Oracle unveils AMD-powered zettascale AI cluster for OCI cloud
The forthcoming zettascale AI cluster is designed to scale up to 131,072 MI355X GPUs, specifically architected to support high-performance, production-grade AI training, inference, and new agentic workloads. The cluster is expected to offer over double the price-performance compared to the previous generation of hardware.
Expanded AI capabilities
The new announcement highlights several key hardware and performance enhancements. The MI355X-powered cluster provides 2.8 times higher throughput for AI workloads. Each GPU features 288 GB of high-bandwidth memory (HBM3) and eight terabytes per second (TB/s) of memory bandwidth, allowing for the execution of larger models entirely in memory and boosting both inference and training speeds.
The GPUs also support the FP4 compute standard, a four-bit floating point format that enables more efficient and high-speed inference for large language and generative AI models. The cluster's infrastructure includes dense, liquid-cooled racks, each housing 64 GPUs and consuming up to 125 kilowatts per rack to maximise performance density for demanding AI workloads. This marks the first deployment of AMD's Pollara AI NICs to enhance RDMA networking, offering next-generation high-performance and low-latency connectivity.
Mahesh Thiagarajan, Executive Vice President, Oracle Cloud Infrastructure, said: "To support customers that are running the most demanding AI workloads in the cloud, we are dedicated to providing the broadest AI infrastructure offerings. AMD Instinct GPUs, paired with OCI's performance, advanced networking, flexibility, security, and scale, will help our customers meet their inference and training needs for AI workloads and new agentic applications."
The zettascale OCI Supercluster with AMD Instinct MI355X GPUs delivers a high-throughput, ultra-low latency RDMA cluster network architecture for up to 131,072 MI355X GPUs. AMD claims the MI355X provides almost three times the compute power and a 50 percent increase in high-bandwidth memory over its predecessor.
Performance and flexibility
Forrest Norrod, Executive Vice President and General Manager, Data Center Solutions Business Group, AMD, commented on the partnership, stating: "AMD and Oracle have a shared history of providing customers with open solutions to accommodate high performance, efficiency, and greater system design flexibility. The latest generation of AMD Instinct GPUs and Pollara NICs on OCI will help support new use cases in inference, fine-tuning, and training, offering more choice to customers as AI adoption grows."
The Oracle platform aims to support customers running the largest language models and diverse AI workloads. OCI users leveraging the MI355X-powered shapes can expect significant performance increases—up to 2.8 times greater throughput—resulting in faster results, lower latency, and the capability to run larger models.
AMD's Instinct MI355X provides customers with substantial memory and bandwidth enhancements, which are designed to enable both fast training and efficient inference for demanding AI applications. The new support for the FP4 format allows for cost-effective deployment of modern AI models, enhancing speed and reducing hardware requirements.
The dense, liquid-cooled infrastructure supports 64 GPUs per rack, each operating at up to 1,400 watts, and is engineered to optimise training times and throughput while reducing latency. A powerful head node, equipped with an AMD Turin high-frequency CPU and up to 3 TB of system memory, is included to help users maximise GPU performance via efficient job orchestration and data processing.
Open-source and network advances
AMD emphasises broad compatibility and customer flexibility through the inclusion of its open-source ROCm stack. This allows customers to use flexible architectures and reuse existing code without vendor lock-in, with ROCm encompassing popular programming models, tools, compilers, libraries, and runtimes for AI and high-performance computing development on AMD hardware.
Network infrastructure for the new supercluster will feature AMD's Pollara AI NICs that provide advanced RDMA over Converged Ethernet (RoCE) features, programmable congestion control, and support for open standards from the Ultra Ethernet Consortium to facilitate low-latency, high-performance connectivity among large numbers of GPUs.
The new Oracle-AMD collaboration is expected to provide organisations with enhanced capacity to run complex AI models, speed up inference times, and scale up production-grade AI workloads economically and efficiently.
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Techday NZ
14 hours ago
- Techday NZ
Watercare uses edge intelligence for smarter sewer monitoring in Auckland
Watercare is deploying 5,000 sewer level monitors across its Auckland wastewater network, utilising edge intelligence to detect issues such as blockages and pipe failures directly at the source. This initiative, carried out in partnership with Kallipr, moves beyond conventional systems which rely largely on cloud-based processing. Instead, the monitors process data locally, enabling real-time responses to potential events and reducing reliance on external communication infrastructure. Edge intelligence Dave Moore, Smart Systems Manager at Watercare, described the importance of the localised approach. "What's unique here is where the intelligence lives. By shifting processing out to the edge, we're cutting down on cloud consumption and enabling the network to respond faster, closer to the source of the issue in near real time. It means less lag, less noise, and far more confidence in how we manage events as they unfold." The local data processing is intended to lower communication overheads, extend the lifespan of monitoring solutions, and identify network issues earlier, reducing the risk of sewage spills and their associated environmental and financial costs. Beyond monitoring The sensors provided by Kallipr form part of Watercare's wider Smart Network Programme, valued at NZD $12 million. This programme focuses on integrating digital technologies and artificial intelligence into existing water infrastructure. The devices are being deployed city-wide, including in regions prone to flooding such as Wairau. Each sensor, installed beneath manhole lids, autonomously monitors sewer conditions, adapting its sampling frequency when it detects anomalies via radar and TinyML-powered logic. The systems' long-life batteries further support extended deployment periods without frequent maintenance. Gerhard Loots, Chief Executive Officer of Kallipr, outlined the broader implications for the utility sector. "Many utilities are still focused on collecting more data but the next leap forward is creating intelligence in the field. This deployment shows how utilities can evolve from data-heavy, cloud-reliant systems to a model that's faster, leaner, and built for real-world complexity. Watercare is leading with a future-ready approach that others can follow." Technology details The core of the deployment is Kallipr's Spectra radar sensor, which is designed to withstand the challenging conditions found in sewers, including exposure to corrosive gases, submersion, potential animal interference, and unreliable network signals. Spectra records data at regular 15-minute intervals, but can increase its activity if local sensors anticipate overflow or blockage situations. These devices use low-power LTE networks - including NB-IoT and Cat-M1 standards - to transmit information, while onboard TinyML algorithms support local adaptation and ongoing battery efficiency. Connectivity challenges in underground settings are mitigated by dual SIM capability, with the devices using edge logic to choose the strongest available network. This ensures stable, continuous communication regardless of external conditions. Device management is centralised through Kallipr Kloud Fleet, a platform built on Microsoft Azure, which allows remote monitoring and control of the distributed monitoring units. Wider impact Edge-based processing reduces the amount of bandwidth required, energy consumption, and load on cloud systems, helping to trim operational expenses and enhance environmental outcomes. According to Watercare, each avoided spill mitigates contamination risks, minimises unnecessary travel to field sites, and prevents excess emissions. The deployment of this network in Auckland is reflective of a global move toward decentralised utility management. Watercare's approach demonstrates that efficiency and responsiveness can be improved by locating decision-making as close as possible to operational events. This strategy is intended to drive cost efficiency, resilience, and the long-term adaptability of critical infrastructure networks.
Techday NZ
4 days ago
- Techday NZ
Oracle launches global Exadata service for AI & analytics
Oracle has launched its Globally Distributed Exadata Database on Exascale Infrastructure, providing businesses with a new approach to deploying mission-critical applications and agentic AI workloads across various cloud regions while meeting local data residency requirements. Service features The recently available service is designed to support organisations as they manage high-availability applications and extensive global operations. Key aspects include automated data distribution and synchronisation across multiple sites, ensuring continued operation during regional outages. The service is intended to enable companies to deploy workloads, including those requiring real-time analytics, transactional capacity, and agentic AI processing, without complex configuration or management tasks. Oracle states that the solution utilises a flexible, serverless architecture, allowing resources to scale in response to fluctuating workloads. Pricing is presented as simple and based on the number and types of virtual machines and storage selected, which the company says gives customers clear visibility of operational costs without hidden charges. Industry adoption PayPal has used Oracle Exadata for years to maintain performance consistency and critical system uptime. Akash Guha, Director of Database Engineering at PayPal, explained the company's plans for the new service: "Providing exceptional customer satisfaction is important to PayPal, so we've been using Oracle Exadata for many years to provide lightning-fast response times and mission-critical availability. As our global business grows, we plan to provide even faster responses by using distributed solutions that are integrated with our core systems of record to provide extreme availability and performance. We look forward to using Oracle Globally Distributed Exadata Database on Exascale Infrastructure's always-on, serverless architecture with built-in Raft replication to accelerate responses, enable greater application resilience, and lower costs with scalable resources." Technical benefits The new offering contains several mechanisms aimed at addressing the main requirements of agentic AI and global applications. These include high performance for large workloads and vector search, always-on availability through Raft replication and fault-tolerant architecture, support for data residency and sovereignty, and cost-efficiency through hyperscale, serverless implementation. Wei Hu, Senior Vice President, High Availability Technologies at Oracle, commented on the challenges faced by customers aiming to distribute databases at scale: "Customers often struggle to deploy and manage distributed databases due to the high cost and complexity involved in operating large numbers of servers across multiple data centres and regions. Oracle Globally Distributed Exadata Database on Exascale Infrastructure's serverless architecture enables customers of all sizes to meet their diverse requirements at a low cost. Today, we are providing a mission-critical distributed database to the masses." Oracle notes that the system naturally supports the full Oracle Database and SQL capabilities, enabling clients to distribute resources to multiple locations with minimal modifications to existing applications. Workload support The Globally Distributed Exadata Database on Exascale Infrastructure has been designed with several specific use cases in mind. Its active/active/active architecture supports mission-critical processes such as payments, eCommerce, and securities trading by maintaining synchronised data across multiple centres. Automated data distribution policies allow enterprises in finance and healthcare to meet regulations on data sovereignty. User proximity features enable companies in sectors like retail, telecommunications, and healthcare to reduce latency by storing data closer to end users. Petabyte-scale analytics functionality provides for real-time streaming and processing, vital for manufacturing and utility industries, and the infrastructure's online transaction processing (OLTP) capability supports millions of transactions with low latency for AI-driven and analytic workloads. AI and data perspectives Holger Mueller, Vice President and Principal Analyst at Constellation Research, said: "In the age of AI, especially agentic AI, customers need a new approach to allow for vector processing across distributed global applications. Oracle Globally Distributed Exadata Database on Exascale Infrastructure incorporates Exadata's extreme performance for AI processing and availability for the core back-end systems that implement agent-initiated tasks, while Exascale's hyper-elastic and pay-per-use capabilities makes it very cost-effective. With this service from Oracle, CIOs can confidently deploy agentic AI and mission-critical applications globally and meet local data residency requirements." According to Oracle, by overcoming barriers related to performance, data sovereignty, cost, and elastic scalability, the service aims to appeal to a range of industries where operational continuity, data compliance, and processing performance are mandatory.
Techday NZ
4 days ago
- Techday NZ
GigaChat AI assistant achieves 93% accuracy in medical diagnoses
SberHealth's GigaChat-powered artificial intelligence assistant has demonstrated a diagnostic accuracy rate of 93% during recent tests conducted by the Artificial Intelligence Research Institute (AIRI). The experiment involved the AI healthcare assistant, which is based on the GigaChat neural network model, diagnosing 30 real clinical cases that were randomly selected from the New England Journal of Medicine. These cases varied in complexity, and the testing methodology used was similar to an experiment conducted by Microsoft to verify its own AI diagnostic orchestrator, MAI-DxO. According to AIRI, the SberHealth system established correct diagnoses in 28 out of 30 cases, while a comparable foreign solution recorded an 85% accuracy rate. The AI assistant operated with limited initial data, receiving only the patient's gender, age and symptoms before interacting through simulated doctor-patient dialogues. It followed a sequence of requesting additional clinical tests, imaging, or consultation information as needed to make diagnoses. The median number of dialogue turns between the AI and the simulated patient was three, indicating a relatively high speed of decision-making. Sergey Zhdanov, Director of the Healthcare Industry Centre at Sberbank, said: "The experiment demonstrated that our technology is not only competitive but also sets new standards in medical diagnostics worldwide. We observe how multi-agent architecture speeds up and enhances the diagnostic process. It's particularly important that the system exhibits flexibility: it revises hypotheses, requests additional data, and even responds to the emotional presentation of clinical scenarios. In the future, this opens up opportunities for interdisciplinary care teams, with AI serving as a reliable assistant to physicians." During the experiment, each clinical case was labelled by level of difficulty. The AI system was able to successfully identify and diagnose several rare conditions, including Whipple disease, which it recognised in one step, aceruloplasminemia, identified in six moves, and rasburicase-induced methemoglobinemia. The assistant's performance was characterised by several features, according to researchers. It typically completed diagnoses in three moves, deployed logical reasoning, and handled both rare and complex pathologies. The system was also noted for its ability to blend clinical accuracy with a dialogue logic that could adapt effectively to different presentation styles, which included effectively responding to emotional cues in simulated scenarios. Ivan Oseledets, Chief Executive Officer of AIRI, commented: "Today, multi-agent systems are capable of confidently identifying rare, masked pathologies that go beyond typical emergency department algorithms. Can a medical AI assistant adjust its hypothesis in time, discarding the most probable but incorrect pathway? The AI assistant proved it could, doing so faster than anticipated by a seasoned observer with 15 years of medical experience." The researchers at AIRI described the experiment as exploratory and indicated that further development is planned. They have proposed expanding the sample size by incorporating additional cases from other medical journals to investigate the capabilities of the assistant more widely. The system's potential uses were not limited solely to practical medicine but also extended to the area of physician training, where it could offer realistic simulations of complex clinical cases. The GigaChat-based assistant is a product of cooperation between AIRI and SberMedAI. Since its introduction, it has been piloted in the SberHealth app and has already been used over 160,000 times in real conditions to assist people seeking medical support. Follow us on: Share on:
