Thank you. I wasn't sure if anyone else noticed that all this complaining is about a higher performance CPU being put into a phone.

This is false.

The SEC specifies no timeline in the rule stating that corporations must "maximize shareholder return". (or whatever the exact wording is) It's often greedy executives (whose bonuses may be tied to quarterly earnings), or greedy shareholders (people who just want a quick buck, and don't care about the future health of the company), who push for quarterly profits above all else.

Even extremely experimental projects, that never turn a profit, would be difficult to prosecute by activist shareholders - there's a burden of proof that company management willfully squandered assets on something that had no chance of working. You'd probably have to prove that the executives materially benefited from the activity as well, e.g. by showing that major contracts went to other companies owned by the execs.

For most companies, the best way to maximize returns "this quarter" is to sell off all assets and fire everyone. Clearly, that won't do anything for next quarter (in most cases), and it's also clear that most companies don't do this. Obviously, there's some part of the business management and planning process that takes time horizons greater than 3 months into account.

GamersNexus has been doing reviews of pre-built gaming machines lately, and the Dell one was referred to as "a warranty with E-waste attached". Partly because they were billed for a warranty they explicitly declined, and partly because the computer sucked so much. A separate review of an Alienware model is labeled "Hilariously Bad Alienware R10 Ryzen PC: $1800 Pre-Built Review" (here).

The rules are pretty benign. The problem is likely that these specific models have large power supplies, which causes them to be subject to this set of efficiency rules (here). Desktops with a power supply under 600W aren't subject to these rules.

The actual rules basically say that if you have certain kinds of computers that use a lot of power, you must use an 80Plus Gold (or better) power supply, and the computer must go to sleep (ACPI S3) after 30 minutes. It doesn't say that the user can't change the sleep settings after they get the computer, it just has to ship with power-saving defaults.

There are some listing requirements (similar to being "Energy Star Certified"), but since the law has been in place for 3 years, Dell could easily have registered these models if they wanted to. Unless they're using crap PSUs that aren't even 80Plus Gold.

In full Internet Pedant Mode, I was talking about the storage medium, not the readout apparatus.

But you have a point :) That's what debounce capacitors and algorithms are for.

Those are the requirements to call yourself a Professional Engineer, but not to call yourself an engineer. The PE accreditation allows you to do things that engineers (lower-case 'e') can't do (like filing construction plans), and it also gives people who don't know you the knowledge that you have a certain level of knowledge and ethics. The PE exam requirements are somewhat different from state to state, and only licensed PEs are allowed to call themselves PEs.

It's explained pretty well here.

FTFY. Digital storage is really analog storage with a wide noise margin (with the possible exception of punch cards - we'll ignore hanging chads for the moment :) ).

There are two better examples in the engineering world: FPGA design and 3D CAD.

The FPGA design process has some parts that are embarrassingly parallel - compiling many source modules into RTL for example. There are also some phases that must be done sequentially, like place-and-route. Medium-sized FPGA designs (something that takes ~70% of an Artix-7 200k gate device) take about 30-40 minutes on my I9-9880H. Larger designs, like something targeting the multi-million gate Virtex and UltraScale chips (and others from Intel/Altera or Lattice) take hours to synthesize. This gets worse when the design uses more of the FPGA fabric, or when there are tight timing requirements. Parts of the process are iterative, with a dependency on data from prior iterations, which prevents parallel processing.

Similarly, CAD software like SolidWorks has to process the feature tree sequentially, because each feature may be dependent on any or all of the features that came before. There's no good way to tell with absolute certainty whether there might be a dependency, and even if there were, the dependency checks would have to be re-done after every change, no matter how trivial. This is because something simple, like changing a fillet radius, could cause a dependency for a seemingly unrelated feature somewhere down the line, like a hole on one of the surfaces that the fillet touches, which would prevent firing off a separate thread to e.g. work out the geometry of that hole.

These examples are the FPGA and CAD equivalents of waiting for code to compile. That's relevant because the engineer is typically waiting for these processes to complete, so overall processing time (ie, latency) is important. Maximum single-core speed makes a big difference because of the sequential processing, maximum multi-core speed also makes a big difference because of the parallelizable parts.

There's no single answer like you're saying, it's a balancing act depending on the workload.

This may just look like the right time to open it up, due to the factors you cited.

Early on, the superchargers were necessary, partly to alleviate range anxiety, and partly because charging was included in the price of the car. Rather than pay another company for power, they paid to build up a charging network that would eventually become a source of income

Now that there are lots more non-Tesla EVs on the road (whose owners will pay for energy), there are lots more charging competitors (available to Teslas as well as other makes), and energy provides an ongoing revenue stream, it seems like about the right time to open it up.

If there are government incentives to build out more charging infrastructure (which I would hope are only available for non-company-specific chargers), that's just icing on the cake.

To put it in perspective, this is about the same transfer rate as 10,000 PCIe gen 5 lanes. Or 20,000 PCIe gen 4 lanes.

You could barely fill this pipe with a dozen racks of Epyc-based servers (at a measly 128 gen4 lanes each). You'd need 150+ computers, and double that if you actually wanted to read the data from somewhere (like an array of 5000 NVMe drives).

You have a good point about keeping the government out of the way, but there is a constitutional mandate (in the US at least) for the federal government to "promote the general Welfare". I would argue that those words include improving the health of everyone (on average) by discouraging individual actions that cause harm to others. It seems that the trend is for people to consider their individual desires more and more heavily rather than the effects their actions have on others.

My preference would be to have an across-the-board levy on all vehicles, with a sliding scale based on vehicle efficiency. Efficiency may not be the best metric for determining how bad a certain vehicle is for the "general Welfare", so maybe there are other metrics that should be used instead. Note that I don't exclude EVs from this levy - less efficient EVs would have higher levies than more efficient ones. The most efficient vehicles might have a levy of zero, but that's an implementation detail.

The plus side to this method is that the government isn't telling people what they may or may not do, it's simply adding some of the externalized costs to the individual decision process. The money raised from programs like this (for vehicle efficiency) should be used to fund some parts of healthcare (since that's the primary effect of pollution on the populace), as well as climate research and/or mitigating the effects of climate change.