Search: “hardware”

AutoLLMResearch: Training Research Agents for Automating LLM Experiment Configuration -- Learning from Cheap, Optimizing Expensive

arXiv:2605.11518v1 Announce Type: cross Abstract: Effectively configuring scalable large language model (LLM) experiments, spanning architecture design, hyperparameter tuning, and beyond, is crucial f

Hardware

BOOST: BOttleneck-Optimized Scalable Training Framework for Low-Rank Large Language Models

DGX agent

arXiv:2512.12131v2 Announce Type: replace Abstract: The scale of transformer model pre-training is constrained by the increasing computation and communication cost. Low-rank bottleneck architectures o

hardwarearxiv-cs-lg

13 May 2026

ChunkFlow: Communication-Aware Chunked Prefetching for Layerwise Offloading in Distributed Diffusion Transformer Inference

arXiv:2605.11335v1 Announce Type: cross Abstract: Layerwise offloading reduces the GPU memory footprint of large diffusion transformer (DiT) inference by prefetching upcoming layers from host memory,

CME Group and Silicon Data to launch AI compute futures market

Silicon Data, the startup that provides market intelligence for artificial intelligence compute infrastructure, will provide the price indexes for a new futures market that will allow investors to hed

hardwaresiliconangle

Efficient Remote KV Cache Reuse with GPU-native Video Codec

arXiv:2602.09725v3 Announce Type: replace-cross Abstract: Remote KV cache reuse fetches KV cache for identical contexts from remote storage, avoiding recomputation, accelerating LLM inference. While i

Fast MoE Inference via Predictive Prefetching and Expert Replication

arXiv:2605.11537v1 Announce Type: new Abstract: The Mixture of Experts (MoE) architecture has become a fundamental building block in state-of-the-art large language models (LLMs), improving domain-spe

How the 'Facebook House' in Los Altos, Mark Zuckerberg's former residence, became a hub for Chinese AI talent, who are a key part of Silicon Valley's AI boom (Viola Zhou/Rest of World)

Viola Zhou / Rest of World: How the “Facebook House” in Los Altos, Mark Zuckerberg's former residence, became a hub for Chinese AI talent, who are a key part of Silicon Valley's AI boom — Chinese-born

NVIDIA, Ineffable Intelligence Team Up to Build the Future of Reinforcement Learning Infrastructure

Reinforcement-learning agents — AI systems that learn by trial and error — can convert computation into new knowledge. That’s the focus of a new engineering-level collaboration between NVIDIA and Inef

hardwarenvidia-blog

Recursive Superintelligence raises $650M to build self-improving AI models

Recursive Superintelligence Inc., a startup that hopes to develop self-improving artificial intelligence models, launched today with 650 million in funding. Alphabet Inc.’s GV fund and Greycroft led t

hardwaresiliconangle

Red Hat and Intel spotlight scalable AI inference as enterprises move beyond the GPU gold rush

As companies move from testing AI to broader adoption, the biggest challenge is building scalable AI inference systems that perform without breaking the budget. The next wave of AI won’t be won on raw

hardwaresiliconangle

Richard Socher's Recursive Superintelligence raised 650M+ from GV, Greycroft, Nvidia, AMD, and others at a 4B valuation to pursue 'recursive self-improvement' (Cade Metz/New York Times)

Cade Metz / New York Times: Richard Socher's Recursive Superintelligence raised 650M+ from GV, Greycroft, Nvidia, AMD, and others at a 4B valuation to pursue “recursive self-improvement” — Recursive S

The Illusion of Power Capping in LLM Decode: A Phase-Aware Energy Characterisation Across Attention Architectures

arXiv:2605.11999v1 Announce Type: cross Abstract: Power capping is the standard GPU energy lever in LLM serving, and it appears to work: throughput drops, power readings fall, and energy budgets are m

To Err Is Human; To Annotate, SILICON? Toward Robust Reproducibility in LLM Annotation

arXiv:2412.14461v4 Announce Type: replace Abstract: Unstructured text data annotation is foundational to management research. LLMs offer a cost-effective and scalable alternative to human annotation,

Transform Video Into Instantly Searchable, Actionable Intelligence with AI Agents and Skills

NVIDIA's video analytics AI agents analyze and process large volumes of video data through natural language tasks to provide critical insights , powered by vision language models, large language model

hardwarenvidia-developer

TriBand-BEV: Real-Time LiDAR-Only 3D Pedestrian Detection via Height-Aware BEV and High-Resolution Feature Fusion

arXiv:2605.12220v1 Announce Type: new Abstract: Safe autonomous agents and mobile robots need fast real time 3D perception, especially for vulnerable road users (VRUs) such as pedestrians. We introduc

AAAC: Activation-Aware Adaptive Codebooks for 4-bit LLM Weight Quantization

arXiv:2605.08692v1 Announce Type: cross Abstract: Post-training weight-only quantization to 4 bits is widely used to reduce the memory and compute costs of large language model inference. Existing PTQ

After studying 300 Leetcode Hards, solving every Jane Street puzzle from the Dwarkesh ads, and watching one Horace He lecture, he finally la…

After studying 300 Leetcode Hards, solving every Jane Street puzzle from the Dwarkesh ads, and watching one Horace He lecture, he finally landed the $400k annualized Jane Street internship. Unfortunat

hardwaredylan-patel--x

CellDX AI Autopilot: Agent-Guided Training and Deployment of Pathology Classifiers

arXiv:2605.10362v1 Announce Type: new Abstract: Training AI models for computational pathology currently requires access to expensive whole-slide-image datasets, GPU infrastructure, deep expertise in

Cluster magicians and GPU whisperers, come join us! We’re looking for supercomputing engineers to build the infrastructure behind real-time …

hardwaresoumith-chintala--x

Cluster magicians and GPU whisperers, come join us! We’re looking for supercomputing engineers to build the infrastructure behind real-time interactive models, Tinker, and large-scale training: schedu

CME Group and Silicon Data announce a futures market for computing capacity, with contracts based on daily GPU benchmarks for on-demand rental rates (Tobias Burns/CNBC)

Tobias Burns / CNBC: CME Group and Silicon Data announce a futures market for computing capacity, with contracts based on daily GPU benchmarks for on-demand rental rates — A new futures market for sem

Did Jensen Huang catch conflict of interest disease from Sam?

Did Jensen Huang catch conflict of interest disease from Sam? HUANG FOUNDATION SIGNS GPU COMPUTE DEAL WITH COREWEAVE $NVDA proxy says the charitable foundation tied to Jensen and Lori Huang entered an

hardwaregary-marcus--x

Efficient Dataset Distillation for Pre-Trained Self-Supervised Models via Statistical Flow Matching

arXiv:2602.05391v2 Announce Type: replace Abstract: Dataset distillation seeks to synthesize a highly compact dataset that achieves performance comparable to the original dataset on downstream tasks.

Energy Consumption of Dataframe Libraries for End-to-End Deep Learning Pipelines:A Comparative Analysis

arXiv:2511.08644v3 Announce Type: replace-cross Abstract: This paper presents a detailed comparative analysis of the performance of three major Python data manipulation libraries - Pandas, Polars, and

FlashSVD v1.5: Making Low-Rank Transformers Inference Actually Fast

arXiv:2605.08314v1 Announce Type: cross Abstract: SVD-based Low-rank compression reduces transformer parameters and nominal FLOPs, but these savings often translate poorly into real LLM serving speedu

Forcing-KV: Hybrid KV Cache Compression for Efficient Autoregressive Video Diffusion Models

arXiv:2605.09681v1 Announce Type: new Abstract: Autoregressive (AR) video diffusion models adopt a streaming generation framework, enabling long-horizon video generation with real-time responsiveness,

Geometric 4D Stitching for Grounded 4D Generation

arXiv:2605.09984v1 Announce Type: cross Abstract: Recent 4D generation methods complete scene-level missing information using generative models and reconstruct the scene into radiance-based representa

GPU-Accelerated Synthesis of Mixed-Boolean Arithmetic: Beyond Caching

arXiv:2605.08243v1 Announce Type: cross Abstract: Synthesizing Mixed-Boolean Arithmetic (MBA) expressions from input-output examples is central to program deobfuscation and also useful for compiler op

Leveraging LLMs to Automate Energy-Aware Refactoring of Parallel Scientific Codes

arXiv:2505.02184v3 Announce Type: replace Abstract: Large language models (LLMs) are increasingly used for generating parallel scientific codes, with a primary focus on generating functionally correct

mHC-SSM: Manifold-Constrained Hyper-Connections for State Space Language Models with Stream-Specialized Adapters

arXiv:2605.08300v1 Announce Type: cross Abstract: Manifold-Constrained Hyper-Connections (mHC) introduce a stability-motivated variant of multi stream residual mixing by constraining residual stream m

Model-Aware Tokenizer Transfer

arXiv:2510.21954v2 Announce Type: replace Abstract: Large Language Models (LLMs) are trained to support an increasing number of languages, yet their predefined tokenizers remain a bottleneck for adapt

Not All Thoughts Need HBM: Semantics-Aware Memory Hierarchy for LLM Reasoning

arXiv:2605.09490v1 Announce Type: new Abstract: Reasoning LLMs produce thousands of chain-of-thought tokens whose KV cache must reside in scarce GPU HBM. The dominant response -- permanently evicting

Novel GPU Boruta algorithms for feature selection from high-dimensional data

arXiv:2605.09950v1 Announce Type: cross Abstract: Most feature selection algorithms, especially wrapper methods, run inefficiently on CPU based platforms because of their high computational complexity

NVIDIA and SAP Bring Trust to Specialized Agents

Announced today at SAP Sapphire — where NVIDIA founder and CEO Jensen Huang joined SAP CEO Christian Klein’s keynote by video — SAP and NVIDIA’s expanded collaboration helps enterprises run specialize

hardwarenvidia-blog

Nvidia says that Jensen Huang is joining President Trump on his China trip; source: the president asked Huang to join after seeing media coverage of his absence (CNBC)

CNBC: Nvidia says that Jensen Huang is joining President Trump on his China trip; source: the president asked Huang to join after seeing media coverage of his absence — BEIJING — Nvidia CEO Jensen Hua

Optimal Transport-Guided Adversarial Attacks on Graph Neural Network-Based Bot Detection

arXiv:2602.00318v2 Announce Type: replace-cross Abstract: The rise of bot accounts on social media poses significant risks to public discourse. To address this threat, modern bot detectors increasingl

PermuQuant: Lowering Per-Group Quantization Error by Reordering Channels for Diffusion Models

arXiv:2605.09503v1 Announce Type: new Abstract: Large-scale visual generative models have achieved remarkable performance. However, their high computational and memory costs make deployment challengin

Qualcomm closed down 11.46% on Tuesday as chip stocks pull back from record AI-driven rally; Intel closed down 6.82%, Sandisk dropped 6%, and Micron 3.61% (Samantha Subin/CNBC)

Samantha Subin / CNBC: Qualcomm closed down 11.46% on Tuesday as chip stocks pull back from record AI-driven rally; Intel closed down 6.82%, Sandisk dropped 6%, and Micron 3.61% — Chip stocks dropped

SkillEvolver: Skill Learning as a Meta-Skill

arXiv:2605.10500v1 Announce Type: new Abstract: Agent skills today are static artifact: authored once -- by human curation or one-shot generation from parametric knowledge -- and then consumed unchang

Sources: Jensen Huang was left out of President Trump's China trip to avoid unwanted scrutiny and awkward conversations about the sale of Nvidia chips to China (Semafor)

Semafor: Sources: Jensen Huang was left out of President Trump's China trip to avoid unwanted scrutiny and awkward conversations about the sale of Nvidia chips to China — THE SCOOP — The Trump adminis

SwiftI2V: Efficient High-Resolution Image-to-Video Generation via Conditional Segment-wise Generation

arXiv:2605.06356v2 Announce Type: replace Abstract: High-resolution image-to-video (I2V) generation aims to synthesize realistic temporal dynamics while preserving fine-grained appearance details of t

The benchmarks show the gap. NVLS all-reduce latency drops from 586.1µs on H200 to 313.3µs on GB200. In MoE prefill at EP=4, combine falls f…

The benchmarks show the gap. NVLS all-reduce latency drops from 586.1µs on H200 to 313.3µs on GB200. In MoE prefill at EP=4, combine falls from 730.1µs to 438.5µs. For decode, GB200 sustains much high

hardwareperplexity--x

The EDA Primer: From RTL to Silicon

The EDA Primer covers the semiconductor design and manufacturing workflow, explaining how Electronic Design Automation tools transform Register Transfer Level (RTL) code into physical silicon through

hardwaresemianalysis

We published new research on how we serve post-trained Qwen3 235B models on NVIDIA GB200 NVL72 Blackwell racks. GB200 is a major step up ove…

We published new research on how we serve post-trained Qwen3 235B models on NVIDIA GB200 NVL72 Blackwell racks. GB200 is a major step up over Hopper for high-throughput inference on large MoE models,

hardwareperplexity--x

An Efficient Hybrid Sparse Attention with CPU-GPU Parallelism for Long-Context Inference

arXiv:2605.07719v1 Announce Type: cross Abstract: Long-context inference increasingly operates over CPU-resident KV caches, either because decoding-time KV states exceed GPU memory capacity or because

Closed-Form Linear-Probe Dataset Distillation for Pre-trained Vision Models

arXiv:2605.07194v1 Announce Type: cross Abstract: Dataset distillation compresses a large training set into a small synthetic set that preserves downstream training utility. While most existing method

Research

Code Generation and Conic Constraints for Model-Predictive Control on Microcontrollers with Conic-TinyMPC

arXiv:2403.18149v3 Announce Type: replace Abstract: Model-predictive control (MPC) is a state-of-the-art control method for constrained robotic systems, yet deployment on resource-limited hardware rem

researcharxiv-cs-ro

Direction-Preserving Number Representations

arXiv:2605.07662v1 Announce Type: new Abstract: Low-precision number formats are widely used in modern machine learning systems due to their efficiency. Accurate direction representation is key to the

Don't Learn the Shape: Forecasting Periodic Time Series by Rank-1 Decomposition

arXiv:2605.07222v1 Announce Type: new Abstract: How few parameters do we really need to forecast a periodic time series? An hourly electricity series, reshaped as a 24-row matrix with one column per d