6:01 Their blockbuster product was indeed a clock generator replacing multiple crystals, but not exactly in the way described. It was a programmable RF PLL frequency generator, which was able to generate several different user selectable pixel clocks for a videocard, while using a single crystal as a reference. Most of their products at the time were focused on video and audio generation for personal computers and professional video gear.
Agilent’s layoffs in 2002-2003 were odd. I was one of those laid off, even though our group was making 50% profit year after year. But, alas, we were part of a division that was doing poorly. So everybody had to share in the pain “of being workforce managed.” Struck me as illogical given the success of the group I was in.
@@foobarf8766I am not conspiracy theory minded but some of the happenings in the economic space over the years makes me wonder if the boom and bust cycles are manufactured to benefit some at the expense of others. Now let me remove my tinfoil hat and go watch more YT videos lol
Hearing about Seagate got their SSD business from LSI is an amazing things. This channel is like a "behind-the-scenes" curtain look at all the actual chips and systems LTT (and others) review and play with. Another excellent video Asianometry!
I hate broadcom for the last 20 years of making net drivers and failing SAN connectivity, that name for me is for Long a nogo, Great to know that Qualcomm shares the shares with broadcom.
What got Agilent is the same thing that sunk Nortel. There was a major slowdown in telecom expenditures on equipment starting in early 2001. Everything got overbuilt and new technologies squeezed even more bandwidth from existing infrastructure (mostly fibre). So aglient starting pushing inventory into their distribution channels, this is called stuffing the channel.. At the time this could be recorded as revenue because the product had shipped even though it had not been paid for. So along comes the next quarter and Agilents distributors tell them their warehouses are full so Agilent slashes prices and offers credit deals to carriers with low or no interest. They clean out the channel (at a loss) and then push more inventory into it. Well eventually no one wants any more hardware period. Nortel was the first to blink and announce they were taking a one time charge which was due to dead inventory. Soon other players followed suit and that was the death of Nortel and drove others into either bankruptcy or right next to it. In the aftermath I can remember seeing cellular infrastructure for sale at about 10% of its original value.
This kind of events seem puzzling, but they are useful to make certain things etched in memory, which in the end made it worth investing, even at a loss (for me). And it's nice to have someone clearly recall what happened back then in 2001.
VMWare was a very poorly run business with a great product. If it wasn't Hock wouldn't have bought it, and the value wouldn't have gone up so much when he did.
There's one more step to the Braedcom cycle of buying a company and then slimming it down by selling parts. They also massively increase prices as seen with VMware recently
@micro-organism-pv5gd None of those mentioned are hedge funds. BlackRock and Adobe aren't private equity either. Blackstone is itself not a private equity fund, it is a management company that sponsors private equity funds. You can invest in stock in Blackstone the management company; or if you're an accredited investor, you can invest in one of their funds. Kinda like how most landlords and HOAs hire a property management company. I guess it would be most accurate to call Broadcom a technology oriented leveraged mutual fund. The closest analog would be Berkshire Hathaway, which is technically an insurance company, but the only time it acts like an insurance company is when filing its taxes.
I was there too. What good times!HP were a major client of mine at a retail advertising agency in Sydney. ...come to think of it though, so were Nokia and Dell! What good times we had. Seinfeld got cropped to widescreen, Minidisc was replaced with MP3s, TVs went plasma. Cameras were cameras and phones were phones. I'd do anything to go back.
@@gus473 I was frequently in Palo Alto. Roughly once a month. I was base in Ft Collins. I no longer deal with tests and measurements. Somewhere in all my HP gear is a brand new never been opened HP iPod, and an HP television. Probably a pile of used iPAQ's (we were making them data logging devices). I've worked with some really great engineers at different corporations over the years... but there was a difference around HP Palo Alto.
Always wondered how Google made its first chips so fast and successful. Never realized that Google's TPU was actually Broadcom, it makes so much sense now.
@@MrTweetyhackNot sure but he said computational lithography, not physical lithography. So nVidia designs and simulates the lithography process, TSMC physically produces the chips.
They appear to be planning to milk that cow until it dies. There are a lot of Broadcom/Brocade chips in enterprise servers, so maybe they make enough money from VMware or bare metal, so in the short term it doesn't matter which. In the long run there are other options.
They did the same to Symantec Endpoint Protection. This video doesn’t cover the way Broadcom acquires companies and then immediately smothers vital documentation and guidance which is critical to any software used in enterprise environments. It’s really heartbreaking and frustrating.
The likely reason for this is that the history of Broadcom isn't really that relevant for the company named Broadcom today. They use the name, but that is basically it
They are bad even for enterprise customers now, see what they do with vmware after they acquired it. I would rather have mediatek wifi modules than broadcom ones in phones.
@@viktorbaresic4180 enterprise will be their only customers going forward and every year theyll tighten the noose a little more until theyve wringed every penny out of the husk
I'm so glad you've started about technology like filters and the science of waves and frequencies, etc. You are very correct that that stuff is unheralded and quite overlooked, as it is the lynchpin of the digital world today.
My lab has an Agilent Technologies LCMS and HPLC machine. I had no idea about there history or connection to HP. Grate video as alwayse and I hope your friends and family are safe after the earthquake!
They're great products, we had a tough time competing with them in GC & HPLC. They didn't get much traction beyond those analytical techniques, however ¯\_(ツ)_/¯
@@gus473At one point they also bought Varian, one of the two leading Nuclear Magnetic Resonance spectroscopy equipment vendors. (The second one was Bruker.) And then they killed it, because NMR machines, expensive as they are, do not sell in very large numbers, unlike HPLC systems.
Observation: The odd company on the list of the top 12 companies displayed at 0:12 is Saudi Aramco (#4) with share price under $9, while the rest listed are all above $140. It's also a non-tech company in a tech dominated list.
The actual stock price is not an indicator of anything. It’s just market cap divided by shares. SA being $9 and others being more doesn’t tell you anything.
I remember Broadcom from the days in the mid-1990s when the tiny company I worked for was doing communications ASIC design - we followed the IEEE papers published by BRCM exec Henry Samueli and his grad students at UCLA for cool digital comms techniques. I also remember the minor scandal when his partner exec Henry Nicholas was discovered to have a "sex dungeon" under his LA home and was eventually drummed out of the company for various reasons. It's been a long strange journey for that company.
I used to work for the 'Inside' company that now seems to be going to hell in a handbasket, while other well run visionary giants eat their lunch; all having had equal access to the same tech playing field, but some playing soooo much better.
Interesting factlet: Sophie Wilson, who designed the instruction set architecture of the ARM, has worked for Broadcom for many years. If they _had_ acquired Qualcomm, it would have been something of a "coming home" for her. But of course not as much as if they had acquired Arm!
Hi, sorry for my poor English, but I like your videos very much and even learn some new words from they. Could you please make a video about Ten Major Construction Projects of the ROC and new Ten, I think it's very interesting subject. Thank you and greetings from Russian Far East!
The focus on 'solutions' looks neat and tidy to senior managers, but holds back technology. Interesting components that enable progress is not there. That includes the iPhone which is just another mobile phone.
Probably a funny anecdote: 20 years ago I had a coworker being obsessed with investing in the stock market. He asked me: You are a computer guy, what's the next big thing you can imagine? I said WiFi. (WiFi was still in the third generation and basically worse than the common 100 MBit or GBit Ethernet.) In my new laptop and computer and home there is this chip from a company called Broadcom, that's my best guess. I felt a bit bad giving that advice and learning a few years after that Linux at the time basically hated these WiFi chips. And the ARM processors. But then the company got richer and richer and went on a shopping marathon of accumulating other companies. I would have loved to see his reaction over the last 20 years, first realizing my terrible judgement, then the incredible luck of picking some company name where plenty of others failed.
Broadcom is onto a "nice little earner" where I hope someone will think of competing with it: a light-emitting diode with the lens moulded so that you can plug a fibre-optic lead into it that costs about a hundred times the price of a normal light-emitting diode.
How did Broadcom win the TPU deal back in 2016? they for sure did not see this to be so huge right now, but i guess still some competition back in 2016
Listening to your excellent commentary makes me realize how unqualified I am to analyze these companies as investment opportunities. They can rocket to the moon and then fall to the depths of the ocean and I would have no idea why they did either. But thanks as always for your excellent analysis.
The companies for AI chips are a fascinating odd-bunch of ASIC focused solution providers. Broadcom with TPUs (Tensor), NVIDIA with a GPU (Graphics) history. AMD with VPU (Video). Sony (aka Sony Semiconductor Solutions Corporation) also has announced AI processors based on its ASIC image sensor technology, and has Microsoft partnership. Samsung announced its ASIC AI chips and a Meta partnership. Broadcom established an early Google partnership. Tesla relied on TSMC to manufacture its Dojo D1 chip, but is not selling the chips. The list of competitors in the ASIC AI space-race is rapidly accelerating! (I likely overlooked many others) It will be interesting to see what architectural standards and nomenclature will emerge to handle LLM's and ML to build on top of a silicon foundation.
"Publicly traded private equity-fund masquerading as a semiconductor-company" sums it up nicely :) No mention of the Raspberry-Pi? (Agreed, that the revenues from the Pi are chump change for Broadcom, but don't underestimate the kudos-value (of being affiliated with the RPi) among the geeks that unknowing rule teh interwebs.
i think my rpi has a chip that has the words, "Broadcom" on it
2 месяца назад
You shouldn't use market cap to conpare company sizes. Use enterprise value instead. Market cap only takes stock into account. You also want to take bonds into account. If tomorrow Apple shifted their capital structure from equity to bomds, nothing about their business wpuld change. But market cap would change, while enterprise value would stay the same.
At 19:25.. the house you show was a Levittown (PA) post war (WW2) type house called the 'NewYorktown" .. and the real big problem is the investor dividend payout.. as were there is none, and a big glitch in the advance or ability of consumption, then they (those stocks of non-paying dividends) collapse like a giant ponzi scheme and the question of rightful success and payment of taxes also comes into question. Broadcom fits into this category of negligent Fascism (as they have curried governmental favoritism).
Indeed SG555 was a Signetics product. But also this product was not representative of the chips which ICS developed. They were making video frequency pixel clock generators, and other similar products, which took a single quartz crystal and allowed to the user to multiply it as needed to produce pixel rate required for a given screen resolution. Since video cards supported multiple resolutions, a variety of pixel clock frequencies were required. Using the programmable PLL solved this problem. One crystal was still needed, to serve as a stable reference.
@@cogoid - yes, the venerable 555 can have its frequency adjusted, but IIRC it couldn’t operate at the frequencies that would be needed by a video clock, but even worse the frequency would not be as stable as a crystal.
@@stevebabiak6997 Yes. Video clock in the era was already getting into 100 MHz range, while the original 555 topped at some 100s of KHz. More importantly, for video applications, the frequency and phase must be very stable, otherwise deviations produce image distortions that are immediately visible. ICS chips offered such stability. I do not think one could do it with an RC based oscillator even if the speed itself were not an issue.
I don't really follow the use of the term "franchise" by Hock Tan, which is even illustrated in this video with a photo of a McDonald's restaurant. How are these franchises? They sound more like divisions of the business than franchises. Who would be the franchisees here?
I was thinking exactly the same thing. My comment was going to be: "Franchises. Tan keeps using that word. I do not think it means what he thinks it means."
@@carterthaxton Yes, I think he is using it like movie "franchises", as in the Spiderman or Batman or other random comic book "franchise". A brand or theme to be milked endlessly until the punters tire of it, VMware customers take note.
17:10 The LSI HBA adapters are also just great. I use an older one in my TrueNAS server, as do many others. However, some of Broadcom's business practices are not stellar; I refer those interested to the VMware acquisition.
If you're interested in a corporate auction saga and how that worked back in the day, read the book "Barbarians at the Gate." Classic. And in it KKR (mentioned in this video) is among the profiled companies.
we'll see how Broadcom does when Apple completely divests .. supposedly by 2025 they will no longer rely on Broadcom, and I'm not so sure theor vision with AI is on par with competition. it will be fun to watch
