Wednesday, November 15, 2023

Microsoft Ignite infrastructure announcements

Microsoft unveiled two custom-designed chips and integrated systems for its Azure hyperscale data centers. The new silicon represents the latest steps from Microsoft to develop core infrastructure in-house, extending from silicon, accelerator boards, custom server boards,  liquid-cooled data center racks, fiber cabling, renewable power facilities, and even subsea cable systems.

Highlights of infrastructure announcements from the Microsoft Ignite event in Seattle.

  • Microsoft Azure Maia, an AI Accelerator chip designed to run cloud-based training and inferencing for AI workloads such as OpenAI models, Bing, GitHub Copilot and ChatGPT. Maia 100, the first generation in the series, boasts 105 billion transistors and is built on 5nm process technology. 
  • Microsoft Azure Cobalt, a cloud-native chip based on Arm architecture optimized for performance, power efficiency and cost-effectiveness for general purpose workloads. Cobalt 100, the inaugural model in the series, is a 64-bit, 128-core chip. It delivers up to a 40 percent performance improvement over current Azure Arm chip generations and powers services like Microsoft Teams and Azure SQL. 
  • Microsoft Azure Boost, a system that makes storage and networking faster by moving those processes off the host servers onto purpose-built hardware and software.  The company cites remote storage acceleration up to 12.5 Gbps throughput with 650K IOPs. Additionally, Azure Boost introduces Microsoft’s proprietary programmable networking interface: MANA (Microsoft Azure Network Adapter). Using MANA allows Azure VM customers to achieve up to 200Gbps networking throughput on select VM sizes. 
  • AMD MI300X accelerated virtual machines (VMs) are now available on Azure. The ND MI300 VMs are designed to accelerate the processing of AI workloads for high range AI model training and generative inferencing, and will feature AMD’s latest GPU, the AMD Instinct MI300X.
  • NVIDIA NC H100 v5 Virtual Machine Series built for NVIDIA H100 Tensor Core GPUs are in preview, offering greater performance, reliability and efficiency for mid-range AI training and generative AI inferencing. 
  • NVIDIA ND H200 v5 Virtual Machine Series in development, an AI-optimized VM featuring the upcoming NVIDIA H200 Tensor Core GPU.
  • Microsoft is manufacturing its own hollow core fiber.  This follows its acquisition of UK-based Lumenisity in December 2022. Lumenisity was formed in 2017 as a spinoff from the world-renowned Optoelectronics Research Centre (ORC) at the University of Southampton
  • Microsoft now has 60+ Azure regions and 300+ data centers worldwide

Cisco reports strong quarter, weak outlook

Cisco reported revenue of $14.7 billion for the first quarter of its fiscal 2024, up 8% year over year;.GAAP EPS was$0.89, up 37% year over year, and Non-GAAP EPS was $1.11, up 29% year over year.

The financial performance marked the strongest first quarter results in Cisco's history in terms of revenue and profitability, however, the company noted  “a slowdown of new product orders in the first quarter of fiscal 2024 and believes the primary reason is that customers are currently focused on installing and implementing products in their environments following exceptionally strong product delivery over the past three quarters. “

Cisco estimates there are one to two quarters of shipped product orders still waiting to be implemented by its customers. 

"We had a solid start to fiscal 2024 with the strongest Q1 results in our history on both revenue and profitability,” said Chuck Robbins, chair and CEO of Cisco. “We are confident in the foundational strength of our business and future growth opportunities fueled by AI, Security, Cloud, and Observability.”

"In Q1, we delivered revenue and EPS at the high end or above our guidance range, generating strong operating leverage," said Scott Herren, CFO of Cisco. “We also saw double-digit year-over-year growth in software revenue, product ARR and total RPO. After customers implement large amounts of recently shipped product, we expect to see product order growth rates accelerate in the second half of the year. We are committed to delivering operating leverage and increasing capital returns to our shareholders.”

Some highlights

  • Total revenue was $14.7 billion, up 8%, with product revenue up 9% and service revenue up 4%. 
  • Revenue by geographic segment was: Americas up 14%, EMEA flat, and APJC was down 3%. Product revenue performance reflected growth in Networking up 10%, Security up 4%, Observability up 21% and Collaboration up 3%.
  • Total software revenue up 13% year over year and software subscription revenue up 13% year over year
  • Total annualized recurring revenue (ARR) at $24.5 billion, up 5% year over year and product ARR up 10% year over year
  • Remaining performance obligations (RPO) at $34.8 billion, up 12% year over year and product RPO up 14% year over year
  • During the quarter, Cisco completed the acquisions of Accedian (network performance monitoring); Working Group Two (a cloud native mobile services platform); Oort (identity threat detection and response technology);  SamKnows (broadband network monitoring company); Code BGP (border gateway protocol monitoring company).

NVIDIA and Amdocs target Gen AI for Telcos

NVIDIA announced an AI foundry service — a collection of NVIDIA AI Foundation Models, NVIDIA NeMo framework and tools, and NVIDIA DGX Cloud AI supercomputing and services.

Amdocs will use this AI foundry service to optimize enterprise-grade large language models for the telco and media industries. By training models on proprietary data, telcos can deliver tailored solutions that produce more accurate results for their use cases.

The companies are exploring several generative AI use cases, including accelerating resolution of customer inquiries by drawing information from across company data.

Argonne's trillion parameter Aurora genAI project leverages Intel GPUs

At SC23 in Denver, Intel is showcased AI-accelerated high performance computing (HPC). 

In partnership with Argonne National Laboratory, Intel shared progress on the Aurora generative AI (genAI) project, including an update on the 1 trillion parameter GPT-3 LLM on the Aurora supercomputer that is made possible by the unique architecture of the Max Series GPU and the system capabilities of the Aurora supercomputer.  Intel Data Center GPU Max Series delivers improved support for AI models, including multiple large language models (LLMs) such as GPT-J and LLAMA2.

The Aurora genAI project is a collaboration with Argonne, Intel and partners to create state-of-the-art foundational AI models for science. The models will be trained on scientific texts, code and science datasets at scales of more than 1 trillion parameters from diverse scientific domains.

New supercomputer deployments with Intel Max Series GPU and CPU technologies include systems like Aurora, Dawn Phase 1, SuperMUC-NG Phase 2, Clementina XX1 and more. New systems featuring Intel Gaudi2 accelerators include a large AI supercomputer with Stability AI as the anchor customer.

QuantLR and PacketLight demo Quantum Key Distribution over DWDM

QuantLR, a provider of Quantum Key Distribution (QKD) technology, and PacketLight Networks accomplished a quantum-safe encrypted network by integrating the QKD services provided by QuantLR with the OTN encrypted transport provided by PacketLight.

The companies validated the compatibility of quantum communication systems with optical networking infrastructure through rigorous testing to ensure optimal performance, stability, and reliability of the integrated solution.

"We are extremely pleased with the outcome of this collaboration between QuantLR and PacketLight," said Nir Bar-Lev, CEO of QuantLR. "This successful integration project paves the way for new opportunities in quantum communication. The joint solution enables customers to enjoy the highest level of secure communication with the easiest integration, and optimal cost effectiveness for a variety of use cases and network topologies. Especially in metro, last mile and other short distance setups."

"The integration of QuantLR's technology with PacketLight’s DWDM/OTN devices opens up new avenues for secure data transmission,” added Koby Reshef, CEO of PacketLight. “We are proud to have been a part of this project and are excited about the possibilities it brings to the market."

Viavi intros frontal verification solution

Viavi announced the expansion of its NITRO® Wireless portfolio with the addition of XhaulAdvisor, a scalable software solution offering real-time data for fronthaul verification, analysis, emulation and channel utilization. 

Vial’s XhaulAdvisor enables Open RAN vendors, operators and OTIC labs to accelerate interoperability testing (IOT) and End-to-End testing and troubleshooting. The solution enables a realistic testing and validation environment in real time. Vendors and operators rapidly and accurately correlate test anomalies to network events and troubleshoot fronthaul and transport timing, packet flows and channel utilization, as well as analyze M-plane, S-plane and C/U-plane sessions.

"The complex Open RAN supplier ecosystem demands reliable testing of different network components early in the development cycle. The XhaulAdvisor significantly reduces test time by delivering the industry's first real-time fronthaul analysis," said Ian Langley, Senior Vice President, Wireless Businss Unit, VIAVI. "This solution is part of the VIAVI commitment to increasing confidence in deployment of Open RAN."

Digital Realty brings on partner for new data centers in No Virginia

Digital Realty and Realty Income Corporation established a joint venture to support the development of two build-to-suit data centers in Northern Virginia. 

Realty Income invested approximately $200 million to acquire an 80% equity interest in the venture, while Digital Realty maintains a 20% interest.  Each partner will fund its pro rata share of the remaining $150 million estimated development cost for the first phase of the project, which is slated for completion in mid-2024.  

The build-to-suit facilities were 100% pre-leased to an S&P 100 investment grade client prior to construction and are expected to generate a 6.9% initial cash lease yield upon lease commencement in mid-2024.  The facilities are subject to a 10-year initial lease term with extension options and 2.0% annual rent escalators.

The two build-to-suit data centers commenced construction in the fourth quarter of 2022 and are slated to deliver 16 megawatts (MW) of initial data center capacity, which is expandable up to 48MW at the client's option.  The budget for the first phase of these yield-on-cost developments is approximately $400 million.  The client maintains the option to expand the projects up to 48MW of total capacity during the initial lease term, which could increase the budget up to $800 million, based on current development cost estimates.