The market is ready for a new classification of Switching in the Data Center
Historically, the portion of the Ethernet switch market geared toward programmability was limited to high-frequency trading (HFT) and some niche Telco use cases. For HFT, Arista and Cisco used a combination of ultra-low latency ASICs and FPGAs.
That market has historically had approximately $200M in annual revenue. The programable switching market, led by Intel's Barefoot Tofino ASIC leveraged by nearly a dozen vendors, also does about $200M a year in revenue.
Network devices have lagged in innovation compared to servers, with the Smart NIC and DPU accelerators running various workloads ranging from storage, virtualization to security. It's time for the network to create a similar innovating leap into accelerated networking.
Aruba, a Hewlett Packard Enterprise Company has launched the CX 10000, which is the first product to enter the market for this class of device. It combines HPE Aruba's networking IP, such as Fabric Composer and switch expertise, and Pensando's DPU to address various new deployment models.
We expect that by 2025, the market for these accelerated class devices will exceed $10B. More than fifteen years ago, the market made a transition to purpose-built products for the data center and campus. Today's data center requirements are similarly pushing for new product classes and purpose-built products to address the need in a hybrid-cloud world compared to the legacy approaches used for the past decade-plus.
If we look at Aruba's CX 10000 in a little more detail, several use cases geared toward colocation, hybrid-cloud, and security take center stage. With each use case, we note that the enterprise could no longer scale with the number of applications, the pools of data, nor the new multi-cloud environments. Contemporary architectures are needed to support applications and to allow the human in IT to scale. At the same time, we are in the early stages of netOps and secOps merging. Today, this leads to early themes, but the next decade of networking must support this broader trend, and we view products, like Aruba's CX 10000, as looking forward to the next ten years of networking design.
Our supply chain checks across the technology and cloud sector continue to turn pessimistic during the early days of October. As a result, orders for semiconductors that were thought to be secured and guaranteed are being cut. The implications are clear that most segments we track are in for several bumpy months, with all efforts to bring 2022 back to normal.
Component suppliers are cutting or negotiating back existing commitments – Factory shutdowns, lack of fab capacity, and logistics gridlock continue to hamper the ability of system vendors to get adequate supplies. Vendors continue to burn through their stock, and no vendor can ship a 99% complete product. Vendor resources, which should be on 2022 product launches and securing capacity for future product cycles, are instead spent figuring out what to ship and adjusting components in existing product lines. In general, vendors are shipping fewer varieties of SKUs now than at any point in the past 20 years.
Prices are increasing for components and, in some cases, will never return to the previous price – While some price increases could be classified as temporary or transitionary like rush charges, air freight, and redesigns, many events are leading to more permanent price increases. At the very least, vendors will dual source suppliers, manufacturing across many geographic regions leading to lower volume with each supplier. As a result, we view some semiconductor components as never returning to pre-COVID price levels. So the big question remains when products will return to historic price erosion?
2022 lacks new product designs – In 2021, there were very few new product introductions and downsizing in SKU offerings. In some markets, we track there was minimal impact, but in other areas like 400/800G, the lack of clarity is causing many hyperscalers to reevaluate their speed transitions.
We expect a different tone in earnings season as vendors, component suppliers, and customers adjust to the new normal of not getting every type of product they want over the next several quarters. As a result, some projects will be delayed, others scaled back, and spare capacity at AWS/Azure/GCP will be put to the test as many enterprises embrace capacity to buffer shortfalls in premises-based hardware.
When Cribl announced its monster $200M financing round, we wondered what is coming next for the observability infrastructure company. The company has revealed some of what got investors excited in its fall release (Version 3.2.0) announcements. We think the two most important elements of the fall release announcements include is Cloud Visibility & Management and its Quick Connect features, both of which enhance Cribl customers’ implementation of the flagship product, LogStream.
In addition to this 3.2.0 release, we think another reason investors are so enthusiastic to back Cribl is the growing demand for observability systems. We expect that, over time, the networking, security, and observability industries will see increasing overlap between each other. This overlap will drive customers to demand solid third-party alternatives to the trend where prominent players from each of the three industries acquire smaller, best-of-breed category leaders. While there is able competition in the overall Observability market, Cribl’s newfound war chest positions it to vie for leadership in the emerging observability infrastructure market.
The company’s main fall release enhancements include:
The second important function of Cloud Visibility and Management is that it:
Nokia Unleashed Next-Gen Routing with FP5 ASIC
On September 21st, Nokia launched the company's fifth-generation IP routing silicon (FP5). Along with support for 800 Gbps routing ports, the company also announced new Fixed form factors, which are becoming popular as Routing supports new uses cases around supporting more distributed IP edge routing architectures. Along with the increase in performance came a significant power reduction in the cost per bit of bandwidth and an increased set of security features. The company also announced ANYsec to extend MACsec encryption capabilities, which are becoming increasingly important to customers and SPs as more mission-critical traffic flows on networks.
Since the launch of Nokia's previous generation ASIC (FP4), Nokia grew its market share position in SP Routing via a combination of custom Silicon innovation and merchant silicon to address more use cases. The company also expanded further into enterprise and Cloud use cases. 800 Gbps and the 112 Gbps SERDES technology will begin shipping in 2022, and we view the technology as an essential enabler to how networks need to evolve in Telco SPs and the Cloud.
We expect that by 2025 over half the revenue and nearly three-fourths of the bandwidth in Routing will come from 400 Gbps and 800 Gbps products with operators using a mix of traditional Modular systems and smaller Fixed form factors.
For many, the DPU is associated with the Smart NIC and the boundary between computing and networking. Therefore it has a finite market size tied to the number of servers and appliances in the market. DPU history focused on offloading the processor with storage, basic security, and virtualization allowed the CPU cores on a server to be 100% utilized for workloads and not for tasks viewed as overhead.
Today the market for DPUs is very different and in the early stages of redefining many markets. Today’s DPU metrics are impressive and altering the conversation. For example, two-three years ago, it was about getting to 100% utilization and defined to a handful of use cases. Today the DPU accelerator can do more, not only offloading basic tasks but accomplishing processing that would take multiple servers to achieve in one card. This is a game-changer and becomes a building block to future data centers looking and acting differently than the current paradigm.
The DPU, and general accelerators, create new markets now and are at the forefront of several technology innovation cycles. They are not only disrupting adjacent markets with new architectures but are developing new use cases and applications. DPUs in the Telcom market was an adjacent natural market, and we highlight a set of new announcements below on several new areas the DPU is taking over:
-- NVIDIA and Palo Alto Networks (May 2021) announce high-end next-generation firewall(NGFW): Palo Alto’s firewall, running on BlueField-2, opens up the entire firewall space to the DPU market. DPUs can replace custom silicon and achieve better performance. This becomes critical as the security market transitions to Cloud and as-a-Service models.
-- NVIDIA and Marvel (MWC 2021 and 2022) announce multiple edge computing, NFV, RAN initiatives: Over the past two years, Marvell and NVIDIA have made significant strides in addressing almost the entire Telecom infrastructure hardware stack. For example, today, a Telecom SP can run RAN baseband directly on the DPU based system instead of on proprietary ASICs and hardware. In many cases, the DPU can achieve better cost, power, and performance than previous hardware versions offered by Ericsson, Huawei, and Nokia.
The DPU also lays the hardware building block for Open RAN.
There are many other examples of the DPU moving beyond basic offload and compute tasks. The above examples highlight new markets for the DPU that are direct replacements of legacy products. To date, much of 2021 has highlighted Telco markets and been SP-focused. That brings us to the enterprise which has several opportunities ahead for DPUs and is more forward-looking:
-- AI/ML workloads in the enterprise: This market remains in its infancy, mostly around basic inference in a few verticals. However, all enterprises will go through a digital transformation around AI or go out of business. AI in the enterprise will be many new applications ranging from business intelligence and customer interaction to the design of new products.
-- Server architectures are changing: Server designs in the enterprise and the overall march towards disaggregated pools of computing, storage, memory, and networking resources will define the next decade. The DPU will play an essential role in this transformation, and it will occur in many parts of the new architecture.
Overall the market for DPU is poised for growth as new workloads and adjacent markets become prime use cases for the technology.
By Alan Weckel, Founder and Technology Analyst at 650 Group.
Marvell to Add Innovium to Its Portfolio to Enhance High-End DC Switching
On August 3rd, Marvell announced its planned acquisition of Innovium for $1.1B. The addition has good synergies. Innovium had several design wins in the hyperscaler segment and Marvell provides additional scale, desire to expand custom, semi-custom, and merchant networking offerings in the Cloud. While the deal is pending regulatoriy approval with a planned close date by the end of 2021, we do not see any impediments to the deal closing or being delayed.
We have been tracking merchant silicon performance for over four years and for the entire 12.8 Tbps sampling and production curve and see many positives to this deal. The current state of 12.8 Tbps in 2021 remains the tale of two sets of customers. Over 1M ports of 400 Gbps a quarter ship into three hyperscalers in each quarter already.
These customers use a mix of copper and fiber, and Innovium’s market share is impressive as a challenger to Broadcom. The rest of the market is mainly in trials, delayed by COVID-19 and multiple semiconductor shortages. As the market moves to mainstream 12.8 Tbps adoption, those three hyperscalers will move to 25.6 Tbps, a chip that Innvoium is currently sampling.
This market dynamic creates an opportunity for Marvell. Not only are there synergies, but Marvell obtains a more substantial portfolio for what lies ahead in networking (chiplets, onboard optics, and silicon photonics) and back-to-back opportunities to win ASIC share. Marvell’s recent product announcements and Inphi acquisition highlight several go-to-market opportunities and new scale as a data center supplier that Innovium can now participate in.
650 Group's early hyperscaler executive leadership interviews indicate positive feedback regarding the announced deal. While this is a consolidation amongst ASIC suppliers, we view the long-term state of Marvell with Innovium as leading to more OEM/ODM ASIC choices and increased innovation and R&D for this market. The deal paves the way for the market to be more robust when we discuss 51.2 Tbps which is not that far out. It’s important to note that the future state of data center switching involves significant innovation and increasing importance inside the data center. Ethernet switches will be critical to deploying AI/ML and other future workload technologies to move beyond plumbing to a key interconnect technology.
Turning to vendors, we see an increased interest in using ASICs as a way to differentiate product offerings, and our OEM/ODM interviews indicate a similarly positive view of this announcement. Product launches in 1H21 were slow as vendors took a conservative approach to launch new products and not ship. We view the timing as suitable for OEM/ODM as they look to early 2022 for new product launches.
We look forward to learning more when the deal closes and during Marvell’s analyst day later this year.
- Alan Weckel, Founder and Technology Analyst at 650 Group.
In the world of networking and security, intelligence and attack mitigation means little if it cannot keep pace with the speed of connection and transactions. Many questions can arise as a raft of technologies, some traditional and others new, have entered the market. There is a mix of chips that power firewalls, including FPGAs, DPUs, and ASICs.
How large is the gap between network performance and firewall performance?
The gap between network performance and firewall performance has been closing somewhat, in that we could argue that the virtual instances can go as fast as desired, leveraging, for instance, DPUs or racks and racks of standard servers.
We’ve looked into this question recently because Palo Alto Networks announced its NVIDIA-based DPUs with its VM-Series last month. In the acceleration game, DPUs are more power-efficient than FPGAs, at least in part because they use leading-edge process geometries from fabs like TSMC.
For context, network performance, as measured at an Internet Exchange Point, might be measured by 100 Gbps Ethernet or maybe at 400 Gbps, which today compares to the throughput of an ultra high-end firewall from the likes of the Palo Alto Networks PA 7080. But architectures are shifting because firewalls are being put elsewhere, like in co-location points, interexchange points, and in SaaS delivery nodes, not just at the WAN link, as was traditional.
Should security vendors invest in a ‘competitive edge” by building their own chips to enable their firewalls to run ‘at scale’ along with the fastest moving traffic coming through the cloud, data centers, or branch offices over the WAN?
Special purpose chips have their place and will always, we expect. However, leading-edge process geometries associated with high-volume, high-performance systems like GPUs may well have an edge at the very high-end compared to ASICs that are purpose-built.
The GPU performance advantage may simply exist because the ASICs ship in lower volume than a more universally accepted chip systems like a GPU or a multi-ARM processor that has many uses.
What edge does a security vendor have in developing its own hardware chips/FPGAs to accomplish wire-rate speeds that keep pace with the data firehose coming from the network?
A well-known security vendor makes its own ASICs, Fortinet. On the other hand, Marvell offers a hedge in the form of semi-custom multi-ARM processors dubbing them as “custom.” In the case of Marvell, these are considered DPUs. With DPUs, perhaps 90-100% of the semiconductor intellectual property is developed by the DPU maker, which for this example is Marvell (leveraging ARM CPU designs, as well).
And, as mentioned earlier, NVIDIA has begun using its GPU-enhanced DPU systems to accelerate Palo Alto VM-Series. If the chip is a standard chip, like a DPU, then the system vendor will differentiate by developing unique software to run on top of the chip. And, over time, we think security vendors can also differentiate based on system-level design advantages and by developing specific systems targeted at certain markets.
Extreme Networks announced its intention to acquire the Ipanema SD-WAN operations of Infovista. The company expects the deal to close in October 2021 and that in Extreme's Fiscal 2022, the revenue contribution will be "material." We think Extreme's entry to the SD-WAN market gives the company exposure to a critical part of the networking industry, the wide area network market. In the past couple years, we've seen enterprise customers are demanding a wider range of networking products that will be managed using a "single pane of glass" management system, and with the addition of SD-WAN to Extreme's portfolio that includes Ethernet Switches and Wireless LAN, Extreme's customer appeal will grow.
Ipanema is based in France and after the acquisition, almost a quarter of Extreme's workforce will be Europe based. Ipanema currently has reseller agreements in place with British Telecom and Orange Business Services and half of its revenues are from the UK and France. The other half of revenues are from other European countries. As part of the SD-WAN operations, Ipanema has a relationship with firewall leader, Check Point Software, that allows Ipanema to offer Firewall as a Service (FWaaS). Over time, Extreme says it plans to address the SASE market, which according to our research is a very high growth opportunity.
Extreme has been acquisitive in the past several years and has been focusing its efforts on offering cloud-managed services as the "glue" that binds its hardware products together. The plans for Ipanema is that it'll be integrated into Extreme's cloud-managed services system as soon as 6-9 months after its acquisition closes. Extreme also said it will continue to offer SD-WAN as a standalone offering.
Wireless Internet Service Providers (WISPs) and Managed Service Providers (MSPs) have a big opportunity coming up as wireless quickly evolves with LTE and 5G technologies.
650 Group analysts have observed the mobile market from a telecom and enterprise edge perspective for the past two decades. Today, there is more action that ever, as “Radio Access Networks” RANs have taken many forms to meet differing needs, from IoT, to regular voice connectivity.
The 5G era represents many things to many audiences. For purposes of this article, we will focus on just two ‘nuts and bolts’ aspects of 5G: (a) the enterprise opportunity and (b) fixed wireless broadband. These 5G’s are different from the 5G that most mobile operators and consumers think about.
First, from an IT perspective, Enterprise networking means Wireless LAN and Ethernet Switching. WLAN operates in unlicensed spectrum and, as a result, suffers when too many devices attempt to use it. Generally, new 4G and 5G enterprise cellular systems operate in licensed spectrum or shared spectrum like CBRS, and as a result, they don’t suffer when too many unlicensed devices like iPhones are present.
To us, this is why 5G Enterprise Cellular is most compelling. And there are many arguments in favor of using LTE instead of 5G, including device compatibility.
Naturally, there is an exciting opportunity developing for Managed Service Providers (MSPs) who will manage the complexity of operating 5G Enterprise Cellular. Some MSPs will purchase technology from vendors, some will make their own, and some could potentially leverage open source technology. We think some of these MSPs will use the business approach of Network as a Service (NaaS).
Second, earlier this year, there was a surge in activity associated with 5G customer premises (CPE) broadband access devices, many of which were supplied to a Tier 1 operator by Nokia. This 5G BB CPE surge is the start of something big, which in this case is from a mobile network operator looking to leverage its mobile 5G network to a greater degree than if only mobile users connect. We do, however, see that a large number of wireless operators won RDOF auctions, and this could drive some 3GPP 5G units, but also other types of non-3GPP fixed wireless systems as well.
Two particular cases in point are Geolinks and Etheric. Geolinks, the southern California Internet Service Provider who recently acquired a lot of Verizon’s LMDS/28 GH spectrum intended for fixed use – we would expect this will be 3GPP equipment. San Francisco Bay Area Internet Service Provider Etheric, like Geolinks, did well in the RDOF auctions, has acquired significant CBRS spectrum.
We believe some of the FWA opportunity will be served with 3GPP 5G, while some others will be met with 5 GHz and 60 GHz equipment from companies like Ubiquiti and Cambium, which won’t meet the 3GPP definition for 5G.
However, as WISPs market their services to customers, we have seen some use the 5G moniker, despite using non-3GPP-compliant equipment. We have raised the forecast for FWA in response to US auctions and the expectation that the Biden administration could push through more broadband programs.
We see a coming trend that other countries will follow the United States’ lead and look to further broaden mobile broadband coverage to the underserved. Expect much of this activity will be called “5G.” And this is a big win for rural customers who will suddenly have terrific “WFH” access technologies they can rely on.
Today, Rakuten announced that has acquired the rest of Open RAN startup, Altiostar. It already owned 67% of the pioneering vendor, and paid another $370 million to acquire the rest, at a valuation just over $1 billion. The company is adding Altiostar to its Rakuten Communications Platform (RCP) capabilities now known as Symphony, that it is selling to mobile network operators. The company also announced its first commercial customer for RCP Symphony, German startup mobile operator called 1&1. Now that RCP Symphony has expanded to include RAN, the total system is rather comprehensive (shown below).
The company shared what is included in the Symphony suite, which can be broken into five major pieces: 1) Internet & Ecosystem Services, 2) Digital Experience, 3) Intelligent Operations, 4) Network Functions, and 5) Unified Cloud. Compared to traditional telecom equipment companies like Ericsson, Nokia and Huawei, who have significant parts of their revenue exposure in the last 4 of the major pieces that Rakuten offers in Symphony. What’s different is Rakuten will be offering Membership & Loyalty Platforms, Media & Gaming Platforms, Payment & Finance Platforms, and Marketplace. Additionally, Rakuten can offer Cloud Infrastructure, similar to what Amazon Web Services, Microsoft Azure and Google Compute offer. We think Rakuten is taking on a lot by offering Symphony, but it has proven that many, if not all, parts already work – that’s because it has 4 million Japanese subscribers on its Rakuten Mobile Network already. That is a big endorsement.
The 1&1 deal includes the entire RCP full stack and is structured as a “Design-Build-Operate” contract. The plan is to start work “next quarter,” and all engineering is done in-house by Rakuten and the team will manage installation done by others. By 1Q22, Rakuten will be deploying commercially for 8 years, and its operation plans last 10 years. The mix between base stations, servers and network software is 70% radio access network and 30% on the rest. Hardware and software mix is 65% hardware, rest software and services. RCP provides an “open-book” approach on purchasing of hardware to its customer.
Rakuten claims it is in discussion with many potential customers and partners around the world about RCP (this includes Altiostar discussions), which include in the Americas, 27 client interactions; In Europe, 19 client interactions; in Middle East & Africas, 11 client interactions; Russia & CIS, 5 client interactions; and in Asia Pacific, 22 client interactions.