blastdoor

Mike Wuerthele said:

neilm said:

The (few) people who need the particular features the Mac Pro has to offer probably know that, and aren’t seriously considering a Studio.

The M3 Ultra Studio is faster at most tasks than the M4 Max, but not by all that much. I don’t think that most buyers would choose to pay the premium for that fairly slim edge. Where the M3 Ultra Studio comes into its own is for workloads requiring massive amounts of RAM. These include AI development, large scale data science, complex engineering simulations, and other things I know nothing about. Again, those users will know that’s what they need and find the choice straightforward. All they have to figure out is how to get it paid for!

That's the thing, though. It's not faster at most tasks, it's faster at nearly all of the massively parallel ones.

The M4 Max is faster at most daily tasks, which are single-core ones.

If someone is buying a computer for "most daily tasks, which are single-core ones" then they should be buying a Mac mini or a MacBook Air. They have no need for the Max and certainly not the Ultra. 

This is the problem with most reviews of the Mac Studio. They are written by people who are not in the target audience for the Mac Studio, make little effort to understand that audience, use benchmarks that aren't relevant to that audience, and so they always end up concluding that the Mac Studio isn't worth the price for most people. Which is true, but one doesn't need a review to know that. Geekbench is a great benchmark for "most users", so if you fall into the "most users" audience, just go look at the geekbench browser. 

The Mac Studio is the Mac analog to a Threadripper type system. Places like Anandtech (before they died) and Phoronix are good at reviewing those types of systems. Anandtech used to do reviews that are relevant to the Mac Studio target audience. Phoronix only does linux systems, so not very helpful to Mac folks. 

If anybody knows of a website that is able to do a relevant review of a Mac Studio class product and compare it to relevant alternatives -- like Linux threadripper systems -- please let me know. 

Current MacBook models rely on Qualcomm-based Wi-Fi chips to connect the devices to wireless internet. 

doesn't Broadcom make the wifi chips? 

proline said:

It would be a horrifically bad idea for Taiwan to export their latest process to America, matched only by Ukraine's idiotic idea of giving up their nukes in the 1990s. What incentive would the USA have to protect Taiwan if their tech was already in the USA? 

Chips are not nukes and chips will not protect Taiwan. It’s highly unlikely that the US will or even can defend Taiwan now. Taiwans best bet is to get actual nukes. Second best bet is to invest in a huge number of missiles and drones. But it’s probably too late if they haven’t already started. 

randominternetperson said:

To make matters even worse, Trump's suggestion that the US can annex Greenland for national security reasons greatly strengthens China's arguments regarding Taiwan.

However, China would be crazy to invade Taiwan now since it's impossible to predict how Trump would react. Anything from a congratulatory post on Truth Social to a nuclear strike on China is conceivable.

The notion that we'd annex Canada is another example. At least China can say that Taiwan was kinda-sorta part of China at one time (even though the CCP never controlled it). There's zero justification for what Trump is saying. 

I agree that invading Taiwan is crazy, but it's less crazy now, because with Trump the probability of the US successfully defending Taiwan is much lower than with Biden. There's no way Trump is going to use nukes. Trump is a bully, and bullies are fundamentally cowards who never start fights that they could potentially lose.

My guess is that China will not invade Taiwan (at least not right away) but instead will blockade it. That blockade is highly likely to lead to surrender by Taiwan unless the US leads a sustained effort to run the blockade (which might fail even if competently executed, but will definitely fail if incompetently executed or not even tried).  The reason that the blockade will eventually work is that Taiwan needs food from abroad. Either the Taiwanese will surrender to avoid starvation or the blockade will be followed by an invasion after the population is severely weakened by starvation. 

If you live in Taiwan and don't want to die of starvation or live under CCP rule, my suggestion is to get out while you still can. 

The silicon shield won’t work. Trump won’t be able to defend Taiwan even if he wanted to because he has filled the national security leadership with fools and broken the civil service.