Yes, you can have too many CPU cores - Ampere's 192-core chips break ARM64 Linux kernel in two-socket systems, company requests higher core count support
namingthingsiseasy @ namingthingsiseasy @programming.dev Posts 6Comments 294Joined 2 yr. ago
Normally, you’d run a cluster of multiple servers to host such workloads, but imagine if all those resources were available on one physical hosts - it’d be a lot more effecient, since at the very least, you’d be avoiding all that network overhead and delays.
Exactly! Imagine you have two services in a data center. If they have to communicate a lot with each other, then you would prefer them as close to each other as possible. Why? Well it's because of the difference between sending a request over a network vs. just sending it to another process on the same host. It's much more efficient in terms of latency and bandwidth. There are, of course, downsides and other other costs (like the fact that the cores that are handling the requests themselves are much less powerful), so you have to tailor your hardware allocation to your workloads. In general, if you're CPU-bound, you would want more powerful CPUs (necessitating fewer cores per host for power reasons), and if you're I/O bound, you want to reduce network latency as much as possible.
Now imagine you have thousands of services. The network I/O can get pretty extreme. Plus, occasionally, you have requirements like the fact that any data traveling from one host to another must be encrypted. So if you can keep as many services as possible on a single host, you reduce a lot of that overhead as well.
tl;dr: everything comes down to trade-offs and understanding the needs of your workloads, but in general, running 300 low power cores is probably indicative of an I/O-bound application and could hypothetically be much more efficient and cost-effective.
I'm not the guy you responded to, nor am I a kernel expert, but I have a few suggestions:
- Sites like phoronix and lwn will go into pretty low-level kernel details like this from time to time. You could consider subscribing to their RSS feeds or something like that
- Review a few open university courses on either Operating Systems or Computer Architecture. Short of that, you can also just browse wikipedia for articles on these kinds of topics. I find it enjoyable to read them from time to time
- Subscribe to the LKML (which is probably a lot more information than any single person can process, but sites like lwn and phoronix highlight/summarize from time to time)
I would also say that there are a lot of people out there who have made contributions to the Linux kernel, including this specific portion of the Linux kernel. The person you're responding to may even do it as a part of his/her day job (and it certainly reads like he does). It's not that uncommon.
And the last thing to keep in mind is that learning knowledge like this doesn't happen overnight. You learn a lot more by learning small things over several years, compared to learning a lot in a short time. Don't make it a goal to learn things like this - instead, try to make it something you enjoy doing, so you keep doing it over the years and learning more and more small bits of knowledge over time. Eventually, all the different pieces start fitting together and you too could mash out an excellent post like GP's!
at least if Microsoft wants to be in compliance with EU rules on tracking
"if" doing a lot of work in that sentence. Even if the EU comes down on them for this, the fines usually end up being less than the cost of doing business. And it's not easy to prove in a court in the first place.
I think companies know and understand this, so they just end up doing it anyway and pay the inevitable fine. And that assumes that the fine comes at all - even if they pay a fine for this practice, there are probably so many others that they're not being punished for that it still makes sense for them to ignore it.
I really hope this is something that gets addressed though, as things are getting absurdly out of hand by this point.
This is why you should use your own domain. If you want to change who's handling your email, you just change your DNS MX record to a new, different host and all your mail ends up there instead. The services don't have to know a single thing about what's going on - the next time they send an email out, DNS will simply resolve to the new mail server.
Here is an example of how you would do it with Proton
First thing is to not mount it at all. Any writes to the overwritten partition will corrupt your data.
Second thing: install system rescue cd to a live usb and boot it. Look into testdisk
and photorec
. It's been a while since I've had to use these tools, but I believe testdisk
can restore the partition and photorec
can find files in a file system that has been deleted. I would try running photorec
first to save the recovered files to an external hard disk, and then testdisk
to try restoring them. But disclaimer: it's been a while since I've had to do this, so my memory is foggy here.
Good luck!
"Waste your Money" Day
while it doesn’t have the most of nutrients
You can always add fruit to it though. Frozen berries in oatmeal are fantastic. I also like adding peanut butter too.
both view themselves as the one rightful Chinese government
This is a bit of an outdated view in my opinion. If you're a KMT voter (ie. 60+), then sure, this is a common view. Younger generations (DPP voters) however don't really view themselves as Chinese. I think this view will die out eventually.
Of course, for the most part, this is all off the official record because of the implications. Chinese nationalists will argue that this is wrong because it's still written in the Taiwanese constitution or whatever, but the truth is that regular people in Taiwan couldn't give less of a shit about China. De facto, most Taiwanese consider themselves their own country with no legitimate claims to China.
Yup, he's the absolute worst. I can't think of a single product that Google has even improved during his tenure as CEO... let alone a product that he successfully launched on his own. All he can do is make existing products more expensive.
Look at how he completely lost his shit when chatgpt was released - probably a huge part of the reason he lost it is cause he realized he'd have to actually do something useful instead of just squeezing more blood from the collective stone of all Google's existing products. His claim to fame is creating Chrome. What fucking good is that? Web browsers have existed since the time he was born. There's nothing to innovate there, and there never has been. It's clear: he's not an innovator.
Whoever takes over after he's gone is going to be in for a hell of a time. The only thing he's created for Google is a shit-ton of anti-trust lawsuits. The company is an empty husk at this point. There's nothing left for them to become.
Likewise sending my support. I've always admired NL, especially for its political system and will undoubtedly continue to do so. My hope is that VVD and NSC do the right thing and keep this piece of trash out on the sidewalk where he belongs! Wishing you all the best....
Wow, these new TLDs are terrible! ICANN has really lost it this time!
Yeah, "monthly active users" does not necessarily mean "unique monthly active users".
I would much rather see activity statistics like posts+comments or something like that. As long as those are looking good, then new unique users will continue to join gradually.
I'm not sure I would agree with that. ISO-8601 is ambiguous, and very difficult to parse. For example, here are a couple valid ISO-8601 strings. Could you let me know what they mean?
P1DT1H R10/2021-208/P1Y T22.3+0800 22,3 2021-W30-2 2021-W30-2T22+08 P1Y 20
Taken from here. My favorite is the last one (20
). If someone just wrote 20
and told you to parse it using ISO-8601, what would you get? Hour? It could even be century (ie. 2023%100
)!!
So I would argue that ISO-8601 is just a wee bit too flexible. Personally, I like RFC 3339 just a bit more...
Edit: that said, I would definitely agree that something along the lines of 2021-07-27T14:20:32Z
is better than any regularly accepted alternative, and I pretty much format my dates that way all the time.
I love it when someone sends me a message like this:
Hey there! What are you doing on 4/5?
????
You're absolutely right, but we haven't even touched on the worst part of ads, which is how they utterly poison your brain with annoying jingles, annoying colors, and stupid catch-phrases, all psychologically engineered to get stuck in your head.
And let's not even go into how they prey on your fears and insecurities, or deceive you into thinking you need things that you actually don't. How they prey on vulnerable children, or the elderly, or brainwash small children into manipulating their parents against their best interests. Or how privacy has been shredded since the advent of behavioral tracking.
I'm not exaggerating at all when I say that advertising is one of the world's biggest psychological hazards. I would rather sit in an empty room with no stimulation whatsoever than let that poison into my brain.
The inevitable fate of any useful software that's not GPL.
When will people learn???
Edit: Ironically, KHTML was originally LGPL. So modifications to KHTML were required to be open source by the license, but Chrome itself isn't required to be open source (at least as far as I understand it, I am not an expert here). Nevertheless, if it were stronger GPL, then it probably wouldn't have been impossible to write features like DRM in chrome. So I would have been a bit of an idiot to say that KHTML isn't GPL (because LGPL is a weaker version of GPL), but in effect, the outcome is the same - all because of that big fat L at the beginning.
Same shit, different day. Companies love denying people basic needs and then blaming it on "the computer" (to borrow a '90s parlance). They love it because you can't blame a person, rather you can just say it's some inherent fault of the universe rather than some trash executive's decision to let people die in the name of profits. When their profits are in danger, the universe can be bent to anyone's will to save them; when their profits are rosy, well, chaotic universe theory, what are you gonna do? :shrug:
The whole "AI" aspect of it is nothing new. "AI" is just regular software with different (and highly non-deterministic) algorithms running under the hood. Can't wait till this stupid term is relegated to the dustbin of history.
I’d really like to do some personal projects, both to learn new stuff and scratch my digital itches.
Likewise. There are so many things I'd love to do. Hell, I'd even love to just practice and improve at things that I do in my day job. My current job requires golang and while I'm competent, I'm nowhere near the level of comfort that I want to be. And that's just the language, there are other libraries, technologies, etc. where I'd like to improve too.
But unfortunately, I have to pace myself. And that basically means I can never do anything computer-related in my spare time. A part of me hopes that some day, I can reduce my hours, or just find a very chill job that still pays decently so I can do more important things in my spare time. But for now, this is the choice that I've made.
Yeah, the biggest problem with Firefox is that its engine is so hard to embed. Chrome has endless clones because it's just so damn easy to embed. And Firefox just has some weak forks like Librewolf.
I'd really rather see Mozilla focus on this rather than all their other stupid endeavors....
GPUs are still pretty bad at handling conditional logic and are more optimized towards doing mathematical operations instead.
But you are right in the sense that people are exploring different kinds of hardware for workloads that are getting increasingly specific. We're not in a CPU vs GPU world anymore, but more like a "what kind of CPU do I need?" situation.