331–340 of 693 posts
Earlier quoted context omitted.
Even SSDs throttle under heavy use.
Sure. The NVMe that gives you 2GB/s write may throttle to only 500 MB/s after a few seconds of writes, but-- How often is that a problem, really?
Having a fast machine that can sustain throughput and I/O is a perfectly rational desire, and for some of us, need.
> I know that new "open" laptops are still available, but most of them--the System76's, the Purisms, etc.--are overpriced in my opinion This quote suggest to me the author is an integral part of the problem. Companies are ALWAYS trying to increase their profitability. If they can lock down the computer, they can drop the price while _increasing_ CLTV. That’s why System76 is more expensive. They’re NOT overpriced. The…
How much? I bet it's not much.
> That’s why System76 is more expensive. They’re NOT overpriced. They’re priced as “general purpose computers” must be to be competitive.
They're so much more expensive because they're niche. Not because of anything inherent to the product or producing it.
So "overpriced" sounds fair to me.
> I agree with this author, but this suggests that the strength of his conviction is maybe $400.
Is that not enough? If you can spread the word enough, and get enough people that are willing to spend $50 extra, you'll find plenty of companies begging for that money. The profit margin for building computers is bad.
Earlier quoted context omitted.
I’m from Denmark and when power went out I was television worthy news. I don’t actually recall it happening more than once. I never experienced a power loss in the five years I lived in France. Now I’m in silicon valley, the tech hub of the world - with 3rd world internet and power that goes out every time the wind blows. Just reporting facts.
There was a ~45 minute power cut in the centre of Copenhagen last month, starting at about 1:30am. There was another (several hours) in 2018, which I think is the one that made the news. (I was at home, awake, for both.) They're a bit more common than you remember, but I suspect most of these Tweets are affecting fairly small areas, e.g. construction work cutting cables etc: https://twitter.com/radiuselnet I have som…
I'm not sure what you mean by "most of these Tweets"; I'm reporting on my experience living in Silicon Valley, California, where power goes out a couple of times a month on avg.
I have all servers etc on UPSes but wish I could find bigger UPSes as most can only hold up about 1 hour which often isn't enough.
I agree with the point that modern devices like smartphones ad tables aren't a PC replacement. But having to buy a pre-2012 laptop just to be able to install a Linux distro is false: nowadays mayor distros support secure boot and if this isn't the case of your distro or it complicates the install process you can easily disable it in most of the cases. Actually I was unaware of the existence of motherboards preventing…
Now getting some of that configured to use the cheap hardware is a more complicated story.
Earlier quoted context omitted.
Microsoft has always required (on x86, ARM is different) that for Windows logo certification Secure Boot must be able to be disabled, you must trust Microsoft’s CA and the third-party CA that they run (and that Red Hat/Ubuntu/Debian use to sign their builds) and that the user must be able to load their own keys. This means that any x86 device sold with UEFI Secure Boot can still boot Linux (or if it can’t, it’s not S…
That used to be the case for x86, but not anymore: https://www.zdnet.com/article/microsoft-to-stop-linux-older-...
I prefer a desktop because: 1. I like big, big monitors. 2. I prefer a full size keyboard. 3. I prefer a separate mouse. 4. I prefer big freaking disk drives installed. 5. I put the desktop under my desk, and with a wireless keyboard and wireless mouse, there is much less of a snarl on my desk. 6. The desktop has an optical drive I still use. 7. The desktop has lots of USB ports and they're all in use. 8. I can repla…
I suspect a lot of the movement towards laptops was and is due to the desire to run macOS. It just so happens that Apple desktop computers are either way underpowered or way expensive. An MBP is basically the sanest way to run macOS. So MBPs tend to become makeshift portable desktops, with external screens, keyboard, mice / trackpads, power bricks, and cooling pads.
If your work processes are not beholden to Apple software, building (or ordering a customized) desktop is a refreshing gulp of power. You suddenly can afford much more, or spend way less.
I'm writing this from a Linux desktop, and I'm quite content with it.
Earlier quoted context omitted.
There are obviously many ways to do this comparison, but going off Wikipedia's pricing for most recent Intel processors. Their highest end mobile processor is currently the i7-1185G7 which apparently retails for $426. Let's compare it with the closest desktop processor to that price, the i9-10900 ($438). Cores (Threads): 4(8) vs. 10(20). Base Freq: 3.0ghz (At 28w TDP) vs. 2.8ghz. Max All Core Turbo: 4.3ghz vs. 4.8ghz…
A MacBook Pro can have an i9 and 8 cores, isn't that better than the i7-1185G7?
CAD, for example, barring new geometry work in progress now, is single threaded.
People working on high surface count models want these things:
Big, fat, fast cache
Sustained sequential compute performance
Sustained I/O
GPU that focuses on geometry and precision. This is not generally an issue today, but can be on laptops.
Desktop machines with active cooling are where it is at.
I know I'm probably a dinosaur. I run a lot of desktop applications for work, and run most my models locally. I use the shell to automate tasks, and rarely use the "cloud" and try to limit browser tasks.
What's funny about this post is that I'm actually in the market for a new laptop, and am quite paralyzed in the process. I can't get myself to spend $1500 on some mediocre machine where everything is soldered on the mobo, I can't replace the battery, and that will become garbage in 2 years.
It's frustrating to me (as a UEFI engineer) to see and read the fear-mongering on blogs like these. Much of what I do is focused on open-sourcing and making the industry as a whole better. I tend to side with others in the thread that the "end of general purpose computing" isn't a result of manufacturers pushing in a direction but rather following consumers. The percentage of computer scientists, programmers, and ent…
This is an even stronger claim than that consumers don't care if their hardware is closed. Consumers aren't begging for their hardware to be closed, and companies aren't passively trying to please them. The vast majority of consumers don't care if their hardware is closed until they discover something that their hardware is capable of doing that they're not allowed to do with it or something that their hardware is doing that they're not allowed to keep it from doing. Companies close as much as they can while trying not to upset consumers enough that they leave or complain to the government - which means either that they have to find ways to 1) add or limit functionality in a closed way, or that 2) they have to get all competitors to agree to prevent or allow the same things (to eliminate alternatives), or 3) they have to advocate a POV (and lobby for it in government) that whatever functionality is missing or required is dangerous to add or remove.
This is following consumers in a way; like building a wall behind them every time they move in a particular direction could be considered following them.