Loading…
Enable Real-Time AI for High-Speed Data Acquisition with DAQIRI
NVIDIA Developer BlogCara Laasch
Summary
Scientific instruments and high-bandwidth sensors generate immense data volumes, but traditional architectures rely on a store-then-analyze model that bottlenecks real-time discovery and forces extensive data discarding. NVIDIA DAQIRI shifts data acquisition to a software-centric pipeline that connects streaming detectors directly to GPU computing. Built as part of the NVIDIA Holoscan Platform, DAQIRI uses DPDK to bypass the Linux kernel and move UDP and RoCE v2 traffic directly from the NIC into GPU DMA buffers at line rates exceeding hundreds of gigabits per second. The system automates packet reordering, flow steering, and on-GPU data conversions such as mapping wire int4 payloads into fp16 tensors using YAML configuration files and simple C++ or Python APIs. This enables real-time edge processing, filtering, and model inference across experimental pipelines like CERN ATLAS.
Context
Modern scientific instruments and detectors, such as LCLS-II and HL-LHC, produce extreme data rates that overload traditional collect-store-analyze architectures, forcing online systems to reject significant fractions of raw data due to kernel bottlenecks and CPU overhead.
Approach / What changed
NVIDIA DAQIRI provides a high-performance networking library using DPDK kernel-bypass to deliver zero-copy UDP and RoCE v2 packet streams directly from ConnectX NICs into GPU DMA buffers, assembling them into GPU-ready tensors via YAML configurations.
Takeaways
- DAQIRI completely bypasses the Linux kernel using DPDK to deliver zero-copy, line-rate data transfers from NICs directly to GPU DMA buffers.
- The architecture handles automated packet reordering, flow steering, and in-flight payload conversion, such as transforming wire int4 data to fp16 tensors on the GPU.
- The A-GHOST project at CERN uses DAQIRI to run AI models on ATLAS collision streams that would otherwise be discarded by nominal selection systems.
Related reading
NVIDIA Blackwell Tops MLPerf Training 6.0 with Industry-Leading Scale and Performance
NVIDIA submitted results across all MLPerf Training v6.0 benchmarks, setting performance records on workloads including the DeepSeek-V3 and GPT-OSS-20B Mixture of Experts models. To scale training up to 8,192 Blackwell GPUs, NVIDIA combined hardware cluster designs with scale-out networking via Spectrum-X Ethernet and Quantum InfiniBand. The engineering team eliminated CPU-GPU synchronization bottlenecks in token-dropless MoEs by implementing full-iteration CUDA graphs with synchronization-free operators and paged stashing. Further software optimizations included CuTe DSL kernel fusions, an MXFP8 attention block in the Transformer Engine, and transitioning router elementwise math to FP32. Across these benchmarks, the GB300 NVL72 platform achieved the fastest time to train at scale and the highest normalized per-accelerator performance.
Farshad Ghodsian