Applied Compute 支持 Nemotron 3.5 Lightning 了，并发翻 16 倍延迟不涨，训练迭代快多了。
Applied Compute 平台现已支持 NVIDIA 的 Nemotron 3.5 Lightning 进行训练和推理。该模型在 agentic coding benchmark 中，当并发和总 token 吞吐量扩展 16 倍时，解码吞吐量、首 token 时间和用户中位延迟保持不变。其 LatentMoE 和 Mamba 架构能以最小开销扩展稀疏性、上下文长度和批大小，显著提升后训练迭代吞吐量。
https://t.co/JYcidGNmvS
x.com/appliedcompute… Applied Compute @appliedcompute Nemotron 3.5 Lightning by @NVIDIAAI is now supported for training and inference on the Applied Compute Platform. On our agentic coding benchmark, decode throughput, time to first token, and median user latency remained effectively unchanged as concurrency and total token throughput scaled 16x. Its LatentMoE and Mamba architecture lets us scale sparsity, context length, and batch size with minimal overhead, dramatically increasing iteration throughput across post-training runs. 🔗 View Quoted Tweet 💬 0 🔄 0 ❤️ 1 👀 88 ⚡