Application Of The Complete Bifurcation Groups Method: Fill & Download for Free

PrologueFirst, let me start off by saying that I agree with Benedict Evans that 5G as a technology isn’t all that earth-shattering. It’s really just a continuation of a well-established trend: fatter and fatter data pipes. Imagine being able to take your home Wi-Fi everywhere and that pretty much describes 5G.Getting excited about 5G, or talking about amazing new applications it enables, is pretty much like getting excited about a new version of DSL or DOCSIS.— Benedict Evans (@benedictevans) March 14, 2018This is not to say that 5G is not important, or diminish the work done by hundreds of thousands of engineers, scientists and other wireless industry professionals around the world … or that it won’t catalyze the development of a host of cool new applications bearing all of the latest buzzwords and acronyms.It’s just more that I find the underlying economic and geopolitical story far more interesting and meaningful. Sort of like the 2006 film Babel starring Brad Pitt, it is a multiple-storyline epic featuring two main protagonists that lead completely separate lives for the first four acts while gradually converging … until the climactic moment when their paths smash into each other.As the curtains open on Act V, we find the two protagonists having finally taken the stage at the same time. And while we can make some guesses as to how things unfold from here, the reality is that the story is still being written.The implications are enormous and bigger than the wireless industry itself. Indeed, this is perhaps the most important area to pay attention to in today’s increasingly tech-driven geopolitical arena.But we are getting ahead of ourselves; to fully appreciate the saga we need to start at the very beginning … where we find ourselves on a deserted Hamptons beach at the break of dawn, sometime in the mid-80s …Act I“It was the best of times, it was the worst of times, it was the age of wisdom, it was the age of foolishness, it was the epoch of belief, it was the epoch of incredulity, it was the season of Light, it was the season of Darkness, it was the spring of hope, it was the winter of despair, we had everything before us, we had nothing before us, we were all going direct to Heaven, we were all going direct the other way — in short, the period was so far like the present period, that some of its noisiest authorities insisted on its being received, for good or for evil, in the superlative degree of comparison only.”Opening paragraph to A Tale of Two Cities by Charles DickensI remember the iconic scene[1] in 1987 film Wall Street when Gordon Gekko officially brings Bud Fox, an ambitious young broker, “inside” the curtain. It is a critical scene in the movie, made even more dramatic by use of what was then a novel piece of modern technology — the cellular phone. Gekko delivers the coup de grâce to the young broker by expounding — in real-time on that phone — on the beauty and awe of the sunrise from his beachfront palace as a metaphor for a new world of hitherto unimaginable wealth that he was about to enter.The first cellular phones were analog radio devices that would connect to a local tower that oversaw a fixed area, or “cell”, on a dedicated frequency. The radio-frequency (RF) technology was pretty much the same as that powering walkie-talkies — the trick there was figuring out how to connect the walkie-talkie to the circuit-switched phone network.Call capacity was limited because there are only so many slices of frequency into which you could divide spectrum before you run into quality issues. As a result, early cellphones and their related service plans were extremely expensive and generally limited to wealthy moguls like the fictional Gordon Gekko.But while Gekko extolled the “virtues” of unmitigated greed, scientists and engineers were working on the next generation of wireless standards, and trying to solve the fundamental problem of how to cram more channels into the same allotment of limited spectrum. It is essentially the same problem that they continue to try to improve on today.At the time, there were two competing methods on how to do this. The first was something called time-division multiple access (TDMA)[2]. With TDMA, you could have multiple users share the same frequency by dividing the signal into fixed time slots that were assigned to each active user.The second method was code-division multiple access (CDMA).As with TDMA, the goal of CDMA was to permit multiple users from sharing the same slice of frequency but instead of having fixed, assigned time slots to differentiate between users, CDMA used unique codes to identify each user (hence the name). These codes could switch and hop across multiple channels, making it more flexible than TDMA.From a technology perspective, CDMA was better because it was more scalable especially as the world became more digital and less analog over time. But as we saw in the battle between VHS and Betamax[3], sometimes it is not just about technological superiority.Act IIThe race was on between the two competing standards.Western European countries latched onto the TDMA method and a generally open, collaborative approach, releasing Global System for Mobile Communications (GSM)[4] in 1991.The world’s first GSM call was made by Finnish Prime Minister Harri Holkeri on July 1st, 1991 and commercially deployed at the end of the year on a network built by German conglomerate Siemens and a then-relatively unknown conglomerate subsidiary called Telenokia. It would later drop the prefix, adopt the name of its conglomerate parent and become widely known simply as “Nokia”.Helsinki, Finland (Photo: Paasitorni)The competing CDMA method was not entirely novel — it had been pioneered as early as the 1930s by scientists from the Soviet Union. Interestingly, wireless phones based on the CDMA method were used in Moscow as early as 1963. However, it wasn’t until a former electrical engineering professor from MIT named Irwin Jacobs latched onto the technology that it found mainstream, commercial applications.In 1985, Jacobs launched Qualcomm — which stood for “Quality Communications” — based in the Southern California paradise of San Diego. The new company was initially focused on mobile satellite communications and because satellite bandwidth was so expensive and precious, there was an intense focus on bandwidth efficiency, which is what had led Jacobs to CDMA.The company went public in September 1991, raising $68 million to fund its CDMA research and later an additional $486 million to help commercialize a CDMA-based ecosystem. Qualcomm was perhaps the highest flier in the high-flyin’ 90s, ending the decade with its stock price increasing around 180x from its IPO price eight years earlier.Knowing nothing else but Qualcomm’s stock chart in the 1990s, one could have reasonably concluded that CDMA and its superior technology had won.But that was not to be, at least here in Act II.One issue for Qualcomm and its CDMA-based “cdmaOne” standard was that GSM had gotten a big head start.The “cdmaOne” standard was not adopted as a standard until 1995[5] at which point GSM networks in Western Europe and the United States had already reached 10 million active subscribers. By the time cdmaOne networks were deployed at scale, GSM networks had already reached over 100 million active subscribers.The other issue is that for voice, the technical advantages of CDMA were not that significant. TDMA did a fine job of transmitting voice and capacity constraints could be alleviated by adding additional wireless radios or reducing the size of each cell, especially if those radios could be purchased at affordable rates.Taking a more open, collaborative approach, GSM had also incorporated certain features such as a standard ID schema that allowed cellphones to be used across multiple networks by simply switching out the SIM card — which was much more important in Europe with its multiple country networks vs. the United States where people tended to travel internationally far less frequently.Ultimately, GSM won decisively by achieving scale and driving down cost. Because GSM networks were first to market, equipment manufacturers were able to deploy networks more quickly and inexpensively. Because GSM operators reached scale, handset manufacturers designed handsets around GSM standards. Because GSM was developed with a more open, collaborative approach, its technology licensing fees were lower. And because costs were lower, active subscribers tended to go with GSM networks vs. cdmaOne when given a choice.In September 2001, shortly after 9/11[6], I moved out to Hong Kong, which had deployed a GSM network.I was amazed at how much cheaper and better my cellphone service was compared to the United States. It was incredibly convenient to be able to simply switch out a small SIM card and start using your phone on another network. I loved my Nokia 8310 handset[7]. And I still distinctly remember how one annoying thing about work trips to South Korea — one of the few markets that had chosen CDMA over GSM — was having to use a clunky loaner Sanyo handset that didn’t have my address book or Snake[8].My Nokia 8310 handset (circa December 2003)GSM and Nokia had won the 2G war. CDMA-based technology was expensive and clunky and few people wanted it. By the early 2000s, Nokia was a giant, one of the world’s most valuable companies, at one point accounting for 21% of Finland’s exports and 70% of the Helsinki stock exchange market capitalization.But we were really just getting warmed up.Act IIILong before Apple unlocked “Smartphones” on the Technology Research Tree in 2007[9], wireless industry executives had suspected that data and not voice was going to be the long-term future of wireless. Fresh off the release of GSM in 1991, the various industry groups that set wireless standards had already begun trying to figure out how to transmit data at high speeds over the airwaves.Most had already known that GSM’s TDMA approach — perfectly adequate for voice communications — was just not going to cut it for data. While data could be transmitted over GSM networks, the transmission rate was capped at speeds reminiscent of the early days of dial-up modems. As nostalgic as I was for the halcyon days of the mid–90s, it was just not practical for anything outside of short-form messaging (i.e. SMS/texting).As wireless industry executives tried to find solutions for this technical issue, every path seemed to lead back to San Diego.San Diego, California (Photo: PV Magazine)It’s not enough to just have a good idea — you need to execute.While wireless operators worked 24/7 to deploy mostly GSM mobile networks around the world in response to the surge in active subscriber growth, Qualcomm was busy executing … and betting its future on CDMA. It too worked round-the-clock — frankly, an amazing accomplishment considering San Diego’s gorgeous year-round weather — to solve fundamental issues related to implementing wireless networks using the CDMA approach.Its main approach was to patent specific methods on how to perform various functions that were important in enabling wireless communication. For example, US Patent No. 5,280,472[10], issued on January 18, 1994, called for a “CDMA communication system in which cellular techniques are utilized in a distributed antenna system environment”. This particular one would cover instances where wireless signals need to be split up and re-routed and amplified within large buildings that remote tower-generated wireless signals would have difficulty penetrating.This was just one of an estimated 16,000 patents filed by Qualcomm over the years[11], of which at least 6,000 are related to wireless. In addition to building its IP portfolio, Qualcomm took a lead role in fostering eco-system development, including at various points producing handsets, network equipment and designing RF chips and chipsets.Photo: Gizmodo: Qualcomm's Amazing Wall of PatentsAs various 3G standards — represented by confusing acronyms like UMTS, W-CDMA, TD-SCDMA, CDMA2000 — emerged and were implemented, it became abundantly clear that CDMA was the common technology tying all of them together. With such a large patent portfolio around this method, it also became clear that Qualcomm was going to be collecting a recurring, steadily increasing stream of royalty payments for the foreseeable future.As 4G standards (LTE) rolled around in the mid- to late-2000s, cementing data as the key focus of the wireless industry, Qualcomm emerged as the dominant toll collector in one of the largest and most strategic industries on the planet.Act IV — Part I:For most of the first three acts, China is a mere after-thought, a minor character that is largely relegated to watching the main action from backstage:While Gordon Gekko was recruiting Bud Fox into his insider trading cabal, China was figuring out how to motivate its farmers to really put their backs into it so the nation could avoid teetering so close to the edge of starvation.While Nokia was busy deploying early GSM networks in Western Europe, China was figuring out how to dismantle its centrally planned industry without uprooting the lives of urban workers to the point where they would pour out into the streets by the millions like they did that fateful spring of 1989.While Qualcomm’s scientists were patenting thousands of wireless patents, China was figuring out how to open its doors so it could actually start trading the things that it had in abundance — e.g. inexpensive labor — for the things that it lacked, like wireless technology.In 1987, Ren Zhengfei — a former mid-level officer in the People's Liberation Army engineering division — founded Huawei in Shenzhen, the city bordering Hong Kong which was at the front lines of China’s economic reform program. At this point, China was 100%-reliant on foreign telecom equipment for its landline industry and most major international telecom equipment companies had established a presence in the country on the promise of tapping into China’s billion-person market.Shenzhen in the late 80s / early 90s (Photo: Shenzhen Municipal Government)At first, Huawei focused on re-selling imported telephone switches and fire alarms from Hong Kong. But for whatever reason, its founders decided very early on that the company should develop its own technology in-house vs. the “easier” path taken by others like Shanghai Bell to form a joint venture with multinationals to access foreign technology via transfers. Ren believed that “foreign companies were unlikely to transfer their cutting-edge technology and that Huawei would be better served by performing its own R&D”[12].Starting from a technology base of virtually nil, Huawei nonetheless prioritized R&D from its early stages. As a private company (vs. state-owned enterprise), Huawei suffered from lack of access to capital and was forced to borrow at extremely high rates in the early years. Despite these challenges, by 1993 Huawei had released its first significant in-house developed product — an electronic switch that could handle 10,000 lines, unprecedented for a domestic company at the time. It was a mature product and comprised almost entirely of foreign components but it was still quite impressive for the six-year old company.Huawei C&C08 Circuit Switch (Photo: DIY Trade, Shenzhen Huaxinzhihe Technology Co.)One of its strategies was to focus on market segments that were ignored by foreign technology suppliers. For example, international telecom companies preferred to focus on the rapidly growing urban centers while ignoring the poor, rural areas. Seeing this, Huawei adapted foreign technology to deal with “frontier market” issues — problems such as unreliable power grids and rats that like to gnaw on cables. Its business practices were “controversial” and by international standards probably textbook “corrupt” but in China at this time, function prevailed over form.Huawei began to separate itself from its domestic peers. By 1996, less than a decade after founding, it had secured its first international customer, selling circuit switches to Li Ka-shing’s telephone company in Hong Kong. By 2002, Huawei had overtaken Shanghai Bell, the largest Chinese-international JV at the time. Around this time it began expanding into adjacent markets like Internet and data communications, which was dominated by companies like Cisco.February 5th, 2003 marked the day that the name “Huawei” was formally introduced to the American lexicon (outside of a small group of telecom industry insiders). This was the day that Cisco sued Huawei’s American subsidiaries for copying code from its routers[13]. It marked the first major instance where a Chinese technology company had brushed up against an American one — not to mention the beginning of what I can only describe as a “lengthy and systematic effort by Americans to devise ever-increasingly creative and sophisticated ways to butcher the pronunciation of its name”.The suit was settled in 2004 but the damage had already been done. By this time, Huawei had captured one-third of China’s enterprise market and has never looked back.By the mid-2000s, Huawei was pushing hard into developing markets with an increasingly sophisticated array of products and services for both landline and wireless communications. Like its foray into China’s rural markets in the early 1990s, Huawei adapted mature products for developing countries facing problems that China had dealt with the prior decade such as non-existent or unreliable power grids and inexperienced technical staff.An example from one of my early Quora answers[14] was a low-power base station that could run on solar power, targeted at African countries that lacked reliable power infrastructure. In another early answer[15], I also discuss the important role the China Development Bank played in helping Huawei expand into overseas markets.RuralStar Base Station (Photo: Huawei)By 2011, Huawei had overtaken Ericsson as the largest telecom equipment supplier in the world with approximately $33 billion in revenue and industry-leading profit margins.It was around this time that Huawei had started aggressively pushing into consumer electronics[16] as well, piggybacking on the smartphone revolution and its now massive R&D operation to vault into the Top 10 of smartphone OEMs. By 2017, Huawei was pushing $100 billion in revenue, largely driven by growth in its consumer devices division which was now challenging Samsung for the top spot in smartphone market share (by unit volume). Today, the company has around 180,000 employees worldwide with 80,000 of them involved in R&D[17].Act IV — Part II:While Huawei was pushing forward at breakneck speed (even compared to the rapidly evolving Chinese economy), China’s state-owned telecom operators were plodding along slowly, trying their best just to keep up with the rapid and accelerating march of communications technology.Prior to 1994, the state held a monopoly on the provision of telecommunications services through the Ministry of Posts and Telecommunications and its operational arm, China Telecom. In 1994, to kick off reforms, the first competitor was established (China Unicom) and in the following years, there would be a series of reforms as Chinese policymakers tried to mold these former government ministries into modern corporations.It was around this time that Qualcomm had first reached out to China. Although the Chinese government had already selected GSM for commercial use in 1994 — attracted by lower cost and ease-of-deployment — Qualcomm set up a partnership with the People’s Liberation Army (kind of crazy when you look back and think about it) to use its CDMA technology for military communications. However, in 1998, Chinese President Jiang Zemin “shocked the world” when he announced[18] that the PLA would no longer be allowed to engage in civilian activities, swiftly killing off the joint venture plans.The Chinese government was initially hesitant to partner with Qualcomm until they would address three priority issues:It wanted to be able to deploy phones that could work on both GSM and CDMA networksIt did not want to pay the royalty fees or structure that Qualcomm was demanding for its CDMA technologyIt wanted access to the design of Qualcomm’s CDMA chipsetHowever, as detailed excellently by MacroPolo[19], in the backdrop of late-90s negotiations to enter the World Trade Organization (WTO), Chinese policymakers decided to drop most of these demands and, under pressure from the US government, agreed to allow Qualcomm and its CDMA technology into the Chinese market. This decision would prove very costly in later years but for now, China was more focused on WTO accession.Source: MacroPolo: From Windfalls to Pitfalls: Qualcomm’s China Conundrum - MacroPoloFollowing this decision, over the next decade Qualcomm’s revenue in the Chinese market grew from zero to nearly $2.5 billion and came to represent almost one-fifth of the company’s revenue. And this was just the beginning — as China began to commercially deploy 3G networks in 2008, this number was set to explode even higher.Source: Company Filings via Capital IQIn the most recent fiscal year (12 months ending September 30, 2018), Qualcomm’s revenue from China had increased to over $14 billion and represented over two-thirds of its revenue stream.A large part of this revenue stream, especially in the earlier years, was paid by foreign smartphone OEMs like Apple[20] but as Chinese smartphone OEMs (incl. Huawei) took market share in China and around the world, they began to realize how much Qualcomm was making off its intellectual property — because they were now the ones paying these royalty fees in increasing amounts.But just as Americans are about to break out the champagne and “USA! USA!!” chants, the latest missive from the Debbie Downer-in-Chief[21] himself flashes across our feed …We are not in a trade war with China, that war was lost many years ago by the foolish, or incompetent, people who represented the U.S. Now we have a Trade Deficit of $500 Billion a year, with Intellectual Property Theft of another $300 Billion. We cannot let this continue!— Donald J. Trump (@realDonaldTrump) April 4, 2018Somewhere between China’s reputation as the world’s most rapacious “intellectual property thief” and the tens of billions of dollars per year it pays to international technology companies like Qualcomm … lies reality.Act V is where we are going to find out what that reality is.Act VOn November 8th, 2016, Donald Trump pulled off a surprise win over Hillary Clinton in the United States presidential election. Eight days later, a far less publicized political battle was taking place, this time over a topic that only a handful of people in the world really understand at a deep, technical level.Remember the industry groups that we met in Acts I to III that played such a critical role in choosing and setting wireless standards?Well, they are still around and playing just as critical a role. Depending on which technologies are incorporated, the respective IP holders may be richly rewarded, just as Qualcomm had for the better part of the last three decades.On November 16th, 2016, members of this standards body, 3GPP[22], met in Nevada to decide whether something called “polar coding” would be incorporated into official 5G canon. It was up against an alternative approach called “low-density parity check”. Intense debate ensued over which one was better.To a casual observer, the debate of “polar coding” vs. “low-density parity check” may have appeared to be a Nerd Fight of Epic Proportions but behind all of the computer science and technical jargon was something much deeper — what it was really about was control over the next-generation of communications technologies.As you may have guessed, this is where the paths of Huawei and Qualcomm finally began to converge.You see, China was getting weary from paying tens of billions of dollars every year in licensing and royalty fees for technology invented 15–20 years ago at a time when they did not have the capability or resources to even have a seat at the standards-setting table. While they had been late to the standards-setting game for even 4G/LTE standards, the country’s leaders had committed to making sure that this would not be the case with 5G. And Huawei was the main horse that they were betting on.As Huawei had grown through the years, it had continuously re-invested this growth back into R&D. By 2017[23], close to RMB90 billion ($13.8 billion) per year, out-spending Qualcomm by two and a half times in absolute terms (i.e. before adjusting for the approximately 3x[24] difference in wages between Shenzhen and San Diego).In doing so, it had quietly built up its very own patent wall:One of these patents was around the aforementioned “polar coding” method while Qualcomm held patents around the competing “low-density parity check” method. During the 3GPP debate, Western companies largely backed Qualcomm’s method while Asian manufacturers favored Huawei’s. In the end, both were accepted into as viable alternatives in the 5G standards book and each side moved on to battle over other (likely even nerdier) topics.While accumulating the most patents is still an important part of the game (as we saw in Act III with 3G), commercialization is an equally important consideration (as we saw in Act II with 2G).And on this front, China is racing ahead. Not only is it already the world’s largest wireless market by far, with 10x the number of base stations as the United States (and 40% of global sites[25]), its wireless operators are already well into the roll-out schedule and plan to be fully commercialized (for “standalone” or “full” 5G; see Note i) by the end of 2020[26][27]:The 3GPP debate in Nevada presaged the fault lines that we are now beginning to see, not only for 5G but other technologies as well. The elections of President Trump and the rise of other right-wing political parties in Western European countries has only increased the politicization trend.On April 16th, 2018, ZTE, the second-largest Chinese communications equipment supplier after Huawei, was hit by the U.S. Department of Commerce with an export ban[28]. The ban would prevent it from accessing critical components provided by U.S. suppliers (e.g. optical chips) and force it to re-design its equipment. It was a crippling blow to the company and while later reversed, was one of the first clear signs of this increased politicization.Then, just a few weeks ago on December 1st, 2018, Sabrina Meng, CFO of Huawei and daughter of its founder, was arrested in Canada at the request of the U.S. government in what was viewed by most as a politically motivated escalation. President Trump essentially confirmed it several days later[29].And that pretty much brings us to the present.The key protagonists, Huawei and Qualcomm stand together on stage, surrounded by a host of supporting cast members. The crowd watches with rapt attention, eagerly awaiting the next twist in the story …EpilogueAs I sit here and write in the last few days of 2018, it is quite clear that we are still very much in the middle of Act V — and it looks like there will be plenty of more excitement and fireworks to come.I also must admit that I am not 100% sure how Act V and the “race for dominance” will ultimately play out between Qualcomm and Huawei, not to mention all of the other actors on stage.As you saw through the first four acts, there were many twists and turns along the way, with new characters entering the space and old ones fading away with each successive generation of wireless standards. Add to that the increasing politicization of technology and the oft-times capricious nature of geopolitics and my crystal ball is quite foggy at the moment.But I do think understanding how we got to this point is very important if we want to think about the possible future scenarios and where we go from here — and that is why I took you through this fairly expansive review of the history of wireless.That said, I do want to leave you with some final thoughts on the topic:The emergence of Huawei as a major IP holder will inevitably cut into Qualcomm’s wireless market dominance and position as the favored toll collector.Opening quote to Act I notwithstanding, this is actually not just a Tale of Two Companies; it is also about existing players like Ericsson, Nokia and Apple that have long chafed at Qualcomm’s licensing fees and dominant market position[30].As I wrote in a recent answer[31], Qualcomm collects upwards of $30–40 on each iPhone that was sold — on top of any chips it provides — due to its “double-dipping” licensing structure. For 5G, Qualcomm announced that it would charge “up to $16.25” in royalties for every phone — much lower, an indication of lower negotiating leverage.The battle between commercialization vs. technology will be another area to watch.I do not know enough of the technical minutiae — stuff like “polar coding” vs. “low-density parity check” — to fully assess but my gut tells me that the differences between Huawei’s approach and the one supported by Qualcomm may not be that material and certainly not like the difference between TDMA and CDMA during the 2G and 3G mobile standards wars.We cannot rule out the possibility (as unlikely as it may seem at this point) that Qualcomm and Huawei end up collaborating or working together out of pure self-interest (an “if ya can’t beat him, join ‘em” type situation).The likelihood of global wireless standards bifurcating into different camps seems to be increasing, although it is far from inevitable at this point.If this happens, there are two clear camps — China and the “Five Eyes” Anglophone group. If you throw the European Union and Japan into the Anglophone group (let’s call it the “U.S. Alliance”), you are talking about a combined population of around 1 billion (that is significantly wealthier on a per capita basis) compared to 1.4 billion in China — all things considered, fairly balanced.But we cannot forget about the other 5 billion+ people out there — and places like Southeast Asia, India and Africa are where the front lines of the battle for technology dominance will take place.From the perspective of these 5 billion plus, the entrance of Huawei into the fray is seen as a positive development, insofar as providing them with another option and greater leverage to negotiate on fees.This bifurcation trend may also play out in other areas of technology, not just wireless standards.Semiconductors are another strategic (and related) industry. Chips are how you take the IP from the patents and convert into real-world use cases. They are critical components in network equipment, as ZTE was reminded in April 2018.The U.S. Alliance dominates the semiconductor industry, especially upstream (i.e. semi capital equipment). Certain specialty equipment like extreme UV lithography[32] is dominated by European like ASML and Japanese players like Canon/Nikon and can be easily controlled through measures like export bans over “dual-use” technology.However downstream production is dominated by Asian manufacturers, notably Taiwanese and South Korean foundries. Moreover, the consumer electronics supply chain is deeply entrenched in China and the East Asia region.So it is very complicated, and this is what makes predicting how the various points of negotiating leverage play out so hard.National security concerns are very valid. But I think they can be addressed without forcing others to have to split into camps that are non-interoperable. That would be a shame for everyone.Finally, the one thing that I do know for sure is that we’ve come a long way since the days of Gordon Gekko and his massive brick of a cellular phone.Explanatory Note[Note i] There is a bit of confusion out there as to what constitutes “5G”. Part of the reason is that there are essentially two different levels of 5G implementation:The first is something called “non-standalone” which means augmenting the existing 4G network with 5G hardware that will focus on ultra-high-bandwidth data services.The second is called “standalone” which means everything can go on the 5G network.It is somewhat analogous to the difference between a plug-in hybrid vehicle like the Chevy Bolt and an electric-only vehicle like Tesla.Roll-outs for “non-standalone” 5G implementation are happening in 2019–2020 throughout most of the world — for example, Verizon announced that “5G services” would begin in 2019[33]. However, China is planning a particularly aggressive roll-out schedule for “standalone” 5G compared to every other country with scale deployments in 2020.Whether or not this is the right strategy remains an open question.Footnotes[1] Wall Street (1987) - Wake up call (Drop it)[2] Time-division multiple access - Wikipedia[3] Videotape format war - Wikipedia[4] GSM - Wikipedia[5] cdmaOne - Wikipedia[6] Glenn Luk's answer to Are there any survivors of 9/11 on Quora?[7] Glenn Luk's answer to Why is the smartphone industry dominated by the U.S. and East Asian nations (e.g. Japan, South Korea and China)?[8] Nokia 8310, giocando a Snake II / playing Snake II[9] Glenn Luk's answer to Will China become an innovator?[10] CDMA microcellular telephone system and distributed antenna system therefor[11] Which Are the Most Valuable Patents in Qualcomm Patent Portfolio? - GreyB[12] https://csis-prod.s3.amazonaws.com/s3fs-public/legacy_files/files/publication/130215_competitiveness_Huawei_casestudy_Web.pdf[13] https://newsroom.cisco.com/dlls/Cisco_Mot_for_PI.pdf[14] Glenn Luk's answer to Is there an indigenous Chinese product that is the best in the world?[15] Glenn Luk's answer to How does China finance its development projects in Africa and South America?[16] INTERVIEW - Huawei makes aggressive push in consumer devices[17] Caring for Employees - Huawei Sustainability[18] 1998年江泽民宣布“军队不再经商” 震惊世界[19] From Windfalls to Pitfalls: Qualcomm’s China Conundrum - MacroPolo[20] Glenn Luk's answer to Where does the money I pay for an iPhone go?[21] Debbie Downer - Wikipedia[22] 3GPP - Wikipedia[23] https://www-file.huawei.com/-/media/corporate/pdf/annual-report/annual_report2017_en.pdf[24] Cost of Living Comparison Between[25] Blog: How many global base stations are there anyway?[26] Subscribe to read | Financial Times[27] China Mobile Confirms Aggressive 5G Standalone Plan | Light Reading[28] Secretary Ross Announces Activation of ZTE Denial Order in Response to Repeated False Statements to the U.S. Government[29] Trump says he would intervene in Huawei case to help secure China trade deal[30] Apple is still selling iPhones in China despite being ordered not to[31] Glenn Luk's answer to Where does the money I pay for an iPhone go?[32] Extreme ultraviolet lithography - Wikipedia[33] Verizon’s first 5G hotspot will launch in 2019

M.S. Longair: Theoretical concepts in physics, 1986.An alternative view of theoretical reasoning in physics for final year undergrads.Arnold Sommerfeld: Lectures on Theoretical PhysicsSommerfeld is God for mathematical physics.Richard Feynman: The Feynman lectures on Physics (3 vols)Highly recommended texts compiled from the undergraduate lecture course given by Feynman.Jearle Walker: The Flying Circus of PhysicsThere is the entire Landau and Lifshitz series. They have volumes on classical mechanics, classical field theory, E&M, QM, QFT, statistical physics, and more. Very good series that spans the entire graduate level curriculum.The New Physics edited by Paul Davies.This is one big book and it takes time to look through topics as diverse as general relativity, astrophysics, particle theory, quantum mechanics, chaos and nonlinearity, low-temperature physics and phase transitions. Nevertheless, this is an excellent book of recent (1989) physics articles, written by several physicists/astrophysicists.Richard Feynman: The Character of Physical LawIn his unique no-nonsense style, Feynman lectures about what physics is all about. Down-to-earth examples keep him from straying into the kind of metaphysics of which he is often critical.David Mermin: Boojums all the way through: Communicating science in prosaic languageFrank Wilczek and Betsy Devine: Longing for the Harmonies: Themes and variations from modern physicsGreg Egan: Permutation CityThis is a science fiction novel which has more to say about the philosophy of physics than do most philosophers and physicists.Don't read: The Physicist's World, by Thomas Grissom. We include this book as an example of a book that contains mostly incorrect physics. Grissom is a philosopher who has managed to publish a book about physics without knowing much physics, and it's a shame that he has taught the content of this book for some (many?) years to philosophy students, who must've gone out into the big world thinking that physicists must be incredibly dumb if they really believe the naïve concepts that Grissom thinks physics is all about. This book gets all the big tenets of the subject wrong: Grissom thinks that special relativity is all about what is seen with the eye, a mistake that only first-year students are expected to make; he thinks that the Heisenberg Uncertainty Principle concerns the limits of measurement of quantities that are otherwise perfectly well defined; he thinks that the Second Law of Thermodynamics is an actual law thatmust be obeyed. And apparently he thinks that physicists spend a great deal of their time pondering the philosophy of the Ancient Greeks. All completely wrong.Classical MechanicsHerbert Goldstein: Classical Mechanics, 2nd ed, 1980.Intermediate to advanced; excellent bibliography.Introductory: The Feynman Lectures, vol 1.Keith Symon: Mechanics, 3rd ed., 1971 undergrad. levelH. Corbin and P. Stehle: Classical Mechanics, 2nd ed., 1960V.I. Arnold: Mathematical methods of classical mechanics, translated by K. Vogtmann and A. Weinstein, 2nd ed., 1989.The appendices are somewhat more advanced and cover all sorts of nifty topics. Deals with geometrical aspects of classical mechanicsR. Resnick and D. Halliday: Physics, vol 1, 4th Ed., 1993Excellent introduction without much calculus. Lots of problems and review questions.Marion & Thornton: Classical Dynamics of Particles and Systems, 2nd ed., 1970.Undergrad level. A useful intro to classical dynamics. Not as advanced as Goldstein, but with real worked-out examples.A. Fetter and J. Walecka: Theoretical mechanics of particles and continuagraduate level text, a little less impressive than Goldstein (and sometimes a little less obtuse)Kiran Gupta: Classical Mechanics of Particles and Rigid Bodies (1988)At the level of Goldstein but has many more worked-out problems at the end of each chapter as a good illustration of the material. Very useful for preparations for the PhD Qualifying Examination (I presume this is America only — ed.).Classical ElectromagnetismJackson: Classical Electrodynamics, 2nd ed., 1975Intermediate to advanced, the definitive graduate(US)/undergraduate(UK) text.Purcell: Berkeley Physics Series Vol 2.You can't beat this for the intelligent, reasonably sophisticated beginning physics student. He tells you on the very first page about the experimental proof of how charge does not vary with speed.plus... Chen, Min, Berkeley Physics problems with solutions.Reitz, Milford and Christy: Foundations of Electromagnetic Theory 4th ed., 1992Undergraduate level. Pretty difficult to learn from at first, but good reference, for some calculations involving stacks of thin films and their reflectance and transmission properties, for e.g. It's a good, rigorous text as far as it goes, which is pretty far, but not all the way. For example, they have a great section on optical properties of a single thin film between two dielectric semi-infinite media, but no generalization to stacks of films.Feynman: The Feynman Lectures, Vol. 2Lorrain & Corson: Electromagnetism, Principles and Applications, 1979Resnick and Halliday: Physics, vol 2, 4th ed., 1993Igor Irodov: Problems in Physics Excellent and extensive collection of EM problems for undergrads.William Smythe: Static and Dynamic Electricity, 3rd ed., 1968For the extreme masochists. Some of the most hair-raising EM problems you'll ever see. Definitely not for the weak-of-heart.Landau, Lifshitz, and Pitaevskii:Electrodynamics of Continuous Media, 2nd ed., 1984Same level as Jackson and with lots of material not in Jackson.Marion and Heald: Classical Electromagnetic Radiation, 2nd ed., 1980.Undergraduate or low-level graduate.Quantum MechanicsQED: The strange theory of light and matter Richard Feynman.One need no longer be confused by this beautiful theory. Richard Feynman gives an exposition that is once again and by itself a beautiful explanation of the theory of photon-matter interactions. Taken from a popular, non-technical lecture.Cohen-Tannoudji: Quantum Mechanics I & II&, 1977.Introductory to intermediate.Liboff: Introductory Quantum Mechanics, 2nd ed., 1992Elementary level. Makes a few mistakes.Sakurai: Modern Quantum Mechanics, 1985Sakurai: Advanced Quantum Mechanics1967Good as an introduction to the very basic beginnings of quantum field theory, except that it has the unfortunate feature of using "imaginary time" to make Minkowski space look Euclidean.J. Wheeler and W. Zurek (eds.): Quantum Theory and Measurement, 1983On the philosophical end. People who want to know about interpretations of quantum mechanics should definitely look at this collection of relevant articles.C. DeWitt and N. Graham: The Many Worlds Interpretation of Quantum MechanicsPhilosophical. Collection of articles.H. Everett: Theory of the Universal WavefunctionAn exposition which has some gems on thermodynamics and probability. Worth reading for this alone.Bjorken and Drell: Relativistic Quantum Mechanics/ Relativistic Quantum Fields(for comments, see under Particle Physics)Ryder: Quantum Field Theory, 1984Guidry: Gauge Field Theories: an introduction with applications 1991Messiah: Quantum Mechanics, 1961Dirac:a] Principles of QM, 4th ed., 1958b] Lectures in QM, 1964c] Lectures on Quantum Field Theory, 1966Itzykson and Zuber: Quantum Field Theory, 1980Advanced level.Slater: Quantum theory: Address, essays, lectures.Good follow on to Schiff.note: Schiff, Bjorken and Drell, Fetter and Walecka, and Slater are all volumes in "International Series in pure and Applied Physics" published by McGraw-Hill.Pierre Ramond: Field Theory: A Modern Primer, 2nd edition. Volume 74 in the FiP series.The so-called "revised printing" is a must, as they must've rushed the first printing of the 2nd edition because it's full of inexcusable mistakes.Feynman: The Feynman Lectures, Vol. 3A non-traditional approach. A good place to get an intuitive feel for QM, if one already knows the traditional approach.Heitler & London: Quantum theory of moleculesJ. Bell: Speakable and Unspeakable in Quantum Mechanics, 1987An excellent collection of essays on the philosophical aspects of QM.Milonni: The quantum vacuum: an introduction to quantum electrodynamics1994.Holland: The Quantum Theory of MotionA good bet for a strong foundation in QM.John von Neumann: Mathematical foundations of quantum mechanics, 1955.For the more mathematical side of quantum theory, especially for those who are going to be arguing about measurement theory.Schiff: Quantum Mechanics, 3rd ed., 1968A little old. Not much emphasis on airy-fairy things like many worlds or excessive angst over Heisenberg UP. Straight up QM for people who want to do calculations. Introductory graduate level. Mostly Schrodinger eqn. Spin included, but only in an adjunct to Schrodinger. Not much emphasis on things like Dirac eqn, etc.Eisberg and Resnick: Quantum Physics of Atoms, Molecules, Solids, Nuclei, and Particles, 2nd ed., 1985.This is a basic intro. to QM, and it is excellent for undergrads. It is not thorough with the mathematics, but fills in a lot of the intuitive stuff that most textbooks do not present.David Saxon: Elementary Quantum MechanicsA decent undergraduate (senior level) text.Bethe and Jackiw: Intermediate Quantum MechanicsP.W.Atkins: Quanta: A Handbook of conceptsShort entries, arranged alphabetically, emphasis on stuff relevant to quantum chemistry. Concentrates on the intuition and not the mathematics.James Peebles: Quantum Mechanics (1993)Intermediate level, based on lectures given by the author at Princeton. Very lucid exposition of the standard material with outstanding selection of mostly original problems at the end of each chapter.Statistical Mechanics and EntropyDavid Chandler: Introduction to Modern Statistical Mechanics, 1987R. Tolman: Prinicples of Statistical Mechanics. DoverKittel & Kroemer: Statistical ThermodynamicsBest of a bad lot.Reif: Principles of statistical and thermal physics.The big and little Reif statistical mechanics books. Big Reif is much better than Kittel & Kroemer. He uses clear language but avoids the handwaving that thermodynamics often gives rise to. More classical than QM oriented.Felix Bloch: Fundamentals of Statistical Mechanics.Radu Balescu: Statistical PhysicsGraduate Level. Good description of non-equilibrium stat. mech. but difficult to read. It is all there, but often you don't realize it until after you have learned it somewhere else. Nice development in early chapters about parallels between classical and quantum statistical mechanics.Abrikosov, Gorkov, and Dyzaloshinski:Methods of Quantum Field Theory in Statistical PhysicsHuw Price: Time's Arrow and Archimedes' PointSemi-popular book on the direction of time by a philosopher. It has been controversial because of its criticism of physicists such as Hawking for their "double standards" in dealing with the old problem on the origin of the arrow of time. It is thought provoking and clearly written.The following 6 books deal with modern topics in (mostly) classical statistical mechanics, namely, the central notions of linear response theory (Forster) and critical phenomena (the rest) at level suitable for beginning graduate students.Thermodynamics, by H. Callen.Statistical Mechanics, by R. K. PathriaHydrodynamic Fluctuations, Broken Symmetry, and Correlation Functions, by D. ForsterIntroduction to Phase Transitions and Critical Phenomena, by H. E. StanleyModern Theory of Critical Phenomena, by S. K. MaLectures on Phase Transitions and the Renormalization Group, by N. GoldenfeldCondensed MatterCharles Kittel: Introduction to Solid State Physics (ISSP),introductoryAshcroft and Mermin: Solid State Physics,intermediate to advancedCharles Kittel: Quantum Theory of Solids.This is from before the days of his ISSP; it is a more advanced book. At a similar level. . .Solid State Theory, by W. A. Harrison(a great bargain now that it's published by Dover)Theory of Solids, by Ziman.Fundamentals of the Theory of Metals, by AbrikosovHalf of the book is on superconductivity.Many-Particle Physics, G. Mahan.Advanced.Special RelativityTaylor and Wheeler: Spacetime PhysicsStill the best introduction out there.Relativity: Einstein's popular exposition.Wolfgang Rindler: Essential Relativity. Springer 1977With a heavy bias towards astrophysics and therefore on a more moderate level formally. Quite strong on intuition.A.P. French: Special RelativityA thorough introductory text. Good discussion of the twin paradox, pole and the barn etc. Plenty of diagrams illustrating Lorentz-transformed coordinates, giving both an algebraic and geometrical insight to SR. (Seems to be out of print)Abraham Pais: Subtle is the Lord: The Science and Life of Albert EinsteinThe best technical biography of the life and work of Albert Einstein.Special Relativity and its Experimental Foundations Yuan Zhong ZhangSpecial relativity is so well established that its experimental foundation is often ignored. This book fills the gap and will be of relevance to many discussions in sci.physics.relativityParticle PhysicsKerson Huang: Quarks, leptons & gauge fields, World Scientific, 1982.Good on mathematical aspects of gauge theory and topology.L. B. Okun: Leptons and quarks, translated from Russian by V. I. Kisin, North-Holland, 1982.T. D. Lee: Particle physics and introduction to field theory.Itzykson: Particle PhysicsBjorken & Drell: Relativistic Quantum MechanicsOne of the more terse books. The first volume on relativistic quantum mechanics covers the subject in a blinding 300 pages. Very good if youreally want to know the subject.Francis Halzen & Alan D. Martin: Quarks & Leptons,Beginner to intermediate, this is a standard textbook for graduate level courses. Good knowledge of quantum mechanics and special relativity is assumed. A very good introduction to the concepts of particle physics. Good examples, but not a lot of Feynman diagram calculation. For this, see Bjorken & Drell.Donald H. Perkins: Introduction to high energy physicsRegarded by many people in the field as the best introductory text at the undergraduate level. Covers basically everything with almost no mathematics.Close, Marten, and Sutton: The Particle ExplosionA popular exposition of the history of particle physics with terrific photography.Christine Sutton: Spaceship NeutrinoA good, historical, largely intuitive introduction to particle physics, seen from the neutrino viewpoint.Mandl, Shaw: Quantum Field TheoryIntroductory textbook, concise and practically orientated. Used at many graduate departments as a textbook for the first course in QFT and a bare minimum for experimentalists in high energy physics. Chapters on Feynman diagrams and cross-section calculations particularly well written and useful.F.Gross: Relativistic Quantum Mechanics and Field TheoryI am familiar with first part only (rel. QM) which I warmly recommend in conjunction with Mandl, since Klein-Gordon and Dirac Equation are explained in greater detail than in Mandl. One of my professors likes a lot the rest of the book too, but I haven't spent much time on it and can't comment. Published in 1993.S. Weinberg: The Quantum Theory of Fields, Vol I,II, 1995 It's the usual Weinberg stuff: refreshing, illuminating viewpoints on every page. Perhaps most suitable for graduate students who already know some basics of QFT. Unfortunately, this book does not conform to Bjorken-Drell metric.M.B. Green, J.H. Schwarz, E. Witten:Superstring Theory (2 vols)Although these two volumes do not touch the important new developments in string theories they are still the best texts for the basics. To keep up with this fast developing subject it is necessary to download the papers and reviews as hep-th e-prints.M. Kaku: Strings, Conformal Fields and TopologyJust a little more up-to-date than GSWSuperstrings: A Theory of Everything ed P.C.W. DaviesThrough transcripts of interviews with Schwarz, Witten, Green, Gross, Ellis, Salam, Glashow, Feynman and Weinberg we learn about string theory and how different physicists feel about its prospects as a TOE. This also predates the new developments which revolutionised string theory after 1993.A Pais: Inward BoundThis can be regarded as a companion volume to his biography of Einstein (see special relativity section). It covers the history of particle physics through the twentieth century but is best for the earlier half.R.P. Crease, C.C. Mann: The Second Creation 1996Another history of particle physics in the twentieth century. This one is especially good on the development of the standard model.. Full of personal stories taken from numerous interviews, it is difficult to put down.L. Lederman, D. Teresi: The God Particle: If the Universe Is the Answer, What Is the Question? 2006This book describes the search for the Higgs Boson at Fermilab. It describes what the Higgs is and gives some background to the subject of particle physics. It also gives an account of some more general physics history.General RelativityMeisner, Thorne and Wheeler:Gravitation W. H. Freeman & Co., San Francisco 1973Usually referred to as MTW. It has two tracks for different levels. A famous work in the subject whose main strength is probably its various asides, historical and otherwise. While it has much interesting reading, it is not a book to learn relativity from: its approach is all over the place, and it pushes gawdy notation which no one actually uses to do anything useful.Robert M. Wald: Space, Time, and Gravity: the Theory of the Big Bang and Black Holes.A good non-technical introduction, with a nice mix of mathematical rigor and comprehensible physics.Schutz: A First Course in General Relativity.A readable and useful book, to a point. The 1988 edition, at least, unfortunately has a tangled approach to its Lambda index notation that is wrong in places. Schutz goes to great lengths to convince the reader of the usefulness of one-forms, but is clearly unaware that everything he does with them can be done in shorter time using vectors alone. Beware the show-stopping typos in the Riemann components for the Schwarzschild metric on page 315. The discussion about Riemann tensor signs on page 171 is also wrong, and will give you wrong results if you apply it. Indeed, that discussion is indicative of a general naïveté in the book's early mathematics as a whole.Weinberg: Gravitation and CosmologyA good book that takes a somewhat different approach to the subject.Hans Ohanian: Gravitation & Spacetime(recently back in print)For someone who actually wants to learn to work problems, ideal for self-teaching, and math is introduced as needed, rather than in a colossal blast.Robert Wald: General RelativityA more advanced textbook than Wald's earlier book, appropriate for an introductory graduate course in GR. It strikes just the right balance, in my opinion, between mathematical rigor and physical intuition. It has great mathematics appendices for those who care about proving theorems carefully, and a good introduction to the problems behind quantum gravity (although not to their solutions). I think it's MUCH better than either MTW or Weinberg.Clifford Will: Was Einstein Right? Putting General Relativity to the TestNon-technical account of the experimental support for GR, including the "classic three tests", but going well beyond them.Kip Thorne: Black Holes and Time Warps: Einstein's Outrageous LegacyAn award-winning popular account of black holes and related objects with many historical anecdotes from the author's personal experiences. The book is famous for the final sections about time travel through wormholes.Ignore Dirac's small book on lectures in GR, unless you like reading books that have almost no discussion of their mathematical content (and almost no discussion of anything else, either). It's a sure bet that this book was only published because Dirac wrote it.Mathematical MethodsMorse and Feshbach: Methods of Theoretical Physics. This book used to be hard to find, but can now be bought atfeshbachpublishing.com.Mathews and Walker: Mathematical Methods of Physics. An absolute joy for those who love math, and very informative even for those who don't. [This has been severely disputed!--ed]Arfken: Mathematical Methods for Physicists Academic PressGood introduction at graduate level. Not comprehensive in any area, but covers many areas widely. Arfken is to math methods what numerical recipes is to numerical methods — good intro, but not the last word.Zwillinger: Handbook of Differential Equations. Academic PressKind of like CRC tables but for ODEs and PDEs. Good reference book when you've got a differential equation and want to find a solution.Gradshteyn and Ryzhik: Table of Integrals, Series, and Products AcademicTHE book of integrals. Huge, but useful when you need an integral.F.W. Byron and R. Fuller: Mathematics of Classical and Quantum Physics (2 vols)is a really terrific text for self-study; it is like a baby version of Morse & Feshbach.Nuclear PhysicsPreston and Bhaduri: Structure of the NucleusBlatt and Weisskopf: Theoretical Nuclear PhysicsDeShalit and Feshbach: Theoretical Nuclear PhysicsThis is serious stuff. Also quite expensive even in paper. I think the hard cover is out of print. This is volume I (structure). Volume II (scattering) is also available.Satchler: Direct Nuclear ReactionsWalecka: Theoretical Nuclear and Subnuclear Physics (1995)Covers advanced topics in theoretical nuclear physics from a modern perspective and includes results of past 20 years in a field which makes it unique. Not an easy material to read but invaluable for people seeking an updated review of the present status in the field.Krane: Introductory nuclear physicsIntroductory-to-intermediate level textbook in basic nuclear physics for senior undergraduates. Good, clear and relatively comprehensive exposition of "standard" material: nuclear models, alfa, beta, gamma radioactivity, nuclear reactions. . . Last edition issued in 1988.CosmologyJ. V. Narlikar: Introduction to Cosmology.1983 Jones & Bartlett Publ.For people with a solid background in physics and higher math, THE introductory text, IMHO, because it hits the balance between mathematical accuracy (tensor calculus and stuff) and intuitive clarity/geometrical models very well for grad student level. Of course, it has flaws but only noticeable by the Real Experts (TM). . .Hawking: A Brief History of TimeThe ghost-written book that made Popular Science popular, but an odd mixture of easy physics and very advanced physics.Weinberg: First Three MinutesA very good book. It's pretty old, but most of the information in it is still correct.Timothy Ferris: Coming of Age in the Milky Way and The Whole ShebangMore Popular Science, and very readable.Kolb and Turner: The Early Universe.At a more advanced level, a standard reference. As the title implies, K&T cover mostly the strange physics of very early times: it's heavy on the particle physics, and skimps on the astrophysics. There's a primer on large-scale structure, which is the most active area of cosmological research, but it's really not all that good.Peebles: Principles of Physical Cosmology.Comprehensive, and on the whole it's quite a good book, but it's rather poorly organized. I find myself jumping back and forth through the book whenever I want to find anything.Black Holes and Warped Spacetime, by William J. Kaufmann III.This is a great, fairly thorough, though non-mathematical description of black holes and spacetime as it relates to cosmology. I was impressed by how few mistakes Kaufmann makes in simplifying, while most such books tend to sacrifice accuracy for simplicity.M.V. Berry: Principles of Cosmology and GravitationThis is very well written, and useful as an undergrad text.Dennis Overbye: Lonely Hearts of the Cosmos The unfinished history of converge on Hubble's constant is presented, from the perspective of competing astrophysics rival teams and institute, along with a lot of background on cosmology (a lot on inflation, for instance). A good insight into the scientific process.Joseph Silk: The Big BangI consider Silk's book an absolute must for those who want a quick run at the current state of big bang cosmology and some of the recent (1988) issues which have given so many of us lots of problems to solve. [of course that's eons out of date now--ed.]Bubbles, voids, and bumps in time: the new cosmology edited by James Cornell.This is quite a nice and relatively short read for some of the pressing issues (as of 1987-88) in astrophysical cosmology.T. Padmanabhan: Structure formation in the universeA no-nonsense book for those who want to calculate some problems strictly related to the formation of structure in the universe. The book even comes complete with problems at the end of each chapter. A bad thing about this book is that there isn't any coverage on clusters of galaxies and the one really big thing that annoys the hell outta me is that the bibliography for each chapter is all combined in one big bibliography towards the end of the book which makes for lots of page flipping.P.J.E. Peebles: The large-scale structure of the universeThis is a definitive book for anyone who desires an understanding of the mathematics required to develop the theory for models of large scale structure. The essential techniques in the description of how mass is able to cluster under gravity from a smooth early universe are discussed. While I find it dry in some places, there are noteworthy sections (e.g. statistical tests, n-point correlation functions, etc.).Andrzej Krasinski: Inhomogeneous Cosmological ModelsIf you are blinded by the dogma of the cosmological principle, this book is a real eye opener. A technical, historical and bibliographical survey of possible inhomogeous universes from solutions of general relativity.Alan Lightman and Roberta Brawer:Origins: The lives and worlds of modern cosmologists, 1990Transcripts of interview with 27 of the most influential cosmologists from the past few decades. This book provides a unique record of how their cosmological theories have been formed.AstronomyHannu Karttunen et al. (eds.):Fundamental Astronomy.The very good book covering all of astronomy (also for absolute beginners) AND still going into a lot of detail for special work for people more involved AND presenting excellent graphics and pictures.Pasachoff: Contemporary AstronomyGood introductory textbook for the nontechnical reader. It gives a pretty good overview of the important topics, and it has good pictures.Frank Shu: The physical universe: an introduction to astronomyThis is a really grand book, which covers a huge sweep of physics in its 600-odd pages. Not only does it describe the field of astronomy in great detail, but it also covers in detail the laws of classical and quantum mechanics, atrophysics and stellar evolution, cosmology, special and general relativity; and last but not least, the biochemical basis of life. In fact the last few chapters would make a great addition to a biochemist's library!Kenneth R. Lang: Astrophysical formulae: a compendium for the physicist and astrophysicistHere is everything you wanted to know (and more!) about astrophysical formulae on a one-line/one-paragraph/one-shot deal. Of course, the formulae come complete with references (a tad old, mind you) but it's a must for everyone who's working in astronomy and astrophysics. You learn something new every time you flip through the pages!Plasma Physics(See Robert Heeter's sci.physics.fusion FAQ for details)Numerical Methods/SimulationsJohnson and Rees: Numerical AnalysisAddison WesleyUndergraduate level broad intro.Numerical Recipes in X (X = C, Fortran, Pascal, etc.) Tueklosky and PressYoung and Gregory: A survey of Numerical Mathematics Dover 2 volumes.Excellent overview at grad. level. Emphasis toward solution of elliptic PDEs, but good description of methods to get there including linear algebra, matrix techniques, ODE-solving methods, and interpolation theory. Biggest strength is it provides a coherent framework and structure to attach most commonly used numerical methods. This helps understanding about why to use one method or another. 2 volumes.Hockney and Eastwood: Computer Simulation Using Particles Adam HilgerGood exposition of particle-in-cell (PIC) method and extensions. Applications to plasmas, astronomy, and solid state are discussed. Emphasis is on description of algorithms. Some results shown.Birdsall and Langdon: Plasma Physics via Computer SimulationsPIC simulation applied to plasmas. Source codes shown. First part is almost a tutorial on how to do PIC. Second part is like a series of review articles on different PIC methods.Tajima: Computational Plasma Physics: With Applications to Fusion and Astrophysics Addison Wesley Frontiers in physics Series.Algorithms described. Emphasis on physics that can be simulated. Applications limited to plasmas, but subject areas very broad, fusion, cosmology, solar astrophysics, magnetospheric physics, plasma turbulence, general astrophysics.Fluid DynamicsD.J. Tritton: Physical Fluid DynamicsG.K. Batchelor: Introduction to Fluid DynamicsS. Chandrasekhar: Hydrodynamics and Hydromagnetic StabilitySegel: Mathematics Applied to Continuum Mechanics Dover.Nonlinear Dynamics, Complexity, and ChaosPrigogine: Exploring ComplexityOr any other Prigogine book. If you've read one, you read most of of them (A Poincaré recurrence maybe?).Guckenheimer and Holmes: Nonlinear Oscillations, Dynamical Systems, and Bifurcations of Vector Fields SpringerBorderline phys./math. Advanced level. A nuts-and-bolts "how to" textbook. They let the topic provide all the razzmatazz, which is plenty if you pay attention and remember the physics that it applies to.Lichtenberg, A. J. and M. A. Lieberman (1982): Regular and Stochastic Motion. New York, Springer-Verlag.Ioos and Joseph: Elementary Stability and Bifurcation Theory. New York, Springer.Heinz Pagels: The Dreams Of ReasonHe is a very clear and interesting, captivating writer, and presents the concepts in a very intuitive way. The level is popular science, but it is still useful for physicists who know little of complexity.M. Mitchell Waldrop: ComplexityA popular intro to the subject of spontaneous orders, complexity and so on. Covers implications for economics, biology etc and not just physics.Optics (Classical and Quantum), LasersMax Born and Emil Wolf: Principles of Optics: Electromagnetic Theory of PropagationStandard reference.Sommerfeld:For the more classically minded.Allen and Eberly: Optical Resonance and Two-Level Atoms.For quantum optics, the most readable but most limited.Goodman: Introduction to Fourier Optics.If it isn't in this book, it isn't Fourier optics.Quantum Optics and Electronics (Les Houches Summer School 1963 or 1964, but someone has claimed that Gordon and Breach, NY, are going to republish it in 1995), edited by DeWitt, Blandin, and Cohen- Tannoudji, is noteworthy primarily for Glauber's lectures, that form the basis of quantum optics as it is known today.Sargent, Scully, & Lamb: Laser PhysicsYariv: Quantum ElectronicsSiegman: LasersShen: The Principles of Nonlinear OpticsMeystre & Sargent: Elements of Quantum OpticsCohen-Tannoudji, Dupont-Roc, & Grynberg: Photons, Atoms and Atom-Photon Interactions.Hecht: OpticsA very good introductory optics book.Practical Holography by Graham Saxby, Prentice Hall: New York; 1988.This is a very clear and detailed book that is an excellent introduction to holography for interested undergraduate physics people, as well as advanced readers, especially those who are interested in the practical details of making holograms and the theory behind them.Mathematical PhysicsLie Algebra, Topology, Knot Theory, Tensors, etc.These are books that are sort of talky and fun to read (but still substantial--some harder than others). These include things mathematicians can read about physics as well as vice versa. These books are different than the "bibles" one must have on hand at all times to do mathematical physics.Yvonne Choquet-Bruhat, Cecile DeWitt-Morette, and Margaret Dillard-Bleick:Analysis, manifolds, and physics (2 volumes)Something every mathematical physicist should have at his bedside until he knows it inside and out--but some people say it's not especially easy to read.Jean Dieudonne: A panorama of pure mathematics, as seen by N. Bourbaki, translated by I.G. Macdonald.Gives the big picture in mathematics.Robert Hermann: Lie groups for physicists, Benjamin-Cummings, 1966.George Mackey: Quantum mechanics from the point of view of the theory of group representations, Mathematical Sciences Research Institute, 1984.George Mackey: Unitary group representations in physics, probability, and number theory.Charles Nash and S. Sen: Topology and geometry for physicists.B. Booss and D.D. Bleecker: Topology and analysis: the Atiyah-Singer index formula and gauge-theoretic physics.Bamberg and S. Sternberg: A Course of Mathematics for Students of PhysicsBishop & Goldberg: Tensor Analysis on Manifolds.Dodson & Poston: Tensor Geometry.Abraham, Marsden & Ratiu: Manifolds, Tensor Analysis and Applications.M. Nakahara: Topology, Geometry and Physics.Morandi: The Role of Topology in Classical and Quantum PhysicsSinger, Thorpe: Lecture Notes on Elementary Topology and GeometryL. Kauffman: Knots and Physics, World Scientific, Singapore, 1991.C. Yang and M. Ge: Braid group, Knot Theory & Statistical Mechanics.D. Kastler: C-algebras and their applications to Statistical Mechanics and Quantum Field Theory.Courant and Hilbert: Methods of Mathematical Physics WileyReally a mathematics book in disguise. Emphasis on ODEs and PDEs. Proves existence, etc. Very comprehensive. 2 volumes.Cecille Dewitt is publishing a book on manifolds that should be out soon (maybe already is). Very high level, but supposedly of great importance for anyone needing to set the Feynman path integral in a firm foundation.Howard Georgi: Lie Groups for Particle Phyiscs Addison Wesley Frontiers in Physics Series.Synge and Schild.Atomic PhysicsMax Born: Atomic PhysicsA classic, though a little old.Gerhard Herzberg: Atomic spectra and atomic structure, Translated with the co-operation of the author by J. W. T.Spinks. — New York, Dover publications, 1944Old but good.E. U. Condon and G. H. Shortley: The theory of atomic spectra, CUP 1951G. K. Woodgate: Elementary atomic structure, 2d ed. Oxford: New York: Clarendon Press, Oxford University Press, 1983, c 1980Introductory level.Alan Corney: Atomic and laser spectroscopy, Oxford, New York: Clarendon Press, 1977Excellent,fairly advanced, large experimental bent, but good development of background. Good stuff on lasers (gas, dye)Low Temperature Physics, SuperconductivityThe Theory of Quantum Liquids, by D. Pines and P. NozieresSuperconductivity of Metals and Alloys, P. G. DeGennes A classic introduction.Theory of Superconductivity, J. R. SchriefferSuperconductivity, M. TinkhamExperimental techniques in low-temperature physics, by Guy K. White.This is considered by many as a "bible" for those working in experimental low-temperature physics.Given you more than you demanded

Here is a complete list starting from zero, assuming you have completed high school level mathematical courses: algebra, geometry, and trigonometry. This is a (almost) complete list assuming you want to fully understand Machine Learning (ML) algorithms from a mathematical, theoretical, and implementation standpoint. ML has more to do with math than anything else.If you want to simply become a ML hobbyist (Kaggle competitions, personal projects, basic data analysis, etc) then most of this is not needed. Before starting to learn ML I’d recommend learning Python and then getting a quick introduction to the math for ML (basic Calculus, Linear Algebra, Probability & Statistics). For example, a good introduction can be found in this Coursera Specialization:Mathematics for Machine Learning | CourseraEdit: several subtopics listed are not directly needed in machine learning but they help form a strong theoretical foundation. Additionally, mastering every topic listed (including optional ones) is over kill; I made this list for breadth. The equivalent of 1–2 college semester of each required topic is enough.Required Mathematics:Calculus: techniques of integration; separable differential equations; infinite sequences and series; Taylor's Theorem and other approximations; curves and vectors in the plane. Vectors in space; partial derivatives; Lagrange multipliers; multiple integrals; vector analysis; line integrals; Green's Theorem, Gauss's Theorem.Linear Algebra: Solution of systems of linear equations, matrices, vector spaces, bases, linear transformations, eigenvalues, eigenvectors.Differential Equations: Homogeneous and non-homogeneous linear differential equations, existence and uniqueness theorems, Gronwall's inequality; systems of first order linear differential equations; autonomous first-order systems: critical points, stability, bifurcation; series and periodic solutions, Fourier series and their convergence. The analytical and geometric theory of ordinary differential equations.Statistics: Populations and random sampling; sampling distributions; theory of statistical estimation; criteria and methods of point and interval estimation; theory of testing statistical hypotheses. Least square principles in multiple regression and their interpretations; estimation, hypotheses testing, confidence and prediction intervals, modeling, regression diagnostic, model selection, and analysis of variance and covariance.Probability Theory: Probabilities of events on discrete and continuous sample spaces; random variables and probability distributions; expectations; transformations; simplest kind of law of large numbers and central limit theorem.Algorithms: Algorithms for searching, sorting, manipulating graphs and trees, scheduling tasks, finding shortest path, matching patterns in strings, cryptography, matroid theory, linear programming, max-flow, etc., and their correctness proofs and analysis of their time and space complexity. Strategies for designing algorithms, e.g. recursion, divide-and-conquer, greediness, dynamic programming. Limits on algorithm efficiency are explored through NP completeness theory.Discrete Mathematics: Sets, propositions, induction, recursion; combinatorics; binary relations and functions; ordering, lattices and Boolean algebra; graphs and trees; groups and homomorphisms.The following are very advanced but optional mathematical topics to add extra strength to your background. This reaches MS/PhD level understanding of ML; you will be able to understand the latest published ML research:(optional) Numerical Methods: Representation of numbers and rounding error; polynomial and spline interpolation; numerical differentiation and integration; numerical solution of nonlinear systems; Fast Fourier Transformation; numerical solution of initial and boundary value problems; Monte Carlo methods.(optional) Optimization: Linear programming, integer programming, stochastic programming, and nonlinear programming using an algebraic modeling language. Convex analysis, unconstrained and constrained optimization, duality theory, Lagrangian relaxation, and methods for solving nonlinear optimization problems, including gradient descent methods, Newton methods, conjugate gradient methods, and penalty and barrier methods.(optional) Bayesian Statistical Topics: Maximum likelihood estimation; Gibbs sampling; Markov chain Monte Carlo; Metropolis Hastings algorithm; variational inference; expectation maximization; latent Dirichlet allocation; Gaussian mixture models; hidden Markov models; matrix factorization; Bayesian networks.(optional) Graph Theory: Sequences and recursive formulas; counting formulas; bijections; inclusion/exclusion; the Pigeonhole Principle; generating functions; equivalence relations. Graph theory topics include trees, connectivity, traversability, matching and coloring.(optional) Game Theory: A mathematical analysis of how people interact in strategic situations. Applications include strategic pricing, negotiations, voting, contracts and economic incentives, and environmental issues(optional) Control Theory: State-space representation of linear systems, linearization, time-variance and linearity properties of systems, impulse response, transfer functions and their state-space representations. Dynamic analysis of linear feedback systems in the time and frequency domain, with emphasis on stability and steady-state accuracy. Major analytical tools: signal-flow graphs, root-locus methods. Nyquist plot, Bode analysis. Cascade compensation techniques.(optional) Computability Theory: enumeration, index and recursion theorems, various models of computation and Church's Thesis, uncomputability results, introduction to reducibilities and their degrees (in particular, Turing degrees, or degrees of uncomputability), computable operators and their fixed points.(optional) Stochastic Processes: Generalization of the Poisson process; renewal theory, queueing, and reliability; Brownian motion and stationary processes; stopping times; martingales; Markov processes; Brownian motion; stochastic calculus; Brownian bridge, laws of suprema; Gaussian processes.(optional) Dynamic Programming: sequential decision making under uncertainty, exact solution algorithms, and approximate methods adapted to large-scale problems. Value iteration, policy iteration and lambda-policy iteration are introduced and analyzed using fixed-point theory.(optional) Statistical Signal Processing: Hypothesis Testing as applied to signal detection. Various optimality criterion including Bayes and Neyman-Pearson. Optimum and locally optimum detection schemes for Gaussian and non-Gaussian noise. Estimation of unknown signal parameters.(optional) Information Theory: Development of information measures for discrete and continuous spaces study of discrete-stochastic information courses, derivation of noiseless coding theorems, investigation of discrete and continuous memoryless channels, development of noisy channel coding theorems.Computer Science:Programming and Data Structures: Algorithmic design and implementation; object oriented programming. Classes, subclasses, recursion, searching, sorting, linked lists, trees, stacks, queues.Software Engineering: The software lifecycle; lifecycle models; software planning; testing; specification methods; maintenance.Database Systems: Design of large databases: integration of databases and applications (SQL); performance tuning; data mining and data warehouses.Some Introduction to Data Science: Exploratory data analysis; focusing on the computational analysis of data to extract knowledge and insight. Collection, storing, preparation, analysis, modeling, and visualization of data.Python, Python, Python, Python, Python!