
Oracle And AMD Collaborate To Deliver Breakthrough Performance In AI Workloads
'To support customers that are running the most demanding AI workloads in the cloud, we are dedicated to providing the broadest AI infrastructure offerings,' said Mahesh Thiagarajan, executive vice president, Oracle Cloud Infrastructure. '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.'
To support new AI applications that require larger and more complex datasets, customers need AI compute solutions that are specifically designed for large-scale AI training. The zettascale OCI Supercluster with AMD Instinct MI355X GPUs meets this need by providing a high-throughput, ultra-low latency RDMA cluster network architecture for up to 131,072 MI355X GPUs. AMD Instinct MI355X delivers nearly triple the compute power and a 50 percent increase in high-bandwidth memory than the previous generation.
'AMD and Oracle have a shared history of providing customers with open solutions to accommodate high performance, efficiency, and greater system design flexibility,' said Forrest Norrod, executive vice president and general manager, Data Center Solutions Business Group, AMD. '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.' AMD Instinct MI355X Coming to OCI
AMD Instinct MI355X-powered shapes are designed with superior value, cloud flexibility, and open-source compatibility—ideal for customers running today's largest language models and AI workloads. With AMD Instinct MI355X on OCI, customers will be able to benefit from: Significant performance boost: Helps customers increase performance for AI deployments with up to 2.8X higher throughput. To enable AI innovation at scale, customers can expect faster results, lower latency, and the ability to run larger AI workloads.
Helps customers increase performance for AI deployments with up to 2.8X higher throughput. To enable AI innovation at scale, customers can expect faster results, lower latency, and the ability to run larger AI workloads. Larger, faster memory: Allows customers to execute large models entirely in memory, enhancing inference and training speeds for models that require high memory bandwidth. The new shapes offer 288 gigabytes of high-bandwidth memory 3 (HBM3) and up to eight terabytes per second of memory bandwidth.
Allows customers to execute large models entirely in memory, enhancing inference and training speeds for models that require high memory bandwidth. The new shapes offer 288 gigabytes of high-bandwidth memory 3 (HBM3) and up to eight terabytes per second of memory bandwidth. New FP4 support: Allows customers to deploy modern large language and generative AI models cost-effectively with the support of the new 4-bit floating point compute (FP4) standard. This enables ultra-efficient and high-speed inference.
Allows customers to deploy modern large language and generative AI models cost-effectively with the support of the new 4-bit floating point compute (FP4) standard. This enables ultra-efficient and high-speed inference. Dense, liquid-cooled design: Enables customers to maximize performance density at 125 kilowatts per rack for demanding AI workloads. With 64 GPUs per rack at 1,400 watts each, customers can expect faster training times with higher throughput and lower latency.
Enables customers to maximize performance density at 125 kilowatts per rack for demanding AI workloads. With 64 GPUs per rack at 1,400 watts each, customers can expect faster training times with higher throughput and lower latency. Built for production-scale training and inference: Supports customers deploying new agentic applications with a faster time-to-first token (TTFT) and high tokens-per-second throughput. Customers can expect improved price performance for both training and inference workloads.
Supports customers deploying new agentic applications with a faster time-to-first token (TTFT) and high tokens-per-second throughput. Customers can expect improved price performance for both training and inference workloads. Powerful head node: Assists customers in optimizing their GPU performance by enabling efficient job orchestration and data processing with an AMD Turin high-frequency CPU with up to three terabytes of system memory.
Assists customers in optimizing their GPU performance by enabling efficient job orchestration and data processing with an AMD Turin high-frequency CPU with up to three terabytes of system memory. Open-source stack: Enables customers to leverage flexible architectures and easily migrate their existing code with no vendor lock-in through AMD ROCm. AMD ROCm is an open software stack that includes popular programming models, tools, compilers, libraries, and runtimes for AI and HPC solution development on AMD GPUs.
Enables customers to leverage flexible architectures and easily migrate their existing code with no vendor lock-in through AMD ROCm. AMD ROCm is an open software stack that includes popular programming models, tools, compilers, libraries, and runtimes for AI and HPC solution development on AMD GPUs. Network innovation with AMD Pollara: Provides customers with advanced RoCE functionality that enables innovative network fabric designs. Oracle will be the first to deploy AMD Pollara AI NICs on backend networks, providing advanced RoCE functions such as programmable congestion control and support for open industry standards from the Ultra Ethernet Consortium (UEC) for high-performance and low latency networking.
Hashtags

Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles


Gulf Today
3 days ago
- Gulf Today
In conversation with Armin Moradi, CEO and Co-Founder of Qashio on how Qashio is solving real CFO challenges
As the financial technology landscape in the Middle East evolves, homegrown businesses like Qashio are leading the charge. Gulf Today sat down with Armin Moradi, CEO and Co-Founder of Qashio, to discuss how the company is leveraging AI, building tailored solutions for key industries like travel, and helping finance teams stay compliant and agile in a fast-changing regulatory environment. 1. With increased interest in AI-driven financial tools, is Qashio exploring or integrating any AI technologies to enhance automation, insights, or user experience? Qashio has been at the forefront of innovation and was an early adopter of AI, using it to streamline manual processes like scanning receipts and automating transaction categorisation. We are now exploring additional AI applications cautiously, with a strong focus on compliance and data privacy. Since UAE regulations mandate local data storage, we partner exclusively with UAE-based AI providers. Fortunately, the UAE's heavy investment in AI has created world-class technologies we are evaluating, particularly for delivering deeper insights and enhancing user experience. Any AI-driven features we roll out will strictly adhere to local privacy and compliance standards. 2. What's driving Qashio's focus on the travel industry, and how does your platform cater to the specific financial needs and pain points of travel-related businesses? Travel is one of the UAE's largest industries, with the UAE receiving over 18.7 million visitors in 2024 alone. As a UAE-based company, we're committed to supporting businesses across all sectors. It was a natural step to design Qashio for Travel, a solution built specifically for the unique needs of the Travel industry. Travel companies can instantly create unlimited corporate cards, leverage multi-currency wallets to save on high FX fees, and integrate seamlessly with GDS platforms. This tailored approach helps streamline financial operations and supports travel agencies book the 5.4 million UAE residents who travelled last year. 3. How is Qashio differentiating its offering to stand out among regional and global players? Innovation is at the core of Qashio's DNA. We want to ensure we're offering our clients the most functionality and value for their business. We are the only spend management solution in the region with a corporate loyalty programme, enabling businesses to earn rewards on their spending, redeemable for air miles, hotel points, and retail benefits. Built by ex-CFOs, Qashio offers unmatched functionality tailored to real-world finance challenges, including WhatsApp-based receipt collection and hyper-segmented accounting abilities. To support our clients fully, we've assembled a UAE-based team of experts with experience at ERP giants like Oracle, Odoo, and SAP, ensuring hands-on, localised support for our clients. 4. How is Qashio supporting finance teams in navigating today's fast-changing regulatory and economic landscape across the GCC? Qashio is committed to helping UAE businesses thrive in the fast-changing regulatory and economic landscape. Our fully local team monitors shifts in GCC regulations and proactively advises clients on compliance requirements, ensuring they remain agile and informed. At the same time, we continuously refine our platform based on client feedback. Designed by finance professionals, for finance professionals, Qashio empowers teams to navigate complexity with automation, visibility, and control at their fingertips.


Arabian Post
09-07-2025
- Arabian Post
Blender 5.0 Brings HDR Viewport to Linux Wayland Users
Blender 5.0 is introducing experimental High Dynamic Range rendering support for Linux users working in a Wayland session with Vulkan acceleration. This enhancement aims to offer creators precise brightness, contrast, and colour fidelity in the viewport, mirroring the output of HDR‑enabled monitors. HDR viewport support in the alpha build of Blender 5.0 must be explicitly enabled under Developer > Experimental settings, requiring a Vulkan graphics backend on Wayland. Testers have successfully run the feature on GNOME and KDE Plasma environments across Ubuntu 25.04 and Fedora setups, demonstrating compatibility with NVIDIA, AMD, and Intel GPUs using monitors such as the ASUS ROG PG27U and Samsung Odyssey OLED G8. Blender's Jeroen Bakker highlighted on DevTalk that this feature remains experimental due to limited system testing, and the development team is actively seeking feedback to determine readiness for stable inclusion in the final 5.0 release. Community testers corroborate this request; one user reported flawless performance on Arch Linux with AMD 7900XTX and a 32‑inch HDR monitor, while another observed UI inconsistencies such as bright region overlays conflicting with UI elements. ADVERTISEMENT Phoronix's assessment affirms the limited initial testing but confirms functional results in alpha stage across Vulkan‑on‑Wayland configurations. Wayland's HDR protocol—absent in X11—and Vulkan's floating‑point buffer support are essential prerequisites, making the feature inaccessible without those components. Professional and hobbyist artists stand to benefit from HDR's ability to display a broader dynamic range, ensuring views in the viewport better match final renders. Users on Reddit commented that the feature is highly anticipated and a compelling incentive to invest in HDR hardware. Some, however, caution that industry‑wide support for HDR remains uneven, with software ecosystems often lagging behind hardware capabilities. Support across GPU vendors appears robust: developer testing includes NVIDIA alongside AMD and Intel. This suggests mature driver compatibility within the Vulkan stack, though users have noted occasional driver‑specific anomalies. Discussions also emphasise that Windows support for the feature is not yet scheduled, sparking interest in extending HDR viewport to other platforms. To experiment with HDR, users should download the latest Blender 5.0 alpha, launch a Wayland session with HDR‑enabled display output, switch to Vulkan in Preferences → System, enable Developer Extras and turn on Vulkan HDR support under Experimental settings, then restart Blender. Within Scene Color Management, HDR must also be activated for accurate viewing. The developer's roadmap hinges on accumulating comprehensive user feedback to address UI integration issues—such as overlay visibility—and ensure stable behaviour across diverse hardware and compositors. If successful, HDR support will graduate from experimental to enabled by default in the 5.0 stable release. This HDR initiative arrives amid a broader shift, as major Linux distributions migrate desktops to Wayland to leverage modern graphics features, enhanced security, and performance improvements. Artists seeking more lifelike previews stand to gain from this update. Those with HDR‑capable displays and Wayland desktops are encouraged to test the feature and share experiences on Blender DevTalk, helping refine HDR delivery ahead of the official 5.0 rollout.


Arabian Post
07-07-2025
- Arabian Post
TikTok Unveils Stand‑alone U.S. App to Secure Sale
TikTok is preparing a standalone version of its app for U. S. users, slated for launch on 5 September 2025, as part of its strategy to comply with the U. S. law mandating a sale or ban of its American operations. The current app will remain functional until March 2026, although officials say this date is subject to change. Under the Protecting Americans from Foreign Adversary Controlled Applications Act, ByteDance must divest TikTok's U. S. unit by 17 September 2025 or face prohibition. The creation of the so-called 'M2' app would separate code and infrastructure from its Chinese parent, aiming to address national‑security concerns raised in Washington. Donald Trump indicated 'the United States pretty much has a deal' and that talks with Beijing would commence early this week, involving President Xi or his envoys. The sale is expected to transfer U. S. operations to a consortium of American investors—potentially including Oracle—while ByteDance retains a minority stake. Yet technical-and-political tension between the partnering governments looms; China has resisted ceding control over the algorithm, deemed strategic technology. ADVERTISEMENT Industry insiders say the 'M2' version represents more than a mere rebrand. It reportedly involves fully separate source code and systems designed to satisfy U. S. regulators and reassure users that data and operations are controlled domestically. But building and migrating some 170 million users carries risks: fragmentation of creator and brand ecosystems, potential loss of content anytime users delay updating, and disruptions for advertisers reliant on consistent U. S. reach. One senior analyst at eMarketer, tracking platform growth, noted that while TikTok's U. S. revenue had slowed in early 2024, its user base remains resilient. Still, the shift could alter engagement patterns. Advertisers are reportedly waiting for regulatory clarity before reallocating ad spends—many are wary of campaign uncertainty tied to a dual‑app environment. On the political front, the bill forcing divestment won bipartisan support in 2024 to curb data security threats, citing TikTok's potential vulnerability to influence operations by foreign actors. The law originally prescribed a January 2025 deadline, but President Biden's enactment in April 2024 and subsequent legal disputes shifted the compliance date. President Trump's extension to September 17, 2025, accommodates final negotiations. Expected approval from Australia and Canadian regulators for a pilot 'Notes' app—which decouples photo-sharing features—demonstrates TikTok's evolving architecture and ability to modularise code for geopolitical compliance. The planned U. S. pivot appears to follow similar design logic. ByteDance, parent of TikTok, declined to comment. U. S. officials, including the Committee on Foreign Investment in the United States, are reviewing proxy structures to ensure full operational and data separation. Final approval will require both Washington and Beijing to accede, with the latter retaining leverage through export controls on algorithmic technology. Major bid participants are still emerging. Oracle reappears as a likely contender given prior discussions. Meanwhile, Project Liberty—a consortium led by Frank McCourt, backed by Kevin O'Leary—has pitched to replace TikTok's algorithm with one focused on user data sovereignty and could resurface in late-stage talks.