Marvell Blog

Featuring technology ideas and solutions worth sharing

Marvell

Archive for the ‘Internet of Things’ Category

January 21st, 2020

HPE Reseller Relies on Marvell FastLinQ & QLogic for I/O Connectivity

By Graham Forrest, Server and Storage Practice Lead, Enterprise Group

gforrest@dtpgroup.co.uk
DTP Group, Leeds UK

Being a trusted advisor to our customers means making sure what we recommend, configure and install works as expected. That’s why here at DTP, we recommend Marvell® QLogic® Fibre Channel and Marvell FastLinQ® Ethernet I/O for all our server and storage connectivity solutions.

The DTP Group has over 30 years of experience in delivering technology and solutions to our customers. We only recommend HPE for server, storage and networking solutions because we know the technology, get great support from the HPE organization and trust the HPE brand. Within the HPE portfolio, there are technology choices that need to be considered, especially when it comes to I/O connectivity. HPE makes a variety Ethernet and Fibre Channel connectivity options available that they source from several different manufacturers. After evaluating all of them, the DTP PreSales and Technical teams have chosen to standardize on 10/25/50GbE based on Marvell FastLinQ technology and on 16GFC and 32GFC based on Marvell QLogic technology.

We find that Marvell’s advanced features help us future-proof our customers to ensure the products they are buying today have the connectivity features needed for deploying next-generation NVMe-enabled storage, hyper-converged infrastructure (HCI) and software-defined storage solutions. Features like dual-port isolation in QLogic Fibre Channel HBAs from Marvell ensure predictable and reliable SAN performance. Marvell FastLinQ Ethernet technology for HPE includes features like storage offload for iSCSI and universal RDMA, with support for both RoCE and iWARP protocols, ensures maximum flexibility in the Nimble storage and HCI solutions that we design and deliver to our customers.

The choice of I/O technology goes well beyond the feature set and the speeds and feeds on the data sheets or marketing literature.  We find that not only does Marvell technology “just work” in all the mainstream enterprise operating system environments and applications, but should we have a problem (which isn’t very often), the Marvell team supporting HPE partners like us are readily available to help. The HPE team at Marvell engages with us in the field and at all the major technology events. Having solid relationships and knowing that when you email or call with a question or issue, you will get a quick response, this is very important in the world of solution providers. Marvell clearly understands this and their HPE-focused team makes themselves readily available to help us and even helps us get things escalated within HPE when required.

In addition, Marvell goes above and beyond what other technology providers make available in the way of education and training. Their Follow the Wire program is a best-in-class platform for providing up-to-date information, training and educational resources. Marvell’s program provides incentives for our team to stay current, access the latest collateral, view on-demand videos, and attend live webinars and webinar replays. This keeps information about new and emerging features readily available at your fingertips.

Bottom line is, at The DTP Group, we rely on more than what you can find on the data sheet or in a line card when we make our decisions on what technology to recommend and who to partner with. This enables us to be more prepared, understand the technology better and ultimately deliver better solutions, service and support for our customers. That’s why we continue to work closely with Marvell. If you are having challenges getting questions answered, getting trained and educated or getting support in general from your I/O technology partners, you might consider Marvell as well.

 

December 12th, 2019

Marvell and HPE Introduce Industry Standard Adapters for HPE ProLiant and Apollo Gen10 Plus Servers

By Todd Owens, Technical Marketing Manager, Marvell

Innovation can come in many forms. Sometimes it’s with a completely new technology, sometimes with updating an existing product and in other cases, just by changing the approach to how you acquire and deliver a product. It is the latter that is the latest innovation from Hewlett Packard Enterprise (HPE) when it comes to I/O connectivity.

In conjunction with the launch of HPE ProLiant and Apollo Gen10 Plus servers, the HPE Server I/O Options team developed a new approach for sourcing and qualifying Ethernet adapters for these servers. Deploying what they call Industry Standard Adapters, HPE can now better meet the needs of their end customers with an increased number of options when it comes to firmware and driver updates for their Ethernet adapters in HPE servers.

Traditionally, HPE would source I/O technology from OEM suppliers such as Marvell, create customer model numbers and specifications for adapters and make firmware and drives only available from HPE. These “custom” adapters were often referred to as HPE-optimized. Starting with Gen10 Plus servers, HPE is eliminating the customization and using standard adapters that can work not only in HPE servers, but in others as well. Hence the term “Industry Standard Adapters.”

Marvell is glad to be a strategic partner of HPE, providing a wide variety of Marvell® FastLinQ® adapters that are fully qualified and supported by HPE on the HPE ProLiant Gen10 Plus servers. Below are the current offerings from Marvell for HPE Gen10 Plus servers.

HPE Part Number Model Name Product Description
P08437-B21 QL41132HLRJ HPE Ethernet 10Gb 2-port BASE-T QL41132HLRJ Adapter
P10103-B21 QL41132HQRJ HPE Ethernet 10Gb 2-port BASE-T QL41132HQRJ OCP3 Adapter
P21933-B21 QL41132HLCU HPE Ethernet 10Gb 2-port SFP+ QL41132HLCU Adapter
P08452-B21 QL41132HQCU HPE Ethernet 10Gb 2-port SFP+ QL41132HQCU OCP3 Adapter
P10094-B21 QL41134HLCU HPE Ethernet 10GbE 4-port SFP+ QL41134HLCU Adapter
P22702-B21 QL41232HLCU HPE Ethernet 10/25Gb 2-port SFP28 QL41232HLCU Adapter
P10118-B21 QL41232HQCU HPE Ethernet 10/25Gb 2-port SFP28 QL41232HQCU OCP3 Adapter

 

With the new approach, HPE Gen10 Plus customers can see the Marvell model numbers in the HPE product description and identify the Marvell vendor and product IDs at server boot. Then they will be directed to Marvell for detailed specifications, user guides, technical briefs and even firmware and/or driver downloads.  The support model will not change and HPE will continue to provide level 1 – 3 support. The new approach will benefit HPE customers in a number of ways:

  • Transparency of the specific type of product and model information – eliminates having to use complex decoder aides or other documentation to determine real manufacturer.
  • Consistency in product/model naming on Operating System HCLs that saves time cross-checking I/O and operating system compatibility.
  • Faster time-to-market for firmware/driver enhancements and updates to get fixes when they are available from the manufacturer.
  • Ability to streamline maintenance process across multi-vendor server environments and get firmware/drivers direct from manufacturer websites.

For those HPE ProLiant customers who prefer utilizing HPE-specific deployment software and utilities like HPE System Insight Manager (SIM), System Update Manager (SUM) or Service Pack for ProLiant (SPP), HPE will also make firmware and drivers available through their normal quarterly processes.

The Marvell portfolio for HPE ProLiant Gen10 Plus includes PCIe and OCP 3.0 form factor adapters in 1/10GBASE-T, 10Gb SFP+ and 10/25GbE SFP28 variants. All these adapters support Marvell’s industry-leading list of features and capabilities, including:

  • Universal RDMA: allows customers to run RDMA over Converged Ethernet (RoCE) or iWARP RDMA concurrently.
  • SmartAN™ technology: automatically sets bandwidth and forward error correction settings of the adapter to match the switch it is connected to and the cable type being used.
  • Redfish PLDM: provides integrated remote monitoring and management of thermal environment and other adapter data.
  • Marvell FastLinQ QCS Management Utility: available as GUI, CLI, VMware plug-in or PowerShell Kit for maximum flexibility.
  • DPDK Offload: up to 68Mpps bi-directional small packet acceleration.
  • SR-IOV offload for virtual server environments.
  • Tunnel Offloads: VXLAN, NVGRE, GENEVE.
  • NVMe over fabric: for high performance SDS/HCI or shared storage connectivity.
  • IEEE 1588 Energy Efficient Ethernet
  • PCIe 3.0 x8 and OCP 3.0 form factors

For more details on Marvell’s FastLinQ adapters, download the family product brief here.

For more information on Universal RDMA, SmartAN technology and other unique Marvell FastLinQ capabilities, visit our Follow the Wire Video library here https://connect.marvell.com/hpe-videos.

November 5th, 2019

Marvell Completes Acquisition of Avera Semi

By Stacey Keegan, Senior Director, Corporate Communications, Marvell

Today marks the close of the acquisition of Avera Semi.

Avera brings over two decades of expertise developing custom ASIC solutions for the infrastructure market, further enabling Marvell to offer a full suite of leading semiconductor solutions. With this acquisition, Marvell will provide the complete spectrum of product architectures spanning standard, semi-custom to full ASIC solutions. We are proud to offer world class custom ASIC design services to our OEM partners.

To learn more, read our latest press release: https://www.marvell.com/company/news/pressDetail.do?releaseID=11497.

September 19th, 2019

Marvell Completes Acquisition of Aquantia

By Stacey Keegan, Senior Director, Corporate Communications, Marvell

Marvell today announced that it has successfully completed its acquisition of Aquantia.

Aquantia pioneered Multi-Gig technology – now the basis for high speed networking in a broad range of applications from enterprise campuses to autonomous cars.  Their portfolio complements Marvell’s industry-leading PHYs, switches and processors, creating an unparalleled networking platform and enabling customers to develop systems that span megabits to terabits per second. To learn more, read our latest press release https://www.marvell.com/company/news/pressDetail.do?releaseID=11257.

June 25th, 2019

Marvell Powers HPE Servers with Next-Generation Ethernet Technology

By Todd Owens, Technical Marketing Manager, Marvell

Data is the new currency for many businesses today.  The ability to access, analyze and act on data has become a competitive advantage for many companies.  While much attention is paid to the storage devices and compute required to optimize data processing, the I/O infrastructure is often overlooked.  The reality is that I/O technologies are just as important as the core count of the CPU or storage capacity and latency of the storage array.  The time is now to future-proof your network and that’s where Marvell is here to help.

Marvell has been a long-time Hewlett Packard Enterprise (HPE) supplier of I/O technology used in the HPE ProLiant, Apollo, HPE Synergy, and HPE Storage offerings.  Over the past year, a new generation of Ethernet I/O has begun making its way into these HPE platforms.  Based on Marvell® FastLinQ® QL41000 and QL45000 Ethernet technology, these new adapters are allowing HPE customers to “future-proof their network” connectivity for what’s to come in the near future of data centers.

The QL41000 and QL45000 adapter technology provides several new capabilities not found in other I/O offerings for HPE.  Advancements include:

  • Universal RDMA – provides customers with flexibility through concurrent support for both RDMA over Converged Ethernet (RoCE) and iWARP RDMA protocols on the same adapter.
  • SmartAN™ Technology – enables auto-negotiation for transitioning from 10GbE to 25GbE connections. SmartAN automatically configures the adapter for bandwidth and error code correction based on the cabling configuration and switch port settings the adapter is connected to.
  • Storage Offload – Converge Network Adapter offerings include full hardware offload for iSCSI and FCoE protocols which greatly reduces the CPU resources required for transmitting storage traffic compared to the use of software initiators.

 

These are in addition to enhanced DPDK performance (up to 36Mpps bi-directional) and support for SR-IOV, TCP/IP stateless offloads, IEEE 1588 time stamping and more.

The FastLinQ 41000 series technology can be found in next-generation Flexible LOM Rack (FLR) and standup PCIe adapters for HPE ProLiant and Apollo servers.
Models include:

  • HPE Ethernet 10Gb 521-T Adapter
  • HPE Ethernet 10Gb 524SFP+ Adapters
  • HPE Ethernet 10/25Gb 621SFP28 Adapter
  • HPE Ethernet 10/25Gb 622FLR CNA
  • HPE StoreFabric 10Gb CN1200-T CNA
  • HPE StoreFabric 10/25Gb CN1300R CNA

These adapters allow HPE Server customers to future-proof their Rack and Tower servers with RDMA for use in Hyper-Converged Infrastructure (HCI) and Software Defined Storage (SDS) solutions; and make the transition from 10GbE to 25GbE connectivity seamless at the server.  These I/O devices are ideal for customers considering Microsoft Azure Stack HCI or VMware vSAN environments, or the deployment of any latency sensitive application.

The FastLinQ 45000 series technology can be found in next-generation mezzanine adapters for HPE Synergy, including:

  • HPE Synergy 10/20/25Gb 4820C Adapter
  • HPE Synergy 25/50GbE 6810C Adapter

With Universal RDMA support improved DPDK performance and high-bandwidth capability, these adapters are ideal for customers with VMware ESXi or Microsoft Hyper-V deployments, and for Telco or high-frequency trading applications.

Since many applications today will start to require more I/O performance and low latency RDMA, HPE’s next-gen Ethernet adapters will go a long way in future proofing networking connectivity for server customers.

For a complete list of Marvell FastLinQ Ethernet adapters for HPE Servers and the features they support, download our HPE FastLinQ Ethernet Quick Reference guide.  If you would like to discuss I/O technology or customer needs in more detail, contact our HPE team .  You can also visit the Marvell HPE microsite at www.marvell.com/hpe.

June 20th, 2019

Marvell’s ThunderX2 Server Ecosystem Expands with NVIDIA GPU and Software Support

By Larry Wikelius, Vice President, Ecosystem and Partner Enabling, Marvell

ISC High Performance, which just wrapped up today in Frankfurt, Germany, is one of the most significant server events of the year and is often a catalyst for key major industry announcements.  This year’s event was no exception with NVIDIA announcing its support for servers based on the Arm architecture.  With this move, NVIDIA will make its full stack of AI and high-performance computing software available to the Arm ecosystem by the end of 2019.  The stack includes all NVIDIA CUDA-X AI and HPC libraries, GPU-accelerated AI frameworks and software development tools such as PGI compilers with OpenACC support and profilers. NVIDIA’s full software suite support will enable the acceleration of more than 600 HPC applications and AI frameworks on Marvell® ThunderX2® systems.

NVIDIA’s support for Arm CPUs marks continued growth of the Arm-based server ecosystem.  Marvell has been a leading driver in the establishment of a standard, complete and competitive ecosystem around the Arm architecture ranging from low level firmware through system software to commercial ISV applications.  The Marvell ThunderX2 processor is the most widely deployed Arm server in the market today and the only Arm server on the prestigious top 500 super computer list with the Astra system at Sandia National Laboratories.

NVIDIA’s announcement underscores the growing momentum of Marvell ThunderX2 in both high-performance computing and cloud deployments.  The entire industry is very excited about the ability to combine the computational performance and memory bandwidth of ThunderX2 with the parallel processing capabilities of the GPU.  NVIDIA’s commitment to the complete software stack is particularly important and is yet another high value solution option in the broadly supported software offering on ThunderX2.  Most ThunderX2 systems have been designed with GPU support in mind from the beginning which enables a simple upgrade for today’s installed base.

Marvell welcomes NVIDIA to the ThunderX2 ecosystem and we look forward to working with customers on this exciting server solution.  See the press release here.

Read more about what the industry is saying about the announcement, and what this means to high-performance computing:

Forbes – NVIDIA Gives Arm a Boost In AI And HPC

The Next Platform – Nvidia Makes Arm A Peer To X86 And Power For GPU Acceleration

HPCwire – Nvidia Embraces Arm, Declares Intent to Accelerate All CPU Architectures

April 29th, 2019

RoCE or iWARP for Low Latency?

By Todd Owens, Technical Marketing Manager, Marvell

Today, Remote Direct Memory Access (RDMA) is primarily being utilized within high performance computing or cloud environments to reduce latency across the network.  Enterprise customers will soon require low latency networking that RDMA offers so that they can address a variety of different applications, such as Oracle and SAP, and also implement software-defined storage using Windows Storage Spaces Direct (S2D) or VMware vSAN.  There are three protocols that can be used in RDMA deployment: RDMA over InfiniBand, RDMA over Converged Ethernet (RoCE), and RDMA over iWARP.  Given that there are several possible routes to go down, how do you ensure you pick the right protocol for your specific tasks?

In the enterprise sector, Ethernet is by far the most popular transport technology.  Consequently, we can ignore the InfiniBand option, as it would require a forklift upgrade to the I/O existing infrastructure – thus making it way too costly for the vast majority of enterprise data centers.  So, that just leaves RoCE and iWARP.  Both can provide low latency connectivity over Ethernet networks.  But which is right for you?

Let’s start by looking at the fundamental differences between these two protocols.  RoCE is the most popular of the two and is already being used by many cloud hyper-scale customers worldwide.  RDMA enabled adapters running RoCE are available from a variety of vendors including Marvell.

RoCE provides latency at the adapter in the 1-5us range but requires a lossless Ethernet network to achieve low latency operation.  This means that the Ethernet switches integrated into the network must support data center bridging and priority flow control mechanisms so that lossless traffic is maintained.  It is likely they will therefore have to be reconfigured to use RoCE.  The challenge with the lossless or converged Ethernet environment is that configuration is a complex process and scalability can be very limited in a modern enterprise context.

Now it is not impossible to use RoCE at scale but to do so requires the implementation of additional traffic congestion control mechanisms, like Data Center Quantized Congestion Notification (DCQCN), which in turn calls for large, highly-experienced teams of network engineers and administrators.  Though this is something that hyper-scale customers have access to, not all enterprise customers can say the same.  Their human resources and financial budgets can be more limited.

Going back through the history of converged Ethernet environments, one must look no further than Fibre Channel over Converged Ethernet (FCoE) to see the size of the challenge involved.  Five years ago, many analysts and industry experts claimed FCoE would replace Fibre Channel in the data center.  That simply didn’t happen because of the complexity associated with using converged Ethernet networks at scale.  FCoE still survives, but only in closed environments like HPE BladeSystem or HPE Synergy servers, where the network properties and scale are carefully controlled.  These are single-hop environments with only a few connections in each system.

Finally, we come to iWARP.  This came on the scene after RoCE and has the advantage of running on today’s standard TCP/IP networks.  It provides latency at the adapter in the range of 10-15us.  This is higher than what one can achieve by implementing RoCE but is still orders of magnitude below that of standard Ethernet adapters.

They say, if all you have is a hammer, then everything looks like a nail.  That’s the same when it comes to vendors touting their RDMA-enabled adapters.  Most vendors only support one protocol, so of course that is the protocol they will recommend.  Here at Marvell, we are unique in that with our Universal RDMA technology, a customer can use both RoCE and iWARP on the same adapter.  This gives us more credibility when making recommendations and means that we are effectively protocol agnostic.  It is really important from a customer standpoint, as it means that we look at what is the best fit for their application criteria.

So which RDMA protocol do you use when?  Well, when latency is the number one criteria and scalability is not a concern, the choice should be RoCE.  You will see RoCE implemented as the back-end network in modern disk arrays, between the controller node and NVMe drives.  You will also find RoCE deployed within a rack or where there is only one or two top-of-rack switches and subnets to contend with.  Conversely, when latency is a key requirement, but ease-of-use and scalability are also high priorities, iWARP is the best candidate.  It runs on the existing network infrastructure and can easily scale between racks and even long distances across data centers.   A great use case for iWARP is as the network connectivity option for Microsoft Storage Spaces Direct implementations.

The good news for enterprise customers is that several Marvell® FastLinQ® Ethernet Adapters from HPE support Universal RDMA, so they can take advantage of low latency RDMA in the way that best suits them.  Here’s a list of HPE Ethernet adapters that currently support both RoCE and iWARP RDMA.

With RDMA-enabled adapters for HPE ProLiant, Apollo, HPE Synergy and HPE Cloudline servers, Marvell has a strong portfolio of 10Gb or 25GbE connectivity solutions for data centers.  In addition to supporting low latency RDMA, these adapters are also NVMe-ready.  This means they can accommodate NVMe over Ethernet fabrics running RoCE or iWARP, as well as supporting NVMe over TCP (with no RDMA).  They are a great choice for future-proofing the data center today for the workloads of tomorrow.

For more information on these and other Marvell I/O technologies for HPE, go to www.marvell.com/hpe.

If you’d like to talk with one of our I/O experts in the field, you’ll find contact info here.