SLM is a frontier compute infrastructure company in California focused on advanced-capability systems. We are currently operating in stealth, and this site's copy has intently been written for opacity.
We're developing architectures for environments where the physics and operational requirements are especially forbidding. Our first-generation timelines are targeted at high-performance specifications in constrained packages combining advanced heterogeneous integration process nodes with physically resilient design and error correction.
This is a new class of high-performance architectures that we've been able to work on in hyper-stealth for just over three years of applied R&D in miniaturized supercomputer, advanced materials, circuit complexity, and superchip design.
These systems target complex and compute-intensive problems with unmatched compute, higher parallelism, fault-tolerance, and scalability than current state-of-the-art architectures that cannot maintain the same performance across both conventional deployments and outside geomagnetic shielding, beyond Earth's magnetosphere, and in the most demanding terrestrial operating conditions.
Be part of our teamOur focus
We’re working on advanced specifications and want to find exceptional, mission-focused engineers, researchers, and generalists who want to do important work that pushes real hardware systems to their physical limits and operate them at the processing edge. During our protocol timelines, we’re scaling R&D, testing, and manufacturing across computer architecture, nanofabrication and advanced materials, hardware-software codesign, and operations.
We fixate a lot on advanced systems and precision, and are particularly intent on finding people who share this felt rigor, obsession, and optimism for advanced-capability systems.
Some of our heuristics
System architecture: Specialized processor designs combining massive parallelism, accelerated math operations, and novel algorithm and domain-specific workloads. This includes custom circuit topologies optimized for computational density, advanced numerical methods, and adaptive performance under physical constraints.
Hardware integration: Integrated circuit layers with high-bandwidth memory systems, efficient data handling, hardware-software cryptography and encryption, and advanced-material interconnect fabrics.
How we operate
We are a frontier technology company with systems and capabilities scaled through targeted research and engineering.
Our mission has very high-budget, hard inputs. This requires focused, timeline-directed R&D, and we marshal significant resources to build and scale our capabilities. Some of our teams are focused on near- and medium-term systems, while others specifically concentrate on self-operated infrastructure projects that run over very long-term but rapid cycles. Both our near- and long-term priorities are targeted at sustained deployments and introducing critical, connected capabilities through our systems.
Our private, national-interest structure allows us to vertically integrate deeply, do institutional interfacing closely, and focus on continuously producing unparalleled models for real complex missions with our customers and partners.