New Server Hopes to Break Through AI’s “Memory Wall”
- Published
- Source
- IEEESpectrumAI
Majestic Labs, an AI hardware startup, is addressing the memory limitations of large language models (LLMs) with its upcoming server, Prometheus, set to launch in 2027. This innovative server will feature up to 128 terabytes of memory, significantly surpassing the capabilities of Nvidia’s current offerings. Co-founder Sha Rabii emphasizes that this substantial memory increase will enhance performance and efficiency, particularly as models grow larger.
Prometheus employs a unique DRAM-centric architecture, utilizing LPDDR6 memory and a proprietary memory interface with miniature copper cables that allow for greater memory placement flexibility. This design aims to overcome the “memory wall” that hampers LLM performance, providing a memory bandwidth of up to 25.6 terabytes per second.
To complement its memory capabilities, Prometheus will incorporate the Ignite AI processing unit, which combines ARM application cores with RISC-V vector and tensor cores on a single chip. This integration allows for seamless handling of LLM inference tasks without the need for processor handoffs.
Majestic Labs is also focused on ensuring compatibility with existing AI frameworks like PyTorch and OpenAI’s Triton, allowing customers to run their models without modifications. The server, designed in compliance with the Open Compute Project, will be modular, enabling future memory upgrades.
Despite the advanced technology, Majestic Labs aims to offer competitive pricing by leveraging DRAM instead of more expensive high-bandwidth memory. Rabii claims that this approach could reduce customer capital expenditures and power consumption significantly, potentially by 10 to 50 times, depending on the workload.
Original report
New Server Hopes to Break Through AI’s “Memory Wall”
IEEESpectrumAI · Matthew S. SmithThis briefing is an independently written summary based on publicly available reporting and is provided for industry information and news discovery. The original report and source publication are credited and linked where applicable. RobotToday does not claim ownership of third-party source material.
Rights concerns? Contact [email protected] with the relevant URL and details. We will review the matter and take appropriate action where warranted.
More related news
- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9
- 10
- 11
- 12
Related Suppliers
BioRobotics Laboratory, Korea University of Technology and Education
Korea, Republic of (South)
South Korean university research laboratory focusing on haptics, exoskeletons, rehabilitation robotics, and autonomous ground vehicles.
McMaster Manufacturing Research Institute
Canada
Hamilton, Ontario applied manufacturing research center at McMaster University; runs a 21,000 sq ft facility with seven research labs.
Robotics Education & Competition Foundation
United States
The REC Foundation is a US nonprofit that runs VEX-based robotics and drone competitions reaching 1.1 million students across 70 countries.
Ortelio Ltd
United Kingdom
Ortelio Ltd is a UK cloud robotics research company founded in 2011 that develops AI platforms giving robots intelligence for service and healthcare tasks.
Parata Systems
United States
Durham, North Carolina pharmacy automation maker acquired by BD in 2022; builds robotic vial-filling and packaging systems.
Pensa Systems
United States
Austin-based AI retail shelf platform using autonomous drones and CV to detect stockouts; customers include J&J, AB InBev, and Circle K.
Datanomix
United States
No-Operator-Input production monitoring software that turns CNC machine data into real-time visibility and prescriptive coaching for manufacturers.
Datatang 数据堂（北京）科技股份有限公司
China
Beijing AI training-data supplier founded 2010 and quoted on NEEQ; its Nexdata arm sells egocentric, teleoperation and hand-pose datasets for robot learning.
Anyscale
United States
Anyscale is a San Francisco AI infrastructure firm behind Ray, the open-source engine for large-scale model training; Nscale agreed to buy it for $1.65B.
Robotics needs a service framework.
RSF defines a common language for robot service capability, lifecycle operations, certification pathways, and service-provider networks.
Join the RobotToday community on LinkedIn
Daily robotics news, in-depth analysis, conference highlights, and discussions with professionals worldwide.
