Linux Kernel Engineer
Hyderabad, India
Full timeOn-site
Job description
About Us
We are a tech company specializing in the design and development of cutting-edge, customized server hardware solutions optimized for artificial intelligence and machine learning applications. Our mission is to empower businesses and researchers to accelerate their AI initiatives by providing them with high-performance, scalable, and energy-efficient hardware infrastructure.
As a rapidly growing company at the forefront of AI hardware innovation, we are constantly seeking talented and motivated individuals to join our team. We offer a dynamic and challenging work environment, with opportunities to make a significant impact on the future of AI technology.
You'll Collaborate With
Firmware, silicon, runtime, datacenter software and architects teams to deliver a robust kernel foundation that integrates seamlessly with modern datacenter environments and AI workloads.
What You'll Own
-
PCIe Foundation: Design and implement robust PCIe kernel drivers for AI accelerator enumeration, configuration, and management that form the critical host-device communication backbone.
-
High-Speed Data Movement: Develop DMA engines and memory management systems that deliver zero-copy, high-throughput data movement between host and accelerator for massive AI workloads.
-
Firmware Bridge: Build interrupt handling (MSI/MSI-X) and mailbox communication protocols that create seamless coordination between kernel space and our firmware stack.
-
System Observability: Create comprehensive sysfs/debugfs interfaces for device configuration, telemetry, and diagnostics that give operators deep visibility into accelerator operations.
-
Multi-Die Orchestration: Support multi-die topologies through device discovery, link management, and topology coordination that scales from single-card to cluster deployments.
-
Power & Thermal Intelligence: Develop power management and thermal control integration with Linux PM frameworks that optimize performance while maintaining reliability.
-
Reliability Engineering: Contribute to device health monitoring, error handling (RAS), and recovery mechanisms that deliver the 99.99% uptime datacenter customers require.
-
Performance Optimization: Optimize kernel-space performance for AI workloads, focusing on latency, throughput, and scalability that directly impacts model training and inference speed.
-
Time Synchronization: Support PTP/PHC time synchronization for distributed training and inference that ensures coordination across multi-node AI clusters.
-
Runtime Partnership: Collaborate with userspace runtime teams on kernel-userspace interfaces and APIs that enable efficient scheduling and resource sharing.
-
Future-Ready Architecture: Lay the groundwork for device virtualization and multi-tenant isolation, ensuring our platform can evolve with customer needs.
Minimum Qualifications
-
5+ years developing Linux kernel drivers and subsystems in C for complex SoCs or accelerators
-
Proven experience with PCIe device driver development (enumeration, BARs, DMA, interrupts)
-
Strong knowledge of Linux memory management, DMA mapping, and IOMMU integration
-
Hands-on experience with kernel synchronization primitives, workqueues, and interrupt handling
-
Solid understanding of Linux device model, sysfs/debugfs, and driver lifecycle management
-
Familiarity with kernel debugging tools: ftrace, perf, crash, kprobes, and hardware debuggers
-
Ability to read hardware specifications and work with RTL/hardware teams on register interfaces
-
Excellent documentation habits; comfortable with kernel development processes and upstream contribution
Preferred Qualifications
-
Experience with AI/ML accelerator drivers or GPU compute drivers (NVIDIA, AMD, Intel)
-
Implemented RDMA/RoCE drivers or high-speed networking kernel subsystems
-
Background in PTP/PHC time synchronization or distributed system timing
-
Multi-die/chiplet device driver experience and topology management
-
Linux kernel upstream contributions and familiarity with kernel community processes
-
Experience with kernel security features, SELinux integration, and device isolation
-
Knowledge of container/runtime integration with kernel drivers (Docker, Kubernetes)
-
Performance optimization experience for high-throughput, low-latency kernel subsystems
-
Familiarity with RAS concepts, ECC handling, and system reliability in kernel space
What Success Looks Like (First 6–9 Months)
-
PCIe driver is stable and reliable across multiple system configurations and stress tests
-
DMA and memory management achieve target throughput with low CPU overhead
-
Multi-die topology management and inter-card communication is functional and performant
-
Integration with firmware mailbox protocols and telemetry collection is operational
-
Performance benchmarks meet targets for key AI workloads (inference and training)
-
Documentation and test coverage enable team-wide development and deployment
Ready to accelerate the future of AI from the firmware up?
Submit your application with links to relevant firmware projects, papers, or upstream contributions demonstrating secure boot, update/recovery, PCIe bring-up, or runtime management. Show us how you think systematically about firmware architecture and what you've built that scales.
Join us in our mission to democratize AI compute — where your firmware expertise becomes the bedrock of tomorrow's AI breakthroughs.