- Last Active
lkrupp said:Does this article mean to imply that Apple is unaware of this and is not tracking it? Seems so. Only Microsoft cares about Apple users, not Apple?
Google’s annual revenue:
2021: $258 billion
2020: $183 billion
2019: $162 billion
2018: $137 billion
2017: $111 billion
2016: $90 billion
2015: $75 billion
2014: $66 billion
2013: $56 billion
2012: $46 billion
2011: $38 billion
2010: $29 billion
2009: $24 billion
While Google clearly benefited from the pandemic if you look at their rate of growth prior to it ($75 billion in 2015 to $162 billion in 2019) they would have still surpassed $200 billion in 2021 without it because they were already growing by $20 billion annually before then.
Another thing: unlike Facebook, Apple's App Tracking Transparency only had a marginal effect on Google.
Another thing: were the DOJ to separate "Google" (search/ads) from "Alphabet" (everything else) and prevent the two from coordinating, they would both still be massive companies. Alphabet, for example, would have had at least $60 billion in revenue last year.
So yeah, definite justification for their stock going up.
No use arguing with this crowd. For them, "patent troll" means "anyone who sues Apple" even if Apple is actively, knowingly using the vital IP of a major, practicing competitor without paying for it i.e Qualcomm LTE modems without which iPhones would be worthless as mobile devices and NPEs can include entities that certainly perform original research but actually cannot legally manufacture and market their IP such as ARM CPU IP owned by the University of Wisconsin's engineering program. For them, any IP that Apple wants to exploit is only worth what Apple is willing to pay - including nothing - because Apple's decision to include the IP in their products is what gives their IP any value in the first place. If the same IP is in billions of products sold by Apple's competitors? Then why are they bothering Apple in the first place! They are making plenty of money already so just trying to get more money from Apple is just being greedy.
But Apple's IP? The hard-earned result of decades of expert engineering. No one should be able to use it or any version of it under any terms or circumstances. Apple should be encouraged to come out with their own versions of competitors' products because said products are generic and bad anyway and Apple improves them in ways that are useful to consumers and the industry. But anyone who comes out with a version of Apple's products, even if the implementation is completely different AND there are major differences in design, appearance and functionality? Block them from being manufactured and sold. Sue the companies that did it to the hilt - and their managers and investors into financial ruin - as a warning message to everyone else.
A rational perspective - such as the fact that Blackberry deserves to be compensated for the IP that they developed even if the only method of doing so is by selling it and that the owners of any IP have every right to license it - is beyond what Apple diehards are capable of. Instead, they want an IP system that maximizes the benefit to Apple and screws everybody else, even if the obvious result is no one - including Apple - having any incentive to innovate.
wood1208 said:It was very good acquisition for Nvidia to push ARM in data centers(low power consumption with higher performance) but UK not giving up it's crown jewel, few good companies UK have left
1. Not even Apple claims that the M1, M1 Pro and M1 Max has higher performance than Intel Core i9 and AMD Ryzen 9 chips.
And for good reason:
Apple has - from day one - only claimed that their chips will have the best power-per-watt.
2. And the main reason why Apple offers the best power-per-watt is always being on the latest process. Apple is on 5nm currently with 4nm chips coming later this year. AMD is on 7nm and HOPES to get to 5nm this year. Intel was stuck on 14nm for 6 years and is finally on 10nm. But even on 7nm, the AMD Aerith SOC used for the Steam Deck will draw an average of 7W (4W min, 15W max). If loaded with Windows 11 and given a USB-C dock, it would be a more powerful PC than most productivity workers have despite being no bigger, heavier or thicker than the original iPad Mini.
3. Another reason why Apple offers the best power-per-watt: the CPU, GPU and RAM are all in the same SOC and they have written macOS to take maximum advantage of it. Please realize that servers and even high performance workstations cannot use a similar architecture or OS. They require much more RAM and GPU power than is currently possible to fit on a single die. Apple's unified approach means that it will generally have advantages over systems that have the exact same configuration (i.e. CPU cores, system RAM, integrated GPU) but at the cost of flexibility. The x86 workstation and server systems will be able to add CPU, RAM and GPU power to base systems, and do so rather cheaply. Apple Silicon doesn't offer that flexibility. You can only pay a ton of money to add CPU, GPU and RAM, and even that can only be added to a certain point.
Of course, ARM servers don't use this approach: they also have CPU, RAM and GPU separate. But similarly, they don't benefit from the advantages of having them all on a single die - RAM in particular is a huge power draw, which is why smartphone batteries needed to get a lot bigger once RAM increased from 2-3 GB to up to 18 GB for some Android gaming phones now - which narrows the gap between them and x86 even further.
Both Amazon and Microsoft are increasing their use of ARM servers in AWS and Azure, but they are nowhere close to ditching their Intel Xeon or AMD Epyc servers yet. Google meanwhile is doubling down on x86 and just did a massive order of AMD Epyc CPUs, which is big because they heavily relied on Intel for quite awhile.
Look folks. This plant doesn't have squat to do with China. This plant is being built so Intel can catch TSMC. Even if Intel doesn't catch TSMC getting "close enough" - like staying one generation behind is fine. Not everyone needs to be on the latest, best node. Qualcomm switched to Samsung - which is one generation behind TSMC due to inferior transistor density - for their flagship chips and it hasn't affected them one bit. MediaTek - until this year - avoided TSMC's latest nodes to save money and have now surpassed Qualcomm as the #1 chipseller in the world. AMD has never used TSMC's latest nodes either. Their current chips are 7nm while TSMC is releasing 4nm chips for their mobile partners. By the time the TSMC backlog FINALLY allows AMD's 5nm Zen 4 chips on the market 4Q2023, Apple's 4nm M2 chips will be out for the 2nd gen MacBook Air, 13' MacBook Pro and Mac Mini refreshes as well as for the 1st gen iMac Pro and Mac Pro.
Getting "close enough" would definitely allow Intel to get business from Nvidia, who also doesn't use the latest process node: they are a generation behind on TSMC and have also used Samsung. It would also be good enough for Qualcomm. Where their flagship chips are currently being made on Samsung's 4nm (which at least allows them to compete with TSMC's 4nm on paper) their non-flagship chips are being made on Samsung's 5nm and TSMC's 7nm. Qualcomm would be able to shift that to Intel, save themselves a lot of money and get their products to market faster.