Broadcom announced VMware Tanzu Platform 10, a cloud-native application platform, with standardized golden paths to production for developers and platform engineers. Developers can use simple operations to automate secure container builds, bind services to apps, deploy code with a single command, and scale applications. Tanzu Platform 10 also automates both application and platform management tasks such as patching vulnerabilities, performing rolling upgrades, and enforcing policies with broad visibility and AI-powered insights.
…
China’s DapuStor has flexible Data Placement (FDP) technology for QLC and TLC SSDs using Marvell’s Bravera SC5 controller, minimizing write amplification down to near zero to extend the SSD’s endurance – or so we’re told. The FDP drives also feature error correction and wear leveling techniques. Zstor is DapuStor’s distributor in Europe.
…
HPE has built a sixth-generation high-performance computing cluster, Iridis 6 with 26,496 AMD computing cores in 138 server nodes, with 3TB of memory, for the University of Southampton. HPE says it provides four times the performance at 70 percent of the power consumption compared to the Iridis 4 (12,000 cores) it replaces. It will operate with 1 petaflop HPL performance. Iridis 6 uses 138x HPE ProLiant gen 11 servers with gen 4 AMD EPYC 192-core CPUs, each with 3TB of memory and 15TB of local storage. It uses InfiniBand HDR100 Interconnect for high-speed data transfer. Iridis 5, with its >25,000 cores (464x Intel Xeon nodes with 2x Gold 6138 CPUs), 74 GPU cards, and 2.2PB of IBM Storage Scale parallel file system capacity, continues in operation. Iridis 6’s storage software has not been revealed but Storage Scale looks well placed.
…
IBM has integrated Ceph 7.1 Object Storage with its Storage Protect v8.1.18 product. It said: “Integrating IBM Storage Ceph Object Lock allows organizations to implement WORM (Write Once, Read Many) protection on critical data, preventing tampering or deletion. This is crucial for compliance with regulations like GDPR, SEC Rule 17a-4(f), and CFTC Rule 1.31(c)-(d) … This integration includes object storage and object lock validation, ensuring your enterprise data is efficiently stored and safeguarded against unauthorized changes and enhancing your data resilience strategy. … IBM Storage Protect integrates seamlessly with IBM Storage Ceph, optimizing storage use by reducing data redundancy and improving overall storage efficiency through features like deduplication and compression.” Read more here.
…
Ocient, whose software analyzes exceedingly big data, has released a standard TPC-H Star Schema Benchmark (SSB) flat version benchmark against Apache Druid, and says Ocient “was faster across the board, with the Ocient advantage growing larger as the compute load increased.” Ocient was also more parsimonious with storage, needing 81TB for a 180TB raw dataset with 536 billion rows. Druid needed almost 250TB. Register here to download a copy of the benchmark report.
…
Attendees at the Open Infrastructure Summit Asia in South Korea learned that the OpenStack market is estimated at $22.81 billion in 2024 and is expected to reach $91.44 billion by 2029, growing at a CAGR of 32 percent. A white paper published last week by the OpenInfra Foundation – Migrating from VMware to OpenStack: Optimizing your Infrastructure to Save Money and Avoid Vendor-Lock-in – documents many reasons to move from VMware to OpenStack, ranging from technical to business reasons. A second white paper entitled “Open Infrastructure Blueprint” discusses users deploying container-based, virtual machine-based, and bare metal-based workloads in the same infrastructure instead of dividing up workloads across multiple infrastructure providers. In particular, using OpenStack and Kubernetes together produces multiple opportunities to enhance the functionality of both platforms beyond what they each provide on their own.
…
PNY has 1 and 2TB portable RP60 SSDs with a PCIe 3.2, 2 x 2 interface, for $103.99 and $161.99 respectively. They offer water, dust, and drop resistance (up to three meters), along with read and write speeds of 2,000MB/sec and 1,800MB/sec.
…
Pure Storage will support the next generation of VMware vSphere Virtual Volumes (vVols) when it becomes generally available, currently planned to be part of VMware Cloud Foundation 9. A VMware blog says next-gen vVols will have three main goals: deliver a consistent, optimized user experience across storage platforms; evolve vVols for modern, large scale use cases such as AI/ML workloads and Cloud Service Provider deployments; fully integrate with VMware Cloud Foundation (VCF). Planned customer benefits of next-gen vVols will include:
- Full integration with VMware Cloud Foundation;
- Install, deploy and manage storage with VCF;
- Usable as primary and supplemental storage for all workload domains;
- Prescriptive VASA model to deliver consistent, optimized user experience across all storage platforms;
- Support for high fan-in, cloud-scale use cases;
- Enable high availability and stretched cluster deployments;
- Seamless migration from legacy vVols platform without any downtime;
- Scalable, immutable snapshots for both traditional and modern applications.
A Pure blog has a little more information.
…
A Seagate disk marketeer claims HDDs will not be replaced by SSDs. Read what Jason Feist, Seagate’s SVP of products and markets, says here. Note we don’t find commodity SSD suppliers saying SSDs will replace HDDs and no new HDD systems will be sold after 2028 – which is the Pure Storage pitch we heard last year. IBM, which also makes proprietary flash drives along with Pure, hasn’t made this claim either.
…
The SNIA Developer Conference (SDC 24) will take place in Santa Clara, CA, over September 16–18. SDC is an educational conference dedicated to providing vendor-neutral information on a broad range of industry trends and technologies. SDC 24 takes a deep dive into exploring CXL. Here are some of the sessions you will find at the event:
- SDXI + Computational Storage Overview and Panel Discussion
- Designing and Optimizing Complex CXL Configurations
- CXL for LLMs and Ray: A Practical Guide to Unleashing the Potential of AI and Machine Learning
- Optimized Resource Allocation for CXL Tiered-Memory Systems
…
China’s TerraMaster announced an upcoming release of its TOS 6 system, integrated with a BBS (Business Backup Suite) full-scenario backup offering. BBS combines a “professional storage management system, a suite of backup software tools including Centralized Backup, TerraSync, Duple Backup, CloudSync, and Snapshot, and a hardware platform, achieving deep integration without the need for additional software costs. Backup tasks and storage capacity are unlimited.”
…
VAST Data has published a 34-page eBook – Optimizing AI Strategy in the Modern Data Landscape – written by Colleen Tartow, field CTO and head of strategy. It concludes that “organizations face an unprecedented explosion of structured and unstructured data in today’s rapidly evolving digital landscape. As the volume, variety, and velocity of data continue to grow, traditional data architectures are struggling to keep pace, leading to silos, complexity, and sub-optimal performance. Intentionally and thoughtfully evolving in terms of data maturity is a key value driver for organizations as they look to stay at the forefront of extracting value from data. In this book we have shed light on these challenges and emphasized the need for a paradigm shift in data management – adopting a unified data platform that seamlessly supports BI and AI workloads.” VAST, of course, supplies a unified data platform that seamlessly supports BI and AI workloads. Get your copy by registering here.
…
Privately held cloud computing platform supplier Vultr, which supplies Nvidia GPU-based services, says data and analytics acceleration platform supplier SQream has has joined the Vultr Cloud Alliance – a partnership program enabling composable cloud services. Through this, Vultr says, “AI-driven enterprises can fast-track their data analysis and machine learning projects without the burden of traditional data processing limitations.” SQream joins other Alliance members, including Qdrant, a high-performance vector database with retrieval-augmented generation (RAG) capabilities; Console Connect, which facilitates private, high-speed networking for secure, low-latency data transfer; DDN for sophisticated AI storage infrastructure; and Run:ai, an advanced AI workload orchestration platform.
