The NVIDIA CUDA 4.0 Toolkit is designed to focus on parallel programming, and enable developers to port their applications to GPUs and x86 CPUs. The three new main features of the new CUDA includes. GPUDirect 2.0 (for peer-to-peer communication among GPUs within a single server or workstation.) Unified Virtual Addressing and Thrust C++ Template Performance Primitives Libraries. You can read more about CUDA 4.0 here.
Related posts
AIRISC-VTools
Codasip announces L110 for power-efficient applications
New ‘bounded’ customization in tools.
AR/VR/XRCES 2025hologramsPhotonics
Help me, Swave Photonics, you’re my only hope…
Interview with CEO Mike Noonen.
ARMDual-architecture mainframe chipIBMIMB ZLinuxONE
IBM unveils dual-architecture mainframe chip
One core runs Arm and IBM Z.
Recent products
AI Processor Management Information Tracker Organization Primer
The JPR AI Processor Management Information Tracker Organization primer explains how JPR's AI Processor Management Information Tracker is structured into five major market segments and assigns each market to endpoint- and workload-specific subcategories, with details about each segment.
This organization guide is offered free of charge.
Dashboards in the AIP Management Information Tracker
Jon Peddie Research’s free Dashboards in the JPR AI Processor Management Information Tracker organizes the rapidly expanding AI processor market.
The dashboards are devised to assist users in navigating the extensive data contained within the tracking system.
This dashboard guide is offered free of charge.
China Robotics White Paper
Jon Peddie Research’s free China Robotics White Paper examines China’s strategic advance in robotics and automation and the effect this is having on the country’s move toward AI chip independence.
The China Robotics White Paper lists the top 10 robot AI processor suppliers, their chips/products, and the Chinese humanoid robot customers using those processors.
This white paper is offered free of charge.