Infrastructure & Hardware
B
Showing 151–180 of 210
-
Your Robot’s Safety Functions Already Work. What If the Input Lies?IEEE Spectrum asks what happens if a robot's safety inputs are falsifiedRobot safety functions such as slowing near people depend on inputs like distance readings, position estimates or stop signals. This IEEE Spectrum piece, sponsored by VicOne, asks what happens when those inputs are false.
-
From Interaction to Capability: Task Advancement as State and Supervision for GUI Agents
-
Sponsored: Ground control: Structuring land deals for data center developmentSponsored: structuring land deals for data center developmentA sponsored piece arguing that AI and unprecedented demand for compute have fundamentally changed how investors, developers and operators evaluate data center opportunities.
-
Fine-Tuning NVIDIA Nemotron to Know Local Rules, Not Just Local LanguagesFine-tuning NVIDIA Nemotron to know local rules, not just languagesAs AI use spreads across languages and cultures, general-purpose models serve many needs. This NVIDIA post covers fine-tuning Nemotron models so they understand local rules as well as local languages, going beyond language support alone.
-
M4Bench: Evaluating Procedural Specification for Clinical EHR Derivation Agents
-
An Input Skip Can Reparameterize a VAE Without Improving the Generator
-
Test-Time Communication in Trading Agent Swarms
-
CRISP: Co-designing Pruning and Scheduling for Efficient MoE Inference on Edge Servers
-
Beyond the Readout: A Latent Numerical Frame Shared Across Modality and Operation
-
Displacement Moments and Path-Terminal Separation in Brownian Interpolant Sampling
-
SoleCrypt ups size of planned data center campus in Bizerte, TunisiaSoleCrypt enlarges planned Tunisia data center campusSoleCrypt has increased the size of its planned data center campus in Bizerte, Tunisia. The first 7MW phase is due live in early 2028 and will host Vera Rubin GPUs.
-
JEPA-Inspired Latent Field Prediction for Sparse-Pilot MIMO-OFDM Channel Reconstruction
-
Channel Selection for Mixture-of-Experts Decoding Needs No Calibration Data
-
Agent-Conditioned Routing: Diverse Reasoning Paths Within One Frozen Model
-
The Geometry of Manifold Learning: Singular foliations via Deep Learning models
-
TRIDENT-Seq: Profile-Matched Prefix-Future Contrast for Label-Efficient Payment Risk Modeling
-
STTGDC secures $590m green loan for Tokyo data centersSTTGDC secures $590m green loan for Tokyo data centersSTTGDC has secured a $590m green loan to finance the fit-out and operation of its 70MW campus in Tokyo.
-
Broadcom organizes $63bn deal with Bank of America, Blackstone, Citigroup, and Morgan Stanley to fund Anthropic chip purchasesBroadcom arranges $63bn financing for Anthropic chip purchasesBroadcom organized a $63bn deal with Bank of America, Blackstone, Citigroup and Morgan Stanley to fund Anthropic's chip purchases. Project North Star will finance the purchase of Broadcom-developed Google TPUs.
-
3 Ways to Boost MoE Inference on NVIDIA DGX SparkNVIDIA outlines three ways to speed up MoE inference on DGX SparkAn NVIDIA developer blog post describes three ways to boost mixture-of-experts (MoE) inference on DGX Spark. It argues that faster token generation starts with identifying what limits inference performance, using a batch-size-1 evaluation of Inclusion AI's Ling-3.0-flash as its example.
-
Online Interval Selection with Binary Predictions: Optimal Consistency and a Continuous Robustness Tradeoff
-
AirTrunk secures AU$4.25 billion green loan for Sydney data center campusAirTrunk secures AU$4.25bn green loan for Sydney campusAirTrunk has secured a AU$4.25 billion green loan for its Sydney data center campus. The current expansion phase at the SYD3 site is due live in 2028.
-
First building at Project Jupiter data center campus in New Mexico topped outFirst building at Project Jupiter in New Mexico topped outThe first building at the Project Jupiter data center campus in New Mexico has been topped out. Structural work on it is complete, with three more buildings to go.
-
Your Policy Is Secretly a Harness: The Decision Geometry of Agent Interfaces
-
Decision Models as Judges for Hallucination Detection
-
Mechanistic Failure Modes of Weight Pruning in Transformer Language Models
-
Prospector: learning a compact graph traversal policy for scientific reasoning
-
Self-Regularized Flow Language Models
-
UK ministers had plenty of time for AI datacenter bosses, but none for environmental groupsUK ministers met AI datacenter bosses, not environmental groupsAn investigation found that UK ministers had plenty of time for AI datacenter bosses but none for environmental groups. Businesses dominated government meetings on AI Growth Zones while communities were left out of decisions.
-
JevRec: Linguistic Encoding with Jev-like Decision Modeling for LLM-based Generative Recommendation
-
Do We Need an LLM to Know Why Sources Conflict? Decision Models for Knowledge-Conflict Typing in RAG