Jensen Groks Groq
On Nvidia's fascinating 'acqui-something' of AI chip startup Groq
December 2025
RSSPi calculation world record shattered at 314 trillion digits with a four-month run on a single server
StorageReview takes the Pi crown again with 314 trillion digits calculated
1,000 computers taken offline in Romanian water management authority hack
No group has claimed the attack yet, and thankfully, water is still flowing in Romania.
Windows Server 2025 gains native NVMe support, 14 years after its introduction
Windows Server 2025 gets native NVMe with massive I/O and CPU gains
TPU Mania
Is GPU vs TPU the CISC vs RISC contest of the modern era?
Dev gambles on 'obviously fake' $8K Grace Hopper system, scores $80,000 worth of hardware on Reddit for one-tenth of the cost
This used Nvidia Grace Hopper system price seemed unbelievable. But it required a lot of cleaning and intricate work to get it running smoothly.
HC2025-T2.2: AI Kernel Programming (part 2)
Tutorial 2, part 2, Hot Chips 2025, Sunday, August 24, 2025. Organizer: Fredrik Kjolstad, Stanford Domain specific languages for GPU kernels and automatic kernel authoring with LLMs Tri Dao, Princeton / Together AI Programming techniques for implementing ML models on GPUs Zihao Jia, CMU
HC2025-T2.1: AI Kernel Programming (part 1)
Tutorial 2, part 1, Hot Chips 2025, Sunday, August 24, 2025. Organizer: Fredrik Kjolstad, Stanford Introduction Fredrik Kjolstad, Stanford Decoupling Performance from Correctness with User-Schedulable Kernel Languages Andrew Adams, Adobe Research Pallas: Using JAX to write custom kernels for GPUs and TPUs Sharad…
HC2025-T1.2: Datacenter Racks (part 2)
Tutorial 1, part 2, Hot Chips 2025, Sunday, August 24, 2025. Organizer: Ralph Wittig, AMD This half of the tutorial consists of three full case studies of rack development at different companies. There is a group Q&A at the end. Case study: Nvidia GB200NVL72 John Norton, NVIDIA Case study: Meta’s Catalina (NVL72)…
HC2025-T1.1: Datacenter Racks (part 1)
Tutorial 1, part 1, Hot Chips 2025, Sunday, August 24, 2025. Organizer: Ralph Wittig, AMD This half of the tutorial is organized by topic. There is a group Q&A at the end. How AI workloads shape rack system architecture Frank Helms, AMD Scaling fabric technologies Darrin Vallis, AMD Liquid cooling with Google…
HC2025-S8: Machine Learning 2
Session 8, Hot Chips 2025, Tuesday, August 26, 2025. Chair: Pradeep Dubey, Intel NVIDIA’s GB10 SoC: AI Supercomputer On Your Desk Andi Skende, NVIDIA 4th Gen AMD CDNA™ Generative AI Architecture Powering AMD Instinct™ MI350 Series GPUs and Platforms Michael Floyd & Michael Steffen, AMD Ironwood: Delivering best in…
HC2025-S7: Machine Learning 1
Session 7, Hot Chips 2025, Tuesday, August 26, 2025. Chair: Ronny Krashinsky, NVIDIA Memory: (Almost) the Only Thing That Matters Mark Kuemerle, Marvell Corsair - An In-memory Computing Chiplet Architecture for Inference-time Compute Acceleration Sudeep Bhoja, d-Matrix UB-Mesh: Huawei’s Next-Gen AI SuperComputer with…
HC2025-S6: Power / Methodology
Session 6, Hot Chips 2025, Tuesday, August 26, 2025. Chair: Jae W. Lee, SNU ECAM Enabled Advanced Thermal Management Solutions for the AI Data Center Michael Matthews, Fabric8Labs Everactive Self-Powered SoC with Energy Harvesting, Wakeup Receiver, and Energy-Aware Subsystem Ben Calhoun, Everactive Taping Out Three…
HC2025-S5: Optical
Session 5, Hot Chips 2025, Tuesday, August 26, 2025. Chair: Borivoje Nikolic, UC Berkeley Celestial AI Photonic Fabric Module (PF Module) - The world’s first SoC with in-die Optical IO Phil Winterbottom, Celestial AI A UCIe Optical I/O Retimer Chiplet for AI Scale-up Fabrics Vladimir Stojanovic, Ayar Labs Passage…
HC2025-S4: Networking
Session 4, Hot Chips 2025, Monday, August 25, 2025. Chair: Sherry Xu, Microsoft Intel Mount Morgan Infrastructure Processing Unit (IPU) Patrick Fleming, Intel AMD Pensando™ Pollara 400 AI NIC Architecture and Application Kevin Chu, AMD NVIDIA ConnectX-8 SuperNIC: A Programmable RoCE Architecture for AI Data Centers…
HC2025-S3: Graphics
Session 3, Hot Chips 2025, Monday, August 25, 2025. Chair: Lavanya Subramanian, OpenAI AMD RDNA 4 and Radeon RX 9000 Series GPU Andy Pomianowski & Laks Pappu, AMD RTX 5090: Designed for the Age of Neural Rendering Marc Blackstein, NVIDIA Specialized SoC enabling low-power ‘World Lock Rendering’ in Augmented and Mixed…
HC2025-S2: Security
Session 2, Hot Chips 2025, Monday, August 25, 2025. Chair: Greg Papadopoulos, NEA Presto: A RISC-V-Compatible SoC for Unified Multi-Scheme FHE Acceleration over Module Lattice Luchang Lei & Hongyang Jia, Tsinghua University Azure Secure Hardware Architecture: Establishing a Robust Security Foundation for Cloud…
HC2025-S1: CPUs
Session 1, Hot Chips 2025, Monday, August 25, 2025. Chair: Gabriel Southern, Qualcomm Cuzco: A High-Performance RISC-V RVA23 Compatible CPU IP Ty Garibay & Shashank Nemawarkar, Condor Computing PEZY-SC4s: The Fourth Generation MIMD Many-core Processor with High Energy Efficiency and Flexibility for HPC and AI…