>Linus "Tech" Tips
absolute moron
>>56344929
>total GHz
why does being alive have to be so painful /g/?
>>imagine calling yourself linus "tech" tips and actually needing tech tips, lol
>>56344929
He's just a merchandise advertiser for his partners. I don't understand why people are shocked he doesn't know anything about the underlying tech he "reviews".
>>56344929
oh my god i'm cringing source?
>>56344957
You realize that's an actual metric used to show capacity in certain enterprise software, yea?
>>56344929
>watching that cancer
>>>/v/
>reverse image search OPs picture
>It's from reddit
Kill yourself
https://www.reddit.com/r/pcmasterrace/comments/50bydq/goddamnit_linus_we_spend_all_our_time_saying_this/
>>56344929
padding-left:0px;
>>56345104
>implying multi-core softwares are a thing
>>56345179
Jesus what is with the gamer cancer on this board?
>>56345212
>implying gaymen need cpu upgrades
No?
>>56345224
Fuck off, child.
>>56344929
He literally says in the video right before/after the chart is shown that it's not normally a meaningful way to compare performance, but he's using in the context of video encoding which should be highly threaded so the chart shouldn't be absolutely meaningless.
i'm not a fan of excessive multi-threading. for some applications like hosting a server and compiling software but oftentimes like with rendering it makes more sense to do it on the gpu rather than the cpu
I hate Linus with passion but there is literally nothing wrong this chart. Fucking retarded gaymer children.
>>56345953
Why do you hate him?
>>56346080
He's a know nothing retard who fucks up constantly? Like, to the point of jeopardizing his entire business on 2 or 3 separate occasions.
>>56346256
Why would you hate that? It's entertainment
>>56346080
Pretty much this >>56346256
He's nothing but an overrated salesman, hes just a bukkake whore for tech companies. He knows fuck all about actual technology, this becomes apparent when he attempts to do anything that isn't standing in front of a camera reading a script.
>LinusCuckTips
>>56346372
I think it's hilarious, but I can see why others would despise him for this. He's followed by idiots who don't know better and only get dumber for watching him.
>>56345156
>reverse image search
You and I both know you just recognized it from reddit
>>56345212
I suppose cad users are cancer to. Back to your duel core you go.
>>56346420
Inventor does a lot of shit multicore, faggot :^).
>>56344929
>total GHz
>>56345104
>4*10=40
wow youre right i needed someone to do the math for me thank you based anon :^)
>>56345168
>Not using a space
Wait that actually quite sounds like Linus.
how many total GHz did it take him to create the Linux kernel ?
>>56345243
ok, so explain whats more childish, trying discuss something related to a technological hardware, or getting mad at someone for discussing said thing, on an anime image board, when 80% sure you're a manchild who still watched anime.
What's wrong with the chart exactly?
>>56344929
This brings me back when /b/ used to ask if their 3GHz quad core equals 12GHz single core.
>>56344929
>Total Ghz
Why, all you need to care about is GHz per core.
>>56344929
>watching e-celebz
What are you, 12?
>>56346540
>Total GHz
it's somewhat relevant for certain things like video encoding which is what the video is about but when taken out of context it's silly
https://en.wikipedia.org/wiki/Amdahl%27s_law
>>56346573
Well, it does, if they're the same architecture.
>>56346438
Wtf. I'm saying multicore proccessing is useful for more softwares than just vidya.
>>56346530
/g/ is not an anigay board
Keep that in mind at all times, thanks
>>56346637
Can't keep track of this shit. I'm out.
>>56346617
no it doesn't, a 12 GHz single core would be much preferred, like how 6700k or maybe 6900k/6950x is ideal for gaming and most general applications because of the single thread performance, whereas a 22 core xeon is more relevant for servers and a more limited set of applications
>>56346665
>no it doesn't, a 12 GHz single core would be much preferred
Sure, for single threaded workloads. But not multi threaded workloads. In that case, they would perform nearly identically.
>>56344929
That's how you measure CPU resources in VMware vCenter and other server software, I don't see the problem? PC is not just for gayman anon, some people do work.
tfw you remember linux being unable to scale past 4 cores
>>56346789
When, 1999?
>>56346692
no you dumb shit, because you can't split cache coherency of _real_ applications across multiple shit cpus, you need a bunch of _really_fast_ cores since your waifu a shit
/g/: techloligy
>>56344929
>total GHz
JUST
ITT: Leet programmers who have absolutely no understanding of infrastructure
>>56346692
Don't forget about turbo boost, on the upper-end chips there can easily be a 50% difference in clocks depending on how many cores are in use, which makes the cores * frequency measurement even more irrelevant.
>>56346920
Yep
Here's a useful chart when I was looking up the very subject.
>>56346892
k tard
>>56346986
>being underage
>>56346985
sauce plz
>>56346985
Oh shit forgot to download the chart thanks m8 for the reminder
>>56347111
>being delusional
you can't magically utilize all your weak cores for all types of applications, it's literally impossible
[spoiler]JIGAHERTZ[/spoiler]
>>56346789
WWWWWWWWWWWWWWwwhat face.
>>56344929
Any actually useful workload is easily parallelization. Video encoding, hosting a database or web server, computing neural networks.
The only time where >muh single threaded performance matters is for playing games meant for manchildren.
>>56349866
>video encoding
LOL no. Just because it's fine on a quad doesn't mean it scales to 1000 cores.
>>56349900
Except encoding does scale very well
>>56344929
Not only physically wrong but also misleading. There are already established and correct metrics like FLOPS, why pull shit out of your ass? To skip actual testing? To fill the screen with something? But the more pressing matter is that there are actual people who actually go to fucking YOUTUBE to get their fucking reviews. It's like people can't read these days.
>>56349866
>Any actually useful workload is easily parallelization
Uh, no.
>>56349866
maybe if you're a faggot youtuber or web fag or """"ai"""" memer those things are "actually useful workloads"
>>56344957
You dumb fucking mouth breathing virgin. If a program can parallelise its a somewhat reasonable metric. However there is a limit to how well things scale in parallel. Read about Amdahls law.
>>56350352
yes anon i really did need someone to tell 4*10 was 40 youre right
>>56344929
But then the 9590 is on par with the i7 6950X with it's total Ghz of 40
>>56350610
Obviously everyone can do basic multiplication. Its relevant for the point they are making and to highlight the relative performance. However a price / total ghz would have been a better inclusion.
>>56346577
So a 2 core processor is as good as a 24 core processor as long as gigahertz are same?