8000
Skip to content
View manishklach's full-sized avatar

Block or report manishklach

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
8000
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
manishklach/README.md

Manish K L

AI infrastructure · GPU runtimes · memory systems · Linux

I build the machinery around models.

Writings Portfolio Patents

I design and prototype AI systems from the Linux and storage layers up through GPU kernels and inference control planes. My work focuses on a practical question: how do we make large-model inference faster, cheaper, and easier to operate?

The repositories here are executable architecture work—code, benchmarks, simulators, kernel RFCs, hardware interfaces, and production-oriented reference systems.

Selected systems

Project What it explores
Inference Factory Simulator Trace-driven LLM serving economics: continuous batching, disaggregated prefill/decode, speculative decoding, MoE balance, tiered KV reuse, GPU cost, and margin.
ExpertMesh A low-cost AMD/ROCm stationary-expert fabric for trillion-parameter sparse MoE inference, with packed-INT4 HIP kernels and activation-only RoCE transport.
KVSSD Attention SSD-resident INT2/INT4 KV cache with bounded asynchronous staging and fused CUDA/Triton dequantizing decode attention.
K3 Inference Platform A production-oriented MoE serving control plane: checkpoint release gates, capacity planning, admission, OpenAI-compatible routing, benchmarks, and observability.
GB300 RL Runtime A close-to-metal C/CUDA runtime with persistent GPU workers, hugepage KV arenas, command rings, and no per-token CPU orchestration.
AttnRes Kernel Lab Fused PyTorch/Triton kernels for routing across model depth and sequence time, including Hydra-2P, KDA, and CADR.
Linux HBF Control Plane An executable Linux memory-tiering RFC with asynchronous range hints, NUMA migration, tracepoints, selftests, and observability.
OpenFlash Controller Lab An executable NAND-controller architecture spanning FTL, scheduling, queue ABI, QEMU PCI emulation, and a Linux blk-mq driver.

The stack I work on

model algorithms     fused attention · routing · low precision
serving systems      admission · scheduling · release gates · economics
GPU runtimes         persistent execution · command rings · fast paths
memory systems       HBM/DRAM/NVMe tiering · KV placement · CXL/HBF
host + kernel        Linux MM · block I/O · NUMA · eBPF · observability
hardware interfaces flash controllers · RTL scaffolds · CPU/GPU fabrics

Current research direction

  • Treating inference as a factory, not a single kernel: utilization, placement, queueing, cost, and operational evidence all matter.
  • Moving fewer bytes: stationary experts, compact activations, quantized KV state, semantic reuse, and predictive residency.
  • Removing work from the token path: persistent GPU execution, bounded queues, preallocation, and explicit control-plane boundaries.
  • Making architecture claims testable through benchmarks, simulators, CI gates, tracepoints, and honest maturity labels.

More work

  • Linux inference fast path — eBPF, cgroups, NUMA/GPU locality, KV memory policy, and TTFT control.
  • AI host observability — Prometheus signals for GPU/RDMA hosts, PCIe, NUMA, memory pressure, IRQs, and kernel events.
  • RL inference scheduler — DQN scheduling against FIFO, SJF, and priority baselines.
  • CPUOpt Kernel — safe, reversible CPU performance profiles across Linux power and thermal backends.
  • MLX Metal Kernels — experimental Apple Silicon kernels for attention, decode, and KV-cache primitives.

Writing and patents

I pair systems code with architecture diagrams, RFC-style documents, and long-form technical writing. The portfolio includes 230+ essays and a record of 68 patents filed or granted worldwide across memory, storage, runtimes, and hardware–software interfaces.

Read the essays · Browse the portfolio · View the patent record · Follow on X

Popular repositories Loading

  1. kernel-dvfs-agentic-latency kernel-dvfs-agentic-latency Public

    Kernel latency control plane for agentic AI: DVFS, cpuidle, IRQ, scheduler, workqueue, I/O, MM, VFS, and cgroup budgets.

    C 4

  2. manishklach.github.io manishklach.github.io Public

    Portfolio hub for systems architecture, AI infrastructure, patents, and technical microsites.

    HTML 3

  3. ai-host-observability ai-host-observability Public

    Linux host observability toolkit for AI/GPU infrastructure, exposing Prometheus metrics for memory pressure, RDMA/NIC health, PCIe/VFIO, NUMA, GPUs, and kernel events.

    Shell 3

  4. mlx-metal-kernels mlx-metal-kernels Public

    Experimental MLX custom Metal kernels for Apple Silicon — fast attention, decode, KV-cache, and future Mac GPU inference primitives.

    Python 3 2

  5. chromelens chromelens Public

    Full-site performance X-ray powered by Chrome DevTools Protocol traces

    Python 2

  6. linux-hbf-control-plane linux-hbf-control-plane Public

    Experimental Linux RFC for an HBF/CXL-era AI memory control plane: runtime hints, prefetch, placement, and tiering.

    C 2

0