#MetaLlama3

[govindhtech]

AMD Instinct MI325X Accelerators Lead AI Performance

The AMD Pensando Pollara 400 NIC, AMD Pensando Salina DPU, and AMD Instinct MI325X accelerators are the newest networking and accelerator solutions that AMD unveiled today. These solutions will enable the next generation of AI infrastructure at scale. Performance benchmarks for Gen AI models and data centers are raised using AMD Instinct MI325X accelerators.

Based on the AMD CDNA 3 architecture, AMD Instinct MI325X accelerators are engineered to deliver outstanding performance and efficiency for challenging AI activities that include inferencing, fine-tuning, and foundation model training. With the help of these components, AMD partners and customers may develop AI systems that are highly efficient and optimized at the system, rack, and data center levels.

AMD Instinct MI325X Continues to Provide Superior AI Performance

Industry-leading memory capacity and bandwidth are provided by AMD Instinct MI325X accelerators; 256GB of HBM3E supporting 6.0TB/s offers 1.8X more capacity and 1.3X more bandwidth than the H200. In addition, the AMD Instinct MI325X has 1.3X higher peak theoretical compute performance for FP16 and FP8 than the H200.

Up to 1.3X, 1.2X, and 1.4X the inference performance on Mistral 7B at FP16, Llama 3.1 70B at FP8, and Mixtral 8x7B at FP16 of the H200, respectively, may be obtained with this leadership memory and compute.

Currently scheduled for production shipments in Q4 2024, AMD Instinct MI325X accelerators are anticipated to be widely available for use in systems from a variety of platform providers beginning in Q1 2025, including Dell Technologies, Eviden, Gigabyte, Hewlett Packard Enterprise, Lenovo, Supermicro, and others.

AMD showcased its upcoming AMD Instinct MI350 series accelerators, continuing its dedication to an annual roadmap cadence. In comparison to AMD CDNA 3-based accelerators, AMD Instinct MI350 series accelerators are built on the AMD CDNA 4 architecture and are intended to provide a 35x increase in inference performance.

The AMD Instinct MI350 series, which offers up to 288GB of HBM3E memory per accelerator, will continue to lead the market in memory capacity. The second part of 2025 is when the AMD Instinct MI350 series accelerators are expected to be ready.

AMD Next-Gen AI Networking

The most popular programmable DPU for hyperscalers is being used by AMD to support next-generation AI networking. AI networking, which is divided into two components the front-end, which provides data and information to an AI cluster, and the backend, which controls data transmission between accelerators and clusters is essential to making sure CPUs and accelerators are used effectively in AI infrastructure.

The AMD Pensando Pollara 400, the first AI NIC in the industry ready for the Ultra Ethernet Consortium (UEC), and the AMD Pensando Salina DPU were introduced by AMD to efficiently manage these two networks and promote high performance, scalability, and efficiency throughout the system.

The third iteration of the most powerful and programmable DPU in the world, the AMD Pensando Salina, offers up to two times the speed, bandwidth, and scalability of its predecessor. Optimizing performance, efficiency, security, and scalability for data-driven AI applications, the AMD Pensando Salina DPU is a crucial part of AI front-end network clusters, supporting 400G throughput for fast data transfer rates.

The AMD Pensando Salina DPU and AMD Pensando Pollara 400 are scheduled to be available in the first half of 2025, and they are now sampling with consumers in Q4 of 2024.

New Generative AI Capabilities Offered by AMD AI Software

In order to provide the AMD ROCm open software stack with powerful new features and capabilities, AMD keeps investing in expanding software capabilities and the open ecosystem.

Among the most popular AI frameworks, libraries, and models, such as PyTorch, Triton, Hugging Face, and many more, AMD is promoting support for AMD compute engines within the open software community. For well-known generative AI models like Stable Diffusion 3, Meta Llama 3, 3.1, and 3.2, as well as the more than one million models at Hugging Face, this work translates to unconventional performance and support with AMD Instinct accelerators.

With the addition of the newest features to support cutting-edge training and inference on generative AI workloads, AMD is further developing its ROCm open software stack outside of the community. Flash Attention 3, Kernel Fusion, FP8 datatype, and other important AI capabilities are now supported by ROCm 6.2. For a range of LLMs, ROCm 6.2 offers up to a 2.4X performance boost on inference and 1.8X on training when compared to ROCm 6.0.

Read more on govindhtech.com

[mariyajames01]

🌟 Exciting News! 🌟

We're diving deep into the world of AI with our latest YouTube video: "What is Meta LLaMA 3 - The Most Capable Large Language Model."

Join us at ValueCoders as we explore the incredible capabilities and potential of Meta LLaMA 3. Don't miss out on understanding the future of AI technology!

📺 Watch now:

🔔 Don't forget to like, comment, and subscribe for more tech insights!

[mysocial8onetech]

Explore the realm of AI with Meta Llama 3, the latest open-source Large Language Model from Meta AI. With its unique features and capabilities, it’s set to revolutionize language understanding and generation.

[govindhtech]

IBM Think 2024 Conference: Scaling AI for Business Success

IBM Think Conference

Today at its annual THINK conference, IBM revealed many new enhancements to its watsonx platform one year after its launch, as well as planned data and automation features to make AI more open, cost-effective, and flexible for enterprises. IBM CEO Arvind Krishna will discuss the company’s goal to invest in, create, and contribute to the open-source AI community during his opening address.

“Open innovation in AI is IBM’s philosophy. Krishna said they wanted to use open source to do with AI what Linux and OpenShift did. “Open means choice. Open means more eyes on code, brains on issues, and hands on solutions. Competition, innovation, and safety must be balanced for any technology to gain pace and become universal. Open source helps achieve all three.”

IBM published Granite models as open source and created InstructLab, a first-of-its-kind capability, with Red Hat

IBM has open-sourced its most advanced and performant language and code Granite models, demonstrating its commitment to open-source AI. IBM is urging clients, developers, and global experts to build on these capabilities and push AI’s limits in enterprise environments by open sourcing these models.

Granite models, available under Apache 2.0 licences on Hugging Face and GitHub, are known for their quality, transparency, and efficiency. Granite code models have 3B to 34B parameters and base and instruction-following model versions for complicated application modernization, code generation, bug repair, code documentation, repository maintenance, and more. Code models trained on 116 programming languages regularly outperform open-source code LLMs in code-related tasks:

IBM tested Granite Code models across all model sizes and benchmarks and found that they outperformed open-source code models twice as large.

IBM found Granite code models perform well on HumanEvalPack, HumanEvalPlus, and reasoning benchmark GSM8K for code synthesis, fixing, explanation, editing, and translation in Python, JavaScript, Java, Go, C++, and Rust.

IBM Watsonx Code Assistant (WCA) was trained for specialised areas using the 20B parameter Granite base code model. Watsonx Code Assistant for Z helps organisations convert monolithic COBOL systems into IBM Z-optimized services.

The 20B parameter Granite base code model generates SQL from natural language questions to change structured data and gain insights. IBM led in natural language to SQL, a major industry use case, according to BIRD’s independent leaderboard, which rates models by Execution Accuracy (EX) and Valid Efficiency Score.

IBM and Red Hat announced InstructLab, a groundbreaking LLM open-source innovation platform.

Like open-source software development for decades, InstructLab allows incremental improvements to base models. Developers can use InstructLab to construct models with their own data for their business domains or sectors to understand the direct value of AI, not only model suppliers. Through watsonx.ai and the new Red Hat Enterprise Linux AI (RHEL AI) solution, IBM hopes to use these open-source contributions to deliver value to its clients.

RHEL AI simplifies AI implementation across hybrid infrastructure environments with an enterprise-ready InstructLab, IBM’s open-source Granite models, and the world’s best enterprise Linux platform.

IBM Consulting is also developing a practice to assist clients use InstructLab with their own private data to train purpose-specific AI models that can be scaled to meet an enterprise’s cost and performance goals.

IBM introduces new Watsonx assistants

This new wave of AI innovation might provide $4 trillion in annual economic benefits across industries. IBM’s annual Global AI Adoption Index indicated that 42% of enterprise-scale organisations (> 1,000 people) have adopted AI, but 40% of those investigating or experimenting with AI have yet to deploy their models. The skills gap, data complexity, and, most crucially, trust must be overcome in 2024 for sandbox companies.

IBM is unveiling various improvements and enhancements to its watsonx assistants, as well as a capability in watsonx Orchestrate to allow clients construct AI Assistants across domains, to solve these difficulties.

Watsonx Assistant for Z

The new AI Assistants include watsonx Code Assistant for Enterprise Java Applications (planned availability in October 2024), watsonx Assistant for Z to transform how users interact with the system to quickly transfer knowledge and expertise (planned availability in June 2024), and an expansion of watsonx Code Assistant for Z Service with code explanation to help clients understand and document applications through natural language.

To help organisations and developers meet AI and other mission-critical workloads, IBM is adding NVIDIA L40S and L4 Tensor Core GPUs and support for Red Hat Enterprise Linux AI (RHEL AI) and OpenShift AI. IBM is also leveraging deployable designs for watsonx to expedite AI adoption and empower organisations with security and compliance tools to protect their data and manage compliance rules.

IBM also introduced numerous new and future generative AI-powered data solutions and capabilities to help organisations observe, govern, and optimise their increasingly robust and complex data for AI workloads. Get more information on the IBM Data Product Hub, Data Gate for watsonx, and other updates on watsonx.data.

IBM unveils AI-powered automation vision and capabilities

Company operations are changing with hybrid cloud and AI. The average company manages public and private cloud environments and 1,000 apps with numerous dependencies. Both handle petabytes of data. Automating is no longer an option since generative AI is predicted to drive 1 billion apps by 2028. Businesses will save time, solve problems, and make choices faster.

IBM’s AI-powered automation capabilities will help CIOs evolve from proactive IT management to predictive automation. An enterprise’s infrastructure’s speed, performance, scalability, security, and cost efficiency will depend on AI-powered automation.

Today, IBM’s automation, networking, data, application, and infrastructure management tools enable enterprises manage complex IT infrastructures. Apptio helps technology business managers make data-driven investment decisions by clarifying technology spend and how it produces business value, allowing them to quickly adapt to changing market conditions. Apptio, Instana for automated observability, and Turbonomic for performance optimisation can help clients efficiently allocate resources and control IT spend through enhanced visibility and real-time insights, allowing them to focus more on deploying and scaling AI to drive new innovative initiatives.

IBM recently announced its intent to acquire HashiCorp, which automates multi-cloud and hybrid systems via Terraform, Vault, and other Infrastructure and Security Lifecycle Management tools. HashiCorp helps companies transition to multi-cloud and hybrid cloud systems.

IBM Concert

IBM is previewing IBM Concert, a generative AI-powered tool that will be released in June 2024, at THINK. IBM Concert will be an enterprise’s technology and operational “nerve centre.”

IBM Concert will use watsonx AI to detect, anticipate, and offer solutions across clients’ application portfolios. The new tool integrates into clients’ systems and uses generative AI to generate a complete image of their connected apps utilising data from their cloud infrastructure, source repositories, CI/CD pipelines, and other application management solutions.

Concert informs teams so they can quickly solve issues and prevent them by letting customers minimise superfluous work and expedite others. Concert will first enable application owners, SREs, and IT leaders understand, prevent, and resolve application risk and compliance management challenges.

IBM adds watsonx ecosystem access, third-party models

IBM continues to build a strong ecosystem of partners to offer clients choice and flexibility by bringing third-party models onto watsonx, allowing leading software companies to embed watsonx capabilities into their technologies, and providing IBM Consulting expertise for enterprise business transformation Global generative AI expertise at IBM Consulting has grown to over 50,000 certified practitioners in IBM and strategic partner technologies. Large and small partners help clients adopt and scale personalised AI across their businesses.

IBM and AWS are integrating Amazon SageMaker and watsonx.governance on AWS. This product gives Amazon SageMaker clients advanced AI governance for predictive and generative machine learning and AI models. AI risk management and compliance are simplified by clients’ ability to govern, monitor, and manage models across platforms.

Adobe: IBM and Adobe are working on hybrid cloud and AI, integrating Red Hat OpenShift and watsonx to Adobe Experience Platform and considering on-prem and private cloud versions of watsonx.ai and Adobe Acrobat AI Assistant. IBM is also offering Adobe Express assistance to help clients adopt it. These capabilities should arrive in 2H24.

Meta: IBM released Meta Llama 3, the latest iteration of Meta’s open big language model, on watsonx to let organisations innovate with AI. IBM’s cooperation with Meta to drive open AI innovation continues with Llama 3. Late last year, the two businesses created the AI Alliance, a coalition of prominent industry, startup, university, research, and government organisations with over 100 members and partners.

Microsoft: IBM is supporting the watsonx AI and data platform on Microsoft Azure and offering it as a customer-managed solution on Azure Red Hat OpenShift (ARO) through IBM and our business partner ecosystem.

IBM and Mistral AI are forming a strategic partnership to bring their latest commercial models to the watsonx platform, including the leading Mistral Large model, in 2Q24. IBM and Mistral AI are excited to collaborate on open innovation, building on their open-source work.

Palo Alto Networks: IBM and Palo Alto now offer AI-powered security solutions and many projects to increase client security. Read the news release for details.

Salesforce: IBM and Salesforce are considering adding the IBM Granite model series to Salesforce Einstein 1 later this year to add new models for AI CRM decision-making.

SAP: IBM Consulting and SAP are also working to expedite additional customers’ cloud journeys using RISE with SAP to realise the transformative benefits of generative AI for cloud business. This effort builds on IBM and SAP’s Watson AI integration into SAP applications. IBM Granite Model Series is intended to be available throughout SAP’s portfolio of cloud solutions and applications, which are powered by SAP AI Core’s generative AI centre.

IBM introduced the Saudi Data and Artificial Intelligence Authority (SDAIA) ‘ALLaM’ Arabic model on watsonx, bringing language capabilities like multi-Arabic dialect support.

Read more on Govindhtech.com

[govindhtech]

Dell AI Factory: Accelerating Your AI Innovation Today

Dell AI Factory

AI is revolutionizing organisations by delivering unprecedented efficiency and competitive advantage. Data sovereignty, skills shortages, and the cost of deploying and growing AI in a corporation make navigating the AI opportunity tough.

This week at Dell Technologies World, They’re talking about how the Dell AI Factory tackles these most important issues by providing a range of AI technologies, an accessible partner ecosystem, integrated, validated solutions, and knowledgeable services to help you get AI results more quickly. Dell is pleased to announce a number of new infrastructure, ecosystem partnerships, and services that will further enhance the capabilities of the Dell AI Factory and contribute to the world’s largest range of AI solutions. Their objective is to assist you in reducing risk associated with your priceless data and intellectual property, increasing control over your budget and resources, and speeding up the time to business outcomes.

AI Infrastructure Provides an Adaptable IT Base

Businesses are realising that the data centres of today were not designed to fully utilise artificial intelligence. The new paradigm is the AI factory, and the infrastructure that supports AI is what makes it feasible. In order to provide you with even more alternatives for the computing power, scalability, data accessibility, and dependability you need for AI success, Dell are expanding dell’s range of AI infrastructure across data centre components and client devices.

Dell’s array of AI infrastructure additions gives you better options for data storage, data security and speedy access. Dell’s largest capacity all-flash NAS system, the Dell PowerScale F910, is the newest model in dell’s next-generation file storage lineup and was designed with artificial intelligence in mind. With this most recent system, it offers up to 127% better performance to deliver AI insights faster, speeding up every stage of the AI pipeline. In addition to PowerScale, they’re excited to introduce Project Lightning, which will enable you to meet the demands of AI factories by providing a scale-out NAS and parallel file system. Dell PowerScale performance will soar up to 20 times faster than competitors using only flash for scale-out files.

The Dell Solution for AI Data Protection provides a structure for enterprises to follow when incorporating strong data security measures into their AI frameworks. Along with new features for Dell’s Enterprise SONiC distribution, Dell is also expanding dell’s networking offerings with the introduction of Dell SmartFabric Manager for SONiC, a new management tool that streamlines deployment, orchestration, and lifecycle management with a single view of SONiC fabrics. These tools combined will enable secure, feature-rich, scalable fabrics with optimised traffic flows for demanding workloads like GenAI. Additionally, AI applications benefit from the new Dell PowerSwitch Z9864F-ON, which can handle large volumes of data and provides two times the switching capacity and bandwidth with state-of-the-art Ethernet technology.

Leveraging AI’s potential goes beyond the data centre; it might also mean arming your team with strong instruments and apparatus to spur innovation and expediency. They’re excited to present Dell’s inaugural fleet of Copilot+ PCs, which are powered by the Qualcomm Snapdragon X Elite and X Plus. These devices support critical AI use cases, such as running Windows 11 AI experiences locally. The performance and efficiency of new on-device AI capabilities are revolutionary, enabling previously unheard-of levels of worker productivity. With lightning-fast speeds and all-day battery life, these PCs come equipped with a specialised AI engine that can run generative AI (GenAI) large language models (LLMs) with over 13 billion parameters on-device.

Partnerships in the AI Ecosystem Facilitate the Adoption of Reliable, Integrated Solutions

Advanced technologies from numerous technology sources are advantageous to AI factories. Dell’s extensive AI ecosystem, which has a track record of fostering industry collaborations, is intended to provide you with increased access to innovation and choice. Dell’s industry-leading AI ecosystem partners work with them to offer seamless compatibility and integration throughout the stack of AI infrastructure. For instance, our pre-validated and pre-tested AI solutions eliminate uncertainty and dramatically reduce deployment time with reliable, full-stack configurations tailored for certain GenAI applications and use cases.

Dell AI Factory with Nvidia

Expanding upon dell’s March launch of the Dell AI Factory in partnership with NVIDIA, a number of new developments leverage the most recent NVIDIA and Dell products and services to boost AI innovation and adoption. Personalised retail experiences, industrial automation, digital assistants, and other use cases are all supported by portfolio expansions. View the significant changes to Dell AI Factory with NVIDIA that are the outcome of dell’s close partnership with NVIDIA.

Data scientists and the developer community benefit from dell’s collaboration with Hugging Face, and dell understand the value of fostering an open ecosystem of AI technology partners. Dell becomes the first infrastructure provider to collaborate with Hugging Face to offer optimised on-premises deployment of generative AI models with the launch of the Dell Enterprise Hub on the Hugging Face platform. In the end, this new portal provides a simplified method for popular LLMs to be deployed and customised on Dell infrastructure for Dell clients.

Furthermore, the open-source AI foundation model space has advanced significantly with the publication of Meta Llama 3 models. Together, Meta and Dell will continue to present benchmarking results, performance information, and deployment recipes that show off how simple it is to set up Meta Llama 3 models on Dell’s reliable infrastructure. With the use of specially made containers and scripts, these LLMs are also accessible on the Dell Enterprise Hub.

To ensure seamless integration and compatibility across various components of the AI infrastructure stack, Dell is expanding dell’s pre-tested and proven Dell Generative AI Solutions in collaboration with other ecosystem partners. For instance, to streamline the delivery of AI services like voice transcription and translation, a new Dell AI Solution for Microsoft Azure AI Services combines Microsoft Azure AI with Dell APEX Cloud Platform for Microsoft Azure. One potential use case where the system can facilitate easier and more effective communication between doctors and patients who speak different languages is the healthcare industry.

AI Services Provide Professional Advice and Direct Assistance to Quicken IT Results

While GenAI is predicted to be the most popular sector for technology investment in 2024, CEOs see talent and AI capabilities as the biggest obstacles to enterprise adoption. Here, Dell’s entire package of professional services for AI fills the gap by helping establish strategies and accelerating the implementation of GenAI.

Dell have added a number of additional services to dell’s portfolio, such as Implementation Services for Microsoft Copilot Solutions, which include GitHub Copilot, Microsoft Copilot for Security, Microsoft Copilot for Sales, and Microsoft Copilot in Windows. To optimise the return on your investments and boost productivity and efficiency throughout the company, these services assist you in planning, implementing, and promoting the adoption of new GenAI copilots. Dell’s extensive service offering now includes new Accelerator Services for RAG on Precision AI workstations and Dell Enterprise Hub on Hugging Face, enhancing the developer experience for quick GenAI development.

Find Out More About The AI Factory at Dell

As a result of dell’s ongoing innovation, Dell is prepared to assist you in deploying AI strategically and swiftly with a wide range of market-leading solutions and alliances, giving you the ability to obtain favourable results and a competitive edge as you apply various AI models to your data across PCs, on-premises IT, multiple clouds, and edge locations.

Read more on Govindhtech.com