arXiv:2607. 11368v1 Announce Type: cross Abstract: Reported serving speedups from quantized kernels typically bundle the weight format, the kernel, and the inference runtime into one number.
Serving an LLM efficiently depends on the engine, how the model is split across GPUs, precision, scheduler settings, speculative decoding, KV-cache management, and prefill-decode disaggregation. The b...
arXiv:2608. 12123v1 Announce Type: cross Abstract: LLM-agent services repeatedly execute small deterministic transitions between model and tool calls: route an outcome, update state, and emit the next effect.
arXiv:2610.09372v1 Announce Type: new Abstract: Mixture-of-Experts (MoE) layers replace the feed-forward block of a Transformer with E expert networks, and each token is routed to k of these experts....
Mixture-of-Experts (MoE) layers replace the feed-forward block of a Transformer with E expert networks, and each token is routed to k of these experts. Under expert parallelism (EP) the experts are di...
arXiv:2606. 09682v1 Announce Type: new Abstract: AutoMegaKernel (AMK) compiles a HuggingFace Llama-family model into a single persistent cooperative CUDA kernel that runs the whole forward pass in one launch, with no per-model hand-written CUDA.
arXiv:2608. 05944v1 Announce Type: cross Abstract: We report operational experience full-fine-tuning a 32.
The paper investigates the performance impact of using a remote KV cache tier for long‑context language model serving. In a disaggregated deployment of GLM‑5.3 on 8xB300 nodes, enabling the remote tier initially increased the 16k‑token p99 latency from 3.96 s to 17.3 s, and synchronous loading caused severe queue instability. By sharding the KV store across context‑parallel ranks, partitioning the remote tier, adding asynchronous loading, and fixing client bugs, the authors reduced the p99 latency from 83.32 s to 0.67 s and increased prefill throughput by 14–34 %. whyItMatters":"The study demonstrates that careful sharding and asynchronous handling of remote KV caches can dramatically improve latency and throughput in large‑scale LLM serving systems."
arXiv:2607. 23264v1 Announce Type: cross Abstract: Fine-grained, device-initiated communication lets persistent GPU kernels in distributed diffusion transformer (DiT) inference issue remote stores and overlap data movement with Tensor Core computation.
The study evaluates the portability of INT8 post‑training quantization across seven hardware platforms, including CPUs, GPUs, and vendor NPUs, by keeping the ONNX model and quantization scales constant. It finds that INT8 performance and output consistency vary significantly: CPU dot‑product instructions determine speedup, identical INT8 outputs only occur when integer kernels match, and vendor NPUs require their own quantization pipelines. The authors also show that edge‑NPU latency is dominated by data transfer rather than compute and provide scripts and reports for reproducibility.
arXiv:2607. 14568v1 Announce Type: cross Abstract: A companion study ran a 35B mixture-of-experts model on a 2011 NVIDIA Tesla C2075 (Fermi, sm_20, 6GB) as a GPU-prefill/CPU-decode hybrid, because the 4-bit model did not fit in device memory (arXiv:2606.
ZYFA is a set of prefill kernels for AMD Instinct MI355X that accelerate FP8 FlashAttention by computing probabilities directly, using a ping‑pong thread schedule, and scaling probabilities before rounding. Compared with the vendor’s AITER kernels, ZYFA delivers 1.16–1.76× higher throughput on real model activations and up to 1.99× on benchmarked Gaussian inputs, while maintaining numerical error within a few percent of state‑of‑the‑art alternatives.