Intel has recently revealed a security vulnerability named Downfall (CVE-2022-40982) that impacts multiple generations of Intel processors. The vulnerability is linked to Intel's memory optimization feature, exploiting the Gather instruction, a function that accelerates data fetching from scattered memory locations. It inadvertently exposes internal hardware registers, allowing malicious software access to data held by other programs. The flaw affects Intel mainstream and server processors ranging from the Skylake to Rocket Lake microarchitecture. The entire list of affected CPUs is here. Intel has responded by releasing updated software-level microcode to fix the flaw. However, there's concern over the performance impact of the fix, potentially affecting AVX2 and AVX-512 workloads involving the Gather instruction by up to 50%.

Phoronix tested the Downfall mitigations and reported varying performance decreases on different processors. For instance, two Xeon Platinum 8380 processors were around 6% slower in certain tests, while the Core i7-1165G7 faced performance degradation ranging from 11% to 39% in specific benchmarks. While these reductions were less than Intel's forecasted 50% overhead, they remain significant, especially in High-Performance Computing (HPC) workloads. The ramifications of Downfall are not restricted to specialized tasks like AI or HPC but may extend to more common applications such as video encoding. Though the microcode update is not mandatory and Intel provides an opt-out mechanism, users are left with a challenging decision between security and performance. Executing a Downfall attack might seem complex, but the final choice between implementing the mitigation or retaining performance will likely vary depending on individual needs and risk assessments.

Supermicro, Inc., a Total IT Solution Provider for Cloud, AI/ML, Storage, and 5G/Edge, is delivering a high-throughput, low latency E3.S storage solutions supporting the industry's first PCIe Gen 5 drives and CXL modules to meet the demands of large AI Training and HPC clusters, where massive amounts of unstructured data must be delivered to the GPUs and CPUs to achieve faster results.

Supermicro's Petascale systems are a new class of storage servers supporting the latest industry standard E3.S (7.5 mm) Gen 5 NVMe drives from leading storage vendors for up to 256 TB of high throughput, low latency storage in 1U or up to a half petabyte in 2U. Inside, Supermicro's innovative symmetrical architecture reduced latency by ensuring the shortest signal paths for data and maximized airflow over critical components, allowing them to run at optimal speeds. With these new systems, a standard rack can now hold over 20 Petabytes of capacity for high throughput NVMe-oF (NVMe over Fabrics) configurations, ensuring that GPUs remain saturated with data. Systems are available with either the 4th Gen Intel Xeon Scalable processors or 4th Gen AMD EPYC processors.

Microsoft Azure users can now turn to the latest NVIDIA accelerated computing technology to train and deploy their generative AI applications. Available today, the Microsoft Azure ND H100 v5 VMs using NVIDIA H100 Tensor Core GPUs and NVIDIA Quantum-2 InfiniBand networking—enables scaling generative AI, high performance computing (HPC) and other applications with a click from a browser. Available to customers across the U.S., the new instance arrives as developers and researchers are using large language models (LLMs) and accelerated computing to uncover new consumer and business use cases.

The NVIDIA H100 GPU delivers supercomputing-class performance through architectural innovations, including fourth-generation Tensor Cores, a new Transformer Engine for accelerating LLMs and the latest NVLink technology that lets GPUs talk to each other at 900 GB/s. The inclusion of NVIDIA Quantum-2 CX7 InfiniBand with 3,200 Gbps cross-node bandwidth ensures seamless performance across the GPUs at massive scale, matching the capabilities of top-performing supercomputers globally.

MaxLinear, Inc., a leader in data storage accelerator solutions, today announced the production-release of the OCP 3.0 storage accelerator adapter card for Panther III. The ultra-low latency accelerator is designed to quicken key storage workloads, including database acceleration, storage offload, encryption, compression, and deduplication enablement for maximum data reduction. The Panther III OCP card is ideal for use in modern data centers, including public to edge clouds, enterprise data centers, and telecommunications infrastructure, allowing users to access, process, and transfer data up to 12 times faster than without a storage accelerator. The OCP version of the card is available immediately with a PCIe version available in Q3 2023.

"In an era where the amount of data generated exceeds new storage installations by multiple fold, Panther III helps reduce the massive storage gap while improving TCO per bit stored," said Dylan Patel, Chief Analyst at SemiAnalysis.

AMD today announced revenue for the second quarter of 2023 of $5.4 billion, gross margin of 46%, operating loss of $20 million, net income of $27 million and diluted earnings per share of $0.02. On a non-GAAP basis, gross margin was 50%, operating income was $1.1 billion, net income was $948 million and diluted earnings per share was $0.58.

"We delivered strong results in the second quarter as 4th Gen EPYC and Ryzen 7000 processors ramped significantly," said AMD Chair and CEO Dr. Lisa Su. "Our AI engagements increased by more than seven times in the quarter as multiple customers initiated or expanded programs supporting future deployments of Instinct accelerators at scale. We made strong progress meeting key hardware and software milestones to address the growing customer pull for our data center AI solutions and are on-track to launch and ramp production of MI300 accelerators in the fourth quarter."

IBM has launched a new tool to help enterprises track greenhouse gas (GHG) emissions across cloud services and advance their sustainability performance throughout their hybrid, multicloud journeys. Now generally available, the IBM Cloud Carbon Calculator - an AI-informed dashboard - can help clients access emissions data across a variety of IBM Cloud workloads such as AI, high performance computing (HPC) and financial services.

Across industries, enterprises are embracing modernization by leveraging hybrid cloud and AI to digitally transform with resiliency, performance, security, and compliance at the forefront, all while remaining focused on delivering value and driving more sustainable business practices. According to a recent study by IBM, 42% of CEOs surveyed pinpoint environmental sustainability as their top challenge over the next three years. At the same time, the study reports that CEOs are facing pressure to adopt generative AI while also weighing the data management needs to make AI successful. The increase in data processing required for AI workloads can present new challenges for organizations that are looking to reduce their GHG emissions. With more than 43% of CEOs surveyed already using generative AI to inform strategic decisions, organizations should prepare to balance executing high performance workloads with sustainability.

Micron Technology, Inc. today announced it has begun sampling the industry's first 8-high 24 GB HBM3 Gen2 memory with bandwidth greater than 1.2 TB/s and pin speed over 9.2 Gb/s, which is up to a 50% improvement over currently shipping HBM3 solutions. With a 2.5 times performance per watt improvement over previous generations, Micron's HBM3 Gen2 offering sets new records for the critical artificial intelligence (AI) data center metrics of performance, capacity and power efficiency. These Micron improvements reduce training times of large language models like GPT-4 and beyond, deliver efficient infrastructure use for AI inference and provide superior total cost of ownership (TCO).

The foundation of Micron's high-bandwidth memory (HBM) solution is Micron's industry-leading 1β (1-beta) DRAM process node, which allows a 24Gb DRAM die to be assembled into an 8-high cube within an industry-standard package dimension. Moreover, Micron's 12-high stack with 36 GB capacity will begin sampling in the first quarter of calendar 2024. Micron provides 50% more capacity for a given stack height compared to existing competitive solutions. Micron's HBM3 Gen2 performance-to-power ratio and pin speed improvements are critical for managing the extreme power demands of today's AI data centers. The improved power efficiency is possible because of Micron advancements such as doubling of the through-silicon vias (TSVs) over competitive HBM3 offerings, thermal impedance reduction through a five-time increase in metal density, and an energy-efficient data path design.

Mark Papermaster, AMD's Chief Technology Officer and Executive Vice President of Technology and Engineering announced: "Over the past 50 years, Ethernet has grown to dominate general networking. One of its key strengths is flexibility - the ability to adapt to different workloads, scale and computing environments. One of the places that it hasn't been well-known, though, is in high-performance networking environments.

Now, the Ultra Ethernet Consortium (UEC) was formed by leading technology companies to focus on tuning the Ethernet foundation for high-performance Artificial Intelligence, Machine Learning, and High-Performance Computing (AI/ML/HPC) workloads. This includes work at the Physical, Link, Transport, and Software layers with robust security and congestion protections.

Announced today, Ultra Ethernet Consortium (UEC) is bringing together leading companies for industry-wide cooperation to build a complete Ethernet-based communication stack architecture for high-performance networking. Artificial Intelligence (AI) and High-Performance Computing (HPC) workloads are rapidly evolving and require best-in-class functionality, performance, interoperability and total cost of ownership, without sacrificing developer and end-user friendliness. The Ultra Ethernet solution stack will capitalize on Ethernet's ubiquity and flexibility for handling a wide variety of workloads while being scalable and cost-effective.

Ultra Ethernet Consortium is founded by companies with long-standing history and experience in high-performance solutions. Each member is contributing significantly to the broader ecosystem of high-performance in an egalitarian manner. The founding members include AMD, Arista, Broadcom, Cisco, Eviden (an Atos Business), HPE, Intel, Meta and Microsoft, who collectively have decades of networking, AI, cloud and high-performance computing-at-scale deployments.

NVIDIA RTX is spinning new cycles for designs. Trek Bicycle is using GPUs to bring design concepts to life. The Wisconsin-based company, one of the largest bicycle manufacturers in the world, aims to create bikes with the highest-quality craftsmanship. With its new partner Lidl, an international retailer chain, Trek Bicycle also owns a cycling team, now called Lidl-Trek. The team is competing in the annual Tour de France stage race on Trek Bicycle's flagship lineup, which includes the Emonda, Madone and Speed Concept. Many of the team's accessories and equipment, such as the wheels and road race helmets, were also designed at Trek.

Bicycle design involves complex physics—and a key challenge is balancing aerodynamic efficiency with comfort and ride quality. To address this, the team at Trek is using NVIDIA A100 Tensor Core GPUs to run high-fidelity computational fluid dynamics (CFD) simulations, setting new benchmarks for aerodynamics in a bicycle that's also comfortable to ride and handles smoothly. The designers and engineers are further enhancing their workflows using NVIDIA RTX technology in Dell Precision workstations, including the NVIDIA RTX A5500 GPU, as well as a Dell Precision 7920 running dual RTX A6000 GPUs.

What's New: As part of a collaboration with Intel and Dell Technologies, the United Kingdom Atomic Energy Authority (UKAEA) and the Cambridge Open Zettascale Lab plan to build a "digital twin" of the Spherical Tokamak for Energy Production (STEP) prototype fusion power plant. The UKAEA will utilize the lab's supercomputer based on Intel technologies, including 4th Gen Intel Xeon Scalable processors, distributed asynchronous object storage (DAOS) and oneAPI tools to streamline the development and delivery of fusion energy to the grid in the 2040s.

"Planning for the commercialization of fusion power requires organizations like UKAEA to utilize extreme amounts of computational resources and artificial intelligence for simulations. These HPC workloads may be performed using a variety of different architectures, which is why open software solutions that optimize performance needs can lend portability to code that isn't available in closed, proprietary systems. Overall, advanced hardware and software can make the journey to commercial fusion power lower risk and accelerated - a key benefit on the path to sustainable energy."—Adam Roe, Intel EMEA HPC technical director

Tachyum announced today it will cease taking orders for its Prodigy Universal Processor Field Programmable Gate Array (FPGA) emulation system boards effective immediately. The company releases the final Prodigy build for tape-out. New partners and customers who wish to work with Prodigy FPGAs for product evaluation, performance measurements, software development, debugging and compatibility testing can arrange for private testing in Tachyum's facility. As these are shared systems, they can't be used for classified or proprietary data or data subject to regulatory governance.

The Prodigy hardware emulator consists of multiple FPGA and IO boards connected by cables in a rack. A single board with four FPGAs emulates eight Prodigy processor cores (a small fraction of the final Prodigy product design, which consists of 128 cores) including vector and matrix fixed and floating-point processing units. Deploying more FPGAs will improve test cycles by orders of magnitudes to achieve target quality, a risk reduction mechanism for early adopters.

Samsung Electronics, a world leader in advanced semiconductor technology, today announced its latest foundry technology innovations and business strategy at the 7th annual Samsung Foundry Forum (SFF) 2023. Under the theme "Innovation Beyond Boundaries," this year's forum delved into Samsung Foundry's mission to address customer needs in the artificial intelligence (AI) era through advanced semiconductor technology.

Over 700 guests, from customers and partners of Samsung Foundry, attended this year's event, of which 38 companies hosted their own booths to share the latest technology trends in the foundry industry.

TrendForce reports that explosive growth in generative AI applications like chatbots has spurred significant expansion in AI server development in 2023. Major CSPs including Microsoft, Google, AWS, as well as Chinese enterprises like Baidu and ByteDance, have invested heavily in high-end AI servers to continuously train and optimize their AI models. This reliance on high-end AI servers necessitates the use of high-end AI chips, which in turn will not only drive up demand for HBM during 2023~2024, but is also expected to boost growth in advanced packaging capacity by 30~40% in 2024.

TrendForce highlights that to augment the computational efficiency of AI servers and enhance memory transmission bandwidth, leading AI chip makers such as Nvidia, AMD, and Intel have opted to incorporate HBM. Presently, Nvidia's A100 and H100 chips each boast up to 80 GB of HBM2e and HBM3. In its latest integrated CPU and GPU, the Grace Hopper Superchip, Nvidia expanded a single chip's HBM capacity by 20%, hitting a mark of 96 GB. AMD's MI300 also uses HBM3, with the MI300A capacity remaining at 128 GB like its predecessor, while the more advanced MI300X has ramped up to 192 GB, marking a 50% increase. Google is expected to broaden its partnership with Broadcom in late 2023 to produce the AISC AI accelerator chip TPU, which will also incorporate HBM memory, in order to extend AI infrastructure.

TikTok developer ByteDance, and other major Chinese tech firms including Tencent, Alibaba and Baidu are reported (by local media) to be snapping up lots of NVIDIA HPC GPUs, with even more orders placed this year. ByteDance is alleged to have spent enough on new products in 2023 to match the expenditure of the entire Chinese tech market on similar NVIDIA purchases for FY2022. According to news publication Jitwei, ByteDance has placed orders totaling $1 billion so far this year with Team Green—the report suggests that a mix of A100 and H800 GPU shipments have been sent to the company's mainland data centers.

The older Ampere-based A100 units were likely ordered prior to trade sanctions enforced on China post-August 2022, with further wiggle room allowed—meaning that shipments continued until September. The H800 GPU is a cut-down variant of 2022's flagship "Hopper" H100 model, designed specifically for the Chinese enterprise market—with reduced performance in order to meet export restriction standards. The H800 costs around $10,000 (average sale price per accelerator) according to Tom's Hardware, so it must offer some level of potency at that price. ByteDance has ordered roughly 100,000 units—with an unspecified split between H800 and A100 stock. Despite the development of competing HPC products within China, it seems that the nation's top-flight technology companies are heading directly to NVIDIA to acquire the best-of-the-best and highly mature AI processing hardware.

Supermicro, Inc., a Total IT Solution Provider for Cloud, AI/ML, Storage, and 5G/Edge, is announcing that its entire line of H13 AMD based-systems is now available with support for 4th Gen AMD EPYC processors, based on "Zen 4c" architecture, and 4th Gen AMD EPYC processors with AMD 3D V-Cache technology. Supermicro servers powered by 4th Gen AMD EPYC processors for cloud-native computing, with leading thread density and 128 cores per socket, deliver impressive rack density and scalable performance with energy efficiency to deploy cloud native workloads in more consolidated infrastructure. These systems are targeted for cloud operators to meet the ever-growing demands of user sessions and deliver AI-enabled new services. Servers featuring AMD 3D V-Cache technology excel in running technical applications in FEA, CFD, and EDA. The large Level 3 cache enables these types of applications to run faster than ever before. Over 50 world record benchmarks have been set with AMD EPYC processors over the past few years.

"Supermicro continues to push the boundary of our product lines to meet customers' requirements. We design and deliver resource-saving, application-optimized servers with rack scale integration for rapid deployments," said Charles Liang, president, and CEO of Supermicro. "With our growing broad portfolio of systems fully optimized for the latest 4th Gen AMD EPYC processors, cloud operators can now achieve extreme density and efficiency for numerous users and cloud-native services even in space-constrained data centers. In addition, our enhanced high performance, multi-socket, multi-node systems address a wide range of technical computing workloads and dramatically reduce time-to-market for manufacturing companies to design, develop, and validate new products leveraging the accelerated performance of memory intensive applications."

Synopsys, Inc. today announced an expanded agreement with Samsung Foundry to develop a broad portfolio of IP to reduce design risk and accelerate silicon success for automotive, mobile, high-performance computing (HPC) and multi-die designs. This agreement expands Synopsys' collaboration with Samsung to enhance the Synopsys IP offering for Samsung's advanced 8LPU, SF5, SF4 and SF3 processes and includes Foundation IP, USB, PCI Express, 112G Ethernet, UCIe, LPDDR, DDR, MIPI and more. In addition, Synopsys will optimize IP for Samsung's SF5A and SF4A automotive process nodes to meet stringent Grade 1 or Grade 2 temperature and AEC-Q100 reliability requirements, enabling automotive chip designers to reduce their design effort and accelerate AEC-Q100 qualification. The auto-grade IP for ADAS SoCs will include design failure mode and effect analysis (DFMEA) reports that can save months of development effort for automotive SoC applications.

"Our extensive co-optimization efforts with Samsung across both EDA and IP help automotive, mobile, HPC, and multi-die system architects cope with the inherent challenges of designing chips for advanced process technologies," said John Koeter, senior vice president of product management and strategy for IP at Synopsys. "This extension of our decades-long collaboration provides designers with a low-risk path to achieving their design requirements and quickly launching differentiated products to the market."

Today, at the "Data Center and AI Technology Premiere," AMD announced the products, strategy and ecosystem partners that will shape the future of computing, highlighting the next phase of data center innovation. AMD was joined on stage with executives from Amazon Web Services (AWS), Citadel, Hugging Face, Meta, Microsoft Azure and PyTorch to showcase the technological partnerships with industry leaders to bring the next generation of high performance CPU and AI accelerator solutions to market.

"Today, we took another significant step forward in our data center strategy as we expanded our 4th Gen EPYC processor family with new leadership solutions for cloud and technical computing workloads and announced new public instances and internal deployments with the largest cloud providers," said AMD Chair and CEO Dr. Lisa Su. "AI is the defining technology shaping the next generation of computing and the largest strategic growth opportunity for AMD. We are laser focused on accelerating the deployment of AMD AI platforms at scale in the data center, led by the launch of our Instinct MI300 accelerators planned for later this year and the growing ecosystem of enterprise-ready AI software optimized for our hardware."

With the launch of 4th Gen Intel Xeon Scalable processors in January 2023, Intel delivered significant advancements in performance with industry-leading Intel accelerator engines and improved performance per watt across key workloads like AI, data analytics, high performance computing (HPC) and others. The industry has taken notice: 4th Gen Xeon has seen a rapid ramp, global customer adoption and leadership performance on a myriad of critical workloads for a broad range of business use cases.

Today, after weeks of rigorous and comprehensive head-to-head testing against the most comparable competitive processors, Intel is sharing compelling results that go far beyond simple industry benchmarks.

ASUS today announced ESC N8-E11, its most advanced HGX H100 eight-GPU AI server, along with a comprehensive PCI Express (PCIe) GPU server portfolio—the ESC8000 and ESC4000 series empowered by Intel and AMD platforms to support higher CPU and GPU TDPs to accelerate the development of AI and data science.

ASUS is one of the few HPC solution providers with its own all-dimensional resources that consist of the ASUS server business unit, Taiwan Web Service (TWS) and ASUS Cloud—all part of the ASUS group. This uniquely positions ASUS to deliver in-house AI server design, data-center infrastructure, and AI software-development capabilities, plus a diverse ecosystem of industrial hardware and software partners.

GIGABYTE is exhibiting cutting-edge technologies and solutions at COMPUTEX 2023, presenting the theme "Future of COMPUTING". From May 30th to June 2nd, GIGABYTE is showcasing over 110 products that are driving future industry transformation, demonstrating the emerging trends of AI technology and sustainability, on the 1st floor, Taipei Nangang Exhibition Center, Hall 1.

GIGABYTE and its subsidiary, Giga Computing, are introducing unparalleled AI/HPC server lineups, leading the era of exascale supercomputing. One of the stars is the industry's first NVIDIA-certified HGX H100 8-GPU SXM5 server, G593-SD0. Equipped with the 4th Gen Intel Xeon Scalable Processors and GIGABYTE's industry-leading thermal design, G593-SD0 can perform extremely intensive workloads from generative AI and deep learning model training within a density-optimized 5U server chassis, making it a top choice for data centers aimed for AI breakthroughs. In addition, GIGABYTE is debuting AI computing servers supporting NVIDIA Grace CPU and Grace Hopper Superchips. The high-density servers are accelerated with NVLink-C2C technology under the ARM Neoverse V2 platform, setting a new standard for AI/HPC computing efficiency and bandwidth.

TYAN, an industry-leading server platform design manufacturer and a subsidiary of MiTAC Computing Technology Corporation, will be showcasing its latest HPC, cloud and storage platforms at Computex 2023, Booth #M0701a in Taipei, Taiwan from May 30 to June 2. These platforms are powered by AMD EPYC 9004 Series processors, which offer superior energy efficiency and are designed to enhance data center computing performance.

"As businesses increasingly prioritize sustainability in their operations, data centers - which serve as the computational core of an organization - offer a significant opportunity to improve efficiency and support ambitious sustainability targets," said Eric Kuo, Vice President of the Server Infrastructure Business Unit at MiTAC Computing Technology Corporation. "TYAN's server platforms powered by 4th Gen AMD EPYC processor enable IT organizations to achieve high performance while remaining cost-effective and contributing to environmental sustainability."

Giga Computing, a subsidiary of GIGABYTE and an industry leader in high-performance servers, server motherboards, and workstations, today announced a major presence at Computex 2023, held May 30 to June 2, with a GIGABYTE booth that inspires while showcasing more than fifty servers that span GIGABYTE's comprehensive enterprise portfolio, including green computing solutions that feature liquid cooled servers and immersion cooling technology. The international computer expo attracts over 100,000 visitors annually and GIGABYTE will be ready with a spacious and attractive booth that will draw in curious minds, and at the same time there will be plenty of knowledgeable staff to answer questions about how our products are being utilized today.

The slogan for Computex 2023 is "Together we create." And just like parts that make a whole, GIGABYTE's slogan of "Future of COMPUTING" embodies all the distinct computing products from consumer to enterprise applications. For the enterprise business unit, there will be sections with themes: "Win Big with AI HPC," "Advance Data Centers," and "Embrace Sustainability." Each theme will show off cutting edge technologies that span x86 and ARM platforms, and great attention is placed on solutions that address challenges that come with more powerful computing.

Molex, a company known for making various electronics and connectors, has today announced that the company has developed a first-of-its-kind chip-to-chip connector. Designed mainly for the data center, the Molex 224G product portfolio includes next-generation cables, backplanes, board-to-board connectors, and near-ASIC connector-to-cable solutions. Running at 224 Gbps speeds, these products use PAM4 signaling and boast with " highest levels of electrical, mechanical, physical and signal integrity." As the company states, future high-performance computing (HPC) data centers require a lot of board-to-board, chip-to-chip, and other types of communication to improve overall efficiency and remove bottlenecks in data transfer. To tackle this problem, Molex has a range of products, including Mirror Mezz Enhanced, Inception, and CX2 Dual Speed products.

Future generative AI, 1.6T (1.6 Tb/s) Ethernet, and other data center challenges need a dedicated communication standard, which Molex is aiming to provide. Working with various data center and enterprise customers, the company claims to have set the pace for products based on this 224G PAM4 chip-to-chip technology. We suspect that Open Compute Project (OCP) will be first in the line of adoption, ad Molex has historically worked with them as they adopted Mirror Mezz and Mirror Mezz Pro board-to-board connectors. The new products can be seen below, and we expect to hear more announcements from Molex's partners. Solutions like OSFP 1600, QSFP 800, and QSFP-DD 1600 already use 224G products.

During the ISC High Performance 2023 international conference, Intel announced interesting roadmap updates to its high-performance computing (HPC) and artificial intelligence (AI). With the scrapping of Rialto Bridge and Lancaster Sound, Intel merged these accelerator lines into Falcon Shores processor for HPC and AI, initially claiming to be a CPU+GPU solution on a single package. However, during the ISC 2023 talk, the company forced a change of plans, and now, Falcon Shores is GPU only solution destined for a 2025 launch. Originally, Intel wanted to combine x86-64 cores with Xe GPU to form an "XPU" module that powers HPC and AI workloads. However, Intel did not see a point in forcing customers to choose between specific CPU-to-GPU core ratios that would need to be in an XPU accelerator. Instead, a regular GPU solution paired with a separate CPU is the choice of Intel for now. In the future, as workloads get more defined, XPU solutions are still a possibility, just delayed from what was originally intended.

Regarding Intel's Gaudi accelerators, the story is about to end. The company originally paid two billion US Dollars for Habana Labs and its Gaudi hardware. However, Intel now plans to stop the Gaudi development as a standalone accelerator and instead use the IP to integrate it into its Falcon Shores GPU. Using modular, tile-based architecture, the Falcon Shores GPU features standard ethernet switching, up to 288 GB of HBM3 running at 9.8 TB/s throughput, I/O optimized for scaling, and support for FP8 and FP16 floating point precision needed for AI and other workloads. As noted, the creation of XPU was premature, and now, the initial Falcon Shores GPU will become an accelerator for HPC, AI, and a mix of both, depending on a specific application. You can see the roadmap below for more information.