aiqbench analysis
Daysense's TianGai 300 IXL144 represents a significant advancement in AI infrastructure by packing 144 GPUs into a single cabinet to optimize large model training and inference costs. The solution leverages high-bandwidth interconnects to create a unified memory semantic computing domain, offering substantial efficiency gains over current international benchmarks while maintaining compatibility with existing CUDA ecosystems.
天数智芯 Launches Industry's Highest-Density Single-Cabinet Super-Node, Integrating 144 GPUs per Cabinet
At the 2026 Work Expo, TianShu Intelligent Chips officially launched its industry-leading highest-density single-cabinet super-node solution—the TianGai 300 IXL144—which integrates 144 GPUs and 18 CPUs per cabinet. The IXL144 is compatible with CUDA and other AI ecosystems, allowing customers to reuse existing models and frameworks while significantly reducing migration and deployment costs. By leveraging full-cabinet high-bandwidth interconnects to build a unified memory semantic computing domain, the solution effectively enhances collaborative efficiency for large model training and inference. Estimates indicate that the cost per million effective tokens is 40% lower than international mainstream solutions.
Read the full article at 36KrRelated coverage
News Archive 2026
Quantum Computing Report · Jul 25, 2026

News
Quantum Computing Report · Jun 7, 2026

NVIDIA GTC Taipei at COMPUTEX: Live Updates on What’s Next in AI
NVIDIA · May 21, 2026

[AINews] OpenAI reports Navier-Stokes singularity find in 88 hours using Astra-next, roughly 10,000 agents and 130B tokens (>$40M), a contender for second ever Millennium Prize awarded
Latent Space · Sep 9, 2026