Insanity Fit Test Sheet: Fill & Download for Free

The Bathroom RuleYes, you read right. I have a bathroom rule in school.We have set times to go in and out of the bathroom. Our classes are all 70 minutes, referred to as blocks.The basement has bathrooms open 24/7 for some reason, but I’m never in the basement. The third floor has one bathroom, for staff only- you get how annoying this is becoming. The main floor, or second floor, has 3 bathrooms. The ones by lunch are not opened until lunch. Fourth block, those bathrooms are open the entire time.The bathroom in the main entrance, for the first two or three blocks is supposed to have a teacher present, as the security guard does not get there until 3rd block I want to say. The teachers that stand on bathroom duty are there for 60 minutes, the first teacher is the first 30 minutes of class the second for the last thirty minutes of the class. The same applies for the history bathrooms, however they never get a security guard.Now imagine young Sarah, upstairs for three blocks and has to literally go on a scavenger hunt to find an open bathroom. Its completely absurd.I”m not done yet though.Originally, each bathroom had one sign in/out sheet, similar to our classes.However, my brother told me yesterday that now they have a girls sign in and a boys sign in, with a rule that only 3 of the same gender allowed in each bathroom at a time.I don’t know how much space they have in the boys bathroom, but I know the girls can fit at least five or six students in the bathroom at a time.Crazy rule? I think so, but it is not crazy enough, don’t worry, I have tons more I can find…~SarahEdit (9/20/19)How about another go?Similar to most schools, my school has water bottles banned. But this is not just the high school, it is that way all across the district, because, hey kindergarteners are going to bring alcohol to school…Anyway, a really dumb rule my school just made is “No eating in the library”Dumb, right?No? Well, let me elaborate. We recently got a new librarian, Mrs McLeod. Mrs McLeod is fine with students eating in the library, but for some reason the school is not. The student(s) who eat there are responsible and don’t make a mess. But for whatever reason, the school does not allow us to eat there. Okay, yes, it is natural for schools to have this rule, but again, we don’t make a mess. And how else do we have a lunch bookclub if we cannot eat?Not crazy enough still? Well, none of these for now will top the bathroom one…Okay, I was in honors algebra II and honors geometry the same semester for the first two weeks of school. I took the first geometry test and everything, until my school got dumb.See, Algebra II is a prerequisite for geometry, and technically I was breaking the golden rule. HOWEVER, I was aware of the challenges I would have faced, I even received a 98 on the first geometry quiz, first geometry test, and my first Algebra II test. Mind you, my geometry test was the day after my algebra test.That Monday, my school decided to take me out of geometry and switch my honors chemistry to this semester and geometry to next. Why is this absurd and crazy? It was the third week of school, and get this, the reason was because of “Math Overload.” Yes, yes, yes, I understand it, but hey. Crazy to me.One last rule, in addition to teachers spending half of their prep monitoring lunches and/or bathrooms, they must also monitor the theater.In last times, some…stuff… happened in the theatre between students when they went to the “bathroom “ from what I am told. Consequently, teachers must sit there in front of the theater and grade, while making sure nobody enters without their teacher. Its rather boring, I went with a friend of mine when our teacher was helping us out with cells for our test.Okay, that’s all I can think of!

Simple - Alien Dreadnought.Alien Dreadnought is what Elon calls the machine that builds the machine. When you take apart a Tesla, all you get are what parts there are and how they fit together. It tells you very little about the machine that puts it together.When you see two plates fastened together with a bolt, do you know if that bolt was installed and tightened by a person, a machine or a combination of both?Tesla has invested tremendously in automating the process to make stuff. From basic stuff like car seats to installing wiring harnesses in cars. Most car makers laughed when Tesla had to rip out a large portion of the Model 3 lines when robots couldn’t effectively install wiring harnesses. This missed the obvious point - wiring harnesses are hard, but if you can beat that manufacturing problem with robots, then pretty much the entire process can be done with robots.Thus with Model Y, a redesigned wiring harness with 90% less wire than the Model S and designed from the start to be handled by robots were patented by Tesla. We will have to wait to see if it actually makes it into production in 2020, but if it does, it means Tesla will be another step ahead of the competition.Vertical integration - as part of the machine that makes the machine, Tesla also ensures its part supply. By making its own parts it has total control of the quality and design of them, it can make compromises or absorb higher costs that a parts contractor might not be willing to do. This doesn’t show when you look at a part, all you can do is estimate how much it cost. By controlling all of these aspects, Tesla can make changes quickly and with less red tape and expense than a standard car maker who has to deal with separate parts suppliers.If you find a part that shows unique welding or shape from forming, you don’t necessarily know how it was done. All you can see is the final result. For instance, Tesla has patented a new casting machine to make the frame to the Model Y, a crazy machine that could reduce a car frame from 70 separate parts that need to be formed and assembled to a single piece. Tesla Goes Patent Crazy With Huge Casting Machine For Model YThat said, it doesn’t really matter that you can’t tell this process from just looking at the pieces. Tesla has released all its patents for any company to read and use. Most choose not to do so. All Our Patent Are Belong To YouUPDATE: 01/15/2020Thanks everyone for the Upvotes!With more comments from Elon regarding Full Self Driving being “feature complete” in 2020, the importance of Alien Dreadnought becomes more important. Software, AI and full self driving capabilities are definitely products that competitors can’t replicate simply by taking the car apart. Even looking at the software code, especially for the AI doesn’t tell you how Tesla got there.Tesla is pushing AI faster and farther, not just in the end product Self Driving AI, but in the training techniques and even the hardware used to run the software to train. This is all part of the Alien Dreadnought - build the machines that build the machines.Tesla Quietly Acquired a Machine Learning Startup That's 'Squeezing' A.I.As for the “parts supply” of this portion of the machine, it is driving data. And Tesla has worked hard to make sure it has a nearly endless supply of driving data. With every Tesla equipped with sensors and most equipped with the full self driving suite of sensors such as cameras, radar and ultrasonics, every Tesla on the road is feeding data back to Tesla, billions (with a B) of miles of it every year. To put that into perspective, Waymo just hit 20 million miles of data total, currently running about a million miles per month. Waymo’s autonomous cars have driven 20 million miles on public roadsAI’s train best on lots of unusual data. Having access to billions of miles means Tesla’s have probably seen lots of weird things on the road - flying cars, kangaroos, planes landing on the highway etc. And with Sentry cam, it can record things even while parked. Tesla's millions of cameras are capturing some crazy things - videos - Electrek This is useful for teaching the AI about human behavior, not just things that happen on roads, so perhaps in the future the AI can determine social interactions like “a fight is happening” and predict that people are going to run out into the road.By designing its own processing chip and designing teaching supercomputers like Dojo, Tesla continues to make the machines to make the machines. Something conventional automakers are completely unable to do.Elon Musk hints at Tesla's secret project 'Dojo' making the difference in race to full self-driving - ElectrekUpdate 02/24/2020And the Alien Dreadnought rolls on - Tesla Self Driving Computer, designed from a clean sheet of paper, puts Tesla years ahead of the competition. Designed in house using the best talent in the world and manufactured exclusively for Tesla by Samsung in Austin.Tesla Self-Driving Computer architect Jim Keller is confident about solving autonomous driving - ElectrekThe issue is not that other companies can’t design their own chips to match the technology, but they can’t get rid of their current contracts and manufacturers for years. Again, by bringing manufacturing in-house, Tesla continues to ensure the supply.“The real reason for holding off? Automakers worry that computers like Tesla’s will render obsolete the parts supply chains they have cultivated over decades, the engineer said. Such systems will drastically cut the number of electronic control units, or ECUs, in cars. For suppliers that depend on these components, and their employees, this is a matter of life and death.” Tesla teardown finds electronics 6 years ahead of Toyota and VWUPDATE 4/28/20Additional evidence that “Alien Dreadnought” is the correct path. VW has had to delay the release of the ID.3 due to “software issues.”. From a highly optimistic 2018 prediction that VW would be producing a vehicle to match Tesla for “half the price” by 2020 - VW Claims By 2020, Its Electric Cars Will Match Tesla At Half Price to the stark reality of having thousands of vehicles bricked and sitting in lots - VW ID.3: thousands of electric cars spotted being stockpiled in Germany - Electrek the reality of failing to invest in the machine to make the machine is evident. While the machine in this case is software, VW’s lack of investment in internal software development shows how unprepared the entire company is for the future of vehicle production.“Today, our share [self-developed software as a portion of all software in Volkswagen Group vehicles] is less than 10 percent. That is clearly too small. In the Volkswagen Group, we want to achieve a share in software development of more than 60 percent by 2025.”Volkswagen Software Comments — 10 Months Ago vs. TodayShortages of robots (Tesla bought the company VW was using), shortages of programmers (Tesla produces most of its software in house and already has a massive team of programmers), a lack of integrated design (VW is trying to integrate dozens of different hardware/software suppliers) all prevent VW from producing a viable vehicle at this time. The machine that makes the machine isn’t working.Now in discussions with Daimler-Benz to try to pool software resources, they are very late to the game and late to the realization that software is as important as hardware. Daimler flirtet mit BMW und VW Simply building the physical isn’t enough, what runs the car is as important, if not more important than the physical parts to producing a top rate EV.UPDATE 05/19/20VW is learning that the future of cars IS to treat them like smartphones. Their software issues now extend to a conventional model - The iconic VW Golf (Gen 8). Failure of the emergency call system is halting their delivery of even conventional vehicles because of software difficulties. Despite having 46 years of experience producing the Golf, VW seems very inexperienced with software. It will have a difficult time building the expertise to handle these problems, let alone getting ahead due to the huge demand worldwide for such expertise as EVERY carmaker realizes that software is the future of the car.https://www.motor1.com/news/423689/vw-golf-software-issues-arise/New Volkswagen Golf Mk. 8 Deliveries Stopped Over eCall Software IssueUPDATE 06/30/20Building the factories to house the machines that makes the machine.As proven by the build speed of Gigafactory 3 (Shanghai) going from a muddy field to a producing manufacturing line in less than one year and with the ground breaking and rapid build out of Gigafactory 4 (Berlin), Tesla shows how important Alien Dreadnought is. Knowing what processes and parts that need to be manufactured and scaling the buildings from the beginning allow Tesla to create a factory space rapidly and repeatedly. If Gigafactory 5 (Texas) becomes an operational reality in 2021, then Tesla will have built 3 operating mega factories in less than 3 years.Breaking! Tesla Gigafactory 3 Groundbreaking In China In A Few Hours — UPDATEDTesla’s Chinese factory just delivered its first carsTesla breaks ground on foundation at Gigafactory Berlin, brings in heavy machinery - ElectrekInteresting details during construction, especially regarding the Model Y. For American built Model Y’s, the back frame is two cast parts, not the one part envisioned in the patent.However, for Giga Shanghai and Fremont, new supersized casting machines were ordered.Tesla Giga Shanghai Could Be Constructing The Next Level Casting Machine for MIC Model YSo a single piece rear frame casting could come online this year for Model Y. I would assume that Giga Berlin will also be equipped with the “Gigapress” for European Model Y production.UPDATE 07/08/2020More signs that vertical integration, cross pollination and diverse investment can prove to be innovative. A bonus efficiency when you control all aspects of the Alien Dreadnought.So it appears that Tesla is looking for a visual image engineer to work at the new Fremont Battery Factory. Current theories run that Tesla is going to apply visual AI to the quality assurance process. This is an indicator that Tesla’s investment in AI, particularity visual AI will provide additional benefits other than driving.Quality control is one of Tesla’s biggest current problems. It involves inspecting pieces and work, identifying if they are correct or incorrect and then (sometimes) fixing the problem. This has traditionally been a human job since the variety of problems (paint problems, panel gaps, incorrectly installed wires, misted headlights, etc.) was too much to simply program in with a huge variety of visual cues. With an AI visually learning what a problem looks like and its thousands of variations, it could quickly be able to inspect, flag and note how to correct any imperfections during the manufacturing process, thus saving time and effort at a later date. Since it will be run by robots and AI, this means it will always be vigilant and always provide a complete inspection.“Vision Systems Engineer, Cell EngineeringSAVETeslaFremont, CAApply on LinkedInApply on Careers For Veterans At Tesla - Veteran HiringApply on Jobilize LLCApply on JobsApply on BIGHiredApply on LocalJobs.com25 days agoFull-timeThe RoleTelsa’s cell manufacturing engineering group is looking for a highly-motivated equipment development engineer to accelerate our next-generation battery cell manufacturing program. This role will be involved in the design, development, commissioning and ramping of high-throughput equipment, while working in a cross-functional team to test, validate, and roll-out process and product design changes. The battery cell is a critical component in Tesla vehicles and storage systems. This role will have the opportunity to make meaningful contributions to our products. The work environment is demanding, fast-paced and incredibly exciting. The ideal candidate should be ready to push their limits, as they join in a highly motivated and capable team to make unbelievable things into reality.Responsibilities • Drive design, commissioning ramp of a complex manufacturing tool that is the first of its kind at Tesla • Work collaboratively as the equipment expert to ensure production lines run in a safe, efficient manner producing high-quality parts. • Become owner/subject matter expert to drive equipment-related process improvement including detailed troubleshooting and root cause analysis. • Troubleshoot and recover processes by PLC, HMI, or tool management software, and implement changes/improvements • Build training material, maintenance routines, and related documentation for equipment life cycle management • Lead safety and ergonomic improvements to production equipment which meets or exceeds regulations or best practices. • Identify equipment critical spare parts and ensure sufficient coverage to prevent extended downtime situations • Participate in supplier/ internal equipment buyoffs • Debug and optimize automated processes to maximize cycle time and uptime. • Develop and manage FMEA risk assessments to prioritize equipment improvement projects and ensure sufficient repair strategies. • Provide hands-on support to the maintenance team to troubleshoot and resolve extended downtime situations. Requirements • Either a BS with 6+ years’ relevant experience or an MS with 3+ years relevant experience in interdisciplinary/integrated engineering, manufacturing engineering, or equivalent • Evidence of exceptional ability. • Strong engineering foundations. Aptitude to learn • Demonstrated experience ramping equipment into manufacturing - automation equipment, web handling, design of electromechanical systems, PLCs, motion controls, and data capture • Demonstrated experienced solving complex electromechanical and software systems with limited documentation to debug issues, build process know-how, and kick off improvement projects • Structured thinking and problem-solving abilities. Proven through a number of root cause analysis following a data-driven approach, statistical software (JMP) a plus • 2D ACAD experience in creating/validating layouts, tooling drawings, Pneumatic and electrical schematics • Experience with robot programming, backup, and restoration a plus • Experience in automotive or high volume manufacturing is highly desired”UPDATE 07/29/20Alien Dreadnought - Building edition. Tesla has gained sufficient experience to make even the manufacturing of factory buildings faster and more efficient. Pre-fab concrete panels to accelerate the building of Gigafactory Berlin.Giga Berlin will come together at an impossible-seeming speed. The prefabricated construction method in Germany is extremely impressive.— Elon Musk (@elonmusk) July 27, 2020Tesla Giga Berlin is coming together quickly thanks to prefabricated constructionWill Giga Berlin break the Giga China construction record? With Giga Texas already clearing land, will Texas be built even faster? With talks of a second China Gigafactory and another Gigafactory somewhere in Asia, the machine that makes the factories to hold the machines seems to be running faster and faster.UPDATE 09/23/20 - BATTERY DAYSo much vertical integration, Alien Dreadnought, and evidence that Tesla is not just an EV company. From investing in a new process to extract lithium from clay using water and salt and being able to return the dirt back into the ground, to producing larger, more capable batteries faster, cheaper and in less space than the 2170, utilizing new chemistry and integrating the battery pack as a primary structure, being able to recycle the batteries at the end of their life with minimal processing, Elon and Baglino showed how integrated manufacturing and vertical integration of the supply chain are the core of Tesla.Everything You Need To Know About Tesla's New 4680 Battery CellWhile it will take some time to ramp up production of the new 4680 cell, the pilot plant just down the street in Fremont is designed for 10 GWh of production, 1/3 that of Gigafactory Nevada in a building 12 times smaller. With limited production already beginning, design and testing of vehicles to use the new battery cell is well underway (Model S Plaid).Bigger batteries that are more efficient, cheaper to make, faster to make, more environmentally friendly and providing additional benefits to charging, structure, weight and longevity. All the result of Alien Dreadnought.UPDATE 10/28/20 - FULL SELF DRIVING BETALimited FSD beta releasing on Tuesday next week, as promised. This will, at first, be limited to a small number of people who are expert & careful drivers.— Elon Musk (@elonmusk) October 12, 2020With the recent release of FSD Beta, limited, it again highlights Tesla’s use of Alien Dreadnought. Since every recent Tesla with HW 3.0 chip is capable of FSD, Tesla has already built a test fleet of hundreds of thousands of vehicles. By slowing releasing beta to “expert & careful drivers” Tesla has recruited dozens if not hundreds of Tesla owners to test drive FSD, for free. Already on YouTube are hundreds of hours of video of FSD beta in action showing how it acts in all kinds of places, times, weather and situations. There are self reported testers in at least 9 states.Tesla Starts to Rollout Full Self-Driving FSD Beta for Different States' EAP MembersWhile the first days showed definite errors or quirks - always taking the leftmost turn lane even if there were two turn lanes, slowing, but not enough for speed bumps, mistaking an angle parked car as a moving car after a turn, taking longer to turn into traffic allowing a crossing car to come close, etc. patches are apparently fast and furious. With massive uploads of driving data - Tesla is collecting insane amount of data from its Full Self-Driving test fleet - Electrek Tesla is well positioned to acquire the data for cheap or free, have vast quantities of it, and apparently process it, use it to reteach the AI and update FSD within a day, if not hours. The machine to teach the machine is running fast and furious.Many of the “driver interventions” occur because FSD doesn’t drive in a fashion that makes the driver comfortable - staying in the center of a lane when there are parked cars rather than riding the dividing line, staying in the left lane when other cars are trying to pass, hitting speed bumps a bit fast, hitting the brakes at the last second, waiting too long to merge so the lagging car closes the gap quite close, etc. This information will be useful to Tesla in terms of creating a tolerable human interface, as many passengers, while quite safe with this driver, would find the experience unnerving.UPDATE 02/10/2021 -Elon Interview, Asking the right question.In an interview with Sandy Munro, Elon described why the first Model 3 wheel wells were a mess. As he mentions, they had many engineers working on the project and each one probably got the right answer for the part they were working on. One piece needed to be corrosion resistant, so the engineer picked the best corrosion resistant material, One part had a complex shape so that engineer picked a material that could be formed easily. One part was a heavy duty part that saw a lot of stress so that engineer picked a strong material. Individually each engineer came to the correct answer for her/his question. But the result was a mess - many parts, different materials, new problems in how to attach them together, new problems on how to seal the gaps due to difficulties in fitting them together well, etc. These questions had answers and they were implemented. It was a disaster. As Sandy noted, he hated how the early Model 3 wheel wells were constructed - too many parts, too many materials, too much labor. So Tesla had asked the wrong questions, and even though they had correct answers to those questions, it didn’t really provide the right solution. So Tesla engineers took a step back and asked the right questions - what does this part do, what does it need to made of and using Elon’s favorite design strategy - no part is the best part - how to eliminate the part. The result? Model Y rear structure megacasting - a single piece that replaces 70, is corrosion resistant, strong, and using the gigacasting machines, able to form complex shapes. As an added bonus, the removal of so many parts meant that 300 less robots were need to work on that section. Ask the right questions, get the right answer. Being vertically integrated allowed Tesla to move forward and scrap what wasn’t working for a whole new direction without dealing with suppliers, contracts or unions.UPDATE 04/13/21 - To the Moon and Beyond - Vertical Integration of Sustainable energyAs Gigafactory Texas takes shape, the first Gigapress is starting preliminary operations using large tanks of Liquid Natural Gas for its furnace to melt Tesla’s proprietary aluminum alloy. It is assumed that this will eventually be piped in through a large natural gas pipeline when the factory is in full operation.Musk’s warming up to gas evident once more at Texas GigafactoryFour tanker trucks belonging to a Houston-based liquefied natural gas company could be seen parked outside the Texas gigafactory construction sitehttps://www.detroitnews.com/story/business/autos/2021/03/30/musks-warming-up-gas-evident-once-more-texas-gigafactory/7060681002/Liquid Natural Gas is methane. Elon has shown a path to make even this process sustainable.Capture Carbon.XPRIZE Carbon RemovalTruly Meaningful Impact The world's leading scientists estimate that we may need to remove as much as 6 gigatons of CO2 per year by 2030, and 10 gigatons per year by 2050 to avoid the worst effects of climate change. For humanity to reach the Paris Agreements goal of limiting the Earth’s temperature rise to no more than 1.5˚(C) of pre-industrial levels, or even 2˚(C), we need bold, radical tech innovation and scale up that goes beyond limiting CO2 emissions, but actually removes CO2 already in the air and oceans. If humanity continues on a business-as-usual path, the global average temperature could increase 6˚(C) by the year 2100. This four-year global competition invites innovators and teams from anywhere on the planet to create and demonstrate solutions that can pull carbon dioxide directly from the atmosphere or oceans ultimately scaling massively to gigaton levels, locking away CO2 permanently in an environmentally benign way. Solutions will be scientifically evaluated across multiple criteria such as; amount of CO2 removed, life cycle analysis of the removal process, energy efficiency, land footprint and sequestration capabilities. Any carbon negative solution is eligible: nature-based, direct air capture, oceans, mineralization, or anything else that sequesters CO2 permanently.https://www.xprize.org/prizes/elonmusk2. Convert carbon to methane, using hydrogen made from water and energy from solar.Correct, pure electric. Wish orbital rockets could be so too, but there’s no escaping Newton’s Third Law. SpaceX is already developing high efficiency CO2 capture with H2O to form liquid CH4 (methane) & O2. Critical for propellant production at Mars Base Alpha.— Elon Musk (@elonmusk) June 10, 2018Sabatier reaction - WikipediaMethanation process of carbon dioxide with hydrogen The Sabatier reaction or Sabatier process produces methane and water from a reaction of hydrogen with carbon dioxide at elevated temperatures (optimally 300–400 °C) and pressures (perhaps 30 bar [1] ) in the presence of a nickel catalyst . It was discovered by the French chemists Paul Sabatier and Jean-Baptiste Senderens in 1897. Optionally, ruthenium on alumina (aluminium oxide) makes a more efficient catalyst. It is described by the following exothermic reaction . [2] CO 2 + 4 H 2 → pressure 400 ∘ C CH 4 + 2 H 2 O {\displaystyle {\ce {CO2{}+4H2->[{} \atop 400\ ^{\circ }{\ce {C}}][{\ce {pressure}}]CH4{}+2H2O}}} ∆ H = −165.0 kJ/mol There is disagreement on whether the CO 2 methanation occurs by first associatively adsorbing an adatom hydrogen and forming oxygen intermediates before hydrogenation or dissociating and forming a carbonyl before being hydrogenated. [3] CO + 3 H 2 ⟶ CH 4 + H 2 O {\displaystyle {\ce {{CO}+ 3H2 -> {CH4}+ H2O}}} ∆ H = −206 kJ/mol CO methanation is believed to occur through a dissociative mechanism where the carbon oxygen bond is broken before hydrogenation with an associative mechanism only being observed at high H 2 concentrations. Methanation reaction over different carried metal catalysts including Ni, [4] Ru [5] and Rh [6] has been widely investigated for the production of CH 4 from syngas and other power to gas initiatives. [3] Nickel is the most widely used catalyst due to its high selectivity and low cost. [2] Applications [ edit ] Creation of synthetic natural gas [ edit ] Methanation is an important step in the creation of synthetic or substitute natural gas (SNG). [7] Coal or wood undergo gasification which creates a producer gas that must undergo methaneation in order to produce a usable gas that just needs to undergo a final purification step. The first commercial synthetic gas plant opened in 1984 and is the Great Plains Synfuel plant in Beulah, North Dakota. [2] It is still operational and produces 1500 MW worth of SNG using coal as the carbon source. In the years since its opening, other commercial facilities have been opened using other carbon sources such as wood chips. [2] In France, the AFUL Chantrerie, located in Nantes, started in November 2017 the demonstrator MINERVE. This methanation unit of 14 Nm3 / day was carried out by Top Industrie, with the support of Leaf. This installation is used to feed a CNG station and to inject methane into the natural gas boiler. [8] It has been seen in a renewable-energy-dominated energy system to use the excess electricity generated by wind, solar photovoltaic, hydro, marine current, etc. to make hydrogen via water electrolysis and the subsequent application of the Sabatier reaction to make methane [9] [10] In contrast to a direct usage of hydrogen for transport or energy storage applications, [11] the methane can be injected into the existing gas network, which in many countries has one to two years of storage caphttps://en.wikipedia.org/wiki/Sabatier_reaction3. Use methane to heat the furnace for melting aluminum alloys.So vertical integration of energy sources - solar to methane to heat, will allow Tesla to continue to pursue its goal of sustainable energy, even while using conventional technologies such as a natural gas furnace.

TSMC was started in the late 1980s. It’s business model was based on a new concept that that DARPA was trying to get US semiconductor vendors to adopt: the separation of IC design from IC foundries.This was desired because though military contracts funded initial semiconductor companies like Fairchild and Texas Instruments, by the 1970s and early 1980s it was clear that commercial volumes would take all the attention of commercial IC companies and they would have no reason nor desire to address the special requirements of military semiconductors and ICs.DARPA saw a need to “piggyback” military designs upon commercial cost structures and volume learning curves without imputing the complexity and technical details of military designs onto commercial fabs.The answer was to create the foundry system but the state of semiconductor manufacturing in the 1960s were not up to snuff in terms of quality or reliability of either products or manufacturing processes. In this video you see yields described in the low 10–30% range (at ~13:00) which is insanely low by today’s standards. It was also common to “loose fab recipes” because of things like “sodium contamination” (e.g. having a shift worker heat pizza in a photoresist oven caused one such incident at Intel in the 1970s). Also note they are using manual design definition methods using large sheets of plastic called red Rubylith rather than computer CAD used today. (“5, 7 or more layers” - LOL - typically you have 200–300 today)This effort led to many of the current standard features of semiconductor manufacturing such as process control, standardized test chips, semiconductor and circuit simulation software (e.g. SPICE, PISCES and SEDAN), standardized design interchange data formats (GDS-II), the invention of CAD software (which lead to supremacy of Cadence and Mentor), etc. These innovations greatly improved all semiconductor manufacturing.The adoption of the foundry system itself by US companies was only lukewarm with companies like Hewlett-Packard and IBM offering foundry services (in part because their own internal IC design groups could use the model perfectly) but larger, pure-play semiconductor companies like Intel rejected the business model. It’s only recently that Intel has offered foundry services and largely doing it poorly because it’s needed to abandon much of its legacy “baggage” to fit in well (Hell freezes over: Intel announces plan to fab ARM processors)In this late 1980s environment, the founder of TSMC, Morris Chang, who was trained as an electrical engineer from MIT, was familiar with the DARPA foundry system and realized it was an amazing business opportunity: he had been working at Texas Instruments and used the IBM foundry services to make an IC and saw how well it worked and realized it would be a winning business model so he founded TSMC when he returned to Taiwan. He picked the right model, was the first “pure foundry play” semiconductor company and the rest is history.(This photo hardly captures what it’s like to visit Hsinchu’s Science Park where TSMC is headquartered - TSMC has dozens of buildings including monster Giga-Fabs like Fab 12).Semiconductor fabs have always been high capital cost so that part isn’t new. However because of the separation of design from manufacturing, the costs to either side of this relationship are still far cheaper than what was had by IDM (full vertical) IC manufacturers such as Intel and others. So the model was ultimately cheaper. And indication of this was illustrated in VC startup funding of IC companies: after 1989 there were no IDM startups funded by VCs in Silicon Valley; only fabless firms were funded since then.