arXiv:2603. 09555v2 Announce Type: replace-cross Abstract: High-throughput Mamba-2 inference is usually tied to fused CUDA and Triton kernels, limiting portability across accelerator backends.
arXiv:2605. 25645v2 Announce Type: replace-cross Abstract: We present the first end-to-end demonstration of fine-tuning and serving Google's Gemma 4 31B model on TPU hardware, providing an empirical comparison of TPU and GPU platforms for large language model adaptation.
arXiv:2609.34818v2 Announce Type: replace-cross Abstract: AlphaFold2 is written in JAX, so the same inference code compiles and runs unchanged on CPUs, GPUs and Google Cloud TPUs. That portability ma...
arXiv:2607. 22785v1 Announce Type: cross Abstract: Apple-Silicon SoCs share CPU, GPU, and Neural Engine over one unified memory system, raising the question of whether transformer inference can be accelerated by splitting single operators across units.
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:2604. 26968v2 Announce Type: replace-cross Abstract: Key-value (KV) cache memory management is the primary bottleneck limiting throughput and cost-efficiency in large-scale GPU inference serving.
arXiv:2606. 28565v1 Announce Type: cross Abstract: As large language models (LLMs) move into production serving, practitioners must rapidly evaluate inference performance across diverse hardware, models, and serving parameters to meet cost and latency targets.
arXiv:2609.17475v1 Announce Type: new Abstract: Capable open-weight models make local coding and reasoning attractive, but their context and execution state strain laptop memory. We present JustFit,...
arXiv:2608.23841v1 Announce Type: cross Abstract: Single-token autoregressive decode on CPUs is bound by memory bandwidth, not arithmetic: a modern CPU sustains roughly 1 TFLOP/s of compute but only...
The paper addresses the problem of non‑deterministic outputs from large language models (LLMs) when run on different GPU architectures, caused by floating‑point non‑associativity and hardware‑dependent kernel choices. It proposes a set of fixed‑configuration fused‑upcast GEMM kernels that load 16‑bit weights, upcast to FP32, and perform IEEE‑754 compliant reductions in a problem‑shape‑dependent order, ensuring identical linear‑layer outputs across NVIDIA Ampere, Ada, and Hopper GPUs. The new approach achieves 1.17–3.1× faster end‑to‑end performance than existing solutions and halves weight‑memory traffic while maintaining cross‑architecture 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.
The paper investigates whether training Mixture-of-Experts (MoE) routers can improve memory‑bandwidth locality on consumer GPUs. Using a new zero‑surgery telemetry tool, the authors measure that a large Qwen3‑235B model is bottlenecked by disk‑based expert access, and that an LRU cache can serve a majority of requests. They pre‑register experiments training 137 M‑parameter MoE models with locality‑aware losses, finding that while cache misses can drop up to 60 % (99 % static‑pin hit rate), every configuration fails to meet a strict 1 % perplexity threshold, indicating a tight coupling between cache efficiency and model quality.
arXiv:2607. 14431v1 Announce Type: cross Abstract: We report a way to make a frozen small language model both more capable and dramatically cheaper at once, without changing any weights.