<?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 Akhilesh Agarwal on Medium]]></title>
        <description><![CDATA[Stories by Akhilesh Agarwal on Medium]]></description>
        <link>https://medium.com/@akhiagarwal?source=rss-61c2680dc833------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*XSrS1F8U5cTN0P3Ggbi7kQ.jpeg</url>
            <title>Stories by Akhilesh Agarwal on Medium</title>
            <link>https://medium.com/@akhiagarwal?source=rss-61c2680dc833------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Thu, 19 Nov 2026 12:13:47 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@akhiagarwal/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[A New Paradigm of Computing: Quantum]]></title>
            <link>https://akhiagarwal.medium.com/a-new-paradigm-of-computing-quantum-988179995792?source=rss-61c2680dc833------2</link>
            <guid isPermaLink="false">https://medium.com/p/988179995792</guid>
            <dc:creator><![CDATA[Akhilesh Agarwal]]></dc:creator>
            <pubDate>Wed, 29 Jan 2025 10:08:18 GMT</pubDate>
            <atom:updated>2025-01-29T10:08:18.681Z</atom:updated>
            <content:encoded><![CDATA[<p>In the ever-evolving field of quantum computing, a new information processing paradigm is emerging, promising to revolutionize computation as we know it. Whether you’re a startup founder, investor, or industry professional looking to learn more about quantum computing, this article delves into quantum computing technology, and a different way to represent, store, and process information.</p><h3>Quantum Information: A New Frontier</h3><p>At the heart of quantum computing lies the qubit, the quantum counterpart to the classical bit. Unlike classical bits, which can only be in one of two states (0 or 1), qubits can exist in multiple states simultaneously, thanks to the principles of quantum mechanics. A physical system — atom, photon, electron, ion, circuit represents a qubit. Quantum information is encoded in the physical properties of these systems such as</p><ul><li>Energy and electronic states of atoms</li><li>The arrival time of photon and the relative phase difference between two light paths</li><li>Electron and nuclear spin states (up or down)</li></ul><h3>Technology Value Chain of Quantum Computing</h3><p><strong><em>Hardware and Components</em></strong></p><p>Qubit states are manipulated and encoded with information by the following tech:</p><ol><li>Lasers: Control lasers to manipulate qubit states and are used with trapped-ion qubits</li><li>Optical Tweezers: Trap and arrange atoms in a reconfigurable array and are used with neutral atom qubits</li><li>Beam Splitters and Phase Shifters: Enable Quantum gates by controlling light interference and are used with photonic qubits</li><li>Electron Spin Resonance (ESR) Spectrometers: Apply radio frequency pulses to manipulate spin states and are used with spin qubits</li></ol><p>These are some physical components that read out, retrieve, or decode the information stored in qubits (electron, atom, etc)</p><ol><li>Single-Photon Detectors: They detect individual photons to retrieve quantum states and are used with photonic qubits</li><li>Microwave Resonators: Control and read qubit states using microwave signals and are used with superconducting qubits</li><li>Fluorescence Detectors: measure quantum states by detecting emitted photons and are used with Trapped-ion qubits</li><li>Josephson Parametric Amplifiers (JPAs): Amplify weak quantum signals for accurate readout and are used with superconducting qubits</li></ol><p><strong><em>System software</em></strong></p><p>Low-level software that manages quantum hardware, enables and optimizes quantum computations. Some crucial functions include:</p><ul><li>Hardware control: Interfaces with quantum processors</li><li>Optimization: Enhances quantum circuit execution</li><li>Resource management: Manages quantum memory and operations</li><li>Middleware: Bridges hardware and application-level software</li></ul><p>Users: Developers, researchers, quantum hardware companies</p><p>Example companies: Q-CTRL, IBM (Qiskit Runtime), HAIQU, Classiq</p><p><strong><em>Application software</em></strong></p><p>End-user applications and quantum algorithms tailored for solving industry-specific problems. Some characteristics include solving real-world problems using quantum algorithms and running on top of system software.</p><p>Users: Enterprises, industry professionals, analysts</p><p>Example applications: Finance, cryptography, molecular simulation, lifesciences</p><p>Example companies: Multiverse Computing, QC Ware, PQShield</p><p><strong><em>Cloud services — </em></strong>Quantum cloud services provide remote access to quantum computing power, allowing users to leverage these advanced systems without significant hardware investments. Customers can rent computational quantum power from a distance and for any amount of time needed. Startups run application software used to solve industry problems on these cloud services to allow customers to solve complex problems using the algorithms on their software by accessing the hardware power remotely.</p><h3>Compare Performance of Quantum Computing Systems:</h3><ul><li>Coherence time: duration a qubit retains quantum information before decoherence disrupts its state. Longer coherence time (~ms to sec) allows more complex computations before errors start occurring.</li><li>Gate fidelity: the probability that gate control mechanisms (known as quantum gates) execute correctly without any errors. High fidelity (&gt;99.9%) ensures accurate quantum operations with low errors.</li><li>Scalability: measures the ability to increase the qubit counts while maintaining control, coherence, and error rates at optimal levels. Easily interconnected qubits are highly scalable, allowing practical fault-tolerant quantum computing.</li><li>Error rates: show us the likelihood of a quantum operation deviating from the expected result due to decoherence and control imperfections. Lower error rates (&lt;1%) reduce the amount of resources needed to mitigate errors.</li><li>Lower fabrication complexity and room-temperature and pressure operating conditions also ensure easier manufacturing and greater reliability in the real world.</li></ul><p>Companies can consider the above metrics while choosing the category of quantum computing system (superconducting, photonic, etc) to innovate upon.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=988179995792" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Navigating the Silicon Frontier: Fabless Semiconductor Startups]]></title>
            <link>https://akhiagarwal.medium.com/navigating-the-silicon-frontier-fabless-semiconductor-startups-c3ec6690d88c?source=rss-61c2680dc833------2</link>
            <guid isPermaLink="false">https://medium.com/p/c3ec6690d88c</guid>
            <category><![CDATA[semiconductors]]></category>
            <category><![CDATA[deeptech]]></category>
            <dc:creator><![CDATA[Akhilesh Agarwal]]></dc:creator>
            <pubDate>Thu, 19 Dec 2024 05:48:07 GMT</pubDate>
            <atom:updated>2024-12-19T05:48:07.123Z</atom:updated>
            <content:encoded><![CDATA[<p>In today’s semiconductor landscape, fabless startups face unique challenges in bringing their innovations to market. This commentary is to put my learning into words and explore proven go-to-market (GTM) strategies, from product development approaches to partnership models. Whether you’re a startup founder, investor, or industry professional, I hope you’ll discover actionable insights for successfully navigating the silicon frontier.</p><h3>The Silicon World Challenge</h3><p>Fabless semiconductor companies face a classic chicken-and-egg dilemma: the entire solution cannot be built until the chip is ready. To overcome this, successful startups employ three approaches:</p><p><strong>Technology at the chip level (blueprint for the product)</strong></p><p>This involves designing the chip’s architecture and defining its functionalities, performance specifications, and integration capabilities. Startups can focus on creating chip designs using HDLs and simulation tools. They may develop prototypes using FPGAs, and use simulations and emulation platforms to validate designs according to customer requirements, before committing to costly fabrication processes.</p><p>For designing a new AI accelerator chip — developing the chip’s architecture, specifying how it processes data and interfaces with other components.</p><p><strong>Product at the board level</strong></p><p>This stage involves integration using PCBs / MCMs / FPCs / ICs to create a functional module or system.</p><p>After designing the AI accelerator chip — create an integrated system that includes memory, power management, and interface connectors, allowing developers to test and integrate the chip into larger systems.</p><p><strong>Solution for the end-user or customer</strong></p><p>This final phase focuses on delivering a complete product or solution that addresses the customer’s needs, incorporating the developed hardware and necessary software.</p><p>Packaging the AI accelerator board with SDKs, documentation, and support services, providing customers with a ready-to-use solution for AI applications.</p><h3>Choosing Your Path: Off-the-Shelf vs. Custom Solutions</h3><p><strong>Off-the-Shelf Approach</strong></p><p><em>What It Means:</em></p><ul><li>The OTS approach involves creating products that can serve multiple customers across a sector without significant customization.</li><li>Products are designed to address common needs within an industry, such as automotive, consumer electronics, or telecommunications</li></ul><p><em>Strategy:</em></p><ul><li>Leverage modular designs and core functionalities applicable to a wide array of applications</li><li>Focus on scalable solutions to optimize manufacturing and distribution efficiency</li><li>Use prototypes and developer kits to demonstrate usability and validate with potential customers before mass production</li><li>Focusing on differentiation through improved performance, lower power consumption, or additional functionalities based on design capabilities</li><li>Exploring cross-selling opportunities within the same OEM’s portfolio. A startup that supplies infotainment processors could cross-sell ADAS chips for the same vehicle platform</li></ul><p><em>Benefits:</em></p><ul><li>Wider Market Appeal: By targeting a broad customer base, startups can achieve rapid market penetration</li><li>Economies of Scale: Producing high volumes reduces per-unit costs, improving profitability and market competitiveness</li><li>Incremental Innovation: Continuous small improvements allow startups to stay ahead without drastic rework or investment</li></ul><p><strong>Custom Approach</strong></p><p><em>What It Means:</em></p><ul><li>The focus is on developing highly tailored solutions in partnership with large players (e.g. Tier-1 OEMs)</li><li>Products are designed exclusively for specific use cases or clients, ensuring a perfect fit for the customer’s requirements</li></ul><p><em>Strategy:</em></p><ul><li>Collaborate with customers during the early design stages, such as block diagram and architecture definition</li><li>Develop unique features that differentiate the client’s end product, such as specialized AI inference for a healthcare device</li></ul><p><em>Benefits:</em></p><ul><li>Deep Customer Relationships: Close collaboration strengthens trust and long-term business prospects</li><li>Higher Margins: Custom solutions often command premium pricing due to their tailored nature</li><li>Strategic Market Positioning: Being a key supplier to major clients builds reputation and credibility</li><li>Possibility of non-exclusive partnerships with multiple players, with certain limitations, allowing startups to mitigate risk while serving multiple clients simultaneously</li><li>Custom projects often have longer sales cycles but result in larger contracts</li></ul><h3>Building a Comprehensive Product Portfolio</h3><p>To attract a broader client base, early fabless companies can focus on creating diverse product offerings. Below is an explanation of some strategies and their benefits:</p><p>Develop multiple SKUs or variants of a base product — creating a modular product design helps achieve variability in offerings while maintaining core design efficiency</p><ul><li>Expand market reach by catering to diverse customer segments (e.g. high-performance vs. cost-sensitive markets)</li><li>Allows for easier scaling of production while reusing the same core design, reducing costs and time-to-market</li></ul><p>Create a range of products with incremental features or removals</p><ul><li>Supports tiered pricing models, enabling customers to select a product that aligns with their budget and needs</li><li>Facilitates upselling as customers scale their applications and require more advanced features</li></ul><p>Allow end customers to choose based on their system configurations</p><ul><li>Increases compatibility with a wide range of devices and applications</li><li>Builds customer loyalty by allowing seamless integration into their systems</li></ul><p>Digital twins can simulate customer-specific configurations, reducing development time</p><ul><li>Reduces development time and cost by identifying issues early in the design phase</li><li>Enhances customer engagement by allowing them to see how a product will perform in their environment</li><li>Xilinx provides adaptable FPGA solutions across industries, enabling customer choice</li></ul><p>FPGAs are like flexible Swiss Army knives for technology. Modular design lets you build different tools using the same base, and digital twins help you test and perfect the tools virtually before building them.</p><h3>Business Approaches</h3><p><strong>Software-Centric Hardware Ecosystems</strong></p><p>Build chips tailored for specific software frameworks, enabling better performance and ease of use. Provide end-to-end solutions, including SDKs, compilers, and application software.</p><p><strong>Hardware-as-a-Service (HaaS)</strong></p><p>Offering hardware as part of a subscription or usage-based model, bundling software for value addition.</p><p><strong>Cost Leadership Through Economies of Scale</strong></p><p>Focusing on reducing per-unit production costs through high-volume manufacturing. Partner with leading foundries to optimize fabrication costs. Target broad markets to achieve high production volumes.</p><h3>Leveraging Partnerships for Market Reach</h3><p>Partnerships are a vital component of the GTM strategy, enabling broader customer access and operational efficiency. Below is a deeper dive into the key partnership models, how they work, and how they benefit startups:</p><p><strong>Distributors and Sales Representatives</strong></p><p>Fabless startups often partner with chip distributors to reach end-users. These distributors employ sales representatives and field application engineers to support customers.</p><p><em>How does this help?</em></p><ul><li>Distributors already have an established network, allowing startups to access markets quickly without building infrastructure</li><li>Startups save on hiring and training in-house sales and support teams</li><li>FAEs act as an extension of the startup’s technical team, ensuring product adoption and customer satisfaction</li></ul><p><strong>Design Houses and ODMs/OEMs</strong></p><p>End-users can outsource their needs to design houses for full solution execution. ODMs and OEMs then pick up these designs for large-scale manufacturing.</p><p><em>How does this help?</em></p><ul><li>End-to-End Solutions: Design houses allow startups to focus on core chip development while outsourcing the integration and productization process</li><li>Faster Time-to-Market: ODMs/OEMs handle manufacturing and branding, enabling quicker product deployment</li><li>Scalability: OEMs produce at volumes startups cannot achieve independently, making it possible to fulfill large customer orders</li></ul><p><strong>Specialized Distribution Channels</strong></p><p>For niche areas, consider working with specialized sales representatives in addition to general distributors. This is particularly important for markets outside one’s home country.</p><p><em>How does this help?</em></p><ul><li>Specialized distributors bring technical know-how, helping startups tailor their chips to meet specific industry standards</li><li>For international markets, specialized representatives ensure that the startup complies with local laws, certifications, and consumer needs</li></ul><h3>Showcasing Your Technology</h3><p><strong>Reference Designs and Demonstration Kits</strong></p><p>R&amp;D teams create reference designs for various application use cases, incorporating all necessary components on a board to demonstrate the chip’s functionality. This is particularly useful for microcontrollers with features such as Bluetooth capability, touch-sensitive interfaces, or AI compatibility.</p><p><strong>Evaluation Kits</strong></p><p>The progression of product development typically follows this path:</p><ol><li>Reference design (initial concept)</li><li>Demo kit (prototype)</li><li>Evaluation kit (customer-ready, bug-free version)</li></ol><p>Evaluation kits are crucial for allowing customers to test the product independently. Customers using evaluation kits often provide valuable insights, helping startups refine their chips or develop additional features.</p><p><strong>Developer-First Approach</strong></p><p>Prioritizing developer engagement by offering robust tools and fostering a collaborative community. This builds an ecosystem around the startup’s chips, increasing their value and long-term market presence.</p><p>SiFive promotes the adoption of RISC-V architecture by providing an open-source software stack along with developer boards and design kits.</p><h3>Market Entry Strategies</h3><p><strong>Multi-Project Wafer (MPW) Approach</strong></p><p>A model where multiple chip designs from different startups or companies are fabricated on the same wafer to produce initial sample units for demo kits.</p><p>Wafer fabrication is one of the highest costs in chip production. MPW allows startups to produce prototypes without committing to a full wafer, saving significant costs.</p><p><strong>Customer-Centric Development</strong></p><p>For established relationships, propose new products based on customer needs, even at the block diagram stage.</p><ul><li>By aligning the product directly with customer needs, startups ensure there is a demand for their offering before fabrication.</li><li>Customers involved in development are more likely to commit to long-term contracts, providing a stable revenue stream.</li><li>Tailoring chips to specific use cases reduces customer integration challenges, accelerating adoption.</li></ul><p>Synopsys collaborates with clients during the block diagram stage to refine ASIC designs.</p><h3>Exploring IP-Based Business Models</h3><p>Many successful companies in the semiconductor space focus on selling intellectual property (IP). Examples include ARM (software IP) and SiFive (RISC-V IP).</p><p>Emerging Trends and Opportunities</p><ol><li>ARM’s Growing Dominance: ARM-based solutions are expanding beyond mobile devices into desktop and server markets.</li><li>RISC-V Potential: Open-source RISC-V architecture is gaining traction, with Chinese players developing RISC-V-based CPUs.</li><li>Open Source Movement: There’s growing support for open-source hardware, including fully open-sourced laptops with RISC-V CPUs.</li></ol><p>Imagination Technologies, known for its GPU IP, licenses technology to major players like Apple.</p><h3>Staying Agile: The Importance of Pivoting</h3><p>Early-stage fabless companies may need to pivot their focus based on learnings and dynamic market needs:</p><ul><li>Adjusting target client segments based on market feedback</li><li>Exploring new applications for their core technology</li><li>Adapting product offerings to meet emerging market needs</li></ul><p>Cerebras, initially targeting supercomputer builders for AI training, pivoted to offer cloud-based AI inference solutions.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c3ec6690d88c" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Fusion and a Journey to Type 1]]></title>
            <link>https://akhiagarwal.medium.com/fusion-and-a-journey-to-type-1-dd3ddfcc6def?source=rss-61c2680dc833------2</link>
            <guid isPermaLink="false">https://medium.com/p/dd3ddfcc6def</guid>
            <category><![CDATA[startup]]></category>
            <category><![CDATA[nuclear-fusion]]></category>
            <category><![CDATA[venture-capital]]></category>
            <category><![CDATA[deeptech]]></category>
            <dc:creator><![CDATA[Akhilesh Agarwal]]></dc:creator>
            <pubDate>Tue, 02 Jul 2024 15:25:23 GMT</pubDate>
            <atom:updated>2024-07-02T15:25:23.035Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*RJrGFZjgSR8N3_HP5EVAJw.png" /><figcaption><em>Generated by Playground</em></figcaption></figure><p>Imagine if humanity could terraform other planets and moons in the solar system or develop interstellar travel capabilities to explore our starry neighbors. It would be understood that we would be able to construct a Dyson Sphere to harness and store 100% of the energy emitted by the Sun (10²⁶ Watts) to build other megastructures or to control the life cycles of stars in our system.</p><p>If this were the case, Kardashev would call us a Type 2 civilization. To put this into perspective, we aren’t even a Type 1 civilization; if we were, we would capture all of the energy that the Earth receives from the Sun, and harness the energy of the oceans and the winds. We would be able to control natural disasters to a degree that they would be called controlled natural events. We would find a way to harness the energy of these occurrences too. Humanity’s power consumption would be 10¹⁶ Watts. A typical work laptop uses 50–100 Watts of power.</p><p>We are currently a Type 0 civilization. To be precise, a Type 0.7276 civilization that has just moved on from utilizing coal as the primary energy source to harnessing renewable sources of energy that supply ~30% of global requirements. To become a Type 1 civilization we would need to capture 100% of ALL of Earth’s power — the sunlight received, oceans, winds, and the core. It might be safe to say we are a few centuries away.</p><p>Crystal ball about 100,000 years into the future and Michio Kaku might tell you that we would be progressing towards becoming a Type 3 civilization — capable of capturing, storing, and utilizing the power of everything in the Milky Way, from a bunny to a black hole. Humanity’s power consumption would be 10³⁷ Watts.</p><p>What can we do to start realizing the science within the fiction of it all and move towards building a Type 1 civilization? The answer could be to build towards seemingly limitless energy solutions within the next 2 decades. The first step is to realize sustained nuclear fusion. We currently consume ~25x10¹² Watts / year across all sources of energy. What do we need to do to potentially improve this by a factor of ~100,000?</p><p>In the coming sections, we will explore what nuclear fusion is, some key challenges to overcome, and finally, what all of this means for our civilization.</p><h3>Part 1: Understanding Nuclear Fusion</h3><p>Fusion is bashing two heavy isotopes of Hydrogen (deuterium &amp; tritium or D-T) together to release energy. When the nuclei of these atoms fuse, a tiny amount of the total mass gets converted into a large amount of energy, following Einstein’s famous equation, E = mc².</p><p>The sunlight we get is an outcome of this process — Fusion powers the stars such as our Sun, where the extreme gravitational pressure and temperatures in their cores allow fusion reactions to occur continuously. To recreate this on Earth, we would need to heat the Hydrogen atoms to at least 100m<strong>°</strong>C and up to 1bn<strong>°</strong>C for other types of fuels (proton-boron-11). At this point, the kinetic energy of the nuclei is high enough to overcome the electrostatic repulsion between them and ‘fuse’ or bind together, releasing Helium 4, neutrons, and their kinetic energies.</p><p><strong>The promise of fusion</strong></p><p>Just 1 gram of D-T fuel could provide 9x10⁷ Watts of energy which is the combustion heat of 11,000 kg of coal.</p><p>Fusion promises to be a revolutionary energy source that is environmentally sustainable, virtually inexhaustible, safe, and capable of meeting Earth’s growing energy demands. We are at least a couple of decades and significant scientific and engineering challenges away from making commercial fusion power a reality.</p><p><strong>Where could controlled energy have started?</strong></p><p>One of our ancestors, homo erectus, perhaps sat looking in the distance and observed a wildfire about a million years ago. In its wake, the fire consumed the forest, charred and killed plants and animals alike. Homo erectus may have scavenged food from the burned remains of the natural fires and decided they liked the taste of cooked meat, perhaps sparking a curiosity. A curiosity to control fire.</p><p>Once homo erectus learned to control it, it changed the way she lived. Fire provided a source of warmth and lighting, protection from predators (especially at night), a way to create more advanced hunting tools and a method for cooking food. Over time, human civilization spread geographically, and cooking improved nutrient intake which perhaps increased their brain size and changed behavior.</p><p>The mastery of fire was one of the most pivotal technological advances for early humans. However, there is one fundamental building block that remains common between a million years and today — the need for more energy and the potential it hides for our civilization to unlock the next mega-technological advancements.</p><p>Next, let’s look at some fundamental challenges standing in our civilization’s way to achieving this milestone.</p><h3>Part 2: Challenges to overcome</h3><p>The conquest of fusion is not a victory that will be realized by any one entity alone. Several organizations will have to come together to the fray and solve individual scientific and engineering challenges that exist. And that too, to get a first working prototype in place. Making it commercializable and then ensuring safe use is going to be a whole other battle.</p><p><strong>The Challenge of Fundamental Plasma Physics</strong></p><p>Once homo erectus figured out that fire could be created by rubbing two dry sticks together or by creating sparks by smashing two pieces of stone together, the rest was history (given this might be a million years ago, perhaps it might be more appropriate to say ancient-history). She no longer needed to keep a fire going by continuously adding wood fuel to a burning flame. Homo erectus could control and create fires at will.</p><p>Solving the challenges of Plasma Physics is something akin to what homo erectus did with fire — once we figure out the fiction behind fusion, it can be made into a repeatable process, albeit initially a crude one to harness the fabled limitless energy.</p><ol><li><strong>Plasma containment and stability: Rocket exhaust in a glass container</strong></li></ol><p>Imagine the hot gas exhaust from a SpaceX Starship’s rocket engine that produces a thrust of 14,700kN or ~1,500,000 kgF. Now let’s go a step further and imagine holding all of this inside a sealed glass container. The gas being extremely hot and full of kinetic energy would likely leak by punching holes in the glass container.</p><p>The plasma in a fusion reactor creates similar problems to the hot gas and presents a key challenge in its containment and manipulation. So homo sapiens came up with an ‘invisible’ container to confine and control the plasma — by using strong magnetic fields. However, there continue to exist several challenges in maintaining the superconductivity of the materials producing the magnetic fields, addressing hot spots in the plasma stream, and containing the high energy plasma at extremely high temperatures among other issues.</p><p><strong>2. Achieving ignition: Losing the match</strong></p><p>Homo erectus sat witnessing the natural fires caused by lightning strikes or lava, arousing curiosity about fire’s power. However, intentionally starting and controlling it was an immense challenge. She had to find or create a way to start the fire, something that could generate enough heat to ignite a fuel source like dry wood. She eventually figured that rubbing sticks together or striking stones could cause an ignition. But dry wood or grass was hard to light with a few sparks, so erectus figured out that dried leaves, pine needles, and grass made for good kindling materials. The challenge then remained to keep the fire going by adding more fuel — too little fuel and the fire would die out, too much and it could go out of control and become a dangerous wildfire. Then, they built structures around the fire to protect it from the wind and rain.</p><p>Today instead of sticks and stones, scientists and researchers experiment with tools such as lasers, microwaves, or particle beams that can provide extremely high temperatures (to reiterate, 100m°C+). Instead of dry wood, sapiens are working on stable D-T fuel to produce a self-sustained reaction. To ignite the self-sustained fusion reaction, the energy to heat up the plasma will have to come from within the reactor. The Helium 4 nuclei that was mentioned earlier hold the potential to be this igniter.</p><p><strong>3. Burning Plasma Physics: Wildfire</strong></p><p>The sun might have been directly overhead when erectus noticed something peculiar — subtle changes in wind caused the raging wildfire to drastically change directions. The size of the fuel, the density of the air, humidity, terrain, and topography are a few of the complex factors that influenced how the wildfire behaved.</p><p>How Helium 4 nuclei energize and influence plasma behavior is far more complex and non-linear. The plasma can either burn steadily and create extremely high outputs of energy or simply extinguish because of subtle instabilities.</p><p>As you guessed, these are tough challenges to overcome within the science of plasma physics. But we wouldn’t be here if that’s what homo erectus thought about fire — she decided to push ahead. Although a significantly smaller challenge than solving fusion, homo erectus was not as intelligent as sapiens are, and figuring out fire took time.</p><p>Luckily for erectus, there were multiple ways to start a fire.</p><p><strong>The Challenge of Reactor Systems and Components</strong></p><p>As the 19th century progressed, humans constructed more complex structures to use the heat from fire in increasingly innovative ways. One such structure, created by Babcock and Wilcox (B&amp;W), laid the foundation for modern steam boilers.</p><p>B&amp;W’s new design overcame material challenges to withstand high pressure and temperatures, and introduced water tubes to prevent explosions, making the structure safer; with continuous design improvements efficiency improved, multiple components such as drums, tubes, furnaces, and control systems were integrated into a cohesive system and efficient systems were developed to manage coal feeding, combustion and by-product removal.</p><p>The challenges to solve in fusion are similar but are an orbital shift in complexity. Developing materials to withstand extreme pressures and temperatures, maintaining long-term reliability, and ensuring safety, all while keeping the system running continuously and efficiently might just prove to be the toughest problem to solve in commoditized energy in the 22nd Century.</p><ol><li><strong>Fuel cycle: Going closed loop</strong></li></ol><p>A factory worker during the Industrial Revolution likely needed to procure raw materials, process them, feed them into machines, use coal-powered steam engines to start the manufacturing process, and then deal with waste and byproducts left behind. All this while, he must have had to maintain a steady flow to ensure machines were supplied with materials for continuous production.</p><p>The fuel cycle in a fusion reactor would need to function in a very similar way to how this worker ran the factory. Once the deuterium and lithium raw materials are acquired, the reactor will produce tritium combined with deuterium to create the fuel. The D-T would be injected into the reactor and ignited, the byproducts would regenerate fresh tritium and be reintroduced to the fusion cycle in a closed loop.</p><p>Undeniably, each of these steps presents immense challenges that a factory worker during the 19th century would likely not have faced.</p><p><strong>2. Reactor design and integration: Doughnut assembly</strong></p><p>Fast forward to the 21st century and one of the modern engineering marvels has been building and perfecting the GE9X jet engine used in Boeing 777x aircraft. The journey to conceptualize, design, and build a complex engine capable of delivering a thrust of up to 510kN was unimaginably hard — the engineers would have to source components designed by separate teams across the world, right from the engine casing, to the choice of material used for the turbines, the fuel-air mixing within the combustion chamber, spray pattern of the fuel injectors, and several other issues of other auxiliary systems.</p><p>A fusion reactor like a tokamak has countless subsystems, each requiring the love and care that was given to the GE9X jet engine. And to make things harder — it has to work at temperatures around 100m°C seamlessly. The challenges in conceptualizing and building the magnetic coils, the vacuum vessel, the plasma heating systems, the fuel injectors, and possibly a tritium breeder all the while being bombarded with high-energy particles are daunting.</p><p><strong>3. Materials: Building reliability</strong></p><p>Let’s go back to the 19th century and the plight of an engineer during the Industrial Revolution. The engineer was likely tasked with building a steam boiler for a new steam engine. The immediate challenge that the engineer faced was to make the boiler such that it didn’t fail while containing high-pressure steam. They improved on designs by utilizing wrought iron and steel — materials that could handle much higher temperatures and pressures. Eventually, the aim became to have the boilers operate reliably for long periods without frequent maintenance or worse, explosions.</p><p>In the case of fusion, the temperatures and pressures are scaled up by a factor of hundreds of thousands. The materials would have to withstand temperatures of the Sun, intense radiation, and high pressure.</p><p><strong>The Challenge of Commercialization</strong></p><p>Generations of inventors worked on the design of steam boilers until B&amp;W kickstarted the steam-generating industry with the convection boiler. Even then, the boilers were fairly small, used lump coal, and were fired by hand.</p><p>Now, modern steam boilers can operate at 92–95% efficiency, compared to 40–60% efficiency of early steam boilers. From the outset, fusion has the promise to be able to deliver many times the output per gram of fuel used once ignition and self-reliability are achieved.</p><ol><li><strong>Energy Input vs. Output</strong></li></ol><p>Starting a fire with wet wood might be fairly hard, you’d imagine! One would likely need to use a blowtorch to set the wood on fire which means that at first, a lot more energy would go into getting the fire going than would be received as heat and light from the fire.</p><p>Fusion faces similar challenges — it requires enormous initial energy input to heat the D-T fuel to extremely high temperatures to overcome the repulsive forces between the positively charged nuclei to get the fusion reaction going. The challenge then is to maintain fusion for more than a few seconds in a self-sustained loop to achieve net energy.</p><p>Although possibly a decade away, the ITER project aims to produce 500MW of fusion power from 50MW of input power, theoretically achieving an energy gain of 10.</p><p><strong>2. Logistical</strong></p><p>Once the science and engineering hurdles are overcome, incredibly high costs (in 2018, Reuters reported the estimated cost for ITER’s fusion reactor to be ~€20bn), long timelines (decades!), and specialized supply chains for materials and components that are not widely available pose the next challenge for fusions entry on the commercial energy scene.</p><p>It took almost 200 years since the first practical steam boiler was invented to make it efficient and safe for reliable commercial use! Perhaps with cutting-edge research globally and with scientists collaborating to share knowledge the timeline for practical commercial fusion can be squeezed significantly.</p><p><strong>The Challenge of Safety</strong></p><p>Containing plasma at 100m<strong>°</strong>C, managing radioactive materials, preventing dust explosions, and developing robotic systems for remote maintenance are but a few of the challenges that arise when designing and building safe fusion reactors.</p><p>Imagine trying to contain the Sun in a small box on Earth — effectively that is what each fusion reactor is! A Sun in a box. Super-hot plasma is contained using magnetic fields since no physical material can withstand such extreme temperatures. Albeit lesser than fission, fusion produces radioactive waste hence solving for any leaks and managing tritium (radioactive isotope of Hydrogen) becomes imperative.</p><p>A larger concern looms as the development of various subsystems for fusion could also be used for weapons programs.</p><h3>Part 3: Conclusion &amp; Outlook</h3><p>The path to achieving practical and sustainable nuclear fusion energy is fraught with unique challenges. Addressing plasma containment and stability, ensuring net energy gain, developing durable materials, managing costs, integrating advanced technologies, ensuring safety, and mitigating proliferation risks are all critical areas that require focused research and development.</p><p>Any one organization will almost certainly not be able to achieve this vision alone though. Various innovations will have to come together and fit like an awkwardly cut jigsaw puzzle at first. However, with iterative refinement, the combined efforts of scientific institutions, private companies, and international collaborations will eventually pave the way for efficient fusion systems and deliver net energy gain.</p><p>Therein also lies the opportunity for early-stage venture investments to propel startups developing individual subsystems to a new orbit and realize one of the building blocks to becoming a Type 1 civilization.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=dd3ddfcc6def" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Paralysis and my Recovery from it.]]></title>
            <link>https://akhiagarwal.medium.com/paralysis-and-my-recovery-from-it-f776288f2a15?source=rss-61c2680dc833------2</link>
            <guid isPermaLink="false">https://medium.com/p/f776288f2a15</guid>
            <category><![CDATA[india]]></category>
            <category><![CDATA[health]]></category>
            <dc:creator><![CDATA[Akhilesh Agarwal]]></dc:creator>
            <pubDate>Tue, 19 Nov 2019 08:30:14 GMT</pubDate>
            <atom:updated>2025-02-03T16:52:27.180Z</atom:updated>
            <content:encoded><![CDATA[<h3>GBS and my recovery from ‘paralysis’</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/750/1*R9A_nT5yXv-iGIzVPaSEmg.jpeg" /><figcaption>You have less than 0.0001% chance of getting GBS</figcaption></figure><p>“…we may have to put him on the ventilator soon. His breathing is slowing and we don’t want to take chances” I heard the neurologist speak softly to my mother.</p><p>I had eaten some improperly cooked meat from a soup joint about a week ago, soon after which I was down with a throat infection chased by a stomach infection. I started a course of antibiotics upon my physician’s advice but didn’t complete the course — I was prescribed a 5 day course but I took meds during the first 4 days and stopped since I felt well again. This was literally the biggest mistake of my life. Turns out that the antibiotics reacted with the infection because I didn’t have enough of the antibiotics in my body to kill the infection completely.</p><p>The result was that I was paralyzed by a virus known as GBS (Guillain Barre Syndrome, with a chance lower than a 0.0001% of getting it) — I lost 20 kgs over the course of a night and the following day as the muscle tissue in my entire body was wiped out. The panel of four doctors looking after me were of the opinion that my nervous system was damaged and the myelin in my nervous system was all but eroded.</p><p>Let’s back up a bit — what is myelin and why is it important? Imagine your nervous system to be wires that transmit information in your body. Every traditional wire that we know has some form of insulation. Myelin is the insulation to nerves and it has the following primary functions:</p><ol><li>Insulating the nerve and protecting it from damage. Damaged nerves do not transmit electrical impulses effectively. This is especially important since nerve fibres do not regenerate themselves (irreparable damage)</li><li>Increasing the rate of information transfer in the body. In simple terms — cognitive functions suffer</li></ol><p>So it is clear that the insulating layer of myelin is very important to the CNS (Central Nervous System) and the PNS (Peripheral Nervous System). My body was void of 70% of the myelin according to my panel.</p><p>And I felt it, too. I couldn’t move a muscle — I was a vegetable, paralyzed in a way. I was lucky that my mother was around to rush me to the hospital at the time that she did. The doctor on duty took a look at me and knew what my condition was and said that he had never seen it progress this quickly, overnight.</p><p>Well, the first time I ever blacked-out was in the car from home to the hospital where we feared the worst as my body had completely stopped responding to any signals that my brain was trying to send it. I was only able to move my eyeballs. I was hoisted out from the car and wheeled into the emergency ward where my condition was diagnosed.</p><p>The doctor had a grim look and suggested that I get admitted immediately. Later that day, I had drips going in to three difference places in my body and calls had gone out to family members.</p><p>The second night in hospital was the worst night of my life. I had never felt such terrible pain before — I could feel the virus crawl up my spinal cord, reaching further up to my body. A feature of the GBS virus is that you lose control of the distal parts of your body like your arms and legs after which it progresses inwards. But in my case the progression was very quick, almost overnight.</p><p>After the night of excruciating pain, my lungs were close to giving up — which is when a panel neurologist suggested that I be moved to the ventilator in the ICU. Not something a parent wants to hear at 2AM but this was the reality. I couldn’t see my parents in any more distress than they already were in so I forced my lungs back to work. How did I do that? I spent the next 6 hours just focussing on my breathing and consciously tried to not fall asleep. I knew if I fell asleep I’d probably wake up in the ICU on ventilator.</p><p>Well, my hard work had paid off through the night and my lungs were showing signs of improvement. 12 of the 40 bottles of the IVIG medication had gone into my body and miraculously I started recovering.</p><p>Seeing signs of recovery, my neurologists assigned a physio to me who would come every day and help me move my arms and legs, and tell me stories of a 6 year old girl with cerebral palsy in the next room.</p><p>Skipping forward a week — I was discharged from hospital and in terms of progress, I was able to wiggle two of my fingers and the little toe on my right foot all on my own. It took me two months of daily physio sessions and early morning yoga lessons from my grandmother to get back on my feet. I had monthly consultations with my neurologist and he mentioned not having seen anyone with such a severe condition recover so quickly before.</p><p>It is a year later now and I am 100% recovered from this unusual form of paralysis and here are my learnings:</p><ol><li>Never be your own doctor. Doctors spend years learning what to prescribe so you better listen to them — at least always complete your antibiotics course.</li><li>Nothing is impossible. No one expected me to walk again, let alone make a 100% recovery, but my family knew that I could do it and that gave me strength to do it.</li><li>Appreciate everything around you — when you are in what seems to be like your last day on Earth you tend to appreciate the little things, kind of like an epiphany. Don’t let that realization happen only when something goes wrong, start practicing appreciation now.</li><li>There are hundreds of people out there who don’t have families to look after them who suffer from conditions like GBS. If you are reading this, you have access to the internet and you probably have an education. You have it relatively easy — so help those who are in distress or dissociate from people who cause distress to discourage it.</li><li>Be thankful to people around you — and show it.</li></ol><p>Thanks for reading this!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f776288f2a15" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>