Apple is the only legitimate mainstream threat to Intel

No

that's where you're wrong kiddo
bloomberg.com/news/articles/2018-04-02/apple-is-said-to-plan-move-from-intel-to-own-mac-chips-from-2020

They cheat at benchmarks, don't expect anything from them.
They barely manage to support 3 devices at a time and still manage to get beaten by android phones.

your wish will come true in about a year when apple ditches jewtel.

>anandtech.com/show/13453/apple-iphone-xs-review-addendum-small-core-and-nn-performance
>anandtech.com/show/12694/assessing-cavium-thunderx2-arm-server-reality/7

A12 ~= 3.8 Ghz Skylake CPU

>lets open the netflix app
>iOS has a long an animation for the netflix logo when android doesnt
>LG wins this one
>comparing 2 completely different applications for speed
These videos really hurt to watch

why not and what about specint2006? literally comparing integer performance...

Someone should make a youtube channel where they post iPhones vs android phones but they post 2 videos. Each video cherry picks apps and tests so that 1 phone dominates the other and you just upload them in seperate videos with clickybaity titles.

>Faster CP

I wish they made bigger MacBooks so that they don't throttle down to a halt.

Attached: 1540821075006.jpg (1076x996, 122K)

I should be easy to port some compute benchmark to run inside an iOS app, why no one does it?
Should put this shit to rest.

>bloomberg

Better cooling means the laptop will be more thicc.

No one will ever bother making a new benchmark for IOS that actually means something because they will just get told they are wrong and geek bench is right and will still float the geekbench score across every news outlet

>kosherbench

Attached: 1529810749748.jpg (300x300, 19K)

>apploo CPUs
>a threat to anyone

Attached: 1520060883359.webm (1280x546, 2.79M)

you're one of these retards sheeping this muh benchmarks not legit

geekbench tests shit like jpeg compression which does literally show that these apple chips are capable in these everyday tasks

specint2006 shows that these chips have good integer performance in general

>muh fixed function hardware boogieman
name one reason why it even matters, if it's fast it's fast

>blaming hardware for poor software

I remember how slow, unresponsive, and prone to freezing the g4 Mac's were despite being "way more efficient" compared to Pentium 4 even after a fresh install of os 10.4. This is the same shit again. Their success in one synthetic benchmark doesn't change anything about subpar general performance.

This. Apple learned their lesson after getting caught cheating PPC benchmarks, so now they cheat in hardware with closed source dedicated geekbench cores that no one can look at. Unfortunately their brain dead customers never learn.

Attached: 1515784842527.jpg (529x605, 164K)

>you're one of these retards sheeping this muh benchmarks not legit
I just can't trust that Apple is not doing something to try to inflate these numbers

>geekbench tests shit like jpeg compression which does literally show that these apple chips are capable in these everyday tasks
If hardware is being used to do these benchmarks its unfair to try to compare to CPUs that have to do these things in software

>specint2006 shows that these chips have good integer performance in general
Is it toss your i7 away and replace with a A12X good like the media is trying to hype?

>muh fixed function hardware boogieman
>name one reason why it even matters, if it's fast it's fast
Fixed function hardware becomes near obsolete so much faster than just having the the raw CPU power to do something
A Galaxy S5 from 2014 for example just manages to fail to software decode 10bit H265 fast enough while a mobile i5 from 2011 is still fast enough to software decode the same file with power to spare.

>it's unfair
why and why does it matter? x86 CPUs have fixed function hardware acceleration for tons of shit too

I won't replace my CPU with an A12X because I can't as it's only in a tablet and it's ARM which means no support without slow emulation

also you're assuming that shit like jped will suddenly disappear, so that the accelerator for that would be useless

by the way my i5 4690k can't manage 4k 60fps youtube videos because it lacks hardware acceleration for VP9

jpeg not jped

sorry I think it's H245, not VP9

anyway whatever the 4K 60fps videos use, maxes my CPU utilization and stutters

Porting the MacOS userland to ARM makes no sense for apple. They're making enough of a killing handling two technically distinct products. If everything was based off the same chipset they'd still be maintaining two different userlands, and i'm sure people would be moaning about that too.

Of course what's far more likely is they will kill the mac with a giant iPad, and a big FU to any devs which don't pay the apple tax to run their own custom software and dev environments on apple hardware.

>why and why does it matter? x86 CPUs have fixed function hardware acceleration for tons of shit too
Maybe a decade or two ago they used to now everything is done in microcode now with CPUs actually becoming very RISC now at least for the clusterfuck of x86 instructions
As for fixed hardware they only really started adding dedicated hardware decoders because of laptops for better battery life

I would almost place the blame for that on just youtube's player rather than the CPU just can't do it. I have the same exact CPU and I wish I could get a copy of a 4k60 video to try locally

>Apple is the only legitimate mainstream threat to Intel
Nope, Apple does not have fabs. AMD also don't have fabs, nor do Nvidia. The threat to Intel is TSMC. Who makes the A12X? Apple? No. They did design it but they didn't make it.

Intel is a chip designer and producer, the others are just designers. Intel failed miserably in the mobile market (they really tried pushing Atom CPUs using all their dirty tricks, giving them away below cost, but it didn't work because they were so bad compared to ARM) so TSMC has been making piles of cash which they've re-invested. AMD and Apple and NVidia now have a huge advantage in terms of access to a more advanced 7nm process node.

I would absolutely not rule out Apple making their own laptop/notebook CPU, it's a move I see as much more likely than not. It's just a matter of their current mobile CPU line becoming powerful enough to do the job and we're pretty much at that point already.

I don't see porting MacOS as a problem, it wouldn't be the first time Apple's moved between architectures. Don't forget that Apple products used to have a PowerPC processor. I do see the possibility that they won't, though, it's equally likely that they simply slap a desktop-like GUI on iOS and use that.

/thread
Apple CPUs are a fucking joke.
EVERY single iToy ever released becomes unusable trash after 2 years that lags and chokes when opening up the fucking calculator.
Shit like that has literally never happened in the entire history of any other CPU architechture.

The only thing Apple's managed to accomplish is cheat at benchmarks in hardware, and suckered gullible morons into actually believing them in spite of all actual performance evidence.

>desktop-like GUI on iOS and use that.
They don't even have to do that.
IOS and macOS use the same kernel its just a matter of recompiling macOS specific stuff to ARM

>I wish I could get a copy of a 4k60 video to try locally
They do it, it's not that hard. Install youtube-dl and grab one of the dozens of 4k60fps videos on youtube like this one,
youtube.com/watch?v=7ivXGUW_NFA
use youtube-dl -F to see formats and -f 315+251 to get 4k60fps with Opus audio

I do believe him when he says the 4690k can't into 4k60fps because it does put a pretty high load on both my CPU and GPU. And there's not that much a difference between playing in chromium and playing with mpv. Network bandwidth is an issue too, btw, you may not be able to play 4k60fps without downloading it first (you can run mpv youtube's-url and play it that way) even though your Internet connection is faster. I have 100mbit but connections to youtube caps out around 26mbit for some reason

/thread

My 4690k can do under MPC HC at least

Attached: Screenshot_1.png (1767x1371, 1.93M)

>Jow Forums forgets mobile risc v is coming

You mean TSMC and yes ARM isn't the x86 architecture.
And yes TSMC will be producing a x86 in the comming months it's called Zen2 designed by AMD at the 7nm node.

Attached: 1540756113681.gif (427x240, 1.83M)

>replies 44
>posters 30
Poo in the loo samefag Poogler

Attached: EB0EF815-8462-4DFF-BB26-94A55BEC84BE.jpg (2312x1156, 854K)

ANDREET BTFO

Attached: 8697F015-FC8A-4717-9B46-8B92EAD7412A.jpg (2048x2048, 407K)

>IBM is like 20% of the server market
>AMD is like 20% of desktop machines
>"they're just for enthusiasts!"
Imagine being this much of a delusional iToddler.

>the Apple-designed A-series chips ship in millions of personal devices every year.
Ahahahaha, didn't Apple have that graph showing worldwide iPad sales at 44 million a year? That's the population of Ukraine or Colombia, or 1/7th of the US population. They aren't selling fuckall, you dumb fruitfag.

Attached: 1537723291967.jpg (1036x919, 190K)

What the fuck, is this a joke?

No, it's not a joke. The FagBook Poors thermal throttle so hard that they perform only slightly better than mobile ARM chips.

The only things here that are throttling are your brain and your testicles, because you're SEETHING.

wow, you're dumb

>>IBM is like 20% of the server market
sauce

>point out genuine and critical flaws about appleshit
>you're poor and a samefag lmao
Only women and burger numales buy apple. Since you're no woman, there's a fat chance you're a mutt that cares what niggers think of you before getting fucked in the ass every day

Attached: 1498235913561.png (420x420, 18K)

>goybench
Almost as good as Nvidias goy-ops measurement.

>be homosexual fruitfag
>see discussion about technology
>talk about testicles
Speaking of testicles, you might wanna let your ThrottleBook cool down a bit before it fries your balls. Those three Safari tabs can put a lot of strain on your brand new laptop.

Google it.

>calling the POTUS a numale
Yep. Mohammed confirmed.

>anything intensive and not bound by splash screen animations is much faster on iPhone
I'm shocked.

Will you now tell me SPEC scores don't matter?

Attached: ipc-1.png (816x275, 32K)

Did no one notice how in the "browser test" he types the sites on the address bar on the LG (with chrome autoloading them before you even tap them), while he selects from bookmarks in Safari?

On the LG site that he had to type in on both phones, the iPhone was noticeably faster.

that's crazy. You could put a RISC-V decoder in front of that and it'd be crazy fast. Apple should do that to avoid having to pay ARM anything.

>Apple should do that to avoid having to pay ARM anything.
They also have backwards compatibility to consider, though. Otherwise, they would also hardly be supporting AArch32 any longer.

>all this intelfag salt

first spectre and meltdown, now apple is ramming intel in the ass with a pozzed strapon

They've done a lot of transitions, what's one more

Holy shit, Intel's getting raped using its blood for lube

Makes you wonder what excuse AMD has.

>170% the ipc of intel on average while having half the threads/core
x86 is done
good riddance, I say

All SPEC tests are single-threaded, though.

>213% advantage at gcc compilation
Arm on desktop when?

If only AMD hadn't mothballed K12. I was looking forward to it so much.

High-performance RISC-V on desktop when?

Yes SPEC2006 is obsolete people move to SPEC2017 and test don't count AVX in Intel processors.

And that's that stupid xeon with 28 shitty cores, so this makes some sense, but how does it stack up to a mainstream laptop/desktop core-for-core?

>SPEC2006 is obsolete
But also the only thing that runs on iOS. Also "obsolete" does not mean that the scores are useless.
>test don't count AVX in Intel processors
Well duh, this is the INT test.

>but how does it stack up to a mainstream laptop/desktop core-for-core?
It's not like the desktop versions have 70+% better IPC.

>Also "obsolete" does not mean that the scores are useless.
Literally SPEC say this

Literally they don't.

spec.org/cpu2017/Docs/overview.html#Q22

Q22. What will happen to SPEC CPU2006?
Three months after the announcement of CPU2017, SPEC will require all CPU2006 results submitted for publication on SPEC's web site to be accompanied by CPU2017 results. Six months after announcement, SPEC will stop accepting CPU2006 results for publication on its web site.

After that point, you may continue to use SPEC CPU2006. You may publish new CPU2006 results only if you plainly disclose the retirement (the link includes sample disclosure language).

I assume you posted that in affirmation of my point, because nowhere in that quote does it say that SPEC2006 results are useless.

>SPEC will stop accepting CPU2006 results for publication on its web site.
>if you plainly disclose the retirement
SPEC stop use SPEC2006 as official comparative scores

Of course they would, they published the new version for a reason, obviously. But just because the new suite is better that doesn't mean the old one is useless. Nothing has changed about the suite itself that makes it intrinsically worse than it was 10 years ago.

Either way, again, SPEC2006 is the last version of SPEC that runs on iOS, so there isn't really much choice.

>Nothing has changed about the suite itself that makes it intrinsically worse than it was 10 years ago.

Yes, but real world workloads has change more complex data and way more data to process begin reason SPEC2017 was made with more complex workload to aproximate real world workloads.

>SPEC2006 is the last version of SPEC that runs on iOS, so there isn't really much choice.
No, SPEC is just code for compilaton but SPEC2017 use way more RAM than best iOS on market.

>No, SPEC is just code for compilaton but SPEC2017 use way more RAM than best iOS on market.
Exactly my point. And which is also why your "real world workloads" doesn't apply here. Clearly, since those workloads are too big to even fit on an iOS device, they are not real-world for that context.

>they are not real-world for that context.
For mobile context just use apps agains android for CPU Intel use SPEC2017 context.

>street shitting iToddlers get BTFO every time a lagToy releases and gets BTFO by """"""weaker""""""" qualcomm CPUs
>Jow Forums gets spammed with Goybench advertising threads
Pathetic.

Attached: 1522037866673.jpg (249x249, 16K)

Kill yourself, retard.

That being said, it is "interesting" how Apple manages to have, by far, the fastest CPUs, and yet also manages to have the slowest phones in practice. Really says something about their so-called whole-system "advantage".

Kill yourself, retard.

Apple slapping a bunch of AES units onto stock ARM cores to cheat at microscopic 2kb workload "benchmarks" is fucking useless for anything other than advertising. They're literally paying Goybench under the table to keep churning out their useless trash benchmark which completely ignores all x86 optimizations.

There is literally no test that isolates single-threaded CPU performance in which the A12 does not win by a very large margin.

>Apple slapping a bunch of AES units onto stock ARM cores to cheat at microscopic 2kb workload "benchmarks"
Proof?
>They're literally paying Goybench under the table
Proof?

>i-i-i-if you use only 5% of your CPU and we get to use 50% of our CPU, WE WIN!!! APEL WINZ AGAN!!11
This is why no one takes iToddler subhumans seriously.

Attached: 1535105476214.jpg (1000x710, 292K)

>single-threaded performance doesn't matter
Interesting to start seeing Intel shills in non-x86 threads.

>iToddler subhuman gets BTFO
>b-b-b-but muh early 2000's era computing paradigm
Unironically gas yourself subhuman.

The A12X is an 8 core you filthy rajeet

A12X has 12 cores you idiot.

>street shitting iPoojeet
>calling anyone else indian
Did Apploo at least give you some toilet paper today when you arrived at One Infinite Loo?

>wccftech.com/apple-a12x-10-billion-transistors-performance/
Retard
>pic related

Attached: 9885F70D-B299-4B02-94F2-CC4FA637B93C.jpg (1088x1088, 162K)

>what is amdahl's law

>it's a self hating street shitter
Yikes.

Attached: 1536566203188.jpg (643x960, 256K)

So tell me again about your computer that uses 128 Pentium cores for absolute maximum performance.

I have money and all the tech I want.
Got no reason to hate myself.
I’ve especially got no reason to mindlessly shill against companies I don’t like, unlike you.

Enjoy being a brainless consumer NPC Rajesh, it suits you.

You don't have money Poojeet. That's why you worship your poverty tier fruit brand. I know you saved up 2 years of your manure shoveling wages to buy your fruit toys so you can lord it over your neighbors in the other shanty, but at the end of the day you still shit in the street.

Attached: 1517607760257.jpg (1165x759, 321K)