Networking

Image: Aussie Broadband
Aussie Broadband has reported its revenue for the first half of FY22 continued the company’s upward trajectory, rising by almost 50% year-on-year to AU$229 million. The sharp revenue increase led to Aussie Broadband seeing an after-tax profit of AU$1.39 million, with the company’s revenue once again closing the gap on its expenses. By comparison, the company’s first-half performance in the previous financial year culminated in an almost AU$10.5 million loss. Earnings before interest, tax, depreciation, and amortisation (EBITDA) for the half year was AU$9.1 million, an increase of 7% compared to the same period a year ago.On the connections front, the company said it had just shy of 495,000 broadband connections and 32,000 mobile connections at the end of 2021, representing 45% and 70% year-on-year jumps, respectively. The broadband connections comprised of 422,034 residential connections, 45,483 business connections, and 27,286 wholesale connections.The company’s migration of white label services also began during the first half of FY22, with 8,725 services being transferred during the period. The remainder of the services are expected to be migrated by the end of March, Aussie Broadband said. Aussie Broadband added its market share for NBN broadband, excluding satellite, reached 5.66% at the end of 2021. Six months prior, the company’s NBN market share had sat at 4.9%.With the growth in connections, Aussie Broadband also increased its staff count, which rose 29% year-on-year to 733.

“It’s been another year of growth for Aussie, and I am extremely proud of the work the whole team has put in to create some great half-year results,” said Phil Britt, Aussie managing director. Looking ahead to the remainder of the financial year, Britt said he expected another 85,000 to 95,000 broadband connections to be added to its customer base, and full-year EBITDA to be in the range of AU$27-30 million. Britt also provided an update on its pending acquisition of Over the Wire, saying the deal was on track to be completed next month following Australian Securities and Investments Commission and Federal Court approval.  Related Coverage More

Netgear’s Orbi Wi-Fi 6E

, aka AXE11000 or RBKE963, mesh Wi-Fi router hardware is expensive with a capital E. It currently costs a cool $1,500 for the router and its two satellites. But, for that money, you’ll get the fastest Wi-Fi networking you’ve ever seen. How fast is it? I used the Ixia’s IxChariot networking benchmark and my

Galaxy S21 Ultra smartphone

to measure its throughput and it beat every other Wi-Fi network I’ve ever tested. And, friends, I’ve tested a lot of network gear in my day.For example, for years now Wi-Fi router companies have been claiming they could break the 1 gigabit per second (Gbps) barrier. Spoiler alert: None of them could. Until now. This new model Orbi cracked the speed barrier at 20-feet with traffic just over 1 Gbps. Color me impressed.

LikeGreat speedGreat rangeEasy setup

Don’t LikeHigh costBarebones software

Of course, at further distances, its speed dropped. But even at 50 to 150 feet I still saw speeds in the 200 Megabit per second (Mbps) to 100 Mbps range. That’s still darn good. Better still, I was seeing these speeds, not across empty space but through the walls and floors of my two building home: A century-old, two-story historic house with thick Wi-Fi unfriendly walls and a brand two-story new office and studio. Together they cover approximately 4,000 square feet and I got good connections from one end of my property to the other. Good luck getting that to work with other Wi-Fi hardware.In addition, my home/office has dozens of computers, half-a-dozen tablets, three smart TVs, and a host of other networked connections. In short, my network gets a real workout. And, the Orbi kept it all running without a moment of trouble. Netgear claims theAXE11000 can cover up to 9,000 square feet and connect up to 250 devices. I see no reason to doubt their claims.

To really get the most from the Orbi AXE11000, though, you need to have the internet bandwidth to feed its need for speed. With a 1 Gbps internet feeding a 10 Gbps wired network, I’ve got that. But, if all you have is say a 300 Mbps connection to the world and older hardware, you’ll be wasting your money with the Orbi.For example, the reason I finally and consistently got a Gigabit of speed from my Wi-Fi is I was able to use the AXE11000’s support for the new short-range 6GHz 6E band.  But, if your equipment can’t support 6E, say with an

Intel Wi-Fi 6E AX210 (Gig+) adapter

, you’ll never see 6E speed. And there are few computers and smartphones which currently support it. If you have a great internet connection, you may want to pay the money to upgrade your PCs and other gear to 6E. For instance, 6E can deliver sub-five-millisecond latency. It’s a gamer’s dream come true. But, even without 6E compliant hardware, the AXE11000 still delivers the bandwidth you crave. That’s because, as Netgear aptly puts it, the world’s first quad-band Wi-Fi mesh router. Besides the commonplace 2.4 and 5GHz bands we all know and use, it comes with a second 5GHz band for backhaul transmissions, and 6E’s new 6GHz band. That backhaul wireless connection is great for delivering 4K video, for example, to otherwise distant Roku streaming devices and the like. Or, if you have a small business you’ll appreciate that it can support three separate Wi-Fi networks: A 2.4-GHz network, a 5-GHz network, and the new 6-GHz network. A backhaul network is also present and used exclusively for connecting with the satellite units. Making all this work, each separate Wi-Fi unit comes with a dozen amplified antennas for the best possible signal. This is all powered by Qualcomm’s best-of-breed Networking Pro 1610 chipset and a 2.2GHz ARM quad-core processor. Each unit also comes with a GB of RAM and 512MBs of storage for firmware, software, and settings. Combine all this with 802.11ax networking and 4K Quadrature Amplitude Modulation (QAM) and you get both great speed and the invisible automatic ability to beamform your signals to make the best possible connection with your hardware. This means you can have up to 16 independent streams of data for the best possible, interruption-less connections. For when I really need speed, I still use wired Ethernet and the AXE11000 delivers here as well. Each device includes a 2.5Gbps LAN port with three additional gigabit Ethernet jacks. In addition, the router’s WAN port supports incoming speeds of up to 10Gbps. Too fast to ever be useful? Think again, we’ll have 10Gbps Internet soon enough.  So, is the AXE11000 for everyone? Heck, no! It’s too expensive and overkill for most users. But, if you’re running a small business, are a serious gamer, or, someone like me, who really can use all the network speed and range he can get, then the AXE11000 is a great investment. Related Stories: More

Image: Cole Bennetts/Bloomberg via Getty Images
Australia’s competition watchdog chair Rod Sims has given his two cents on how the NBN should approach recovering its costs, saying that the NBN should be viewed as a sunk cost and decisions should flow from that starting point. “Now that [the NBN’s] built, I think it’s appropriate to treat its cost as sunk and therefore, what matters for Australia is getting the best use out of the NBN,” Sims told Senate Estimates yesterday.The ACCC chair was speaking to the NBN’s efforts to recoup costs, wherein the company responsible for running the network has previously said it needs to eventually have an average revenue per user of AU$51 to avoid a potential write-down. In NBN’s FY22 first-half results posted last week, the company said its average revenue per user finally shifted from AU$45 to AU$46.Debtwise, as of the end of 2021, NBN has AU$17.3 billion in private debt. It also has a AU$19.5 billion loan from the federal government, with AU$7.5 billion of that amount still outstanding.”Obviously, NBN need enough cash going forward to cover their investment, [it] would be absurd not to do that. But I wouldn’t be personally hung up on getting a commercial return on every last dollar spent because I think that’s just bad economics. What’s the best use we can make of the NBN should drive it provided they’ve got enough money to do all the things they have to do,” Sims said.In providing that view, Sims said the NBN should prioritise generating enough revenue so it can continue upgrading and investing into the network to meet future demand rather than prioritising making the utmost commercial returns.

During Senate estimates, Sims also maintained the ACCC’s view that the 25-50Mbps down and 5-20Mbps up Fixed Wireless Plus plan is sufficient for most families if their needs are working from home and using streaming services at the same time. In recent releases of statistics on fixed wireless performance by the ACCC, those on the supposed 25-50Mbps down and 5-20Mbps up Fixed Wireless Plus plan have been shown to be barely able to crack the 6Mbps mark for upload speeds, and it has been that way for some time.The Regional Telecommunications Review, published on Monday, backed ACCC’s stance that the plan was sufficient. The review noted, however, that the 6Mbps target and other speed targets needed to be significantly strengthened to meet continual demand increases and network growth.”This is insufficient for many of the activities higher-bandwidth users are looking to use the service for and inconsistent with the upload speeds available to fixed line consumers,” the review said.When it fronted Senate estimates earlier in the week, NBN said it would be formally lodging its Special Access Undertaking variation with the ACCC in the coming weeks.”I would expect the ACCC will then consult on that variation. As I understand, it is required by legislation actually, and I would expect them, therefore, to issue a consultation paper and provide a timeline for the process,” NBN CEO Stephen Rue said on Tuesday night.RELATED COVERAGE More

Japanese telecom giant KDDI said on Friday it has deployed the world’s first commercial 5G standalone open radio access network (Open RAN) in Japan.

The network, now available in the city of Kawasaki at Kanagawa Prefecture, is powered by Samsung’s virtualised central units and virtualised distributed units as well as Fujitsu’s massive MIMO radio units. Samsung’s baseband and Fujitsu’s massive MIMO radio units are connected with an open interface, the telecom giant said. The site at Kawaski also has network slicing and multi-access edge computing capabilities, which will offer higher speeds and lower latency for mobile users, the company said. KDDI said the use of virtualisation and Open RAN technologies, which use software that can operate on commercial off-the-shelf servers to replace previously hardware elements, will bring flexibility and agility to its network with deployment being more cost-effective. The launch of the network will also allow the company to accelerate deployment of Open RAN across Japan, including in rural areas, which it will continue to do with Samsung and Fujitsu throughout 2022, the teleco said. Prior to Samsung becoming the 5G network equipment supplier for KDDI, the pair had already been collaborating on related technologies since 2017. Last year, the South Korean tech giant also announced it was supplying its 5G kit to NTT Docomo, Japan’s largest telco.

Samsung, a vocal supporter of vRAN and Open RAN technologies, has also provided its vRAN solutions in the US and the UK for Open RAN rollouts there. Related Coverage More

Networking giant Cisco Systems announced its FY22 Q2 numbers on Wednesday. ZDNet summarized the results in this post, so I won’t go into detail on the numbers. At a high level, the company put up a modest beat, which is impressive given the unprecedented supply-chain constraints that are playing havoc with infrastructure vendors.

Quarterly revenue and non-GAAP EPS were $56 million and $0.03 ahead of street expectations respectively. The $12.7 billion in revenue represents 6.4% year-over-year growth, which is impressive for a $50 billion annual-run-rate company. While the financials give the industry a sense that demand for Cisco is strong, it’s worth looking behind the numbers to get a better picture of where Cisco’s business is and where it’s going.  Demand for company’s products never stronger The numbers show 6.5% year-over-year growth, but that was tempered by the macro supply chain issues. Total product order growth for the quarter was up 33% YoY, making it the third consecutive quarter this metric has topped 30%. Looking at segments, enterprise orders grew 37%, its strongest number in a dozen years. Service providers and webscale grew 42% and 70%, respectively. These are notable because Cisco has struggled in these segments historically, but that ship seems to have turned on the strength of a refreshed ASR 9K and Catalyst 8K portfolios. These include the acquisition of Acacia. Commercial business (SMB) jumped 34% and public sector 22%. This order growth has resulted in an RPO growing (remaining performance obligation) 8.5% to $30.5 billion, with 53% to be recognized in the next 12 months. Also, Cisco reported its backlog is now over $14 billion, which oddly enough includes $2 billion in software backlog, which is unusual but occurs due to the tie to hardware. CEO Chuck Robbins addressed the backlog on the earnings call and noted that supply issues have not gotten worse but also have not improved. The shift to subscription and inventory backlog has put Cisco in a comfortable, predictable position with demand not seen in more than a decade. AA A massive software company Under Robbins, Cisco has been aggressive in its transition to subscription software. This wasn’t an easy thing to do for a company whose value is mostly tied up in hardware. Total software revenue is now $3.8 billion, 80% of which comes via subscription. That number annualizes to more than $15 billion in software, making it a top-5 software company. While many Cisco products are still delivered as hardware, much of the value is now through software. This is an important pivot because it enables the company to innovate new features its customers can use, faster than with a hardware-only model.

Consider how Teslas, iPhones, and other consumer devices are hardware devices that deliver value through software updates. All of Cisco’s newer hardware products work this way. Customers buy the hardware but also purchase a software subscription. This ensures they can run the latest and greatest features without going through a costly hardware upgrade. The network is not a commodity 

Many industry watchers have been calling for the network to be commoditized for the better part of two decades. The bearish outlook on Cisco was that network features were largely becoming standardized, leaving no room for differentiation and causing the bottom to fall out of the industry. Huawei was going to do this to Cisco, the same way software-defined networking was going to do it to white boxes. These trends have come and gone, yet Cisco’s gross margins remain in the mid-60% range, where it has historically been. In technology, it’s always been my belief that no market is a commodity if the vendor can create differentiation. People pay thousands for a MacBook even though Chromebooks are just a few hundred dollars. VMware maintains healthy margins for virtualization despite a strong push from Microsoft. Cisco’s differentiation in networking has been and continues to be driven via its custom silicon. Most network vendors use merchant silicon from vendors such as Broadcom, but Cisco has largely eschewed that model. It has built its hardware, allowing it to often get a first-mover advantage with new features and performance numbers. From an enterprise perspective, the shift to hybrid work has shone a new light on the network, which rarely earned any C-level attention previously. In a recent ZK Research study, a little over 60% of business leaders stated that the network has grown in business value since the pandemic began. Most enabling technologies for digital initiatives, such as cloud, mobile, and IoT are network-centric, making the choice of network vendor a critical one for businesses. With that being said, the competitive landscape in this area is much tougher for Cisco. Arista, HPE, VMware, Juniper, and Extreme Networks are all strong companies. Also, AWS, GCP, and Azure eyeing network services and 5G could shift things to telcos, so it will be interesting to see if Cisco can continue to stay at the front of the innovation train. Security, collaboration in transition While the network business remains robust at Cisco, its security and collaboration businesses are currently in the midst of transitions. Security showed growth of 7%, which is well behind the growth of companies like Palo Alto, Fortinet, and Zscaler. However, Cisco’s security growth number is a mix of declining on-premises hardware and cloud-delivered security centered in its Duo product. Over time, the hardware business will level out and security growth should jump back up to the teens, but right now the legacy products are acting as a drag on the business.Similarly, Cisco’s collaboration products experienced a decline of 9%. I know the core Webex business is growing, but Cisco has a huge base of customers still using on-premises VoIP, Jabber, and Telepresence. Also, many of Cisco’s newer collaboration endpoints were bitten by supply chain issues, limiting availability. Again, over time I expect to see Cisco migrate customers over to Webex and, as this happens, collaboration should return to growth.

Tech Earnings More

Cisco on Wednesday published better-than-expected second quarter financial results, reporting strong product order growth as well as a major product order backlog. All told, Cisco’s Q2 non-GAAP earnings per share came to 84 cents on revenue of $12.7 billion, up 6% year over year.Wall Street was expecting second-quarter earnings of 81 cents per share on revenue of $12.65 billion.

Networking

“We continue to see incredibly strong demand across our portfolio, emphasizing the criticality and relevance of Cisco’s innovation,” CEO Chuck Robbins said in a statement. “Our robust order strength, record backlog and double-digit growth in annual recurring revenue position us well to deliver growth.” Total product order growth in Q2 was up 33% year-over-year, making it the third consecutive quarter of year-over-year product order growth of 30% or higher. Enterprise orders growth accelerated to 37%, and webscale orders grew over 70%.However, Cisco had an all-time high product backlog of $14 billion, increasing more than 150% year-over-year. Within that amount, the software backlog almost doubled to more than $2 billion. “Our incredibly strong demand continues to outpace supply, expanding our backlog of products, software and services,” CEO Chuck Robbins said in on a conference call Wednesday. “Our supply chain team continues to take aggressive action through strong inventory positions, deepening supplier relationships, qualifying alternative components and increased used of expedited freight. There are still significant constraints with semiconductors preventing us from completing manufacturing of some of our products, and that remains a headwind to revenue growth despite very strong demand.”

Product revenue was up 9%. Product revenue performance was led by growth in Secure, Agile Networks up 7%, Internet for the Future up 42%, End-to-End Security up 7%, and Optimized Application Experiences up 12%. Hybrid Work was down 9%.Reflecting Cisco’s transformation to a software and subscription-based business, total Annualized Recurring Revenue (ARR) came to $21.9 billion in the second quarter, up 11% year-over-year. Software revenue grew to $3.8 billion, and 80% of software revenue was subscription-based, up 4 percentage points year-over-year. Total subscription revenue grew to $5.5 billion, representing 44% of total revenue.For the third quarter, Cisco expects revenue growth of 3% to 5% year-over-year and an EPS between 85 cents and 87 cents. The market is expecting an EPS of 86 cents. For the full FY 2022, the company expects revenue growth of 5.5% to 6.5% and an EPS of $3.41 to $3.46.

Tech Earnings More

Soon both Google Chrome, the most popular of all web browsers, and the Firefox web browser will release their 100th version. Now, besides just being a cool number, there are technical issues that come with these anniversary releases. Some of those issues may cause your websites to fail. Yes, fail. Here’s why.

ZDNet Recommends

All web browsers come with a User-Agent (UA). This is a string that browsers send in HTTP headers, so servers can identify the browser.  JavaScript also uses it with the JavaScript navigator.userAgent. Web developers use the UA in all kinds of ways with their server-side programs. The UA’s format is: browserName/majorVersion.minorVersionAs this written typical examples of the latest release versions of browsers UAs are:Chrome: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/94.0.4606.54 Safari/537.36Firefox: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:96.0) Gecko/20100101 Firefox/96.0So, what’s the problem here? It’s an ancient one: Date format misconfigurations. The most famous example is the still not quite dead yet Y2K problem. Then, the problem was that most programs of the late 90s and earlier couldn’t deal with four-figure year dates. This time around our problem is that too many website programs can’t deal with three-figure UAs. Yes, it’s that simple.But, while it may be simple, it doesn’t mean that it’s inconsequential. You see, we’ve already had a sneak preview of this problem when we went from 1-figure UAs (1-9) to 10-figure UAs. For instance, Opera 10 wouldn’t render sites correctly back in 2009 and some sites wouldn’t render at all with Firefox 10 because their scripts read Firefox 10 as the out-of-date Firefox 1.0. We can expect all this and more as Chrome and Firefox 100 arrive.

Google and Mozilla are well aware of these coming browser UA problems. Both are working on finding and fixing the headaches.Some of these problems will escape their efforts. For example, while it’s been known for decades that using UAs to determine what web pages or services should be served to a specific browser is a bad idea, that’s never stopped all too many web developers from misusing them anyway. If your website does this, odds are good your site will end up sending an error message instead of web pages to a version 100 web browsers.You can check today if your site has such a problem using a Chrome feature flag, which forces Chrome to send a three-digit UA. Then, you can check to see if the new UA is being presented properly by visiting the test site, Is Chrome 100 Yet? Then you can use this browser to check out your own sites for problems. Firefox is also offering similar tests.With either browser, if you find something breaks because of the UA before fixing it, file a report on Webcompat. Also, be sure to check that you haven’t uncovered another kind of bug by checking to see if the problem still pops up when you’re using the normal UA.In cases things go more badly than either Chrome or Firefox’s engineering teams expect, both have mitigation plans in place. In Firefox, there’s a site intervention mechanism. With this, the Mozilla webcompat team can hot-fix broken websites. To see what’s being fixed you can type about:compat in the URL bar. And, of course, if a site breaks because it can’t handle the major version being 100, a user can fix it by sending version 99 instead. But, it’s much too much to ask for ordinary users to manually change their UAs. If things go completely haywire and there are widespread site failures, Mozilla plans to temporarily freeze Firefox’s major version at 99 and test other approaches.With Chrome, the backup plan is to use a flag to freeze the major version at 99 and report the real major version number in the minor version part of the UA string. This fall-back code is already available in Chrome’s upstream open-source Chromium browser. In this case the Chrome version UA string will use the following  pattern …. So, for example, the important part might look like 99.101.4988.0. Google’s Chrome developers will decide on whether to resort to this backup option if things go badly wrong. If you want to help make this problem a non-issue–the reason why people thought Y2K wasn’t that big a deal was because of all the efforts made beforehand to make sure it was properly fixed–both Google and Mozilla would welcome your help. And, of course, your own company would appreciate making sure its website doesn’t go up in smoke when the version 100 editions are released.You can do this by setting up your early release browser to report the version as 100 and report any issues you come across. Here’s how to do this. Configure Firefox Nightly to report the major version as 100Open Firefox Nightly’s Settings menu.Search for “Firefox 100” and then check the “Firefox 100 User-Agent String” option.Configure Chrome to report the major version as 100Go to chrome://flags/#force-major-version-to-100Set the option to `Enabled`.Before starting, keep in mind several UA string failures have already been found. If you’re a web developer using an old UA parsing library, you should test to make sure it can deal with UA versions greater than or equal to 100. Early tests show that most recent libraries will do fine. But, as we all know, the web is filled with old code. So it’s all too possible that you’re using an old, incompatible parsing library, and not even know about it until they hiccup on the latest browsers leaving your users wondering what the heck just happened.It’s time to get to work. Chrome 100 is expected to be released in March 2022 and Firefox 100 is scheduled for release on May 3. 2022. Before then, you’ll want to make sure your websites work the way you expect them to come the day,Related Stories: More

Networking technology pioneer Arista Networks this afternoon reported Q4 revenue and profit that both topped Wall Street’s expectations, and an outlook for this quarter’s revenue that was higher as well. The report sent Arista shares up 8% in late trading. Despite the upbeat report and forecast, CEO Jayshree Ullal told analysts during the company’s conference call Monday evening that disruption of the global supply chain that is affecting its sales of network switches will continue into 2023. “And so despite the supply chain obstacles that we now expect to continue into 2023, we have emerged stronger,” Ullal told analysts. Some progress with suppliers in November yielded to further supply disruption as the Omicron variant of COVID-19 emerged, she said.  Supply chain, we felt, improved in November when we met with you all at the Analyst Day, but declined in January, when we started seeing some de-commits from some of our component vendors. So, I would describe our supply chain shortages as 2 steps forward and 1 step backward. We don’t like the 1 step backward, but between the Omicron virus, the labor shortages, the logistics and the component shortages, we’re certainly experiencing another wave of uncertainty in Q1 over here. Q1 isn’t the great indicator of supply chain improving.In the company’s press release, Ullal said in prepared remarks, “I am delighted with Arista’s record 2021 milestones in innovation, diversified customer momentum and earnings. “We have executed well to establish Arista among the fastest growing networking companies in this decade.”

Said CFO Ita Brennan, “The Arista team has shown great resilience and flexibility throughout 2021, maintaining operational excellence in the face of industry-wide challenges and delivering our first billion-dollar cash flow year.”Revenue in the three months ended in December rose to $824.5 million, yielding a net profit of 82 cents a share, excluding some costs.Analysts had been modeling $790 million and 73 cents per share.The company’s gross profit margin, on a non-GAAP basis, was 64.3%, down from 65% a year earlier and 64.9% in the prior quarter. Cash from operations in the year rose to $1.015 billion, while capital investment was $64.7 million, leaving free cash flow of $950 million.For the current quarter, the company sees revenue in a range of $840 million to $860 million, above consensus of $831.6 million.Arista expects gross profit in a range of 63% to 64%, it said.

Tech Earnings More