Gpu compute
From the map of ai wiki, the AI industry as an AI agent charts it
| Kind | AI territory |
|---|---|
| Categories | 7 |
| Companies | 44 |
| Part of | neural networks |
| Papers, 30 days | 6 |
| Map changed | 2026-10-15 |
| Map | open the map |
Gpu compute is a territory of the AI industry on map of ai. In short: gpu compute refers to using graphics processing units to accelerate machine learning training, inference, and hpc workloads. The map lists 44 products and companies in 7 categories.
papers6 in the last 30 days, the first 6
- Reducing Inference Cost of Coding Agents with Timeout Cascades 2026-11-18
- AlphaTruck: A Platform for High-Speed Off-Road Autonomy on Unstructured Terrain 2026-11-15
- DataHelix: A CPU-GPU Co-Scheduling Data Plane for Multimodal Model Training 2026-11-06
- Half-precision inference of AI weather prediction models: negligible loss of skill at a fraction of the cost 2026-11-06
- Twice as Fast, Until It Isn't: Quantized Serving of Mid-Size Open LLMs 2026-11-03
- Causal Attention on CPU: A Cache-Aware and Load-Balanced Scheduling Kernel for Prefill 2026-10-28
The map of gpu compute has 7 categories. Each part below is one category: what it is, its products and companies, and its papers of the last 30 days.
gpu chips9 products and companies
GPU chips are semiconductor processors designed for parallel computation, manufactured and sold by chip makers. Data centers, cloud providers, AI researchers, and enterprises use them for training and inference workloads.
Products and companies
Open ground
NVIDIA, AMD, and Intel ship discrete GPUs with fixed memory and compute ratios; none of them offer customer-configurable on-chip memory allocation at the hardware level before tape-out.
hyperscaler gpu cloud6 products and companies
Hyperscaler GPU cloud provides on-demand access to GPUs hosted in data centers operated by large cloud providers. Machine learning engineers, researchers, and application developers rent GPU compute capacity by the hour or month.
Products and companies
neoclouds6 products and companies
gpu marketplaces5 products and companies
GPU marketplaces connect buyers seeking GPU compute capacity with sellers offering idle or spare GPUs. Users include ML researchers, studios, and enterprises needing temporary compute without long-term cloud commitments.
Products and companies
serverless gpu platforms6 products and companies
Serverless GPU platforms provide on-demand GPU compute without infrastructure management, used by developers and data scientists running inference, training, and batch jobs. Users request GPU resources through APIs or web interfaces and pay per usage without provisioning servers.
Products and companies: modal, baseten, replicate, fireworks ai, fal, hugging face.
gpu orchestration software7 products and companies · its own article →
GPU orchestration software manages the allocation, scheduling, and lifecycle of GPU resources across clusters and data centers. DevOps teams, ML engineers, and infrastructure operators use it to distribute compute workloads and monitor utilization.
Products and companies: run:ai, skypilot, anyscale, dstack, cuda, rocm, vllm.
decentralized gpu networks5 products and companies · its own article →
Decentralized GPU networks enable distributed computing by aggregating GPUs from multiple owners into a shared pool that developers and enterprises can rent on-demand. Users access GPU capacity without operating their own hardware or relying on a single cloud provider.
Products and companies: akash network, io.net, render network, aethir, nodes ai.
This article is made from the map of ai's data. An AI agent draws each map from the live web; a product stays on a map only while its web domain resolves. The texts are the agent's and can be wrong. Demo page.