At Intel, we are hearing from customers that digital transformation is not only accelerating, but organizations are shifting from tactical modernization efforts to more strategic transformation. Infrastructure is a key asset in this strategic transformation that can enable new insights from data, drive improved business outcomes, and enhance customer experience.
But apps, together with the growth of data, are placing more demands on infrastructure. Infrastructure should be based on a consistent technology foundation that enables apps and workloads to be optimally placed and managed, ensuring workload portability between infrastructure on premises, at the edge, and the instance offerings of various cloud providers - avoiding lock-in to single cloud provider and providing the flexibility needed to meet business objectives, as well as privacy and security requirements.
HCI is often the best approach to modernize on premises and edge infrastructure to provide a foundation for your hybrid cloud. As a leader in HCI, Nutanix has been consistently advancing not only their core HCI capabilities, but also their partnerships with leading public cloud providers like AWS through the Nutanix Clusters cloud platform.
A Flexible Infrastructure Foundation
We see organizations’ increasing and changing workload demands as they transform to hybrid cloud, and we hear them asking Intel for more scalable and flexible infrastructure across compute, network and storage. In response to that, we designed the latest additions to our data center portfolio, our 3rd Gen Intel Xeon Scalable platform, with these demands in mind. At the same time, we have been investing in enabling customers to deploy solutions quickly and at scale. More on those solutions in a minute, but first let me recap the new platform capabilities.
In the processor, we significantly increased I/O bandwidth, the number of memory channels and memory capacity, and integrated security capabilities, crypto acceleration, and AI acceleration into the silicon itself - all to deliver more advanced platform capabilities that help our customers solve their business challenges.
It’s important to look at the total system capabilities to evaluate whether it will meet your workload requirements. Memory requirements of modern apps and workloads such as analytics continue to increase. Intel Optane persistent memory 200 series, our next-generation persistent memory modules, deliver up to 6TB of affordable total memory per CPU. Many workloads can benefit from a larger amount of memory and that’s now achievable at a more affordable price point vs. DRAM.
From an I/O perspective, we increased I/O bandwidth, as I mentioned earlier, and integrated PCIe Gen4. When paired with our latest Intel Optane P5800X SSDs, we’re delivering unprecedented endurance of 100 drive writes per day, up to 26x more IOPS/GB and up to 13x lower latency1. This is ideal for the soon-to-be-available cache tiering solution in an upcoming release of Nutanix’s software. You’ll hear more about that cache tiering from Ricky Trigalo, Director, Product Management at Nutanix in the Next Level Hyperconverged Performance Innovations session (BS3) at .NEXT.
Customizable Performance and TCO for Nutanix HCI
The entire platform is designed to deliver flexible performance and capabilities to customers. A broad range of SKUs on the 3rd Gen Intel Xeon Scalable processors allows you to customize the level of performance - for example, with HCI, achieving higher virtual machine density with fewer stranded resources, or supporting more VDI users, or consolidating the number of nodes required to support certain workloads on your clusters. And through the built-in acceleration capabilities in the CPU, as well as the complementary platform technologies like Intel Optane persistent memory, Intel Optane SSDs, and Intel Ethernet 800 Series network adapters, we help customers optimize their total cost of ownership.
We have been partnering closely through our joint innovation lab to optimize Nutanix’s software for 3rd Gen Intel Xeon Scalable processors, Intel Optane technologies, and our Intel Ethernet 800 Series network adapters. These technologies can be mixed and matched to provide the customizable performance your infrastructure requires to support intelligent workload placement and digital transformation on your terms.
Reducing Complexity through Solutions
Customers have a wealth of choices, but that also increases complexity. Customers need technology that’s easy to deploy, pre-validated, battle tested, so they can deploy and execute with confidence. One of the best investments we’ve made over the last several years, one of the ways we’re adding value for our customers is Intel® Select Solutions. Four years ago, we launched our Intel Select Solutions for data center use cases. In that time, we’ve dramatically built a portfolio of solutions that span from the edge to the cloud. Select Solutions go through an additional level of rigor to ensure the recommended configurations deliver not only strong performance, but also price/performance.
We’re currently updating our Intel Select Solution for Nutanix to the latest generation technologies discussed above, so you’ll start to see those systems in the market from multiple server partners. This latest version extends the core HCI capabilities by demonstrating multi-cloud connections and workload placement via Nutanix Clusters to AWS, and support for VDI users through Nutanix Xi Frame to AWS, Azure and Google Cloud. Net-net, these solution offerings accelerate deployments. They simplify the process of choosing and deploying new technology from the edge to the data center to the cloud.
Check out my session at .NEXT, 3 Best Practices for Building an Agile Hybrid and Multi-cloud Infrastructure. I’ll cap off that course with more detail on the latest Select Solutions configurations.
This post was authored by Christine McMonigal, Director of Hyperconverged Marketing, Cloud & Enterprise Marketing at Intel Corporation
Notices & Disclaimers
Performance varies by use, configuration and other factors. Learn more at www.Intel.com/PerformanceIndex.
Performance results are based on testing as of dates shown in configurations and may not reflect all publicly available updates. See backup for configuration details. No product or component can be absolutely secure.
Your costs and results may vary.
Intel technologies may require enabled hardware, software or service activation.
© Intel Corporation. Intel, the Intel logo, and other Intel marks are trademarks of Intel Corporation or its subsidiaries. Other names and brands may be claimed as the property of others.
- P5800X gains vs Intel® SSD D7-P5600 NAND (both Gen4 PCIe). Date tested – March 18, 2021. Workload – FIO rev 3.5, based on random 512B transfer size with total queue depth of 64 (QD=8, workers/jobs=8) workload, 4KB transfer size with total queue depth of 32 (QD=4, workers/jobs=8) workload, 8KB transfer size with total queue depth of 16 (QD=4, workers/jobs=4) workload in most case, except where specified.
LATENCY System configuration: Intel Optane SSD P5800X: 1.6TB: CPU: Intel® Xeon® Platinum 8380 2.30GHz 270W 40 cores per socket, CPU Sockets: 2, BIOS: SE5C6200.86B.3021.D40.2103160200, UCODE: 0X8D05A260, RAM: 32GB @3200 MT/s DDR4, DIMM Slots Populated: 16 slots, PCIe Attach: CPU (not PCH lane attach), OS: Ubuntu 20.04.2 LTS, Kernel: 5.4.0-67-generic, FIO version: 3.16; NVMe Driver: Inbox, C-states: Disabled, Hyper Threading: Disabled, CPU Governor (through OS): Performance Mode. Intel Turbo Mode, and P-states = Disabled; IRQ Balancing Services (OS) = Off; SMP Affinity, set in the OS; FIO with ioengine=io_uring. Test by Intel on 3/18/2021 vs Intel® SSD D7-P5600: see https://www.intel.com/content/www/us/en/products/docs/memory-storage/solid-state-drives/data-center-ssds/d7-p5600-p5500-series-brief.html