PC, Gaming & Electronics Channel News (PCGE)’s Post

The ramp up is on, as Microsoft gears up to unveil its custom Cobalt 100 chips at the upcoming Build conference, showcasing a 40% performance improvement over other Arm chips in the market. These 64-bit chips, based on the Arm architecture with 128 cores, are already being utilized by companies like Adobe and Snowflake. Alongside Cobalt, Microsoft will introduce AMD’s MI300X accelerators to Azure clients, aiming to provide a cost-effective alternative to NVIDIA’s GPUs for AI tasks. Additionally, Microsoft plans to reduce pricing for accessing and running large language models. They'll also introduce a new real-time intelligence system, integrating with various data-streaming platforms like Kafka, AWS Kinesis, and Google Cloud’s Pub/Sub. Let the games being! #chips #datacentre #aiforbusiness #aiforall

Microsoft will debut its custom Cobalt 100 chips in a public preview at the Build conference next week. These 64-bit, 128-core ARM architecture chips promise 40% better performance compared to other ARM chips and are already in use by companies like Adobe and Snowflake. Additionally, Microsoft will introduce AMD’s MI300X accelerators to Azure clients, aiming to offer a cost-effective GPU alternative to Nvidia. The event will also feature a price reduction for running large language models, a preview of a new real-time intelligence system for data streaming, and a partnership with Snowflake to enhance data interoperability. Source: https://lnkd.in/dGR99FGG #MicrosoftBuild #CobaltChips #Azure #AI #CloudComputing #TechInnovation #AMD #Snowflake #DataStreaming #Copilot #CxONews #CxOGlobalForum

Not quite $1T, but $100B for the future of compute as this article from The Information and Anissa G. + Amir Efrati descibes the combo plans of Microsoft and OpenAI to build a #superdatacenter labeled #Stargate. ▶ Article below in comments (sub required) ▶ Sources report it will be US based ▶ Targeted for 2028 ▶ Other significant datacenter will be online as early as 2026 Full details in the article! The focus is providing full runway for the capacity of #LLM and #AI builds in coming years. NVIDIA and other manufacturers most likely will be involved. Do you think this type of investment is too much or not enough?

At #DellTechWorld, NVIDIA Vice President of Enterprise Computing Manuvir Das shared his insights on why a full-platform solution is vital to support #AI deployment across workloads. Watch his talk to learn more: https://nvda.ws/4bHlhXQ

#AI #Microsoft #copilot #dynamics #powerplatform #Azure #machinelearning Microsoft and Nvidia make significant AI announcements at Ignite. Microsoft emphasizes AI infrastructure advancements, introducing breakthrough Hol core fiber technology for a 47% speed improvement in data center networks. Azure Boost, a system offloading server virtualization processes, enhances networking, remote storage, and security. Microsoft unveils Azure Cobalt, its first custom in-house CPU, and expands collaboration with Nvidia, adding the latest GPU AI accelerator H200 and AMD's Mi 300X to the Azure Fleet. Deepening its partnership with OpenAI, Microsoft integrates GPT-4 and GPT-4 Turbo with vision into Azure services, introduces models as a service, and announces S2, a scaled-up version with superior performance. Co-Pilot innovations include Co-Pilot Studio for custom GPTs, plugins, and workflow orchestration, highlighting its role in coding, sales, and mixed reality applications. The vision extends to a future where AI-driven personalized tutors, doctors, and mentors are universally accessible.

Microsoft will publicly preview its Cobalt 100 chips at the Build conference, promising 40% better performance over other ARM chips. AMD’s MI300X accelerators will also be available to Azure clients, as Microsoft seeks cost-effective alternatives to Nvidia GPUs. The event will introduce new pricing for running large language models and features for real-time data streaming in Fabric. #CobaltChips #AMDAccelerators #RealtimeData

𝐌𝐢𝐜𝐫𝐨𝐬𝐨𝐟𝐭 𝐢𝐬 𝐟𝐢𝐧𝐚𝐥𝐥𝐲 𝐦𝐚𝐤𝐢𝐧𝐠 𝐜𝐮𝐬𝐭𝐨𝐦 𝐜𝐡𝐢𝐩𝐬 — 𝐚𝐧𝐝 𝐭𝐡𝐞𝐲’𝐫𝐞 𝐚𝐥𝐥 𝐚𝐛𝐨𝐮𝐭 𝐀𝐈 Source: By Tom Warren via The Verge💡 (https://lnkd.in/gj-vdWza) The rumors are true: Microsoft has built its own custom #AI chip that can be used to train large language #models and potentially avoid a costly reliance on NVIDIA. #Microsoft has also built its own Arm-based #CPU for #cloud workloads. Both custom #silicon chips are designed to power its #Azure data centers and ready the #company and its enterprise customers for a future full of #AI. #Microsoft’s #AzureMaia #AI chip and #Arm-powered #AzureCobalt #CPU are arriving in 2024, on the back of a surge in demand this year for #Nvidia’s H100 #GPUs that are widely used to train and operate generative image tools and large language #models. There’s such high demand for these GPUs that some have even fetched more than $40,000 on eBay. “#Microsoft actually has a long history in silicon development,” explains Rani Borkar, head of Azure hardware systems and infrastructure at Microsoft, in an interview with #TheVerge. #Microsoft collaborated on silicon for the #Xbox more than #20years ago and has even co-engineered chips for its Surface devices. “These efforts are built on that #experience,” says #Borkar. “In 2017, we began architecting the cloud #hardware #stack and we began on that journey putting us on track to build our new #custom #chips.”

Exciting updates from Microsoft Ignite! Microsoft just unveiled two bespoke chips - Azure Maia AI Accelerator and Azure Cobalt CPU - engineered to boost AI capabilities and satisfy increasing compute demands. The Maia AI Accelerator specialises in AI tasks, while the Cobalt CPU, an Arm-based processor, adeptly manages general cloud workloads. These chips, powering services like Microsoft Copilot and Azure OpenAI, will enhance our data centres starting next year. Moreover, Microsoft introduced Azure Boost, a game-changer in storage and networking by offloading these tasks to dedicated hardware. This innovation promises faster, more efficient operations. Expanding industry partnerships, we're offering diverse infrastructure options. Notably, the NC H100 v5 Virtual Machine Series preview, featuring NVIDIA H100 Tensor Core GPUs, promises exceptional AI training and inferencing capabilities. Looking forward, Microsoft plans to integrate NVIDIA's H200 Tensor Core GPU in our lineup next year for seamless large model inferencing. As Scott Guthrie, EVP of Microsoft’s Cloud and AI Group, remarks, "We're redefining our data centres to foster AI innovation and cater to varied customer needs." Rani Borkar, Corporate VP for Azure Hardware Systems, underscores our aim for a flexible Azure hardware system, optimised for power, performance, sustainability, and cost. #MicrosoftIgnite #Azure #AI #CloudComputing #innovation #erptoday

Another first! #Mavenir teams with #Nvidia and #AWS to harness the power of #GenAI for intelligent network management. A benign and helpful Skynet (yep, thats a Terminator reference)!