<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Kay Singh on Medium]]></title>
        <description><![CDATA[Stories by Kay Singh on Medium]]></description>
        <link>https://medium.com/@singhkays?source=rss-122b8af0de61------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*no_geO0s15xJR-x-P41kcA.png</url>
            <title>Stories by Kay Singh on Medium</title>
            <link>https://medium.com/@singhkays?source=rss-122b8af0de61------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 11:06:18 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@singhkays/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Upcoming ARM Chip That’s Faster Than Apple Silicon M1]]></title>
            <link>https://singhkays.medium.com/upcoming-arm-chip-thats-faster-than-apple-silicon-m1-bd7933f12dba?source=rss-122b8af0de61------2</link>
            <guid isPermaLink="false">https://medium.com/p/bd7933f12dba</guid>
            <category><![CDATA[intel]]></category>
            <category><![CDATA[apple-silicon]]></category>
            <category><![CDATA[nuvia]]></category>
            <category><![CDATA[arm]]></category>
            <category><![CDATA[amd]]></category>
            <dc:creator><![CDATA[Kay Singh]]></dc:creator>
            <pubDate>Wed, 16 Dec 2020 06:03:34 GMT</pubDate>
            <atom:updated>2020-12-16T06:05:56.587Z</atom:updated>
            <content:encoded><![CDATA[<blockquote>This post was first published on my blog @ <a href="https://singhkays.com/blog/arm-cpu-faster-than-apple-m1/">https://singhkays.com/blog/arm-cpu-faster-than-apple-m1/</a></blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*fdQcEk6EG1XyY0zq.jpg" /></figure><h4>If you’d like to subscribe for more updates use the <a href="https://www.singhkays.com/index.xml">RSS feed</a> or follow me on Twitter <a href="https://twitter.com/singhkays">@singhkays</a></h4><p>The release of Apple M1 Silicon has laid to rest many questions about whether ARM CPUs can go toe-to-toe with the best of the bunch from Intel and AMD. The question now is not whether ARM chips can perform better than Intel/AMD but whether others can build similarly powerful ARM CPUs. Today, companies like Marvell, Ampere and AWS are shipping ARM CPUs that keep getting better with each generation and close to the level of performance of top-end X86 CPUs. It only feels like a matter of time before ARM CPUs will provide tough competition to the X86 CPUs from Intel and AMD. This feels a bit like standing on the Florida coast and watching the first bands of a hurricane arrive. This reminds me of the iconic scene in the movie “The Dark Knight Rises” when Selina Kyle (<em>Catwoman</em>) tells Bruce Wayne (<em>Batman</em>) that there’s a storm (<em>Bane</em>) coming. The full dialogue goes something like:</p><blockquote><em>Selina Kyle (Catwoman): There’s a storm coming, Mr. Wayne. You and your friends better batten down the hatches, because when it hits, you’re all gonna wonder how you ever thought you could live so large and leave so little for the rest of us.</em></blockquote><blockquote>— The Dark Knight Rises (2012)</blockquote><p>If you read between the lines, there are many parallels that you can draw from the characterization in the movie to real life.</p><ol><li>Batman and Catwoman are frenemies, which is not much different from Intel and AMD’s relationship. They are often competitors, just like Batman &amp; Catwoman. Still, they managed to work together once and ship this baby — <a href="https://ark.intel.com/content/www/us/en/ark/products/130409/intel-core-i7-8809g-processor-with-radeon-rx-vega-m-gh-graphics-8m-cache-up-to-4-20-ghz.html">Intel® Core i7–8809G</a> — a rare Intel CPU with AMD integrated graphics.</li><li>Intel has dominated a large part of the CPU market’s profits for a very long time. In contrast, everyone else has had to contend with using lower cost as a vector to compete against Intel’s hegemony. For e.g. <a href="https://venturebeat.com/2020/02/05/amd-gained-share-against-intel-in-x86-processor-market-in-q4/">Intel dominates about 95.5% of the server chip market</a>.</li><li><strong>(<em>SPOILER ALERT</em>)</strong> Selina ends up joining up with Bane by giving up Batman. This might be very similar to the rumors that AMD is planning its own ARM CPUs in real life.</li></ol><h3>JavaScript is not available.</h3><p>AMD has an M1 competitor in prototype stages, one version with integrated RAM, and one without ithe said &quot;almost ready&quot;but -imo- idkleak is only a few days old, the chip idk</p><h3>The King</h3><p>Apple Silicon M1 has taken the CPU market by storm and, in doing so, answered every doubt whether it can compete with the best of the best. The M1 is currently sitting at the top of Geekbench’s <a href="https://browser.geekbench.com/mac-benchmarks">single-core throne</a>, all while using less power than competing CPUs. The M1 has completely reset the Performance per Watt expectations. It’s even more astonishing to think that as Apple has introduced the M1 for only the lowest-end Macs, this is the slowest SoC that Apple will ever make (<em>for Macs</em>). Based on Apple’s history of consistently delivering year-over-year performance improvements, the follow-ups to the M1 will likely improve upon the current single-core performance very rapidly. It is not clear whether Intel and AMD have something in the pipeline that could shift the balance back in X86’s favor.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/650/0*dJvbwhuRo04pJX3G.png" /></figure><p><em>From: </em><a href="https://www.anandtech.com/show/16252/mac-mini-apple-m1-tested/2"><em>https://www.anandtech.com/show/16252/mac-mini-apple-m1-tested/2</em></a></p><p>However, this could change soon.</p><h3>A Challenger Emerges</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/0*NoF9jmxvRaB-0J3q.png" /></figure><p><strong>Nuvia</strong> is a startup founded in early 2019 that is building an ARM CPU for the server market. The company was founded by John Bruno, Manu Gulati, and Gerard Williams III. They bring a breadth of experience in system engineering and silicon design for more than 20 chips and more than 100 patents. Nuvia’s leadership team holds an impressive resume with various architect and leadership roles at Google, Apple, ARM, Broadcom, and AMD (<a href="https://nuviainc.com/leadership"><em>See here for the leadership team’s full details</em></a>). It is quite possible that Nuvia’s leadership was involved with the previous Apple Silicon designs and bring with them the ideas and strategies that have made Apple’s chips a market-leader.</p><p>Here is <a href="https://medium.com/silicon-reimagined/performance-delivered-a-new-way-8f0f5ed283d5">more on Nuvia’s design goals and philosophy</a> (<strong><em>emphasis</em></strong><em> on key talking points mine</em>):</p><blockquote><em>Our focus at NUVIA is to develop an SoC that will deliver industry-leading performance with the highest levels of efficiency, at the same time. To do this, </em><strong><em>we are creating a server CPU that is built in a new way, with a complete overhaul of the CPU pipeline</em></strong><em>. Our first-generation CPU, code-named “</em><strong><em>Phoenix</em></strong><em>” will be a custom core based on the ARM architecture and central to our “</em><strong><em>Orion</em></strong><em>” SoC.</em></blockquote><blockquote><em>With X86 solutions claiming most of the market, only a small percentage of niche customers are willing to accept a lower per-core performance, high core count product. Arguably the most successful ARM-based design today is Amazon’s Graviton. Graviton is a captive design, aimed solely for a limited portion of AWS that values cost over performance. While there will likely be additional growth in this area, </em><strong><em>the heart of the market clearly demands the highest single-thread performance at TDP and the highest all-core performance at TDP</em></strong><em>. This is the fastest way to improve Performance/TCO for the most demanding hyperscale customers.</em></blockquote><blockquote><em>While these new entrant’s products show improvements, they still fall short of disrupting their X86 incumbents. At NUVIA, we are taking a different approach, with a clean-sheet CPU design that will deliver an elegant balance of performance leadership and power efficiency that maximizes memory bandwidth and core utilization. </em><strong><em>Our solution does not need to add extraneous cores to try and make up for a single-threaded performance deficit</em></strong><em>. Also, there would be </em><strong><em>no need to employ marketing-inflated boost clocks</em></strong><em> that are not achievable in any real-world applications of server SoCs, due to running into TDP constraints. In real-world scenarios, server SoCs are designed to be heavily loaded to make the best use of the capitalized hardware and allocated power and cooling budgets. The optimal solution is one where a workload finishes in the shortest time possible while consuming the least possible power.</em></blockquote><p>I’ve highlighted the key design goals, but the gist of it is that:</p><ol><li>Nuvia is likely working on a custom ARM design like Apple rather than reusing ARM’s reference architecture like some other ARM licensees.</li><li>The goal is the highest single-thread performance possible without the use of boost or turbo clocks.</li></ol><h3>Testing Methodology</h3><p>Nuvia recently <a href="https://medium.com/silicon-reimagined/performance-delivered-a-new-way-8f0f5ed283d5">published their findings</a> on how their Phoenix CPU fares against the latest X86 and ARM CPUs. Here are the systems that were tested:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*kGEWoiZ5mq5uDFoB.png" /></figure><p>The reasons for choosing these CPUs were as follows:</p><blockquote><em>The devices tested demonstrate the current state of the art from the majority of the major players, both ARM and X86 based platforms. Intel’s Core i7–1068NG7 Ice Lake based SoC is the highest performance variant currently available utilizing the new Sunny Cove CPU microarchitecture, based on Intel’s 10nm process node. We are also assuming that Intel’s next-generation Ice Lake Server will utilize a CPU core built upon a similar base architecture as Sunny Cove. The Intel Core i7–8750H is the last generation of the Skylake microarchitecture and is more closely related to the CPU core shipping in today’s latest Xeon processors. AMD’s Ryzen 4700U utilizes the latest Zen 2 CPU core on TSMC’s 7nm process node. AMD uses the same Zen 2 CPU core within the CCX chiplet in the Rome EPYC family of processors. Qualcomm’s Snapdragon 865 utilizes ARM’s latest A77 as its performance core and is implemented on TSMC’s 7nm process node. The latest announced Ampere Altra and Amazon Graviton 2 both use an ARM N1 CPU core that is more closely related to the older A76, and both are built upon TSMC’s 7nm process node. Lastly, the Apple A13 and A12Z demonstrate the current fastest ARM-based processors, also both built upon TSMC’s 7nm process node.</em></blockquote><h3>The Results</h3><p>Nuvia released the expected single-core Performance-Per-Watt curve of their upcoming CPU. To get an idea of how this ranks against the M1, I’ve added the <a href="https://browser.geekbench.com/v5/cpu/5197838">single-core Geekbench score for M1 based Mac Mini</a> to Nuvia’s graph. Nuvia is not disclosing the actual score, just giving estimates for now. The actual scores will be released at a later date. My best estimate based on their current projections is that these are higher than Apple M1’s single-core score. Based on the posted graph, I would put the single-core score between <strong>1900</strong> to <strong>2250</strong> range which is about a <strong>10–30%</strong> improvement on single-core performance.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/0*OwyEceFMuPHJgQlY.png" /></figure><p>Nuvia Phoenix CPU performance compared to X86 and ARM CPUs</p><h3>How to read these results?</h3><p>From Nuvia:</p><blockquote><em>When measured against current products available in-market in the 1W-4.5W power envelope (</em>per core<em>), the Phoenix CPU core performs up to 2X faster than the competition.</em></blockquote><p>The 2X claim should now change since the results were published in August before the Apple M1 and the Zen 3 launch. These products now have higher single-core performance than the products Nuvia tested.</p><p>Nuvia did, however, expect such a scenario:</p><blockquote><em>We realize the companies we have measured against in these tests are not standing still and will have new products in the market over the next 18 months. That said, we believe that even with significant performance gains (20%+) with new CPU architectures, we will continue to hold a clear position of leadership in performance-per-watt.</em></blockquote><h3>Why is Nuvia focusing on this 1W-4.5W per core power envelope?</h3><p>I’ll let Nuvia answer this:</p><blockquote><em>All current and future flagship server SoCs are power constrained, very much like mobile SoCs. As core count increases, what is not increasing is the TDP. TDPs are likely going to remain in the 250W — 300W range, which is the maximum power that can be dissipated in an air-cooled environment in a typical datacenter. As more cores are added, the power allocation per core shrinks significantly. A rough calculation can be made to determine the high and low bounds of the per-core power allocation in servers. We can assume that future flagship SoCs will have a minimum of 64 cores and a maximum of 128 cores. The TDP range is 250W — 300W, and the power outside of the CPU can range between 10W — 120W depending upon the workload. Taking into consideration these factors, the amount of power that each CPU can be allocated ranges between 1W — 4.5W when heavily utilized, as is the case in a datacenter environment.</em></blockquote><h3>Why was Geekbench used instead of an industry standard benchmark like SPEC?</h3><p>Nuvia explains:</p><blockquote><em>We believe Geekbench 5 is a good starting point, as it consists of a series of modern real-world kernels that include both integer and floating-point workloads. It runs on multiple platforms, including Linux, iOS, Android, and Windows. It also gives us the ability to conduct these tests on commercially available products .… You may be wondering how we can make the extrapolation from smartphone and client CPU cores to a server core. In our view, there is no meaningful difference. If anything, these environments now deliver incredibly similar characteristics in how they operate.</em></blockquote><p>As a follow-up, <a href="https://medium.com/silicon-reimagined/performance-delivered-a-new-way-part-2-geekbench-versus-spec-4ddac45dcf03">Nuvia posted comparison of testing Geekbench vs SPEC</a>.</p><h3>How is Nuvia’s Performance Per Watt projection calculated?</h3><p>From Nuvia:</p><blockquote><em>NUVIA’s Phoenix CPU performance is projected using architectural performance modeling techniques consistent with industry-standard practices on future CPU cores.</em></blockquote><h3>Can these early projections be trusted?</h3><p>There’s always a risk with early projections. Still, given the team’s experience at Nuvia, I’m inclined to give them the benefit of the doubt that they know what they’re talking about.</p><h3>Why are we talking about a CPU that’s not even released?</h3><p>Well, for a couple of reasons:</p><ol><li>Nobody else developing ARM chips is projecting similar kinds of performance, which makes me question if anyone else is close to achieving similar performance levels on ARM.</li><li>It is way easier to underpromise and overdeliver. Nuvia could have chosen to keep this a secret and just released the CPU but releasing performance expectations this early in the design cycle sets a public bar they will now be held to.</li></ol><ul><li>Of course, the other side of that argument is that this is good marketing that can drum up investor interest and make it easier to recruit silicon engineers.</li></ul><h3>Conclusion</h3><p>It’s an exciting time to be a technologist! After a decade of incremental improvements in single-core performance, we’re starting to see some real focus on single-core performance. I’m excited to see what Nuvia brings to the server market and hope we see some of those gains trickle down to the consumer market as well.</p><h3>References and Sources</h3><ul><li><a href="https://medium.com/silicon-reimagined/performance-delivered-a-new-way-8f0f5ed283d5">Performance Delivered a New Way</a></li><li><a href="https://medium.com/silicon-reimagined/performance-delivered-a-new-way-part-2-geekbench-versus-spec-4ddac45dcf03">Performance Delivered a New Way Part 2 — Geekbench versus SPEC</a></li></ul><h3>Contact</h3><p>Reach out if you have any questions! Feel free to follow me on</p><ul><li><a href="https://twitter.com/singhkays">Twitter — @singhkays</a></li><li><a href="https://linkedin.com/in/karan-kay-singh-2a897919">LinkedIn</a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=bd7933f12dba" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Fixing “sed: command i expects \ followed by text” error]]></title>
            <link>https://singhkays.medium.com/fixing-sed-command-i-expects-followed-by-text-error-798f9b921fe5?source=rss-122b8af0de61------2</link>
            <guid isPermaLink="false">https://medium.com/p/798f9b921fe5</guid>
            <category><![CDATA[command-line]]></category>
            <category><![CDATA[linux]]></category>
            <category><![CDATA[bash]]></category>
            <category><![CDATA[de]]></category>
            <category><![CDATA[script]]></category>
            <dc:creator><![CDATA[Kay Singh]]></dc:creator>
            <pubDate>Wed, 16 Dec 2020 05:56:49 GMT</pubDate>
            <atom:updated>2020-12-16T05:56:49.209Z</atom:updated>
            <content:encoded><![CDATA[<blockquote>This post was first published on my blog @ <a href="https://singhkays.com/blog/sed-error-i-expects-followed-by-text/">https://singhkays.com/blog/sed-error-i-expects-followed-by-text/</a></blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/667/0*suJkaWhk2egyucDA.jpg" /></figure><h4>If you’d like to subscribe for more updates use the <a href="https://www.singhkays.com/index.xml">RSS feed</a> or follow me on Twitter <a href="https://twitter.com/singhkays">@singhkays</a></h4><p>I ran into a very weird error today. While trying to automate the deployment for this blog, I needed to replace a string in a file. The internet tutorials [<a href="https://linuxhint.com/replace_string_in_file_bash/">1</a>, <a href="https://clubmate.fi/replace-strings-in-files-with-the-sed-bash-command/">2</a>, <a href="https://linuxize.com/post/how-to-use-sed-to-find-and-replace-string-in-files/">3</a>, <a href="https://www.diskinternals.com/linux-reader/shell-script-replace-string-in-file/">4</a>] told me that sed would be perfect for this job and the following command should work:</p><pre>sed -i &#39;s/search_string/replace_string/&#39; filename</pre><h3>The Error</h3><p>However, I kept getting tripped up by the following error:</p><pre>sed: 1: &quot;i2.html&quot;: command i expects \ followed by text</pre><p>I spent a bunch of time trying to make sure I escaped the right characters in my strings but the error was persistent! What a nuisance! 😭</p><h3>The Issue</h3><p>Upon further searching, the issue turns out to be differences in sed on Mac systems vs Linux. As is with every question asked, StackOverflow has an answer. <a href="https://stackoverflow.com/a/4247117">High Performance Mark explains</a>:</p><blockquote><em>Your Mac does indeed run a BASH shell, but this is more a question of which implementation of sed you are dealing with. On a Mac sed comes from BSD and is subtly different from the sed you might find on a typical Linux box. I suggest you </em><em>man sed.</em></blockquote><h3>The Fix</h3><p>It turns out that providing the backup extension is required when using the -i switch. <a href="https://stackoverflow.com/a/14813278">chipiik provides the answer that works!</a></p><blockquote><em>This works with both GNU and BSD versions of sed:</em></blockquote><blockquote><em>sed -i&#39;&#39; -e &#39;s/old_link/new_link/g&#39; *</em></blockquote><blockquote><em>or with backup:</em></blockquote><blockquote><em>sed -i&#39;.bak&#39; -e &#39;s/old_link/new_link/g&#39; *</em></blockquote><blockquote><em>Note missing space after </em><em>-i option! (Necessary for GNU sed)</em></blockquote><h3>Conclusion</h3><p><a href="https://www.howtogeek.com/666395/how-to-use-the-sed-command-on-linux/">HowToGeek described it perfectly</a>:</p><blockquote><em>The </em><em>sed command is a bit like chess: it takes an hour to learn the basics and a lifetime to master them (or, at least a lot of practice).</em></blockquote><p>Happy editing!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=798f9b921fe5" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How to get 4K 60 FPS working on an external monitor with Macbooks]]></title>
            <link>https://singhkays.medium.com/how-to-get-4k-60-fps-working-on-an-external-monitor-with-macbooks-fe2a7245e392?source=rss-122b8af0de61------2</link>
            <guid isPermaLink="false">https://medium.com/p/fe2a7245e392</guid>
            <category><![CDATA[60fps]]></category>
            <category><![CDATA[4k]]></category>
            <category><![CDATA[apple]]></category>
            <category><![CDATA[monitor]]></category>
            <category><![CDATA[macbook]]></category>
            <dc:creator><![CDATA[Kay Singh]]></dc:creator>
            <pubDate>Wed, 16 Dec 2020 05:51:53 GMT</pubDate>
            <atom:updated>2020-12-16T05:53:49.065Z</atom:updated>
            <content:encoded><![CDATA[<blockquote><em>This post was originally published on my blog @ </em><a href="https://singhkays.com/blog/4k-60fps-external-monitor-macbook/"><em>https://singhkays.com/blog/4k-60fps-external-monitor-macbook/</em></a></blockquote><h3>The Problem</h3><p>With many of us knowledge workers working from home in the COVID-19 era, our trusty Apple MacBook’s retina displays often need a bigger friend to take over the display duties. It’s no fault of our MacBook’s though — this is the tradeoff we make for portability.</p><p>The problem starts when we start hooking up an external monitor to match the clarity and DPI of the MacBook’s retina screen i.e., a 4K or 5K monitor. 4K monitors have only recently become mainstream. Now, the external display can be on par with the retina screens’ crispness and clarity. This, though, is the root of the problem. The refresh rate on the MacBook retina displays is 60 Hz, which means the screen updates with new images 60 times each second. A higher refresh rate results in a smoother picture, which is why an iPad Pro screen that refreshes at 120 Hz is smoother to use. To run a 4K monitor at 60 Hz requires the right combination of hardware, cables, and adapters.</p><p>Here are two kinds of problems you might encounter:</p><ol><li>MacBook does not allow you to select a 4K resolution when connected to a 4K monitor</li><li>MacBook only runs the 4K resolution at 30 Hz aka animations and mouse movement is choppy</li></ol><p>To figure out a solution, here’s what I recommend.</p><h3>Step 1: Check if your MacBook supports 4K resolution</h3><p>The first step is to make sure your MacBook supports outputting 4K signal to an external monitor. The best place to check is on <a href="https://support.apple.com/">Apple’s Support website</a>. A search query like the following will usually bring up the correct model.</p><blockquote><em>apple macbook 13 inch 2016 technical specifications</em></blockquote><p><em>NOTE: You’ll need to specify the year and screen size (13, 15, 16 inches etc.) to match your machine.</em></p><p>Then the supported resolution’s are usually specified in the “Video Support” section. Here are two examples of the 2016 models.</p><p><strong>MacBook 2016</strong></p><blockquote><em>Graphics and Video Support Intel HD Graphics 515 Dual display and video mirroring: simultaneously supports full native resolution on the built-in display and up to </em><strong><em>3840 by 2160 pixels at 30Hz</em></strong><em> on an external display, both at millions of colors</em></blockquote><p><strong>MacBook Pro 2016</strong></p><blockquote><em>Video Support Simultaneously supports full native resolution on the built-in display at millions of colors and: One display with </em><strong><em>5120-by-2880 resolution at 60Hz</em></strong><em> at over a billion colors Up to two displays with </em><strong><em>4096-by-2304 resolution at 60Hz</em></strong><em> at over a billion colors</em></blockquote><p>Notice the difference above in the bolded text. Both these 2016 models support 4K resolution (<em>3840 by 2160 pixels or higher</em>), but the 2016 MacBook only supports running an external display at 30 Hz. So, this is the first order of business to tackle and find out if your machine can support 4K external display and, more importantly, at 60 Hz.</p><h3>Step 2: Make sure you’ve got the right adapter</h3><p>In recent MacBook models, Apple has removed the full HDMI ports and replaced them with USB-C Thunderbolt ports that perform video-out duties. This is great if you want to reduce the number of cables because the USB-C cable can also charge the MacBook. Still, the majority of users have legacy type connections i.e. HDMI or Display Port cables. Apple sells a USB-C adapter that lets you plug in an HDMI cable. There are actually two Apple USB-C adapters, and this is where Apple makes it more challenging than it needs to be. Both these adapters look exactly the same as the below image, but only one supports 4K 60Hz.</p><ol><li><strong>Model A2119</strong> is the newest model that supports HDMI 2.0 and up to <strong>4K 60 Hz</strong>.</li><li><strong>Model A1621</strong> supports HDMI 1.4b and <strong>4K up to 30 Hz</strong>.</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/616/0*9WHB88Faity3ojGt.jpg" /></figure><p>If you’re using this adapter and seeing only 4K 30Hz resolution options, make sure you’ve got the newest model. The model number is written on the adapter but is very faint and small, so you might have to squint 👀</p><p><em>(DISCLOSURE: This post includes affiliate links. Should you click an affiliate link and make a purchase I may receive a small commission at no extra cost to you.)</em></p><h3>Official Adapter costs $60–70</h3><p>If you want to buy the official Apple adapter, it’s normally $69.99 on Apple’s website but it’s going for $61 on Amazon at the time of this writing.</p><ul><li><a href="https://amzn.to/2K5CJ0H">Apple USB-C Digital AV Multiport Adapter A2119 MUF82AM/A</a></li></ul><h3>Third-party adapters for &lt; $20</h3><p>If the Apple adapter is too expensive, then I’d recommend the following third-party adapters. You have to read the fine print on these as some of these only work at 4K 30 Hz. Here are a few recommendations with the first one that personally worked for me:</p><ul><li><a href="https://amzn.to/3abLiBD"><strong>CHOETECH USB Type C (Thunderbolt 3 Compatible) to HDMI Adapter Cable</strong></a></li><li><a href="https://amzn.to/3mlHPmH">Anker USB C to HDMI Adapter (4K@60Hz)</a></li><li><a href="https://amzn.to/2JWWo33">Uni USB C to DisplayPort Cable</a></li><li><a href="https://amzn.to/37hSQ3V">Uni USB C to HDMI Adapter</a></li><li><a href="https://amzn.to/2Wg7XET">UGREEN USB C to HDMI Cable 4K 60HZ Thunderbolt 3</a></li></ul><h3>Contact</h3><p>Reach out if you have any questions! Feel free to follow me on</p><ul><li><a href="https://twitter.com/singhkays">Twitter — @singhkays</a></li><li><a href="https://linkedin.com/in/karan-kay-singh-2a897919">LinkedIn</a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=fe2a7245e392" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How Spotify is Killing the Open Podcast Ecosystem]]></title>
            <link>https://singhkays.medium.com/how-spotify-is-killing-the-open-podcast-ecosystem-eed787ea4857?source=rss-122b8af0de61------2</link>
            <guid isPermaLink="false">https://medium.com/p/eed787ea4857</guid>
            <category><![CDATA[podcast]]></category>
            <category><![CDATA[strategy]]></category>
            <category><![CDATA[spotify]]></category>
            <category><![CDATA[product-mangement]]></category>
            <category><![CDATA[ecosystem]]></category>
            <dc:creator><![CDATA[Kay Singh]]></dc:creator>
            <pubDate>Thu, 29 Oct 2020 06:17:44 GMT</pubDate>
            <atom:updated>2020-10-29T06:17:44.974Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="Podcasts vs Spotify" src="https://cdn-images-1.medium.com/max/1024/0*ht3rr9EdkpKR84O8.jpg" /></figure><p>Earlier this year, when Spotify announced that “The Joe Rogan Experience” would stream exclusively on Spotify, it sent shockwaves throughout the podcast ecosystem. This deal — reportedly worth over $100 million — is one of the most lucrative podcast deals. The presence of this deal, let alone the magnitude, is rare for the podcast ecosystem, which is built on open principles <em>i.e. all podcasts are available everywhere</em>. Therefore, the concept of exclusivity does not exist in the vocabulary of podcast listeners nor podcast creators. For Spotify to spend a considerable sum of money on a medium that is not its core business <em>(i.e. music)</em> represents a significant shift in strategy. In this blog, I explore why Spotify is embarking on this journey, and the value Spotify can deliver in this space.</p><blockquote>This article first appeared on my blog at <a href="https://www.singhkays.com/blog/how-spotify-is-killing-the-open-podcast-ecosystem/">https://www.singhkays.com/blog/how-spotify-is-killing-the-open-podcast-ecosystem/</a></blockquote><h3>Unexpected Competition</h3><p>To understand this sudden shift in strategy, we need to understand the current state of the podcast industry. <a href="https://www.statista.com/topics/3170/podcasting/">According to Statista</a>,</p><blockquote>Back in 2006, only 22 percent of the adult population in the United States was aware of podcasting. By 2020, this figure had risen to 75 percent. Podcasting is an increasingly popular pastime in the US, and there were an estimated 88 million podcast listeners in the country in 2019. Forecasts suggest that the number of podcast listeners will surpass 160 million in 2023 after increases of around 20 million each year.</blockquote><p>When consumers are not in Spotify’s app listening to music, they are listening to podcasts in a different app. This represents a threat to Spotify’s ad-supported free product and a loss of ad-revenue from its free users <em>(Spotify’s premium product is not affected by this because the subscription cost per user is fixed regardless of listening time).</em> To serve relevant ads and, in return, increase its CPM for advertisers, Spotify needs to know its users intimately. This means that Spotify needs its users to spend as much of their digital lives as possible in its walled garden. The quickest way for Spotify to enable this is to build a podcast streaming capability and stream all available podcasts through its music app. This is precisely what Spotify started doing in 2018. Since then, Spotify has quickly ramped up the number of podcast titles available on its platform with increasing engagement every quarter.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fdatawrapper.dwcdn.net%2F9DKUY%2F1%2F&amp;display_name=Datawrapper&amp;url=https%3A%2F%2Fdatawrapper.dwcdn.net%2F9DKUY%2F1%2F&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=dwcdn" width="500" height="280" frameborder="0" scrolling="no"><a href="https://medium.com/media/a8cbc85f4fe625afa1a7484def22b269/href">https://medium.com/media/a8cbc85f4fe625afa1a7484def22b269/href</a></iframe><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fdatawrapper.dwcdn.net%2FDxB4K%2F1%2F&amp;display_name=Datawrapper&amp;url=https%3A%2F%2Fdatawrapper.dwcdn.net%2FDxB4K%2F1%2F&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=dwcdn" width="600" height="311" frameborder="0" scrolling="no"><a href="https://medium.com/media/f3a94b6672601e0399cecac4390815fe/href">https://medium.com/media/f3a94b6672601e0399cecac4390815fe/href</a></iframe><h3>How Podcasts Work</h3><p>It is important to understand the open podcast ecosystem before moving forward. Podcasts are delivered through a technology called RSS (Rich Site Summary). If you’ve seen this icon on the internet before, then you have come across an RSS feed.</p><figure><img alt="RSS Feed Icon" src="https://cdn-images-1.medium.com/max/256/0*aKyKA-WbjlEwHEcP.png" /></figure><p>An RSS feed acts as a content-distribution tool and is often used by websites to distribute web content such as new articles, blogs, and podcasts. It is an XML file that holds metadata and location of the podcast content, which in most cases, is the audio file.</p><p>The publisher sets up a public RSS feed URL which is used by the listener in a podcast player to subscribe to the podcast i.e. the podcast player downloads new episodes as they are published. Of course, in modern times, any capable podcast app indexes the most popular podcasts and provides users with search functionality to discover the podcasts rather than inputting podcast URLs. Notice the beauty of this ecosystem because both sides of this publisher-listener transaction are open. On the publisher side, the podcast feed URL is available publicly without any authentication to be consumed by anybody who understands the XML structure of the feed. On the listener side, the listener is free to use their favorite podcast player (<em>there are many out there…maybe too many</em> 😄). The downside to this open nature of the ecosystem is that it creates challenges for measurement and monetization as we’ll see in the rest of this article.</p><h3>Spotify’s Business Model</h3><p>Spotify made it possible for anyone to pay $10 a month and have access to almost any song ever released! By doing so, Spotify might have single-handedly solved the piracy problem for the music industry and proved a business model exists where consumers will pay for music. Buoyed by Spotify’s success, major giants like Apple, Amazon &amp; Microsoft, and many others have tried to replicate Spotify’s strategy — albeit to varying degrees of success.</p><figure><img alt="https://www.statista.com/chart/20826/music-streaming-services-with-most-subscribers-global-fipp/" src="https://cdn-images-1.medium.com/max/1024/0*gVcezPo5SObCbBj0.jpeg" /><figcaption><a href="https://www.statista.com/chart/20826/music-streaming-services-with-most-subscribers-global-fipp/">https://www.statista.com/chart/20826/music-streaming-services-with-most-subscribers-global-fipp/</a></figcaption></figure><p>Spotify has built an impressive business model in a crowded market despite being a late entrant in the music streaming business. Pandora, which once was the darling of the industry, has been continually losing users to Spotify. Pandora, which rose to prominence by helping users discover new music through its recommendation algorithm, was always marred by two fundamental limitations — limited catalog and a limited number of song skips due to licensing restrictions. In comparison, Spotify offered unlimited song skips and a bigger catalog in its free tier and was able to get many users to switch from Pandora. These free users would later go on to become Spotify’s paying users as Spotify delivered better value in its premium product and added restrictions on its free plan (<em>restricted song skips on mobile apps etc.</em>) The chart below is evidence of what happened next. Pandora has been consistently losing users, while Spotify now almost has three times the number of Monthly-Active-Users (MAUs) as Pandora.</p><figure><img alt="Spotify vs Pandora MAUs — Monthly Active Users 2008–2020" src="https://cdn-images-1.medium.com/max/1024/1*FPwFTepQJn0UWlxT20f1Ng.png" /></figure><p>Today, Spotify is seeing increased growth quarter-over-quarter in both its ad-free and premium products with most of its revenue coming from its premium product.</p><figure><img alt="Spotify Quarterly MAUs — Monthly Active Users — Ad-Supported vs Premium" src="https://cdn-images-1.medium.com/max/1024/1*oEIbzMUlLeIRWSxH9QWPtg.png" /></figure><figure><img alt="Spotify Quarterly Revenue — Ad-Supported vs Premium" src="https://cdn-images-1.medium.com/max/1024/1*lIpQ0s-NM__1VT6XGCWyVQ.png" /></figure><h3>Spotify’s Podcast Investments</h3><p>019 was a year that really set the foundations for Spotify’s podcast strategy. Spotify spent nearly $500 million in acquiring — <a href="https://investors.spotify.com/financials/press-release-details/2019/Spotify-Announces-Strategic-Acquisitions-to-Accelerate-Growth-in-Podcasting/default.aspx">Anchor, Gimlet Media,</a> and <a href="https://investors.spotify.com/financials/press-release-details/2019/Spotify-to-Acquire-Parcast-a-Premier-Podcast-Storytelling-Studio/default.aspx">Parcast/Cutler Media</a>. Anchor, which bills itself as <em>“The easiest way to make a podcast”</em>, supplies tools that make it easy for podcasters to create, distribute and monetize podcasts. Gimlet and Parcast, on the other hand, produce some of the most popular podcasts out there, such as StartUp, Reply All, Homecoming, Mogul, Serial Killers, Unsolved Murders, Cults and Conspiracy Theories and many more.</p><p>With these acquisitions, Spotify is pivoting into an “Audio” company — and no longer a music-only company. Here is what <a href="https://newsroom.spotify.com/2019-02-06/audio-first/">Spotify CEO Daniel Ek had to say about these acquisitions</a> in his essay titled “Audio-First”:</p><blockquote>That’s why we announced today the strategic acquisitions of two podcasting companies, Gimlet and Anchor. These companies serve two different, distinct roles in the industry. Gimlet is one of the best content creators in the world, with unique, celebrated podcast shows like <a href="https://open.spotify.com/show/2pWaoB1xgi5x6UgMgwRlsV?si=kggK3qC_Tl2zRGgyoi3OKw">Homecoming</a>, which was recently adapted into a critically acclaimed show on Amazon Prime, and the internet culture hit <a href="https://open.spotify.com/show/7gozmLqbcbr6PScMjc0Zl4?si=tK1XAVQwSvS-LBCH0baQCA">Reply All</a>. And Anchor has completely reimagined the path to audio creation, enabling creation for the next generation of podcasters worldwide — 15 billion hours of content on the platform during Q4. These companies are best-in-class, and together we will offer differentiated and original content. Gimlet and Anchor will position us to become the leading platform for podcast creators around the world and the leading producer of podcasts.</blockquote><p>These acquisitions play in different parts of the podcast ecosystem but help Spotify cater to both sides of the podcast ecosystem — producers and listeners. Spotify believes it can extract value from the podcast ecosystem in a way no one has been able to do until now. An indicator of the importance of the podcast strategy to Spotify’s business can be judged by the number of mentions of the word “podcast” in its recent quarterly earnings releases.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*1tuB-lRPm9YjLGPpZ3NwvQ.png" /></figure><p>So far, 2020 is turning out to be another podcast investment year for Spotify. There’s a good reason why Spotify has spent close to a billion dollars on signing exclusive deals and buying podcast studios. From the <a href="https://s22.q4cdn.com/540910603/files/doc_financials/2019/q4/Shareholder-Letter-Q4-2019.pdf">Q4 2019 quarterly report:</a></p><blockquote>We continue to see exponential growth in podcast hours streamed (up approximately 200% Y/Y) and are now seeing clear indications that podcast usage is driving increased overall engagement and retention. We have seen early indications that our investments in podcasts are having a **positive impact on conversion of free to paid users.</blockquote><p>And from the <a href="https://s22.q4cdn.com/540910603/files/doc_financials/2019/q3/Shareholder-Letter-Q3-2019-%5bFinal%5d.pdf">Q3 2019 quarterly report</a>:</p><blockquote>For music listeners who do engage in podcasts, we are seeing increased engagement and increased conversion from Ad-Supported to Premium. <strong>Some of the increases are extraordinary, almost too good to be true.</strong> We’re working to clean up the data to prove causality, not just correlation. **Still, our intuition is the data is more right than wrong, and that we’re onto something special.</blockquote><p>With these investments, Spotify is aiming to add value, not just for the user but to its bottom-line as well. And it is clear that Spotify is having success that even it did not predict so soon.</p><h3>Spotify’s Podcast Strategy</h3><p>Spotify’s podcast strategy hinges on becoming an “Aggregator” of podcasts, as described in <a href="https://stratechery.com/2015/aggregation-theory/">Ben Thompson’s Aggregation Theory</a>. Ben writes,</p><blockquote>The value chain for any given consumer market is divided into three parts: suppliers, distributors, and consumers/users. The best way to make outsize profits in any of these markets is to either gain a horizontal monopoly in one of the three parts or to integrate two of the parts such that you have a competitive advantage in delivering a vertical solution.</blockquote><p>Today, Spotify acts as the distributor of the audio content from suppliers i.e. the music labels <em>(Warner Music Group, Universal Music, Sony Music etc.)</em> and podcast producers (<em>NPR, iHeartRadio, PRX, Wondery etc.</em>). There are other distributors as well who are distributing the same content.</p><figure><img alt="Spotify’s place in the music universe with Pandora, Amazon Music, Apple Music. Spotify acts as one of the music distributors" src="https://cdn-images-1.medium.com/max/1024/1*ypnS5hnoEotXV_bOBLmpWQ.png" /></figure><figure><img alt="Spotify’s place in the podcast universe where it’s a new distributor like previous podcast apps" src="https://cdn-images-1.medium.com/max/1024/1*9kAAiIwBo60ajOVgqGVDNA.png" /></figure><p>With the 2019 and 2020 investments, Spotify is beginning to differentiate itself from being just a podcast app and positioning itself as a supplier of podcasts.</p><p><strong>Spotify will now be able to offer exclusive content on its platform. If this sounds like something you’ve heard before, then it is 😉 This is exactly what’s happening in the video streaming market with everyone looking to build their library of original content. I expect podcasts to be the next platform where platforms will compete on original content.</strong></p><p><strong>Besides offering exclusive content, Spotify’s building a platform where current podcasts that are published in the open model might become exclusive to Spotify for better monetization opportunities. These are what I call Spotify’s “Killer Features” and the topic of discussion in the next topic.</strong></p><figure><img alt="Spotify is becoming a Podcast supplier + distributor with the 2019 acquisitions" src="https://cdn-images-1.medium.com/max/1024/1*FxbCa2uDYiWnxTwijpzzYw.png" /></figure><h3>Spotify’s Killer Features</h3><p>The inertia of entrenched workflows is such that they are tough to change. Most users simply do not abandon an existing app to try out a shiny new app. I am often surprised at how many people do not take the time to explore all the knobs and dials in the settings of an app they use every day. Therefore, to get a user to adopt a new app, you need to offer significantly more value, and more importantly, that value needs to be observable with the default settings.</p><p><a href="https://stratechery.com/2015/aggregation-theory/">Ben’s Aggregation Theory again</a>,</p><blockquote>This has fundamentally changed the plane of competition: no longer do distributors compete based upon exclusive supplier relationships, with consumers/users an afterthought. Instead, suppliers can be commoditized leaving consumers/users as a first order priority. By extension, this means that the most important factor determining success is the user experience: the best distributors/aggregators/market-makers win by providing the best experience, which earns them the most consumers/users, which attracts the most suppliers, which enhances the user experience in a virtuous cycle.</blockquote><p>This essentially means that Spotify needs to offer both Suppliers and Consumers a much better user experience for Spotify to become the platform of choice. Otherwise, users will continue to listen in their existing podcast apps. On the other hand, publishers have no incentive to publish exclusively on Spotify’s platform. At a high-level Spotify’s will provide the following benefits to both suppliers and consumers:</p><ol><li>Algorithmic Discovery</li><li>Analytics &amp; Measurement</li><li>Targeted Advertising &amp; Monetization</li><li>Original &amp; Exclusive Content</li></ol><h4>1. Algorithmic Discovery</h4><p>Today, a user discovers the podcast title on their own through the internet, friends, family, etc. This model is remarkably like how consumers used to discover music. This model worked fine when there were few good podcasts, but with the explosion of new material on the podcast scene, finding enjoyable content is becoming akin to finding a needle in a haystack.</p><figure><img alt="Traditional podcast discovery and listening behavior. Users subscribe themselves to the podcasts." src="https://cdn-images-1.medium.com/max/1024/1*HJFWj90N98dl2isSX6SkyA.png" /></figure><figure><img alt="Problem with discovering podcasts today. Too many podcasts make it difficult to find good ones." src="https://cdn-images-1.medium.com/max/1024/1*8ekz1mRyDYVBtyos4gDLpA.png" /></figure><p>Today, new music discovery is primarily done through algorithmic recommendations that are based on demographics, genre, mood, similarity, etc. Spotify is no stranger to this model, having transformed music discovery with playlists like <a href="https://newsroom.spotify.com/2019-05-02/five-ways-to-make-your-discover-weekly-playlists-even-more-personalized/">Discover Weekly</a> and <a href="https://newsroom.spotify.com/2018-05-18/how-your-daily-mix-just-gets-you/">Daily Mix</a>. Now, Spotify introducing a similar model to podcast consumption and discovery.</p><figure><img alt="Spotify’s weekly recommended podcasts" src="https://cdn-images-1.medium.com/max/800/0*n099jCkRcHtiE9sO.png" /></figure><p><strong>First</strong>, Spotify recently introduced podcast playlists such as <a href="https://newsroom.spotify.com/2019-11-19/your-daily-podcasts-playlist-makes-finding-your-next-favorite-show-easier-than-ever/">Daily Podcasts</a>, Editorial playlists — <a href="https://newsroom.spotify.com/2020-04-21/spotifys-new-podcast-playlists-will-help-you-discover-your-next-obsession/">Best Podcasts of the Week, Crime Scene, and Brain Snacks</a> and to help with the pandemic blues — <a href="https://newsroom.spotify.com/2020-04-27/daily-wellness-a-new-mix-of-motivational-podcasts-and-personalized-music/">Daily Wellness</a>. Here’s an example from Spotify of how Daily Podcasts playlist makes podcast discovery easier:</p><ol><li>Spotify’s algorithms analyze your podcast behavior — like recent streams and follows.</li><li>Then, based on your listening history and the podcast type, we’ll recommend the next best episodes for you.</li><li>That might be the next sequential episode in a podcast you’re already listening to (think <a href="https://open.spotify.com/show/6Cg6ZBmPpcq8GFgFdGLKxV?si=vXREWiyITS2wFRPxM0qLDQ">Dog Tales</a> and <a href="https://open.spotify.com/show/0P13JasQfVZ1RiDCMZMYNU?si=XJeha62gSYaqWfthc9BRHw">How’s Work with Esther Perel</a>), a recent stand-alone evergreen episode in another show (maybe <a href="https://open.spotify.com/show/2UtdM2mvBQesTqGTyiKn8w?si=NOCAePrSQdK6DEudfQf_iQ">Amy Schumer Presents: 3 Girls 1 Keith</a> or <a href="https://open.spotify.com/show/4iNCNuREKROIipaBnYs6N2?si=6x-DrTtDRO-f9NjcD5EwWw">Certified Buckets</a>), or a timely episode from a daily updating podcast (like <a href="https://open.spotify.com/playlist/3xUqX6SyysyzBLuI5Y0CMA?si=ddJnXV8FSNmtm2xl23p1oQ">Horoscope Today</a> or <a href="https://open.spotify.com/show/0KxdEdeY2Wb3zr28dMlQva?si=_hsRUosgRzCwS774cVi_Jg">The Journal</a>).</li><li>Don’t worry — no spoilers here! If you’ve never listened to a story-driven sequential show we think you’d like, you’ll get the trailer or pilot episode first — to see if it catches your eye (er, ear).</li></ol><p>The importance of these editorial playlists cannot be understated. <a href="https://www.stevebenjamins.com/blog/music-in-the-age-of-algorithms-47syg">As Steve Benjamins’s</a> experience proves that being included in the <strong>“Discover Weekly” playlist is the primary reason for his $800 per month income</strong>. Steve writes:</p><blockquote>Every Monday my music gets a spike in streams on Spotify. You could set a watch to it — it’s that consistent.</blockquote><blockquote>What makes Monday so special?</blockquote><blockquote>Well every Monday Spotify sends out a new Discover Weekly playlist. Discover Weekly is an algorithmic playlist — which means its personalized with songs Spotify thinks the user would like.</blockquote><blockquote>No other promotional tactic in music comes close to Discover Weekly in delivering new listeners in such a low-effort, high volume way.</blockquote><blockquote>Back in 2013 I would spend hours cold-emailing bloggers. I would be lucky if I got a hit and got 1,000 plays on one of my songs. It was labor intensive.</blockquote><blockquote>Now algorithmic playlists like Discover Weekly send 1,000 new listeners every week without any work on my part. This is amazing. Cold outreach sucks. It sucks for the artist and it sucks for the bloggers. Spotify deserves a lot of credit for here.</blockquote><p><strong>Second</strong>, Spotify is borrowing a leaf out of Netflix’s recommendations strategy. Like <a href="https://en.wikipedia.org/wiki/Netflix_Prize">Netflix Prize</a> before it, which aimed at making Netflix’s recommendation engine more accurate, Spotify announced the <a href="https://engineering.atspotify.com/2020/04/16/introducing-the-spotify-podcast-dataset-and-trec-challenge-2020/">“Spotify Podcast Dataset and TREC Challenge”</a>. This dataset holds raw audio and transcripts of 100,000 episodes from thousands of different shows on Spotify. With this research challenge, Spotify is trying to understand the content of podcasts (e.g. what exactly is being covered, by whom, and how?) and, and how to use this information to connect users to shows that align with their interests. This data will help answers questions such as:</p><ul><li>How to identify podcasts that interview Barack Obama, as opposed to those that talk about him?</li><li>What are the most important parts of a 45-minute episode?</li><li>Which podcasts are “high quality” or “informative” or “interesting”?</li></ul><p>With these answers, Spotify will be able to recommend better not just shows but a segment within a specific episode that might align with your interests. These recommendations have the potential to change podcast consumption forever and build a moat for Spotify, which will be hard to overcome or replicate.</p><figure><img alt="Spotify recommendation engine can suggest podcasts or even episodes that might be related to your existing subscriptions" src="https://cdn-images-1.medium.com/max/1024/1*D3XWFdxC2QOTv3i2Fo5aTw.png" /><figcaption><em>Spotify recommendation engine can suggest podcasts or even episodes that might be related to your existing subscriptions</em></figcaption></figure><h3>2. Analytics &amp; Measurement</h3><p>Reporting and measurement of listener demographics are not easy due to the decentralized podcast distribution model through RSS feeds. The podcast feed publishing system acts as a dumb 1-way delivery pipe to the user’s podcast player of choice. This has made publishers produce creative strategies such as audience surveys, coupon codes, vanity URLs, etc. to derive proxy metrics for ad impressions, audience segments, etc.</p><p>To solve this problem, Spotify launched the <a href="https://newsroom.spotify.com/2019-08-13/introducing-spotify-for-podcasters/">“Spotify for Podcasters”</a> dashboard. Here is an excerpt from the <a href="https://newsroom.spotify.com/2019-08-13/introducing-spotify-for-podcasters/">launch blog post</a>:</p><blockquote>At its core, Spotify for Podcasters is a discovery and analytics dashboard. One where you can both submit your show to Spotify and dive deep into engagement and demographic data for your podcast, tracking things like average listening times, episode streams, and total listeners. With so many podcasts out there, it’s more important than ever that you have the data you need to help you understand and grow your audience. That’s exactly what your dashboard is designed to provide.</blockquote><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2F7BeKkcY5y4E%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3D7BeKkcY5y4E&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2F7BeKkcY5y4E%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/071cf930a4b35567cbd30ccbaeffc741/href">https://medium.com/media/071cf930a4b35567cbd30ccbaeffc741/href</a></iframe><p>Spotify has already done the hard work of building the tools for tracking music engagement. Therefore, it is only natural that those same tools can be extended to podcast publishers. “Spotify for Podcasters” dashboard is designed to provide these analytics and engagement tracking tools for publishers to track what kind of content resonates with their listeners and adjust.</p><h3>3. Targeted Advertising &amp; Monetization</h3><p>While other digital mediums such as online retail and social media have figured out advertising models, podcast monetization remains a tough challenge for publishers. To sell ad spots, a publisher needs to build relationships with brands and provide podcast consumption metrics and listener demographics. In the podcast ecosystem, these metrics and listener demographics are not easily discoverable as I described in the last section and how Spotify will add value in this area. <a href="https://www.spotifyforbrands.com/en-US/news/sai-announcement">Spotify compares the current state of advertising</a> in the podcast ecosystem to that in the print medium.</p><blockquote>“…the podcast industry measures audience, reach, and impact much like you would a full-page ad in a magazine. Advertisers generally have a sense for who they’re reaching, based on survey data and the magazine’s target audience. Total circulation measures the number of doorsteps that the issue hits, but there’s no precise data on how many readers actually opened the magazine, let alone reached or acted on the ad.”</blockquote><p>Spotify is in a unique position to solve these challenges because it has more personalized data on its users than any of the podcast apps. More importantly, it has already built this kind of targeted ad technology to monetize its free music users. Spotify has launched two key features to enable podcast monetization.</p><p><strong>First</strong>, Spotify launched <a href="https://www.spotifyforbrands.com/en-US/news/sai-announcement">Streaming Ad Insertion (SAI)</a> technology to dynamically insert ads into user’s podcast streams based on what it knows about the user, like where they are located, type of device they use, age, etc. Spotify posits that SAI can <em>“</em><a href="https://www.spotifyforbrands.com/en-US/news/sai-announcement"><em>offer intimacy and quality of traditional podcast ads with the precision and transparency of modern-day digital marketing</em></a><em>.”</em> So instead of your favorite podcast host talking through the benefits of a product that might be irrelevant to you, you might hear a more personalized ad based on your listening habits and interests, which might be good or bad depending on your take on advertisements.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*rJLzS5GUCDt935Dd.png" /></figure><p>Spotify’s podcast streaming model also helps as the ads can be inserted dynamically into the podcast stream instead of being pre-baked into the downloaded audio file. With SAI, Spotify offers the easiest way to monetize podcasts for creators — very YouTube-like. Theoretically, in the future, a publisher could just tick a checkbox saying they want to opt their podcast into ads, and Spotify would take care of everything — ad bidding, ad insertions, etc. It’s conceivable that the Spotify could even allow publishers to replace the generic ads with more targeted ones and receive a higher payout compared to advertising on other platforms. In the end, you’d have the same podcast stream with generic ads in the podcast apps that download the podcast and targeted ads in the Spotify stream.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*vzmhAJ32C87y_G4ll5e2YA.png" /></figure><p><strong>Second</strong>, Spotify is launching <a href="https://www.spotifyforbrands.com/en-US/news/spotify-introduces-in-app-offers-for-podcast-ads/">In-App Offers</a>. Podcast listening usually happens in parallel with another activity like running, workout, cooking, etc. Therefore, it’s easy to miss the in-podcast advertising offers. To combat this, podcast show hosts repeat these offer codes and unique URLs multiple times to make sure the listener doesn’t miss the offer. In-App Offers is designed to reduce this friction and make it easy to connect users with advertiser offers. Listeners can redeem deals when they open the Spotify app and navigate to the podcast page. Spotify elucidates that this will lead to one fewer ‘w-w-w-dot’ spelling lesson from our favorite podcast creators.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/0*QklhaHY5SEfyyNyh.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/636/0*bKhfj_ZLe-8KEqS_.gif" /></figure><h3>4. Original and Exclusive Content</h3><p>For every song that is streamed on Spotify’s platform, it pays a royalty fee to the music labels who own the rights to the song. As Spotify’s userbase grows, the fixed costs also grow. The only way for Spotify to increase its margins is to negotiate better deals with the music labels, which is easier said than done. At some point, the ROI on such negotiations is not going to be worth it, and Spotify needs to come up with a better way to increase its gross margin. To combat this, Spotify is taking a leaf out of Netflix’s original content strategy. Netflix’s gross margin has seen a significant uptick since it started investing in original content in 2013.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*laNo4AM0CK3XS2Evchqeew.png" /></figure><p>The 2019 acquisitions (Gimlet, Anchor &amp; Parcast) were the first manifestations of this strategy. Since then, Spotify has acquired <a href="https://investors.spotify.com/financials/press-release-details/2020/Spotify-Announces-Strategic-Acquisition-of-Bill-Simmons-The-Ringer-to-Grow-Sports-Vertical/default.aspx">Bill Simmons’ The Ringer</a>, struck deals with President Barack Obama and Michelle Obama’s production company — <a href="https://newsroom.spotify.com/2019-06-06/higher-ground-announces-partnership-with-spotify-to-produce-podcasts/">Higher Ground</a> for an <a href="https://newsroom.spotify.com/2020-07-16/higher-ground-and-spotify-announce-the-michelle-obama-podcast-debuting-july-29/">exclusive podcast</a>, Joe Rogan for the ‘<a href="https://newsroom.spotify.com/2020-05-19/the-joe-rogan-experience-launches-exclusive-partnership-with-spotify/">The Joe Rogan Experience’</a>, Warner Bros for <a href="https://newsroom.spotify.com/2020-06-18/warner-bros-is-bringing-the-dc-super-heroes-and-super-villains-to-spotify-as-an-exclusive-series-of-podcasts/">DC Super Heroes and Super Villains exclusive podcast</a>, Kim Kardashian West for a <a href="https://www.wsj.com/articles/kim-kardashian-west-lands-podcast-deal-with-spotify-11592427337">criminal-justice podcast</a>. If every major media company creating their streaming service in the streaming wars proves anything, it’s that having a strong lineup of original content is the best way to retain and attract users. Spotify is just getting started, but the volume of original content released every quarter is increasing.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*PoVqV8Qif0_RiZ9bo1bNGA.png" /></figure><h3>Conclusion</h3><p>Spotify is in a prime position to solve a lot of creator challenges and gain the first-mover advantage into what I’m calling the Podcast 2.0 era. Spotify’s latest quarterly earnings prove that Spotify is seeing success with its Podcast strategy. Because of this success, I expect others to follow just as they did in streaming music. Amazon has already embarked on a similar strategy by adding podcasts to its Prime Music offering. Early reports indicate competition is coming down the line from Apple as well. Rumors suggest Apple is <a href="https://www.bloomberg.com/news/articles/2020-05-21/apple-ramps-up-original-podcasts-in-part-to-help-promote-tv">currently negotiating deals for original podcasts</a> on its platform. Apple, which would represent significant competition for Spotify because the Apple Podcasts app (previously iTunes) has historically been <a href="https://www.statista.com/statistics/943537/podcast-listening-apps-us/">the top used app for podcast listening</a>, has never capitalized on its podcast popularity so it remains to be seen how potent they can be against Spotify and now Amazon.</p><p>Podcast ecosystem is going to be an exciting space to watch as it evolves over the next few years. There are three key trends to watch as this evolution plays out:</p><ol><li>Our music apps are going to be streaming podcasts and, in the process, more targeted ads.</li><li>An avenue to monetization is about to be opened for the podcast creators. If you have interesting content, this might be the best time to be a podcaster! On the flip side, it might not be the best time for the open podcast ecosystem and podcast apps not named Spotify, Amazon Music or Apple Podcasts.</li><li>We are about to see balkanization of the podcast ecosystem which will resemble the video streaming ecosystem. Until now, every podcast title has been available on every podcast very reminiscent of the early days of Netflix streaming. Now a days, every media conglomerate is focusing on creating a streaming offering around the content they own instead of licensing it out to the established streamers like Netflix, Hulu etc.</li></ol><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=eed787ea4857" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[AV1 Ecosystem Update: July 2019]]></title>
            <link>https://singhkays.medium.com/av1-ecosystem-update-july-2019-bbded4a66a1?source=rss-122b8af0de61------2</link>
            <guid isPermaLink="false">https://medium.com/p/bbded4a66a1</guid>
            <category><![CDATA[xilinx]]></category>
            <category><![CDATA[av1]]></category>
            <category><![CDATA[h264]]></category>
            <category><![CDATA[mozilla]]></category>
            <category><![CDATA[video-encoding]]></category>
            <dc:creator><![CDATA[Kay Singh]]></dc:creator>
            <pubDate>Tue, 13 Aug 2019 00:00:00 GMT</pubDate>
            <atom:updated>2019-08-30T23:35:59.210Z</atom:updated>
            <content:encoded><![CDATA[<p>If May and June were F1 race cars in terms of speed of AV1 development, then July would probably be a Honda Civic. After numerous exciting announcements in the previous month, July felt like everyone’s on vacation, which is probably true because summer months tend to be slow as the school’s out. So, this update will be a short one as I’m out on vacation as well :)</p><h3>Mile High Video 2019</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*o_7fDtsenBUMOht-FKRAhg.png" /></figure><p>Nathan Egge of Mozilla and Brion Vibber of Wikimedia presented a session recapping AV1’s progress in the last year. Here’s a recap of the most important parts of the presentation</p><ul><li>Live encode and playback demo in all major browsers (even Safari!) using the dav1d decoder. The Safari demo was done using a WebAssembly shim called ogv.js. ogv.js includes a build of the dav1d AV1 decoder which works well enough in Safari at low resolutions currently. Once the WebAssembly threading and SIMD support matures, the higher resolutions should also become playable in browsers without native AV1 support</li><li>4K decoding on an iPad Pro showing that it’s possible to decode a 4K AV1 file even in software!</li><li>1 in 8 videos played in Firefox in July 2019 were AV1!</li></ul><blockquote><em>AV1 live encode and playback demo at </em><a href="https://twitter.com/hashtag/mhv2019?src=hash&amp;ref_src=twsrc%5Etfw"><em>#mhv2019</em></a><em> <br>AV1 in </em><a href="https://twitter.com/firefox?ref_src=twsrc%5Etfw"><em>@firefox</em></a><em> (ARM+x86), </em><a href="https://twitter.com/googlechrome?ref_src=twsrc%5Etfw"><em>@googlechrome</em></a><em>, </em><a href="https://twitter.com/MicrosoftEdge?ref_src=twsrc%5Etfw"><em>@MicrosoftEdge</em></a><em>, Safari and </em><a href="https://twitter.com/videolan?ref_src=twsrc%5Etfw"><em>@videolan</em></a><em> VLC, all using dav1d.<br>Encoded live with </em><a href="https://twitter.com/intel?ref_src=twsrc%5Etfw"><em>@intel</em></a><em> SVT-AV1 in </em><a href="https://twitter.com/DASH_IF?ref_src=twsrc%5Etfw"><em>@DASH_IF</em></a><em> (360p to 1080p) format. </em><a href="https://t.co/pyrj2sXfYv"><em>pic.twitter.com/pyrj2sXfYv</em></a></blockquote><blockquote><em>- VideoLAN (@videolan) </em><a href="https://twitter.com/videolan/status/1156612202423623681?ref_src=twsrc%5Etfw"><em>July 31, 2019</em></a></blockquote><p>For the details of other demos shown, I’ve embedded the presentation below.</p><p>Kevin Staunton-Lambert of Switch Media posted his presentation about working with AV1. It has some useful info on how to use AV1 as well as why AV1 achieves better compression. The slides are embedded below.</p><h3>Xilinx acquires NGCodec</h3><h3>libvips now supports compressing images in AVIF format</h3><blockquote><em>Read the rest of the story @ </em><a href="https://www.singhkays.com/blog/av1-ecosystem-update-july-2019/">https://www.singhkays.com/blog/av1-ecosystem-update-july-2019/</a></blockquote><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=bbded4a66a1" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[AV1 Ecosystem Update: June 2019]]></title>
            <link>https://singhkays.medium.com/av1-ecosystem-update-june-2019-d7c92c4d956?source=rss-122b8af0de61------2</link>
            <guid isPermaLink="false">https://medium.com/p/d7c92c4d956</guid>
            <category><![CDATA[dav1d]]></category>
            <category><![CDATA[twitch]]></category>
            <category><![CDATA[video-encoding]]></category>
            <category><![CDATA[av1]]></category>
            <category><![CDATA[cisco]]></category>
            <dc:creator><![CDATA[Kay Singh]]></dc:creator>
            <pubDate>Mon, 22 Jul 2019 00:00:00 GMT</pubDate>
            <atom:updated>2019-08-30T23:32:15.495Z</atom:updated>
            <content:encoded><![CDATA[<p>It’s been a pretty busy for me professionally which is why this month’s AV1 ecosystem update is coming in pretty late. It’s been one of the most exciting months for AV1 development and a lot to catch up on! With so much news this month, it is starting to feel like we’re at the cusp of something big! ( <em>which we hope AV1 will be</em> )</p><h3>Youtube pauses AV1 rollout</h3><h3>Realtek demonstrates first 8K AV1 Decoder and 4K UHD Set-top Box SoCs</h3><h3>Conferencing!</h3><p>There were two conferences in June, <a href="https://allthingsrtc.org/">AllThingsRTC 2019</a> and then <a href="https://bigapple.video/">Big Apple Video</a>, where a lot of new information about AV1 was shared. We heard updates on the following topics which are all covered later in this update:</p><ul><li>libaom AV1 encoder</li><li>Aurora AV1 encoder</li><li>SVT-AV1 encoder</li><li>rav1e AV1 encoder</li><li>NGCodec FPGA encoder for AV1</li><li>EVE AV1 encoder</li></ul><h3>Past, Present &amp; Future of AV1</h3><h3>Visionular’s Aurora AV1 encoder shows promise</h3><h3>SVT-AV1 v0.6.0 released</h3><h3>Rav1e + Vimeo = ❤️</h3><h3>NGCodec shows off AV1 encoding with their FPGA encoders</h3><h3>Twitch talks about their AV1 adoption plan</h3><h3>EVE AV1 encoder</h3><p>Ronald Bultje, who designed VP9 at Google previously, introduced the world to Efficient Video Encoder (EVE) for AV1 at Big Apple Video. EVE is a commercial encoder that promises to provide the best visual quality amongst the AV1 encoders. Here are some interesting slides from Ronald’s presentation especially the quality comparison versus libaom AV1. The whole presentation can also be viewed below.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*2e7Io9PCvQ84yxnz.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*Nd73dkHCqN1aq6vg.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*sdBHyWhKPxlpmH_B.jpg" /></figure><h3>Cisco’s real-time AV1 video encoder</h3><h3>Millicast’s real-time AV1 video encoder</h3><blockquote><em>Read the rest of the story @ </em><a href="https://www.singhkays.com/blog/av1-ecosystem-update-june-2019/">https://www.singhkays.com/blog/av1-ecosystem-update-june-2019/</a></blockquote><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d7c92c4d956" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[AV1 Ecosystem Update: May 2019]]></title>
            <link>https://singhkays.medium.com/av1-ecosystem-update-may-2019-c39c1e46a121?source=rss-122b8af0de61------2</link>
            <guid isPermaLink="false">https://medium.com/p/c39c1e46a121</guid>
            <category><![CDATA[intel]]></category>
            <category><![CDATA[av1]]></category>
            <category><![CDATA[video-encoding]]></category>
            <category><![CDATA[h265]]></category>
            <category><![CDATA[transcoding]]></category>
            <dc:creator><![CDATA[Kay Singh]]></dc:creator>
            <pubDate>Mon, 10 Jun 2019 00:00:00 GMT</pubDate>
            <atom:updated>2019-08-30T23:28:30.040Z</atom:updated>
            <content:encoded><![CDATA[<p>May was another amazing month for the AV1 codec! We saw great progress in the SVT-AV1 encoder, Android update news and some surprising news out of China!</p><h3>SVT-AV1 is making strides!</h3><h3>Android Q gets AV1 support</h3><h3>Visionular Aurora AV1 codec claims it’s faster and better than x265</h3><p>Another piece of interesting news was the Visionular Aurora AV1 codec that I first read about in <a href="https://mux.com/blog/highlights-livevideostackcon-shanghai-2019/">this summary of LiveVideoStackCon Shanghai 2019</a>. In this presentation it is claimed that Aurora AV1 beats libaom encoder by ~33% and x265 by ~42% on VMAF model.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1023/0*j512ROME0ygRLSQA.png" /></figure><p>From Visionular website, Auorara claims to improve encoding speed by 32.2% against x265 veryslow which is very surprising and the first time I’m hearing about an AV1 encoder being better in quality and faster than x265. It’ll be interesting to see if we find out more info and are able to test this encoder in the coming months.</p><h3>BBC compares AV1 &amp; VVC</h3><h3>Amphion Semiconductor Hardware decoder</h3><blockquote><em>Read the rest of the story @ </em><a href="https://www.singhkays.com/blog/av1-ecosystem-update-may-2019/">https://www.singhkays.com/blog/av1-ecosystem-update-may-2019/</a></blockquote><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c39c1e46a121" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[AV1 Ecosystem Update: April 2019]]></title>
            <link>https://singhkays.medium.com/av1-ecosystem-update-april-2019-2149b673c56a?source=rss-122b8af0de61------2</link>
            <guid isPermaLink="false">https://medium.com/p/2149b673c56a</guid>
            <category><![CDATA[videos]]></category>
            <category><![CDATA[av1]]></category>
            <category><![CDATA[intel]]></category>
            <category><![CDATA[netflix]]></category>
            <category><![CDATA[samsung]]></category>
            <dc:creator><![CDATA[Kay Singh]]></dc:creator>
            <pubDate>Mon, 06 May 2019 00:00:00 GMT</pubDate>
            <atom:updated>2019-08-30T23:25:51.239Z</atom:updated>
            <content:encoded><![CDATA[<p>It’s been an exciting time following the progress of the royalty free video codec AV1 over the past year and April might just be the most exciting active month in the amount of news we got about AV1! It’s not every day that media industry giants such as Netflix, Google, Mozilla, Microsoft, Apple, Amazon, ARM, Facebook, Nvidia, Intel, Cisco and many others agree on the same technology stack!</p><h3>Samsung joins the party</h3><h3>Intel and Netflix are joining forces</h3><p>At NAB 2019, <a href="https://newsroom.intel.com/news/intel-netflix-deliver-av1-scalable-codec-power-next-gen-compression-tech-visual-workloads/">Netflix announced they had selected Intel’s SVT-AV1 encoder</a> for their workflow.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/974/0*E3EHy7WHA8Gmlhyz.png" /></figure><p>Although surprising at first, given the gulf in visual quality between the aomenc encoder vs the SVT-AV1 encoder at the moment, Netflix <a href="https://medium.com/netflix-techblog/introducing-svt-av1-a-scalable-open-source-av1-framework-c726cce3103a">posted a techblog few days later explaining the rationale.</a></p><blockquote>At Netflix, we believe that the AV1 ecosystem would benefit from an alternative clean and efficient open-source encoder implementation. There exists at least one other alternative open-source AV1 encoder, <a href="https://github.com/xiph/rav1e">rav1e</a>. However, rav1e is written in Rust programming language, whereas an encoder written in C has a much broader base of potential developers. The open-source encoder should also enable easy experimentation and a platform for testing new coding tools. Consequently, our requirements to the AV1 software are as follows:</blockquote><ul><li><em>Easy to understand code with a low entry barrier and a test framework</em></li><li><em>Competitive compression efficiency on par with the reference implementation</em></li><li><em>Complete toolset and a decoder implementation sharing common code with the encoder, which simplifies experiments on new coding tools</em></li><li><em>Decreased encoder runtime that enables quicker turn-around when testing new ideas</em></li></ul><h3>MSU Codec Comparison 2018</h3><h3>Mozilla’s NAB 2019 presentation</h3><h3>dav1d 0.3.0 release</h3><h3>AVIF POC and libraries</h3><blockquote>Read the rest of the story @ <a href="https://www.singhkays.com/blog/av1-ecosystem-update-april-2019/">https://www.singhkays.com/blog/av1-ecosystem-update-april-2019/</a></blockquote><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=2149b673c56a" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How to identify boot, temp and data disks in Azure Linux VMs]]></title>
            <link>https://singhkays.medium.com/how-to-identify-boot-temp-and-data-disks-in-azure-linux-vms-d24fb4f0c081?source=rss-122b8af0de61------2</link>
            <guid isPermaLink="false">https://medium.com/p/d24fb4f0c081</guid>
            <category><![CDATA[infrastructure]]></category>
            <category><![CDATA[azure]]></category>
            <category><![CDATA[iaas]]></category>
            <category><![CDATA[cloud-computing]]></category>
            <dc:creator><![CDATA[Kay Singh]]></dc:creator>
            <pubDate>Tue, 30 Apr 2019 00:00:00 GMT</pubDate>
            <atom:updated>2019-08-30T23:18:43.671Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/974/1*Iw21gd7czHvQjvNeahbOxA.png" /></figure><p>In Azure, every VM has a persistent OS Disk that is used for booting the VM and contains the operating system. In addition to an OS disk, each VM also has a temporary disk that is present on the compute cluster hosting your VM.</p><h3>Why is my OS disk not @ /dev/sda?</h3><p>In most cases the OS disk is /dev/sda but in some odd cases it is not. This is not an Azure specific issue but a default Linux behavior. The ordering in which Linux enumerates SCSI devices is nondeterministic. It scans SCSI devices in the order they appear on the PCI Bus. Therefore, whichever disk is enumerated first is /dev/sda.</p><h3>OK, so how to identify my disks?</h3><blockquote>Read the rest of the article on my blog @ <a href="https://www.singhkays.com/blog/how-to-identify-azure-linux-vm-disks/">https://www.singhkays.com/blog/how-to-identify-azure-linux-vm-disks/</a></blockquote><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d24fb4f0c081" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[It’s time to replace GIFs with AV1 video!]]></title>
            <link>https://singhkays.medium.com/its-time-to-replace-gifs-with-av1-video-3001fbfcb0e0?source=rss-122b8af0de61------2</link>
            <guid isPermaLink="false">https://medium.com/p/3001fbfcb0e0</guid>
            <category><![CDATA[gif]]></category>
            <category><![CDATA[video-encoding]]></category>
            <category><![CDATA[av1]]></category>
            <category><![CDATA[web-development]]></category>
            <category><![CDATA[web-design]]></category>
            <dc:creator><![CDATA[Kay Singh]]></dc:creator>
            <pubDate>Sun, 28 Apr 2019 00:00:00 GMT</pubDate>
            <atom:updated>2019-08-30T23:22:41.337Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/779/0*FYferpQUK4ov1VQS.png" /></figure><p>It is 2019 and we need to make a decision about GIFs ( <a href="https://bits.blogs.nytimes.com/2013/05/23/battle-over-gif-pronunciation-erupts/">no, not that one! We’re never going to be able to decide that one!</a>). GIFs take up a massive amount of space<em>(often multiple megabytes!)</em> and if you’re a web developer, then that’s completely against your ethos! As a web developer, you want to minimize the bits your users need to download so that your website loads fast. That’s why you minify javascript, optimize PNGs, JPEGs and sometimes turn <a href="https://developers.google.com/speed/webp/">JPEGs to WebPs as well</a>. But then what about the venerable GIF?</p><h3>Where we are going, we don’t need no GIFs!</h3><p>If your goal is to improve your website’s loading performance, then a GIF needs to be yanked! But then how do you have moving pictures? The answer is a video. And in most cases, you’ll get better quality as well as almost 50–90% size savings! In life, most things have a pro and a con. When you’re replacing a GIF with a video, in majority of the cases you’ll be hard pressed to find a con</p><h3>Replace all the GIFs!</h3><p>Fortunately, replacing GIFs with video has now been common for a few years so the tools needed to pull off this trick are already in play. In this blog, I’m not going to break new ground, but improve upon what already exists. So, here’s the lowdown on what you can currently do:</p><ol><li>Take a GIF and convert it into video</li><li>Encode it with H.264 or VP9 codec i.e. compress it and package in a MP4 of WebM container</li><li>Replace animated GIF &lt;img&gt; elements with &lt;video&gt;</li><li>Set the video to play automatically, sliently and in a loop to replicate the GIF functionality</li></ol><p><a href="https://developers.google.com/web/fundamentals/performance/optimizing-content-efficiency/replace-animated-gifs-with-video/">Google has a good documentation page describing the process.</a></p><h3>It is 2019</h3><p>It’s 2019. Tech moves forward and we need to move with it. Until now you’ve had two choices of codecs that are widely adopted across all browsers and platforms for encoding the video:</p><ol><li><strong>H.264</strong> — Introduced in 2003 and the most widely used codec today</li><li><strong>VP9</strong> — Introduced in 2013 and achieves up to 50% more compression compared to H.264 most of the times but it can look worse sometimes according to <a href="https://www.reddit.com/r/AV1/comments/bjzqly/its_time_to_replace_gifs_with_av1_video/emcohnq/">this discussion on reddit</a>.</li></ol><blockquote><strong><em>NOTE:</em></strong><em> Even though H.265 is the next-gen version of H.264 and competes with VP9, I’m not including it in this comparison becuase of lack of browser support as noted on </em><a href="https://caniuse.com/#feat=hevc"><em>https://caniuse.com/#feat=hevc</em></a><em>. Licensing costs is the major reason why H.265 is not as widely adopted as H.264 and why Alliance of Open Media consortium is working a royalty-free codec i.e. AV1.</em></blockquote><p>Remember, our goal is to reduce the humongous GIF to as small a size as possible to improve our webpage loading time. It wouldn’t be 2019 if we didn’t have a new video compression standard we could encode to. This is called AV1. With AV1, we are able to <a href="http://compression.ru/video/codec_comparison/hevc_2017/MSU_HEVC_comparison_2017_P5_HQ_encoders.pdf">achieve ~30% more compression compared to VP9.</a> Exciting! :)</p><h3>Serving you AV1 since 2019!</h3><blockquote>Read the rest of the story @ <a href="https://www.singhkays.com/blog/its-time-replace-gifs-with-av1-video/">https://www.singhkays.com/blog/its-time-replace-gifs-with-av1-video/</a></blockquote><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3001fbfcb0e0" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>