AI TECH,GENERAL AI
Intel | September 17, 2022
Arm, Intel and Nvidia have jointly authored a paper describing an 8-bit floating point (FP8) specification and its two variants E5M2 and E4M3 to provide a common interchangeable format that works for both artificial intelligence (AI) training and inference. This cross-industry specification alignment will allow AI models to operate and perform consistently across hardware platforms, accelerating AI software development.
Computational requirements for AI have been growing at an exponential rate. New innovation is required across hardware and software to deliver computational throughput needed to advance AI.
One of the promising areas of research to address this growing compute gap is to reduce the numeric precision requirements for deep learning to improve memory and computational efficiencies. Reduced-precision methods exploit the inherent noise-resilient properties of deep neural networks to improve compute efficiency.
Intel plans to support this format specification across its AI product roadmap for CPUs, GPUs and other AI accelerators, including Habana Gaudi deep learning accelerators.
FP8 minimizes deviations from existing IEEE 754 floating point formats with a good balance between hardware and software to leverage existing implementations, accelerate adoption and improve developer productivity.
The guiding principle of this format proposal from Arm, Intel and Nvidia is to leverage conventions, concepts and algorithms built on IEEE standardization. This enables the greatest latitude for future AI innovation while still adhering to current industry conventions.
Cycle.io | August 05, 2022
Cycle.io, a low-ops application deployment platform, recently announced the Cycle Partner Program, and today we are happy to add Dev.Pro as one of our new Development Agency Partners. Our partnership empowers Dev.Pro development teams to build and deploy advanced solutions, products, and services efficiently on the Cycle Platform.
Dev.Pro, a US based software, custom development partner, works with customers across many industries using state of the art technologies. Dev.Pro enables their customers to grow by expediting time to market and delivering new products and features in a cost effective and efficient manner. This new partnership with Cycle expands the Dev.Pro portfolio and ensures that customers continue to have access to the latest solutions.
“The Cycle & Dev.Pro partnership is a natural fit for both organizations. Cycle offers a reliable, scalable, and feature rich platform; while Dev.Pro has the engineering expertise and experience to build and support even the most complex environments. ” Brian Agee, Chief Revenue Officer, Dev.Pro.
Cycle’s all-in-one platform has been used to deploy over a million containers, becoming the bridge between our partners' infrastructure and their code. The platform offers increased flexibility to scale their solutions utilizing Cycle’s rich set of features via the portal or through their powerful API.
“We are extremely excited to have Dev.Pro join our partner program. Dev.Pro has a reputation for having impressive engineering skills and outperforming customer expectations. Working with the Dev.Pro team over the course of the past six months indicates that this is only the tip of the iceberg”.
Karl Empey, Head of Sales, Cycle.io.
Cycle is the all-in-one low-ops platform built for deploying, scaling, and monitoring your websites, backends, and APIs. With automatic platform updates, standardized deployments, a powerful API, and bottleneck crushing automation—the platform offers batteries included, and no DevOps team required. Founded in 2015, the company is headquartered in Reno, NV.
Dev.Pro is a software development partner that allows innovative technology companies to amplify their growth ambitions and expedite time-to-market.Result-driven. Quality-obsessed. Dev.Pro builds teams that deliver a custom software development experience to meet any skillset, complexity, or scale. With a carefully curated staff of more than 850 highly skilled specialists, Dev.Pro offers a wide range of technical expertise including: Software Development, Cloud, DevOps, UI/UX Design, System Integration, Reporting/Analytics, and Manual/Automation Testing. Dev.Pro is a globally distributed company operating in 50+ countries spread across 5 continents.
Inspur Information | August 03, 2022
Inspur Information, a leading IT infrastructure solutions provider, is combining AIStation, its unified management and scheduling AI computing resource platform, with NVIDIA AI Enterprise, a cloud-native suite of AI and data analytics software, to provide enterprise users with a convenient and efficient platform for utilizing AI computing resources. This includes professional development, advanced deployment tools, and state-of-the-art components. The combined platform allows enterprises to quickly implement the industry's most advanced AI capabilities and rapidly deploy AI applications to drive intelligent business innovation.
There have been multiple attempts to build AI resource stacks for improving AI innovation and implementation. However, implementing AI technologies is a very challenging task. Enterprises not only need to meet requirements for computing power, but also maximize their computing resource utilization, quickly build and manage AI deployment environments, and efficiently introduce AI capabilities. It requires an AI platform that supports agile development and necessitates easy management and deployment to quickly implement AI applications and accelerate their performance.
Inspur Information's AIStation is an end-to-end platform that provides one-stop support for AI development and deployment. With powerful resource scheduling and management capabilities, it accelerates the development and deployment of AI technologies. NVIDIA AI Enterprise is an optimized end-to-end, cloud-native AI and data analytics software suite. It supports the processes from environment construction, data processing, training, and inference to allow for faster development and implementation of AI technologies.
AIStation's superior scheduling, management and monitoring capabilities are able to fully unleash NVIDIA AI Enterprise's excellent AI development and deployment capabilities. The combined solution expertly handles data storage, computing power and task scheduling, cluster operation and maintenance, and provides enterprises with an agile and easy-to-use AI platform that can fully utilize their AI computing resources. It also enhances the development and deployment of AI-related tools and components. Resulting AI development and deployment is markedly sped up, making AI transformation more efficient, and improving the business value of AI. Detailed benefits of the combined Inspur Information and NVIDIA AI Enterprise solution include:
Maximizing computing resources The solution makes it easier for enterprise customers to use and manage hardware infrastructure including NVIDIA GPUs, and supports Multi-Instance GPU features in the NVIDIA Ampere architecture, enabling users to centrally allocate and monitor computing resources and efficiently allocate those computing resources in large-scale computing power clusters.
Acquiring AI capabilities quickly The cloud-native environment management mechanism based on Docker containers gives users an effort-free experience for building cloud-native development environments. Thanks to the powerful scheduling and management capabilities of Inspur Information's AIStation and the professional frameworks, models, applications and development tools of NVIDIA AI Enterprise, users can build AI environments and run AI tasks with ease to achieve fast delivery and quickly introduce advanced AI capabilities.
Managing AI infrastructure with ease The multi-tenant management policy allows cluster administrators to focus on managing hardware infrastructure and controlling user quotas. Users can manage AI services and perform AI production within quota limits, and developers can quickly obtain AI compute resources without needing to be audited by administrators often. In addition, administrators can monitor key services and performance of the platform in real-time, and improve system availability by leveraging the fault-tolerant mechanisms and high-availability architecture.
Today, AI technologies are changing various industries such as finance, transportation, medical care, education, and scientific research. By introducing AI technologies and applications, enterprises can boost their business capabilities and create greater value. Inspur Information is working closely with NVIDIA to promote the construction of intelligent computing infrastructure and develop AI capability stacks to further accelerate enterprises' AI strategies and AI transformation.
Inspur Information and NVIDIA have a longstanding collaboration, with Inspur AI servers supporting the full lineup of NVIDIA GPUs, including the NVIDIA A100, A30, A10, and A2. Inspur AI servers were among the first in the industry to fully support NVIDIA Ampere architecture-based GPUs and the company has obtained NVIDIA-Certified System status for its AI servers.
About Inspur Information
Inspur Information is a leading provider of data center infrastructure, cloud computing, and AI solutions. It is the world’s 2nd largest server manufacturer. Through engineering and innovation, Inspur Information delivers cutting-edge computing hardware design and extensive product offerings to address important technology sectors such as open computing, cloud data center, AI, and deep learning. Performance-optimized and purpose-built, our world-class solutions empower customers to tackle specific workloads and real-world challenges.