Contrastive decoding
Contrastive decoding (CD) is a decoding strategy for text generation from a large language model that selects tokens by contrasting two models of different sizes.
AI Infrastructure
Explore AI Infrastructure through related topics and the articles other pages reference most.
Explore articles
Browse subtopics (52)
Articles that also belong to these categories. Counts cover all of AI Infrastructure.
All articles (281)AI Hardware (98)AI Companies (57)Data Centers (43)Developer Tools (37)Open Source AI (37)NVIDIA (35)Machine Learning (29)AI Inference (26)Google (13)AI Energy (11)Chinese AI (10)MLOps (9)Training & Optimization (9)AI Agents (8)OpenAI (8)Computer Science (7)Deep Learning (7)Enterprise AI (7)Information Retrieval (7)Large Language Models (7)AI Policy & Regulation (6)People (6)Artificial Intelligence (5)Meta AI (5)Microsoft (4)Robotics (4)AI Code Generation (3)AI Research (3)AI Tools & Products (3)Anthropic (3)Generative AI (3)Mixture of Experts (3)Venture Capital (3)AI Models (2)Model Architecture (2)Model Evaluation (2)Programming Languages (2)AI in Gaming (1)Agentic Commerce (1)Data & Datasets (1)Data Science (1)Edge computing (1)Embodied AI (1)Finance AI (1)Hardware (1)Image Generation (1)Mathematics (1)Neural Networks (1)Physical AI (1)Research Organizations (1)Software Development (1)Video Generation (1)
Showing 61-120 of 281 articles
CoreWeave
CoreWeave, Inc. is an American artificial intelligence cloud computing company headquartered in Livingston, New Jersey, that rents large clusters of NVIDIA GPUs to AI developers and enterprises and builds the…
Cross-Model KV Cache Transfer
Cross-model KV cache transfer is an experimental technique for converting the KV cache produced by one large language model into the cache representation expected by another model.
Crusoe
Crusoe is an American AI infrastructure company headquartered in Denver, Colorado, that builds and operates data centers for artificial intelligence workloads.
DRAM (Dynamic Random-Access Memory)
Dynamic random-access memory (DRAM) is the working memory of nearly every computer built since the late 1970s, and the physical substrate on which modern AI hardware runs.
Dask
Dask is an open-source Python library for parallel and distributed computing that scales the familiar APIs of libraries such as NumPy, pandas, and scikit-learn to process larger-than-memory datasets.
Data Center
A data center is a purpose-built facility, or a dedicated part of a facility, that houses and interconnects information technology and telecommunications equipment together with the power…
Data Parallelism
Data parallelism is a distributed training technique in which the same neural network model is replicated across multiple processing units (typically GPUs), each device trains on a different shard of the input…
DeepEP
DeepEP is an open-source GPU communication library built by DeepSeek for Mixture-of-Experts (MoE) models.
DeepGEMM
DeepGEMM is an open-source library from DeepSeek that provides fast FP8 general matrix multiplication (GEMM) kernels for NVIDIA Hopper GPUs.
DeepInfra
DeepInfra is a serverless AI inference cloud that hosts open-source and open-weight AI models and serves them to developers through a single pay-per-token API.
DeepSpeed
DeepSpeed is an open-source deep learning optimization library, originally developed by Microsoft, that makes distributed training and inference of large models efficient, easy to use, and cost-effective.
Disaggregated serving
Disaggregated serving is an LLM inference architecture that physically separates the prefill phase and the decode phase of text generation onto different sets of GPU hardware.
DriveNets
DriveNets is an Israeli networking-software company that builds large telecom and AI networks out of standard "white box" hardware controlled by cloud-native software instead of traditional proprietary routers.
DualPipe
DualPipe is a bidirectional pipeline parallelism scheduling algorithm created by DeepSeek-AI to almost fully overlap computation with communication and shrink the idle time, called the pipeline "bubble," that…
EAGLE (speculative decoding)
EAGLE (Extrapolation Algorithm for Greater Language-model Efficiency) is a lossless speculative decoding method that speeds up large language model (LLM) inference by 2x to 6.5x by doing autoregression at the…
Edge computing
Edge computing is a distributed computing paradigm that runs computation and data storage close to where data is generated, at the "edge" of the network
Ethernet
Ethernet is the family of wired computer networking technologies standardized by the IEEE 802.3 working group.
Exa AI
Exa AI (formerly Metaphor) is an artificial intelligence company that builds a search engine designed specifically for AI applications.
Expert Parallelism
Expert Parallelism (EP) is a model-parallelism strategy specific to Mixture of Experts (MoE) neural networks in which the individual expert sub-networks (typically feed-forward blocks) are sharded across…
FAISS
FAISS (Facebook AI Similarity Search) is an open-source library from Meta for efficient similarity search and clustering of dense vectors
FPGA
A field-programmable gate array (FPGA) is an integrated circuit whose logic functions and internal wiring are set by the customer after the chip has been manufactured, and can be reset later.
Feature store
A feature store is a centralised data system that stores, serves, discovers, shares, monitors and reuses machine-learning features, separating feature computation from model training and inference so the same…
Fermi America
Fermi America is a United States power and data center developer attempting to build what it calls the world's largest energy and computing complex
Firebase
Firebase is a backend-as-a-service (BaaS) platform developed by Google. It began in 2012 as a real-time database product launched by Andrew Lee and James Tamplin and was acquired by Google on October 21, 2014.
FlashInfer
FlashInfer is an open-source GPU kernel library and code-generation system for large language model inference.
FlashMLA
FlashMLA is an open-source GPU kernel from DeepSeek that accelerates the decoding step of Multi-head Latent Attention (MLA)
Frozen v2 (Google AI chip)
Frozen v2 is the informal internal codename for a specialized artificial-intelligence inference chip that Google is reportedly developing to run its Gemini models more cheaply and with far less energy.
Fully Sharded Data Parallel (FSDP)
Fully Sharded Data Parallel (FSDP) is a distributed training technique implemented in PyTorch that shards a model's parameters, gradients, and optimizer states across data-parallel workers, allowing models…
GE Vernova
GE Vernova (NYSE: GEV) is an American energy and electric-power equipment company, headquartered in Cambridge, Massachusetts, that builds and services much of the hardware used to generate and move…
GEM (Meta)
GEM (Generative Ads Recommendation Model) is a proprietary foundation model for advertising recommendation developed by Meta Platforms.
GPU Cluster
A GPU cluster is a group of servers containing graphics processing units that are connected by high-bandwidth, low-latency networks and operated as one computational pool.
GPU computing
GPU computing is the use of a graphics processing unit (GPU) to perform general-purpose computation that was traditionally handled by the central processing unit (CPU).
Genesis (simulator)
Genesis is an open-source, generative physics simulation platform for robotics and embodied AI, released on December 19, 2024 after a roughly two-year (24-month) collaboration involving more than 20 academic…
Google Axion
Google Axion is Google's first custom Arm-based central processing unit designed for the data center.
Google Cloud
Google Cloud is the enterprise cloud business of Google and the name of a reportable segment in Alphabet's financial statements.
Google Virgo Network
Virgo Network is a megascale data-center fabric introduced by Google at Google Cloud Next 2026 in April 2026.
Groq Hardware
Groq hardware is a family of artificial-intelligence accelerators and multi-chip systems built around a statically scheduled streaming architecture.
HUMAIN
HUMAIN is a Saudi Arabian artificial intelligence company launched on 12 May 2025 and owned by the Public Investment Fund (PIF), the kingdom's roughly trillion-dollar sovereign wealth fund.
Horovod
Horovod is an open-source distributed training framework for deep learning that lets a single-GPU training script scale across many GPUs and many machines by adding only a few lines of code.
Huawei Ascend
Huawei Ascend is a family of AI accelerators designed by Huawei around a custom processor architecture called Da Vinci.
Huawei Ascend 910B
The Huawei Ascend 910B is a data-center AI accelerator designed by Huawei's HiSilicon unit that became China's most widely deployed domestic alternative to restricted NVIDIA data-center GPUs across roughly…
Huawei Ascend 910C
The Huawei Ascend 910C is a data-center artificial intelligence accelerator developed by Huawei and positioned as China's leading domestic alternative to high-end NVIDIA GPUs that are barred from sale to…
Hyperion (Meta data center)
Hyperion is an artificial intelligence data center campus built by Meta in Richland Parish, in northeast Louisiana, United States.
Hyperscaler
A hyperscaler is a company that builds and operates computing infrastructure at a scale far beyond a conventional enterprise IT estate: globally distributed fleets of data centers holding millions of servers…
IREN
IREN Limited is an Australian-founded data center company listed on the Nasdaq Global Select Market under the ticker IREN.
Inferact
Inferact is an artificial intelligence infrastructure company founded by creators and core maintainers of vLLM, the open-source inference engine that emerged from the University of California, Berkeley in 2023.
InfiniBand
InfiniBand is a high throughput, low latency networking interconnect standard used to connect servers, storage, and accelerators inside high performance computing systems and large artificial intelligence…
Internet of Things
The Internet of Things (IoT) is the network of physical objects ("things") embedded with sensors, software, and connectivity that lets them collect data, exchange it with other devices and systems over the…
Invisible Technologies
Invisible Technologies is an American artificial intelligence company that combines a global network of vetted human experts with its own orchestration software to provide AI training data, reinforcement…
Ion Stoica
Ion Stoica is a Romanian-American computer scientist, professor of electrical engineering and computer sciences at the University of California, Berkeley, and a serial entrepreneur whose academic and…
Jalapeño (OpenAI chip)
Jalapeño is a custom AI accelerator designed by OpenAI with silicon implementation and networking technology from Broadcom.
KAI Scheduler
KAI Scheduler is an open-source Kubernetes scheduler that optimizes the allocation of GPU resources for artificial intelligence and machine learning workloads.
KV cache offloading
KV cache offloading is the practice of moving part or all of a Transformer model's KV cache out of accelerator memory (GPU HBM) into a larger, slower tier such as CPU DRAM, local NVMe storage, or remote…
KV-cache quantization
KV cache quantization is a family of large language model inference optimizations that store the attention key and value (KV) cache in low-bit numeric formats, typically 2 to 4 bits per value
Kairos Power
Kairos Power is a United States advanced nuclear reactor developer, founded in 2016 and headquartered in Alameda, California
Kioxia
Kioxia is a Japanese flash memory manufacturer, the direct corporate descendant of the Toshiba division where flash memory was invented in the 1980s.
Kiro (AI IDE)
Kiro is an agentic integrated development environment (IDE) built by Amazon Web Services and released in public preview on July 14, 2025 .
LLM inference engine
An LLM inference engine (also called an LLM serving engine or LLM inference server) is the systems software stack that loads trained large language model weights into GPU or CPU memory and answers user…
LMDeploy
LMDeploy is an open-source toolkit for compressing, deploying, and serving large language models, developed by the MMRazor and MMDeploy teams associated with the InternLM project at the Shanghai AI Laboratory.