading racks, they said. The company closed its $71 million Series A funding round in September, led by Bow Wave Capital. Among its others investors is Lux Capital. Prior to Monday's announcement, the co-founders had been quietly working on the startup since late 2023. Majestic's funding announcement comes as major tech companies raise their capital expenditures, primarily for data center infrastructure. Alphabet, Meta, Microsoft and Amazon each lifted their guidance for capital expenditures in October, and they collectively expect that number to reach more than $380 billion this year. While the majority of large language models and AI workloads have relied on Nvidia's graphics processing units, or GPUs, more companies are entering the fold. Google last week announced Ironwood, its latest tensor processing unit, or TPUs, which artificial intelligence startup Anthropic plans to use for its Claude model.

AI Server Startup Majestic Labs Comes Out of Stealth With $100 Mn Funding | AIM

Its patent- pending architecture promises 1000 times the memory capacity of a conventional enterprise server. Three former Google and Meta silicon executives have raised $100 million to launch Majestic Labs, a startup developing high-capacity AI servers designed to cut data centre costs for hyperscalers. Founded by Ofer Shacham, Sha Rabii, and Masumi Reynders, Majestic Labs is building patent-pending silicon architecture that promises up to 1,000 times the memory capacity of conventional enterprise servers. The company says its system can replace up to ten existing server racks, potentially reducing power, cooling, and space requirements for large AI data centers. Majestic's $71 million Series A, led by Bow Wave Capital with participation from Lux Capital, closed in September. The company has been quietly operating since late 2023 and now employs fewer than 50 people split between Los Altos, California, and Tel Aviv, Israel. Prototypes are expected in 2027, with pre-orders already under discussion. "We're not trying to replace GPUs across the board -- we're solving for memory-intensive AI workloads where the fixed compute-to-memory ratio becomes a constraint," Shacham, Majestic's CEO, told CNBC. The startup is targeting hyperscalers and enterprises in data-heavy industries such as finance and pharmaceuticals. The timing aligns with surging global investment in data centre infrastructure. Tech giants Alphabet, Meta, Microsoft, and Amazon together expect capital expenditures to exceed $380 billion in 2025, much of it driven by AI workloads. Majestic's architecture aims to "collapse" multiple racks of hardware into a single server, which the founders say could dramatically increase efficiency. "NVIDIA makes excellent GPUs and has driven incredible AI innovation," Shacham added, "but our goal is to complement that by solving for memory bottlenecks." The trio's collaboration dates back two decades. Reynders joined Google in 2003 and later became director of product management for silicon. Rabii, who sold his chip design startup Arda Technologies to Google in 201, led the Argos video chip team at YouTube. Shacham, who sold Chip Genesis to Google in 2013, oversaw silicon design for consumer hardware. In 2018, the three moved to Meta, where they founded the Facebook Agile Silicon Team (FAST). After layoffs at Reality Labs in 2023, they regrouped to start Majestic Labs, focusing on one of AI's biggest infrastructure bottlenecks -- memory. "We've been friends and colleagues for a long time, so this notion of working together and doing something exciting has always been in the periphery," Reynders said.

Majestic Labs launches with $100M to build memory-dense AI servers - SiliconANGLE

Majestic Labs launches with $100M to build memory-dense AI servers Artificial intelligence infrastructure startup Majestic Labs Inc. launched today with $100 million in initial funding. The company raised the bulk of the capital, $71 million, through a Series A round led by Bow Wave Capital. Lux Capital led Majestic Labs' earlier seed investment. The hardware maker is also backed by SBI, Upfront, Grove Ventures, Hetz Ventures, QP Ventures, Aidenlair Global and TAL Ventures. Majestic Labs is led by chip engineering executives Ofer Schacham, Masumi Reynders and Sha Rabii. Schacham, the company's Chief Executive Officer, earlier led the Google LLC team that develops chips for the search giant's consumer devices. Reynders and Rabii also held senior roles in the Alphabet Inc. unit's semiconductor group. Large language models generate a significant volume of temporary data while processing prompts. As a result, they have to be deployed on servers with a large memory pool that can accommodate their data. Moreover, LLMs require the ability to move data between a server's memory pool and graphics cards at high speed. Increasing an AI cluster's memory capacity often requires companies to add servers, which in turn necessitates deploying more supporting equipment such as cooling systems. That significantly increases hardware costs. Majestic Labs is tackling the challenge with a server that it says will include as much memory as 10 racks. A server requires a small fraction of the cooling equipment as a rack, which decreases procurement costs and power usage. Majestic Labs says that its system contains a "custom accelerator" chip and a likewise internally-developed memory interface module. The server can be equipped with up to 128 terabytes of RAM. Majestic Labs didn't specify what variety of RAM its system contains but described the technology as "extremely fast, power-efficient, high-bandwidth memory," which hints it might be using HBM memory. HBM is a particularly performant type of RAM that is ubiquitous in AI clusters. A standard RAM chip comprises a single layer of memory cells. An HBM module, in turn, contains upwards of a dozen memory cell layers that are stacked atop one another. The layers are linked together by millions of microscopic, copper-filled electric wires called through-silicon vias. Increasing the amount of memory in an AI cluster's servers can boost LLM performance. However, the size of the performance gain varies based on the inference task that the model performs. The computations an LLM carries between the moment a user enters a prompt and the first output token are more influenced by graphics card performance than memory speed. In contrast, the calculations that are used to generate subsequent output tokens can be significantly accelerated by adding more memory to an AI cluster. Majestic Labs' server can not only run inference workloads but also train new LLMs. "Our systems support vastly more users per server and shorten training time, lifting AI workloads to new heights both on-premises and in the cloud," Rabii said. CNBC reported that Majestic Labs plans to sell its hardware to hyperscale data center operators, financial institutions and pharmaceutical companies. The startup plans to ship its first servers in 2027. In the meantime, Majestic Labs will work to grow its engineering team, enhance the software stack that ships with its server and raise additional funding.