Tag: Principled

 Creating a virtual environment for applications can bring a host of benefits, but it also adds some amount of computing overhead compared to a bare-metal environment. For compute-heavy workloads such as machine learning (ML), the overhead could impact performance and lead to a loss in revenue or delays in production for manufacturers. PT found that a Dell PowerEdge R7525 server, running VMware vSphere 7 Update 3 and with an NVIDIA vGPU, delivered 97.5 percent of the image classification performance that the bare-metal PowerEdge R7525 server did, minimizing performance degradation from virtualization. According to the PT report, “Virtualizing the PowerEdge R7525 ... Read more

 Business leaders need the right information at the right time. For organizations that get their information from cloud-powered data analytics apps, faster VMs can provide insights that enable stakeholders to make decisions sooner. When Principled Technologies compared the performance of Microsoft Azure Lsv3-series VMs to that of older Lsv1-series VMs, they found compelling speed advantages. According to the report, “At each VM size, we found that the Lsv3-series VMs with 3rd Gen Intel Xeon Scalable processors analyzed data more than twice as fast. For small VMs, the Lsv3-series VMs processed the workload with up to 2.72 times the speed of ... Read more

 Moving data from a legacy environment to a new Dell PowerStore array can seem a daunting task. Administrators can choose to plan and complete the migration manually, or engage Dell Data Migration Services. Principled Technologies (PT) completed a file and block migration from a legacy Linux environment both 1) manually and 2) using Dell Data Migration Services and compared the two experiences. According to the report, “Compared to our example manual migration, using Data Migration Services reduced infrastructure downtime by 99.5 percent and helped to ensure we planned the details of data migration, including preserving important user permissions. Dell Data ... Read more

 For organizations that rely on machine learning to process images, saving time during any part of the workflow could translate to getting more done. During the preprocessing phase of the workflow, using Kubernetes environments can make it easy to deploy, scale, and manage image preparation for machine learning algorithms. In their new study, Principled Technologies (PT) found that a four-node cluster of 3rd Gen AMD EPYC 75F3 processor-based Supermicro AS-1124US-TNRP servers took 7.2 percent less time to preprocess a 1,720-image workload and achieved a 7.9 percent higher frames-per-second rate than a four-node cluster of Intel Xeon Platinum 8380 processor-based Supermicro ... Read more

 The fast pace and flexibility required for modern application development have led many companies to choose public cloud solutions for their DevOps teams. While public cloud solutions may meet some of these organizations’ demands, companies may also be able to get the self-service of a public cloud solution via a private, on-premises solution, such as those available in the Dell Technologies APEX portfolio. Principled Technologies set up a solution using Dell Technologies APEX Private Cloud and APEX Data Storage Services running a VMware vSphere with Tanzu environment to validate several management use cases. According to PT, “we successfully used a ... Read more