Research Note: Kioxia’s Open Source AiSAQ ANN Search
Today, Kioxia announced the open-source release of All-in-Storage ANNS with Product Quantization (AiSAQ), an approximate nearest neighbor search (ANNS) technology optimized for SSD-based storage. AiSAQ enables large-scale retrieval-augmented generation (RAG) workloads by offloading vector data from DRAM to SSDs, significantly reducing memory requirements.
CES 2025- Enterprise Tech Was There Too
Oh my, CES 2025 has taken me on one heck of a ride through the tech universe. I am an enterprise IT guy but, I must admit, the non-IT tech at CES had me fully distracted. Automated lawnmowers, high tech indoor garden planters, and my favorite- an ultra-realistic flight simulator. Wowzers- really neat stuff.
Research Note: Marvell Custom HBM for Cloud AI
Marvell recently announced a new custom high-bandwidth memory (HBM) compute architecture that addresses the scaling challenges of XPUs in AI workloads. The new architecture enables higher compute and memory density, reduced power consumption, and lower TCO for custom XPUs.
Research Note: Enfabrica ACF-S Millennium
First detailed at Hot Chips 2024, Enbrica recently announced that its ACF-S “Millennium” chip, which addresses the limitations of traditional networking hardware for AI and accelerated computing workloads, will be available to customers in calendar Q1 2025.
Research Note: AWS Trainium2
Tranium is AWS’s machine learning accelerator, and this week at its re:Invent event in Las Vegas, it announced the second generation, the cleverly named Trainium2, purpose-built to enhance the training of large-scale AI models, including foundation models and large language models.
Quick Take: AWS re:Invent AI Services & Infrastructure Announcement Wrap-Up
AWS has made close to 50 announcements in the first 2 days of re:Invent. This blog takes a look at the most interesting AI services & infrastructure related ones.
Research Note: NVIDIA SC24 Announcements
At the recent Supercomputing 2024 (SC24) conference in Atlanta, NVIDIA announced new hardware and software capabilities to enhance AI and HPC capabilities. This includes the new GB200 NVL4 Superchip, the general available of its H200 NVL PCIe, and several new software capabilities.
Quick Take: AMD Data Center Group Earnings – Q3 2024
AMD this week announced significant revenue and earnings growth in Q3 2024, driven primarily by exceptional performance in the Data Center segment. AMD’s Data Center revenue increased by 122% year-over-year, reaching a record $3.5 billion and marking over half of AMD’s total revenue this quarter. CEO Lisa Su, on the earnings call, attributed this growth to the success of the EPYC CPUs and MI300X GPUs, which experienced strong adoption across cloud, enterprise, and AI applications.
OCP Global Summit 2024: Key Announcements
At the recent OCP Global Summit 2024, the organization unveiled several major initiatives highlighting OCP’s commitment to driving innovation and fostering collaboration in the tech ecosystem. This includes expanding its AI Strategic Initiative with contributions from NVIDIA and Meta, new alliances for sustainability and standardization, and the launch of an open chiplet marketplace.
Raspberry Pi Integrates the Sony Semi Solutions Intelligent Aitrios IMX500 Image Sensor
Sony Semiconductor Solutions Corporation (SSS) and Raspberry Pi Ltd recently announced the release of a co-developed AI Camera designed to accelerate the development of AI solutions for edge processing.
AMD MI300 Gains Momentum with New Vultr & Oracle Cloud Wins
The AMD MI300’s advanced architecture, featuring high memory capacity, low power consumption, and solid performance, is finding a home among cloud providers. While Microsoft Azure previously announced MI300-based instances, there are new announcements from specialty GPU cloud provider Vultr and the mainstream CSP Oracle Cloud Infrastructure of new integrations with AMD’s MI300 accelerator.
Is Ampere Computing Up for Sale?
Ampere Computing, the leading vendor of server processors based on the Arm architecture, is reportedly exploring a sale, signaling a strategic pivot amidst increasing competition in the market for alternative server processors.
Quick Take: Intel Gaudi 3 on IBM Cloud
IBM and Intel announced a partnership to integrate Intel’s Gaudi 3 AI accelerators into IBM Cloud, which will be available in early 2025. This collaboration aims to enhance the scalability and affordability of enterprise AI, focusing on performance, security, and energy efficiency. IBM Cloud will be the first cloud provider to offer Gaudi 3, which […]
Research Note: IBM Telum II & Spyre AI Accelerators
At the Hot Chips 2024 conference in Palo Alto, California, IBM unveiled the next generationj of enterprise AI solutions: the IBM Telum II processor and the IBM Spyre Accelerator. These new technologies should meet the demands of the AI era, providing enhanced performance, scalability, and AI capabilities. Both are expected to be available in 2025.
Research Note: AMD Acquires ZT Systems
AMD announced its strategic acquisition of ZT Systems, a specialty provider of AI and general-purpose compute infrastructure for major hyperscale companies, in a deal valued at $4.9 billion. The acquisition aligns with AMD’s AI strategy to enhance its capabilities in AI training and inferencing solutions for data centers.
Research Note: SiFive P870-D RISC-V Server CPU
The SiFive P870-D is a new cutting-edge RISC-V-based processor specifically designed for data center applications. As part of SiFive’s broader family of high-performance processors, the P870-D is engineered to meet the growing demands for compute density, power efficiency, and system resilience in modern infrastructure environments.
Research Note: Ampere Computing Roadmap Updates
Ampere announced updates to its AmpereOne processor roadmap, highlighting its upcoming product AmpereOne Aurora processor. The company also offered pricing updates and other details about its overall roadmap.
Research Note: Qualcomm FQ3 2024 Earnings
In its fiscal third quarter, Qualcomm demonstrated strong financial performance, driven by its continued success in diversifying its business beyond mobile handsets into sectors like automotive, IoT, and PCs. With its focus on innovation, particularly in AI and advanced computing, Qualcomm is well-positioned to sustain its leadership across various industries.
Research Note: AMD Q2 2024 Data Center Earnings
AMD reported strong Q2 financial results, with total revenue reaching $5.8 billion, a 9% year-over-year increase. The Data Center segment, a key driver of this performance, achieved a record revenue of $2.8 billion, reflecting a 115% year-over-year growth.
Backgrounder: State of CXL
Compute Express Link (CXL) is a high-speed interconnect standard designed to enhance communication between CPUs, GPUs, memory, and other accelerators. It’s designed to address the growing needs for efficient data sharing and coherent memory access in data centers, HPC, and AI-driven workloads.
Research Note: onsemi EliteSIC M3e MOSFET
onsemi’s new EliteSiC M3e MOSFETs address the growing demands for more efficient, reliable, and cost-effective power solutions across various industries. The increasing global focus on mitigating climate change and transitioning to renewable energy sources necessitates significant advancements in power semiconductor technology. onsemi’s latest generation EliteSiC M3e MOSFETs are a major step forward.
Quick Take: AWS Launches Graviton 4
Today, Amazon Web Services (AWS) launched its Graviton4 processors, the fourth generation of its Arm-based custom processors. This new chip is touted as AWS’s most energy-efficient and high-performance solution for cloud workloads, marking a significant upgrade over its predecessor, Graviton3.
Research Note: Micron FQ3 2024 Earnings
Micron Technology delivered a strong performance in Q3 FY2024, significantly surpassing market expectations in several key areas. The company’s financial results reflect robust demand, strategic pricing, and technological advancements that position it well for future growth.
Research Note: AMD Computex MI325x & MI350 Accelerator Announcements
At the 2024 Computex event in Taiwan, AMD CEO Lisa Su revealed details about AMD’s upcoming MI350 and MI325X accelerators, follow-ons to its current MI300x products, highlighting significant advancements in AI performance and memory capacity. The new products are positioned as key components in AMD’s strategy to lead the AI accelerator market.
Research Note: UALink Alliance & Accelerator Interconnect Specification
UALink is a new open standard designed to rival NVIDIA’s proprietary NVLink technology. It facilitates high-speed, direct GPU-to-GPU communication crucial for scaling out complex computational tasks across multiple graphics processing units (GPUs) or accelerators within servers or computing pods.
Research Note: Ampere Computing 1H 2024 Roadmap Update
Ampere Computing released its annual roadmap update, outlining significant progress and strategic initiatives aimed at driving innovation in sustainable, power-efficient computing for cloud and AI applications.
Microsoft Azure Cobalt 100 Arm-based VMs
At its 2024 Microsoft Build Event this week, Microsoft announced the preview of new Azure Virtual Machines powered by Microsoft’s previously announced in-house-design Arm-based processor, the Cobalt 100
Research Note: Google Trillium TPU
The Trillium TPU, Google’s sixth-generation TPU, was announced at Google I/O. It promises unprecedented compute performance, memory capacity, and energy efficiency for generative AI training and inference.
Research Note: Arm FQ4 2024 Earnings Results
Arm reported record revenues for its fiscal Q4 2024, increasing revenue during the quarter by 47% year-over-year, driven by significant gains in royalties and licensing. Arm’s growth was fueled by the rapid adoption of the v9 architecture and a strategic increase in R&D investment, mainly targeted towards harnessing opportunities in AI technologies.
Research Note: NVIDIA H100 Confidential Computing
This week, NVIDIA made its confidential computing capabilities for its flagship NVIDIA Hopper H100 GPU, previewed in August 2023, generally available. This makes NVIDIA’s H100 the first GPU with these capabilities, which are critical for protecting data as it is being processed.
This Research Note looks at confidential computing and how it works on the NVIDIA H100 GPU.
Quick Take: Intel FQ1 2024 Earnings Results
Intel announced solid Q1 earnings, with revenue meeting expectations and EPS exceeding guidance. The company’s results reflect a disciplined approach to cost reduction and steady progress toward long-term goals.
Research Note: Intel Gaudi 3
Intel announced its long-anticipated new Intel Gaudi 3 AI accelerator at its Intel Vision event. The new accelerator offers significant improvements over the previous generation Gaudi 3 processor and promises to challenge Nvidia’s current generation accelerators in training and inference for LLMs and multimodal models.
Quick Take: Google Axion Arm-based Processor
At its Google Cloud Next event, Google announced its new in-house designed Axion processor, a series of custom Arm-based CPUs explicitly designed for data center applications. These processors are part of Google’s continued investment in custom silicon to enhance its cloud computing services’ performance and energy efficiency.
Research Note; Arm Ethos U-65 microNPU
Arm introduced its new Ethos-U65 microNPU (Neural Processing Unit). This state-of-the-art AI accelerator facilitates machine learning (ML) inference in many embedded systems and high-performance devices.
Research Note: AMD Versal Edge Series Gen 2
AMD expands its Versal portfolio with the introduction of its Versal AI Edge Series Gen 2 and Versal Prime Series Gen 2 adaptive SoCs. These next-generation solutions cater to the increasing demands for AI-driven and classic embedded systems, providing a balanced mix of performance, power efficiency, functional safety, and security within a single chip.
Is NVIDIA Lagging in Lucrative Automotive Segment?
Nvidia’s most recent earnings release is a tremendous achievement for the company, with reported revenue of $22.1 billion, up an incredible 265% year-on-year. Earnings grew an equally unbelievable 765% year-on-year.
Its automotive revenue was $281 million.
Research Note: Arm’s New Automotive Cores
Arm recently launched new safety-enabled AE processors incorporating Armv9 technology and server-class performance. These processors are tailored for AI-driven applications to enhance autonomous driving and advanced driver-assistance systems (ADAS).
Research Brief: Inside Arm’s Neoverse CSS N3 & V3 Announcements
Arm brings two new Neoverse compute subsystems to market, each based on its third-generation Neoverse IP, extending its N-Series and V-Series product lines. These new platforms, Neoverse CSS N3, and V3, aim to improve performance-per-watt and support the implementation of new technologies like chiplets.
Quick Take: Intel’s Big Automotive Play
At the CES show in Las Vegas, Intel outlined a bold strategy to expand its AI capabilities into the automotive sector. This includes an agreement to acquire Silicon Mobility, a company specializing in system-on-chips (SoCs) for EV energy management, marking a significant step in Intel’s pursuit of automotive market growth.
Research Note: AMD’S New Embedded+ CPU/FPGA Platform
AMD announced the launch of AMD Embedded+, a novel architectural solution that integrates AMD’s Ryzen Embedded processors with its Versal adaptive onto a single board.
Research Note: ARM’s FISCAL Q3 2024 EARNINGS
Arm Holdings reported a solid financial fiscal 3Q 2024, exceeding expectations and highlighting its robust position in the technology sector. In its earnings call, the company announced record revenues and raised its revenue guidance for the upcoming quarter.
Research Note: Inside AMD’s 4Q 2023 Data Center Group Performance
Advanced Micro Devices (AMD) reported a stellar end to 2023, marking achievements within its Data Center business amidst a fluctuating demand landscape.
This Research Note takes a look at AMD’s performance, focusing on its Data Center Group.
Research Note: Inside Intel’s 4Q 2023 Data Center & AI Earnings
The latest earnings release from Intel Corporation offers a comprehensive overview of the company’s current trajectory and outlook, underscoring significant strides in its ambitious IDM 2.0 transformation.
This Research Note focuses primarily on Intel’s Data Center & AI Group (DCAI) and those elements that impact enterprise infrastructure.
Quick Take: Intel’s New Emerald Rapids Processor
Intel unveiled its new Emerald Rapids processors, part of its 5th-Gen Xeon Scalable lineup. The new processors arrive with multiple features designed to enhance performance across workloads, including AI and HPC.
Research Note: AMD’s new MI300X & MI300A AI Accelerators
AMD recently announced its new MI300x and MI300a AI accelerators, its most significant challenge to NVIDIA yet.
Research Note: Microsoft’s New Processor & Accelerator
At its recent Microsoft Ignite event, the company announced the launch of the Azure Cobalt 100 CPU and the Maia 100 AI accelerator, marking a significant pivot in Microsoft’s approach to its cloud infrastructure and representing the company’s commitment to driving innovation in high-performance computing and AI.
Research Note: Amazon’s New Graviton4 & Tranium2
Amazon introduced generational updates to both its Graviton and Tranium custom silicon solutions at its recent re:Invent conference in Las Vegas.
This Research Note delves into the intricacies and potential impacts of these latest innovations from AWS.
Research Note: Inside Qualcomm’s Snapdragon X Elite
Qualcomm unveiled its highly anticipated Snapdragon X Elite system-on-a-chip (SoC), representing a significant leap in its chip design capabilities. This next-generation Arm-based SoC is designed to power Windows devices and showcases Qualcomm’s commitment to high-performance computing.
Research Note: Data Center Impact from AMD’s Q3 2023 Earnings
Advanced Micro Devices, Inc. (AMD) reported its financial results for the third quarter of fiscal year 2023. AMD showcased strong performance across its key segments during the quarter, including Data Center, Client, Gaming, and Embedded, with a focus on AI and high-performance computing solutions.
In this research note, we provide an overview of AMD’s key announcements and financial performance during the quarter, along with NAND Research’s analysis of what the results mean for AMD’s data center business.
Arm introduces Arm Total Design to Further Simplify SoC Development
Arm this week introduced its Arm Total Design program for Neoverse CSS, an ecosystem approach to facilitate developing and delivering custom system-on-chip (SoC) solutions based on Arm’s Neoverse Compute Subsystems (CSS).
Intel Details New Processors & Looks to the Future @ Intel Innovation
Intel made a sweeping set of announcements at its recent Intel Innovation event, including detailing next generation Xeon processors.
Marvell Sees Momentum in Cloud, AI, and Automotive
Last week, Marvell released its earnings for the second quarter of its fiscal 2024, demonstrating robust performance with $1.34 billion in top-line revenue. While that number was down year-over-year, it surpassed the midpoint of the company’s guidance. Insight: Growth in Cloud & AI Marvell’s data center business was a bright spot. Revenue from that business […]
Arm Enables Faster Silicon Development with Neoverse CSS N2
At this week’s Hot Chips 2023 event, Arm unveiled its Arm Neoverse Compute Subsystems (CSS) offering, which promises to simplify and accelerate the adoption of Arm Neoverse technology into new solutions. It’s a powerful enabler that will reduce the friction of integrating Arm IP into new designs and accelerate the time-to-market for Arm’s partners. News: […]
Qualcomm Grows Automotive Business With Cadillac Win
This month, the momentum continues as Cadillac announces it’s putting Snapdragon Digital Chassis technology into its upcoming 2025 Cadillac Escalade IQ, Cadillac’s first all-electric full-size SUV. The design win includes more than just Snapdragon Digital Cockpit. The new vehicle will consist of Snapdragon Auto Connectivity and Snapdragon Ride.
Ampere’s Momentum Continues with New Oracle Database Support
Oracle and Ampere Computing held an event today to talk about Ampere’s role at Oracle, in Oracle’s cloud business, and in support of Oracle’s enterprise Oracle Database. Today’s announcement goes beyond Oracle Database simply gaining support for a new processor architecture. The news acknowledges Ampere’s credibility, will contribute to Ampere’s growing momentum, and will play […]