Diamanti's previous platforms have been in the form of on-premise hyper-converged infrastructure (HCI) appliances that provide both the HW and SW needed to support containerized applications. With Spektra, Diamanti aims to combine their hardware-accelerated x86 platform with cloud-based infrastructure to provide Kubernetes-as-a-Service, along with built-in disaster recovery of workloads and application data. While Diamanti claims that they will be able to provide a single control pane spanning both their on-premise hardware and public clouds, they have provided no details on what that might look like or how it would work. Someone seems to believe their solution is worthwhile though. On November 7th the company announced the close of a $35 million Series C funding round.
Diamanti Announces Hybrid Cloud Kubernetes Platform
Dell Technologies is expanding its portfolio of Dell EMC Ready Solutions for HPC Storage with two new, turnkey solutions. The first solution is for ThinkParQ's BeeGFS. ThinkParQ's software-defined parallel file system, speeds-up input/output-intensive workloads with the ability to scale from small clusters to enterprise-class systems on-premises or in the cloud. The second solution is for ArcaStream's PixStor file systems. This solution offers a high-performance parallel file system, enabling data management at scale with the ability to perform archive and analytics in place. Both of these new solutions are immediately available. At the same time, Dell EMC is also releasing a 400GbE networking switch called the PowerSwitch Z9332F-ON.
Dell EMC Server Announcements from SC19
The reference design platform includes both hardware and software building blocks and was designed in response to the HPC community’s need for a more diverse range of CPU architectures. As such, it gives supercomputing centers, hyperscale-cloud operators and enterprises the ability to combine NVIDIA’s accelerated computing platform with the latest Arm-based server platforms.
NVIDIA Announces New Reference Design Platform
NVIDIA partnered with industry leaders in networking and storage to develop Magnum IO, including DataDirect Networks, Excelero, IBM, Mellanox and WekaIO. This software suite release is also highlighted by GPUDirect Storage, which gives researchers the ability to bypass CPUs when accessing storage for quick access to data files in applications such as simulation, analysis or visualization.
NVIDIA Magnum IO Software Suite Now Available
To build this new scalable GPU-accelerated supercomputer, Microsoft and NVIDIA engineers used 64 NDv2 instances on a pre-release version of the cluster to train BERT in approximately three hours. This was possible due to the multi-GPU optimizations provided by NCCL, an NVIDIA CUDA X library, and high-speed Mellanox interconnects.
NVIDIA Announces Scalable GPU-Accelerated Supercomputer in the Microsoft Azure Cloud
There have been many announcements around AMD EPYC Rome CPUs today, and now TYAN has released advancements in its Transport HX and SX Product lines. For the HX, TYAN rolled out three new platforms powered by AMD. The Transport HX TN83-B8251 is a 2U server platform that supports up to eight 3.5” hot-swap SATA or NVMe U.2 tool-less drive bays, and is ideal for AI training and inference applications deploying 4 double-width or 8 single-width GPU cards, and 2 PCIe 4.0 x16 high-speed networking cards. The Transport HX TS75-B8252 and Transport HX TS75A-B8252 are 2U server platforms with support for 32 DIMM slots and up to 9 PCIe 4.0 slots. These server platforms are ideal for HPC and virtualization.
TYAN Launches AMD EPYC HPC & Storage Servers At SC19
Since the release of AMD EPYC Rome CPUs, the industry has seen a pretty good adoption. The new CPUs came out and shattered world records and can deliver better performance in a single socket versus its competition’s dual socket setup. A big advantage of the AMD EPYC Rome is that it allows servers to leverage PCIe 4.0 devices. In a world that is leveraging GPUs more and more, this is a huge leg up over the competition. With these Advantages, AMD is pushing its way into the HPC/supercomputing market where the above will be leveraged quickly.
AMD Makes Several Announcements At SC19
GIGABYTE has offered several GPU dense servers in its G-Series of server family. Today they are expanding this with seven new server, including five more 2nd Generations AMD EPYC severs (bringing their AMD total to 28 servers). The newly announced servers leverage GIGABYTE Management Console as standard, which is based on an AMI MegaRAC SP-X web-browser based platform and is compliant with the latest standards of Redfish API. Also available as a free download is GIGABYTE Server Management (GSM), GIGABYTE’s multiple server remote management software platform, and includes both a desktop and mobile APP.
GIGABYTE Releases 7 GPU Servers
With this fourth-generation release, Cumulus believes that customers will have everything they need to meet modern network challenges, as this networking software is specifically designed to run and operate today’s data center and campus environments.
Cumulus Linux 4.0 and NetQ 2.4 Announced
With ZCentral, HP claims that IT departments will gain predictable and dedicated remote performance coupled with an easy way to securely manage these connections. Moreover, ZCentral is designed to lessen both the hurdles and costs that are inherent in managing various platforms and devices while offering secure remote access to users. HP adds that it will boast 16%-33% higher performance, at 59%-72% of the cost compared to “today’s virtualized workstations”.
HP Announces HP ZCentral
While TrilioVault is a well-established software solution, we haven’t previously covered it in depth so it is worth briefly reviewing what it is. The core TrilioVault technology is an agentless, software-only model that is self-service for multi-tenant environments that support high degrees of scale. It uses an open and incremental point-in-time backup scheme, and the company boasts it can support any storage target.
TrilioVault Achieves Red Hat’s Container Certification
This collaboration is part of the U.S.A. Department of Energy’s National Renewable Energy Laboratory (NREL) ongoing efforts to reduce the amount of energy, and water used by data centers. In 2016 NREL even installed a prototype closed-loop cooling system in their own datacenter. You can see the real-time performance of the cooling system on NREL’s website. At the time of writing, the closed-loop system was providing about half of the cooling provided by the evaporative cooling towers, effectively cutting the water used at the facility by a little less than a third.
HPE Collaborates With National Renewable Energy Laboratory
HPE appears to be attempting to provide a package suitable for all enterprise environments and any conceivable use case with their HPE Container Platform. According to the company, the platform will be able to deploy application environments on any infrastructure, any public cloud, and at the edge. HPE specifically lists support for bare-metal deployments, virtual machine deployments, and cloud instance deployments. Astonishingly, HPE also claims their upcoming new platform will provide a unified Kubernetes-based software platform for both traditional and cloud-native application workloads. The company even says they'll be able to modernize non-cloud-native monolithic applications and elevate them to modern cloud standards. Which would be an astonishing achievement if true.
Hewlett Packard Enterprise Announces Container Platform
Kubernetes has been widely popular since it was announced in 2015. Since its release it has seen wide and quick adoption across the board. However, many of the touted benefits tend to overlook issues that are still present when leveraging the technology. In order to deal with these arising challenges D2iQ has released Kommander. Kommander provides role-based visibility and unified operational control for a wide variety of in-the-cloud and on-premises Kubernetes distributions, including D2iQ’s Konvoy. Kommander also enables multi-cluster governance and centralized supervision without interfering with the day-to-day business functions and individual team requirements that different clusters support.
D2iQ Introduces Kommander