Threadripper: confirmed underperformance

>>61666766
>32 threads at once
>clock speed 1ghz faster than Intel's
>"it's underperforming"
Pajeets on suicide watch

>>61666766
Go away BR glueeater, we know it's you.

>the absolute fucking state of JIDF

they are afraid

File: IMG_20170711_122047.jpg (153KB, 942x1024px) Image search: [Google]

153KB, 942x1024px

at least you put effort into the bait

>>61666796
>Update
And absolutely garbage performance for light threaded works. Strange how for AMD there is only 100% multitaksing

>>61666766
>he thinks the higher core parts will have the same clocks
Lmao

>>61666955
>4.1 GHz XFR is now "garbage"
You can try harder than this, glueeater.

>>61666766
Not a very convincing damage control, jew nigger

File: Intel-CoreX-Specs.jpg (192KB, 1200x729px) Image search: [Google]

192KB, 1200x729px

>>61666960
Yeah, the law of diminishing returns kicks in pretty hard for Intel CPUs. Check out that 2.6 GHz base! Boosts up to 2.9 GHz if you're lucky! lol

>>61666824
Wait that shitposter is a BR monkey?

>>61666876
Context if you don't mind?

>>61667032
Yes it was revealed during one of his previous autistic anti-AMD shitposting outbursts. https://rbt.asia/g/thread/61550845/#61572265

>>61667064
>BR monkey is shitposting about pajeets
Wow.
Really makes me think.

Intel is this desperate...

>>61667092
Bet he's shitposting from some dumpster tier Pentium 4.

>>61667064

>caring that much about a shitposter that you keep links to archives to show he's Brazilian

How's that autism going lad?

File: 1408251127948.jpg (48KB, 437x471px) Image search: [Google]

48KB, 437x471px

>>61667064
UMA DELICIA

>>61667110
>bentium4
>in brasil
I wonder how he's not dead from the heat yet.
>>61667114
Fucking BR monkey kill yourself.

>>61667114
Search .com.br in archives it's not that hard to pull up the first result. He's been shitting up the board so much it's hard to ignore.

>>61667114
Back to the favela, chimp.

>>61667127
>>61667135

Cunt I'm not the Brazilian shitposter, just found it amusing.

I'm Australian

>>61666968
with what 1 toy core?

>>61667146
>I'm Australian
What's the difference, nigger?

>>61666766
im 43 years old.
AMD has been pulling the same shit for 20 years.

>>61667146
>dusty fucking Ausfag protecting a jungle ape
Sad desu senpai

>>61667146
>I'm Australian
EVEN FUCKING WORSE

>>61667155

t. seppo

>Turns out the faggot incel shitposter was juan from semiaccurate forums all along

i wish i could say i was surprised

>>61667155
Austr*lians like kangaroo soup

File: poo-hands.jpg (31KB, 600x400px) Image search: [Google]

31KB, 600x400px

>>61667158
>>61667158
When did AMD not lie?

>>61667173
>semiaccurate
CHARLIIIIIIIIIIIE FIX YOUR FUCKING WEBSITE

>>61667195
Fucking favela nigger kill yourself.

>>61667038
https://www.cjr.org/q_and_a/jared-yates-sexton-trump-jr-tweets-russia-qa.php

i got you desu

>>61667038
>I just crawled out of that rabbit hole
Trump Jr tweeted out all the info needed behind Trump's dealings with Russian lawyers/informants on how to win against Hillary.

JYS was a journalist working on the case.

>>61667004
6+ cores for less than 30% extra performance
Its like intel cores are 100% more powerful.


Stay mad its okay to use the inferior product if you are poor. Dont get angry.

>>61667148
>>61667232
Go away BR glueeater.

>>61667232
BANANA

>>61667232
>AMD
>inferiror

I mean sure; it cant burn a house down as fast, drain a bank account as fast, etc but like. Come on.

File: 11875170.jpg (107KB, 1366x768px) Image search: [Google]

107KB, 1366x768px

>>61666766
I really don't understand /g/
Here we have a comparison between processor A and processor I
A has better performance? check
A costs less? check
A gives more cores and threads and allow better multitasking? check

And yet /g/ is such a contrarian that they will shill for processor I, _knowing_ that I has been doing anti consumer practices and such (explained in full detail here https://www.youtube.com/watch?v=osSMJRyxG0k)

What's wrong with /g/? Seriously?

>>61667265
Poor BR monkey is shitposting.

File: 1497319224332.png (3MB, 1440x1920px) Image search: [Google]

3MB, 1440x1920px

>>61667265
You should know that I employees post here frequently

>>61667265
This

File: 89375.png (41KB, 650x350px) Image search: [Google]

41KB, 650x350px

Meanwhile TR at 3.9 GHz is already a beast. Good luck bringing up an Intel 16 core to that speed without melting a hole in the socket.

>>61667265
>ttps://www.youtube.com/watch?v=osSMJRyxG0k
I saw that video, never going to purchase intel CPU again

>>61667313
Shoving chopsticks in my ears would be more pleasant than listing to that shite

>>61667243
Who's that?

>>61667265
You forgot the PCIE lanes. i9 is basically neutered down xeon, while zen series are completely new architecture and AMD worked hard for it. Intel is in full panic, it's painfully obvious

>>61667331
Quality argument, really, 4chan worthy

>>61666766
Yes, it performs better, draws the same amount of power, has more IO, and costs less. So what's the problem?

>>61667350
It's not an argument, but an observation. Learn to language.

>>61667351
The 12+ core Intel CPUs will somehow beat it despite the clockspeeds dropping like a rock, because Intel is always better. Or something.

>>61667362
So no argument? I'm not surprised

>>61666766
SOPA MACACO DELICIA

>>61667162
DE MADEIRA

>intel
>panic inside

>>61667308
>Muh cinebench
It seen like everything AMD can do is only Cinebench and winzip

>>61666766
> faster on everything
> just 2% lower perf/watt
> AMD BTFO

>>61667403
DELICIA

>>61667370
Is all you come here for is to argue? I would really like to watch it, but it's nails on a chalkboard. Listing to groundskeeper Willie for 30 minutes? No thanks.

Even HUEs can buy Intel CPUs while /g/ can't, how poor someone can be? Are AMDfags literally indian?

File: 1494774850330.jpg (30KB, 216x250px) Image search: [Google]

30KB, 216x250px

>>61667423

>>61666766
How much does intel pay you?

>>61667403
All these shills don't even have Cinema installed, or much less use it. e-peen.

File: mad.png (40KB, 1262x883px) Image search: [Google]

40KB, 1262x883px

>>61667417
Wait til this shit toy ripper is released.
Will be the same shit as zen: shitty memory support, abysmal single thread performance, glue everywhere, faux cores + disabled cores for maximum latency

>>61667403
Hey maybe if you're lucky the 16 core Intel i9 will beat it at something by 10%, and will only cost $700 more.

>>61667442
Oy vey, payment discussion aren't allowed in public

>>61667442
4$/day

>>61667423
You own a P4 you rescued from a landfill, Joao. You don't get to talk about modern chips.

>>61667423
They need to save some money for curry

File: 1432992895398.gif (2MB, 400x487px) Image search: [Google]

2MB, 400x487px

>>61667424
>>61667440

File: hype-train.jpg (966KB, 2661x1638px) Image search: [Google]

966KB, 2661x1638px

>>61667423
Just for you to see.
/g is a bunch of basement dwellers stroking to their phenons and when something not so shite like zen comes they belied it is the second coming.

While here im comfy with my intel cpus. Top performance.

>>61667265
Intel: Holding back human progress because money.

File: intel glue.png (81KB, 244x274px) Image search: [Google]

81KB, 244x274px

>>61667446
>>61667467
Yum yum.

>>61667473
You can't help but admire their business acumen, alpha af.

>>61667467
>tfw switched from 6900k
>never been comfier

>BR monkey that can't even into fucking Java is shitposting about pajeets

>>61667490
Did you try sucking one less dick today?

>>61667490
(((business)))

>>61667509
>beta cuck detected

>>61667493
Pajeet pls, you can't keep using the same strawman without people noticing

>>61667519
Eбyчaя БP мapтыхa, yбeйcя oб cтeнy.

>>61667517
Jew cock sucker spotted

File: 1469655959613.jpg (14KB, 320x320px) Image search: [Google]

14KB, 320x320px

>>61667004
>TDP
>165W

OP, your life must be really miserable to shill intel this badly.
Just admit intel lost this generation, may be intel will do better in the next one(s), who knows.

>>61667528
cyka blyat to you too

>>61667515
CRUSH was masterful marketing, maybe nothing like it until Apfel "Think different" came along or something
>>61667534
I don't have to support something to admire it.
>>61667551
tdp is not what you think it is.

>>61667555
He's a fucking BR chimp.
His entire life is miserable.

>>61667551
vs toyripper 180w.
Will be nice when muh overheating meme come back to plague amd

>>61667561
(((masterful marketing)))
Oh the amount of jewery. Can't wait till isr*el gets nuked the fucked out of the face of the earth

hi /pol/

File: 1501245890732.png (944KB, 1228x1502px) Image search: [Google]

944KB, 1228x1502px

>>61666766
>2017
>shilling for intel
I rank you below ISIS monkeys.

>>61667573
They have nukes also.

>>61667587
Well until a few months ago, they were literally the only choice.

File: 11975ubuntu_at_best_buy_550.jpg (54KB, 550x723px) Image search: [Google]

54KB, 550x723px

>94 replies
>26 IPs
is that the power of autism

Intel tried to stop Ryzen by forcing certain tech media to give bad reviews and they did that at first, when almost every tech press shitting on Ryzen but adored and some tech media saying the truth about Ryzen performance. But now maximum tech media is supporting Ryzen I really really hope amd comes back with a big bang this time

File: aHR0cDovL21lZGlhLmJlc3RvZm1pY3JvLmNvbS9MLzEvNjg1MDQ1L29yaWdpbmFsLzEyLVRlbXBlcmF0dXJlLURpZmZlcmVuY2VzLnBuZw==.png (21KB, 711x533px) Image search: [Google]

21KB, 711x533px

>>61667566
lol.

>>61666766
you're wrong

>>61667618
>That steady p̶e̶r̶f̶o̶r̶m̶a̶n̶c̶e̶ temperature scaling

>>61667617
zen was a hot mess at launch. Can't blame them, really.

>>61667641
Not exactly.
X99 was a hot mess though.

>Intel performs better
>pajeets still shilling for AMD
This is some pure kek

>>61667638
Keeps your cores nice and toasty and your water cool.

>>61667651
Yeah it was flawless, wasn't it?

>>61667655
BANANA
>>61667662
Not exactly flawless, but nowhere near as bad as X99 launch.

>>61667655
>>Intel performs better
Correct me if I'm wrong, the graph itself shows intel is inferior?

File: i-use-this-want-to-clean.jpg (64KB, 1280x720px) Image search: [Google]

64KB, 1280x720px

>>61667655
If AMD was that GOOD, like a Gods send no one would be so rabid defending those toy cores. But no, its the same passion we see like those crappy vega vgas.

File: 1499957247669.png (84KB, 653x726px) Image search: [Google]

84KB, 653x726px

Sure Threadripper has better price/performance, better performance/watt and better performance outright but um I forget how I was going to finish that sentence but still

AMDRONES BTFO FOR ETERNITY

>>61667682
basically, current state of OP

>>61667672
I dont know how is burger education, if you read AMD needs 60% more cores to win intel.

>Muh top cinebench performance

Right because when you need single thread performance you just wait and cry while intel offer a rounded product

>>61666766
Don't see what the big deal is. AMD has never gone after the best performance crown in the market, they've always gone after value. Threadripper, Ryzen, and Vega seem to deliver in value.

Need proper benchmarks to truly decide.

I don't think AMD can really compete with Intel's clock speed and the only way they will is if they significantly increase IPC which seems unlikely. They keep going from really good value/performance to almost bankrupt in 3-5 year spans.

>>61667673
>>61667690
Kill yourself glueeater.

>>61667690
>you read AMD needs 60% more cores to win intel.
Does that cost extra? Don't think so, it costs LESS than intel. The fact that it can distribute workloads between cores is a good thing. Having better thread counts means better multitasking.

Also you yourself are admitting that intel has been won, that with lower costs.

>>61667693
>I don't think AMD can really compete with Intel's clock speed
See >>61667004
By the time Intel reaches 16 cores their clockspeeds are a joke compared to AMD.

>>61667717
Furthermore, as a programmer, having more threads is a big relief and intel has been actively holding back the technology

AMD is allowing us to create better performing applications with more threads

File: 1501491714987.jpg (123KB, 638x599px) Image search: [Google]

123KB, 638x599px

>>61667682
INTEL is still the king right?

amd cannot possibly be beating us after the billions we spent buying brand loyalty to every possible retailer in the world?

>>61667717
All they are making so far is faggy gamur mobo's. Guess I'll have to go epick instead.

>>61667740
Time's changed. Even normies are more towards AMD these days.

>>61667743
What? Did you reply to the wrong person?

>>61667716
That's some awesome argument, exactly what's expected from an angry pajeet

>>61666766
>When 7920X hits the market we will know who is the king. With just 2 extra cores intel will spank this abortion
You mean melt it, right?

Because the only things the 7920X will do are empty your wallet and incinerate the contents of your house

>>61667734
7700k can overclock to right around 5GHz on air. It took N2 on threadripper for that.

File: novidea.png (462KB, 411x564px) Image search: [Google]

462KB, 411x564px

>>61667740
Yes, they just added one full quarter of AMD sales to their revenue guidance.
>>61667756
No, I thought you should know.

>>61667737
t. java programmer

>>61667769
Do you even know how to read, dipshit?

>>61667779
D, C++ and Rust actually. Java is good for server-side but I don't do servers.

>>61667759
You aren't worth taking seriously or replying to seriously. You've shit up enough threads on here to turn you into another joke meme shitposter like IMG_xxxx or one of the more ridiculous tripfags.

>>61667791
>D
>Rust
kek

>>61667791
>D
>Rust
whew lad, one failed language, and another stillborn

>>61667446
>glue everywhere
glue is only acceptable between die and heatspreader, right moishe?

>>61667800
>>61667804
>D
My team is actually considering some python codebase to D actually.
>Rust
Already in production.

Are you interested?

>>61667797
kek, mad ask fuck

>>61667791
java's good cus you can shove any ass java program onto your server and it just werks

use native programs on your loonix and you find yourself with a billion broken dependencies

>>61667824
Dependency problem doesn't really exist anymore, 2bh

>>61667815
>already in production
Didn't know enterprise took on languages that weren't standardized yet. Are you a Mozpoz fag, by chance?

>>61667848
There are a very few language that's not ``standardized``.

>>61667816
Can you at least learn english before trying to look like a normal human being and not a fucking chimp?

>>61667769

Unless you got a golden egg. Kaby Lake/Skylake need water-cooling at minimal to go beyond 4.5Ghz even then it is not guaranteed.

>>61667878
>says Pajeet Poo in Loo

File: Screenshot from 2017-08-01 10-53-32.png (246KB, 1920x1080px) Image search: [Google]

246KB, 1920x1080px

Part of our runtime :)

File: 1348500855709.gif (594KB, 239x270px) Image search: [Google]

594KB, 239x270px

>>61667308
>3.9 to 5.2 = 33,33% increase
>3099 to 4122 = 33,01% increase

Holy shit, just think when IBM 5ghz magic 7nm hit next year

File: 1497881707324.jpg (123KB, 653x726px) Image search: [Google]

123KB, 653x726px

>these intel shill tears.
>the serious ammount of butthurt and back peddling

DETELE....

>>61667934
How would you write this in rust?

public abstract class Widget {
    private Theme _theme { get; set; }
    private int _fontSize { get; set; }
    public int GetFontSize() {
        return (_fontSize < 0) ? _theme.GetStandardFontSize() : _fontSize;
    }
}

>>61667937
Holocaust.

>>61667969
Depending on your case you might just create and implement traits for it.

>>61667922
Go eat some monkey soup you cannibal.

>>61667991
A trait has no knowledge of the underlying implementation. So a trait can define abstract functions, but can't access any underlying fields. OOP was a mistake.

>>61667993
kek, trying so hard
Calm down Pajeet, go out and take a shit, maybe it will help with the butthurt

>>61667737
>as a programmer

Unless you do very specialized things that involve very distributed loads you know that paralellism is a hell to implement and several loads cant be parallelized at all

>>61668070
No, it just means I can spawn crazy amount of threads and the users won't have their PC frozen

>>61668020
Let your brain stew on this a moment, and you might find it easier to just keep using C, or C++.

>>61667265
>Muh anti consumer
Fuck off to reddit currynigger

>>61667769
This is my favorite meme

>>61667878
Op here. Its just funny when your AMD droids just starts to vent on anyone

File: 1455954055666.jpg (49KB, 600x900px) Image search: [Google]

49KB, 600x900px

>>61668112
>Run out of argument
>le reddit
>le currinigger
No wonder no one takes /g/ seriously

>>61668107
C++ can get complex too. It all boils down to how good are you in using your own tool.

>>61668140
Sorry, bud Intel was the only company for the last 8+ years that gave us a product that wasn't total dogshit.

>>61668163
Intel sabotaged the CPU market and held back technology

>>61668179
They're only holding back because they have literally no competition. You should be more mad that AMD hasn't made anything worth a damn to get Intel's backs against the wall.

Let's clear this up.

You have data, and you have functions. You create functions that work with your data. You don't combine the two and call them an object.

Also the type system in rust as far as traits are concerned is nothing new and is overall pretty basic.

Basically the powerscale goes like this:

Level 0: Interfaces (C#, JAVA, etc..)

these are pretty useless and you might as well just cast your stuff to objects. First of all you have to add them before the fact, for instance if we wanted an Additioninterface it has to be added to the Int type on the int type itself, meaning the type quickly becomes outdated or have to be updated all the time, or gets bloated to insanity because it has to care for every use case. They also lose their own type when used i.e. if a is an int and int has the interface Addition which has a function +. then c = a + a is an Addition and not an int.

Level 1: Traits/Statically Resolved Type (Rust, F#)

I like to think of these as type of kind order 0. These types can now be added after the fact, and doesn't lose knowledge of their own type i.e. a + a is int and not addition. They are just barely above interfaces and are somewhat useful, for instance you can define a library that works with both floats and double, and you don't have to duplicate the code. However as they are not higher kinded you can't have M<A> where both M and A are polymorphic. Such that the function would work for any M and any A where M could be types like List<A>, Set<A> or Graph<A>. I should note that traits are a bit more powerful than the Statically Resolved Type of F# as F#'s doesn't always work when added after the fact, and is overall more like a hack to a problem than a feature in the language.

ctd.

>>61668162
There is no good way to solve that problem, unless you make bindings to C, but why write bindings to C when I could just as well be writing in C? Or C++?

Level 2: Modules (Ocaml)

Or first order kinded type, you can now have types with 1 level polymophism. Meaning you can make the M<A> function but you can't make an M<C<A>> function. This type system is pretty good and will get you through most problems however there will always just be those few times where you have to hack a bit to get something working.

Level 3: Type classes (Haskell, purescript)

Finally we reach the level of higher order kinded types or n order kinded types, the sky is the limit and there is milk and honey flowing in the street, however sadly we get articles like the ones above missing basic points of a type system and preferring to reverting back to something that is completely useless.


>>61668212
I'm more happy that AMD is finally doing something good but too bad intel shills are here to control the damage and spread FUD

>>61668222
Sorry I can't really relate to the code snippet you posted, I probably would design my program around it but have you spoken to the folks at #rust? They would probably be helpful.

>>61668179

>It's their fault nobody can offer a better product
You're delusional if you buy your microchips according to what's "morally right"

>>61668305
>It's their fault nobody can offer a better product
actually it is. Intel bribed HP a.d other companies and threatened lawsuit if they sold AMD

>>61668360
ooooh hp, acer, and lelnovo, such quality manufacturers!

>>61668382
you are in denial

>>61668179
Do you think CPU manufactures work for free? Research and development is far from cheap. If no one could offer a better product, why even bother making anything? Truth is AMD was the one who held the market with their Bulldozer abomination that no one wanted

>>61667797
I have been curious about it for a while, is IMG_xxxx coming from cuckphones?

>>61668248
Ask your team leader, if he says anything other than bindings, he's an idiot.

>>61668396
>why even bother making anything?
So that AMD comes back and can't grab intel by their balls like this. Karma.

>>61668389
How is paying those guys any different than a grocery store coupon?

>>61668417
I accept your defeat

>>61668402
He knows Rust doesn't have the same model of OOP as C++

>>61668417
Intel were paying them billions and billions of dollars. It was a substantial portion of their revenues. AMD offered HP a million CPUs for free and HP said no because it would impact their Intelbux too severely.

>>61668468
>Free bulldozers
Who could say no to that...

>>61668457
wew, lad you said traits, but to get that without compiler error, you need a "hack" where by adding a function to the trait that is not implemented in the trait itself but on an actual object, then use that function to do your function with.

trait Widget {
    fn get_theme(&self) -> Theme;
    fn get_internal_font_size(&self) -> i32;
    fn get_actual_font_size(&self) -> i32 {
        if self.get_internal_font_size() < 0 {
            return self.get_theme().get_standard_font_size();
        } else {
            return self.get_internal_font_size();
        }
    }
}

// playground

which goes back to a point I made about making a public function that you #still have to implement in all your concrete types.

>>61668396
I think Intel works for free, I mean they offered hundreds of millions in rebates every quarter to Dell alone, to say nothing of similar arrangements they made with dozens of other manufacturers across the United States, Europe and Japan on the condition that the OEM not offer any product lines featuring objectly superior CPUs like AMD's K7 and K8 architectures.

>>61668530
This was back in the Pentium 4 days. AMD was shitting all over Intel and Intel bribed OEMs to only sell their inferior parts.

File: 398567.png (544KB, 736x408px) Image search: [Google]

544KB, 736x408px

oldfags remember this

>>61668544
>by adding a function to the trait that is not implemented in the trait itself but on an actual object,
That's how trait works though

>>61668621
AMD didn't do anything great until k8, I don't know why anyone ever brings up k7

>>61668697
But now you have a public function (trait functions act like an interface, it is currently impossible to mark a trait function as mod-only) that you still have to implement in all your concrete types. So you either don't use abstract functions and have a lot of code duplication or use the setup of the example and have slightly-less-but-still-too-much code duplication AND a leaky API. Neither is acceptable. Neither is an issue in standardized languages such as C++, C# and heck, even Go has a suitable answer.

>>61668705
K7's were better than P3's and offered much better value than P4's.
What's wrong with it?

>>61668732
single channel ram, sse2, p4 clocks went higher and outclassed k7, 180nm p4 was a turd, but 130nm northwoods was good, better than AMD. They released a few more k7's to keep up with the p4 with more cache, 130nm, but it was all in vain until the k8. K8 was somewhat of a gamechanger tho. Serious question: how old were you in y2k?

>>61668727
>still have to implement in all your concrete types.
You only implement traits you require, that's literally how traits work, anon

File: image-10.jpg (55KB, 628x537px) Image search: [Google]

55KB, 628x537px

>>61667230
A single meeting with a russian lawyer and a lobbyist (with American citizenship who had emigrated in the early 90's), in a meeting where the lawyer promised to reveal ties between Russia and Hillary, that was just a bunch of conspiracy bullshit, then they turned the conversation to about sanctions.

Bait and switch, everyone walked out, big ole nothing burger. Even if legitimate info was conveyed this would not violate any law at all.

>>61667882
Nah Kaby will hit 4.8 on a decent air cooler, 4.9 if you really push it, 5 if it's a golden goose. 6700K take off about 300mhz for each.

>>61668805
Old enough to fuck your bitch

>>61667969
>OOP
>Rust
Why do people do that?

>>61667004
>$2000 for a 2.6GHz cpu

holy shit lmfao Kikeripper won

File: 1490871261371.jpg (35KB, 369x387px) Image search: [Google]

35KB, 369x387px

zen 2 cores are gonna be even better

Will AM4 support 7nm Ryzen?

>>61667195
those look like BR hands

>>61670046
AM4 will support up to Ryzen Revision 3

>>61667444
>he doesn't have any games made by futuremark installed

>>61666766
>60%+ cores for less than 30% extra performamce
with all of those cores and the performance coming at the same price as your office heater, making it a much better deal either way.
>but muh same tdp
Yet Intel has 60% less cores
>overclocking on a hedt
HAHAHAHAHAHA
Maybe for idiot gaymers like you OP
>AMD cherry picked perf/watt numbers
And you think Intel doesnt do the same?
>Oy vey, he took their best results from their tests, just like us!
>They are scammers, trust us with your shekels! We are more trustworthy goy!
The toy glued together cores are better than your professional mayonnaised together cores

When the fuck are the non-X versions of Threadripper going to come out?

If they fix the segfault on compilation and can offer a 1950 16c/32t for 850$ they will completely destroy any hope intel ever had for a very long time.

I hope they don't ruin their launch by only releasing the X versions first.

File: 1497990958532.jpg (112KB, 691x771px) Image search: [Google]

112KB, 691x771px

>>61666766
>more core than any intel sku
>better perf/$ than any intel sku
b-b-but they don't have premium cores goy

>>61668874
Incorrect

>>61670585
at the end of the day, HEDT for business use comes down to one thing: performance/dollar in the desired application. lower core counts for single threaded workloads don't really make sense on X299, since Z270 does that for a fraction of the price and similar lane count.
TR and the i9 lines both cut into the Xeon line's market, which has higher profit margins.
that said, no business will use either X299 or X399 right now, they'll buy X99 because it's a mature platform with long term field testing. the real test is X299 v. X399 in a year when the microcode and prices have leveled out.

ITT: assblasted Intel fanboys are mad that a competitive chip is about to released that takes a massive royal shit on anything Intel currently offers in the same manner that ryzen did. Instead of failing to see the obvious consequences of what competition means for their favorite brand (better pricing and performance gains from Intel offerings forced to compete), they make the decision to actively shitpost misinformation based on fundamental misunderstandings of how CPUs function and the benchmarks used in order to feel better about intels currently inferior price/perf, overall build quality, anti-consumer business practices, and fatal design errors that cause catastrophic failure of both chip and motherboards.

>>61668942
>decent
You need at least a 180W TDP cooler, regardless of whether it's air or water. Most of those coolers are $65+.

>>61666766
>beats its competitor in ever single metric including price
>"it's underperforming"
Wew fucking lad

>>61666955
>100% multitaksing
That's not for AMD, that's for any HEDT CPU you clown.

File: vega-gpu.jpg (37KB, 630x354px) Image search: [Google]

37KB, 630x354px

>>61667195

File: 1492139297054.jpg (282KB, 752x548px) Image search: [Google]

282KB, 752x548px

>>61666766
YOUR DESPERATION IS DELICIOUS

>>61671396
The actual difference between a 212 Evo and a D15 is shit all. Using a 150 watt test platform with LGA115# mounting and measuring temp over ambient, the D15 is 13.3 C vs 15.9 C on the 212 Evo. As a comparison the Intel stock heatsink is 24.5 C, and a CLC gets you to around 10C.

>>61670747
Inform me on what part is incorrect.

>>61670718
Intelcuck logic:

>Multi-core doesn't matter!
>Price/performance doesn't matter!
>Power usage doesn't matter!
>Temperatures don't matter!
>Soldered dies don't matter!
>Stutters don't matter!
>Streaming doesn't matter!
>Data centers don't matter!
>OEMs don't matter!
>Hyperscalers don't matter!
>Upgradeability doesn't matter!
>*NEW* Locked platform features don't matter!
>*NEW* Synthetic loads don't matter!
>*NEW* Burnt pins don't matter!

>>61670835
Jewtel never lowers prices.

File: 1500260377912.jpg (96KB, 638x710px) Image search: [Google]

96KB, 638x710px

>>61667937
>>61667990
DELET

>>61666766

Two X299 RAID keys have been dispatched to your home location (total value of $600 MSRP).

Thanks again for correcting the record.

File: untitled-2.png (10KB, 682x336px) Image search: [Google]

10KB, 682x336px

>>61666766
same tdp

who knew amd created an equal housefire!
oh wait this is /g/ they are so blind they cant recognise numbers anymore

>>61666766
Notice price was never mentioned.
Notice that despite the higher core count, the fact that TPD is the same is a boon; was never mentioned.

>>61666955
So what?... Its illegal to show where a product does best?
>Its okay when Intel does it. Intel shows single threaded performance because thats what they're best at.

Are Intel shills the most stupid creatures on earth? At least Nvidiots have some actual basis to shitpost and shill about but Intel got literally nothing going on at all.

>>61672439
they have light threaded workloads!
like literally now they invent segments just to say that they have one......last......castle.....to.....be.....raided.....

>>61668163
Fx8320 and 6300 were shit?
Haha no, Intel couldn't compete with the value offered back then.

File: 1480313620386.png (275KB, 465x450px) Image search: [Google]

275KB, 465x450px

>>61666766
>>61667114
>>61667232
>>61667655
>>61667759
>>61667816
>>61667922
>>61667922
para de postar merda autista favelado do caralho

>>61672619
Is this a new ebin meme?

>>61672439
They're literally braindead: >>61672123

>>61673425
as expected from intel users

File: 15008206598292s.jpg (12KB, 218x250px) Image search: [Google]

12KB, 218x250px

>>61666766

>>61668874
DELETE THIS

>>61672371
>AMD created an equal housefire
Where's the horse cum TIM?

>>61667146
In brazil nerf guns are legal, the same cannot be said for a penal colony

>>61666766
Remember to sage and report any thread this filthy huemonkey trash opens.

File: 1501595997620.png (589KB, 1609x859px) Image search: [Google]

589KB, 1609x859px

lol.

>>61676097
This kills the dresdenlake-x.

>>61676097
DELET

>>61668727
>Go
Are you FUCKING dumb?
Go is literally zero OOP, less than Rust since it doesn't even have traits.You can do the same thing described there in Rust by having a base struct, embedding the base struct in any child structs (composition over inheritance), and implementing Deref/DerefMut on the child if you like.

Go has the exact same set of features Rust has here, no more: struct composition or interfaces/traits.

Regarding graphs: The best solution is option 3, using a crate on crates.io (petgraph is the most popular, but rust-forest is fine too). The preferred solution to having multiple types of nodes in the same tree is to use an enum or a trait. (Servo uses a trait for its flow tree, with downcast methods like as_block() if needed.)

I'm puzzled by the contention that the fact that all nodes have to be the same type is a problem with Rust. That's always true in any language. Even in C, you'd have to have your next/previous pointers have some type, or else you couldn't traverse the list!

>>61676097
>128gb ram
10 tabs max in google chrome.

>>61676353
Also, I know where you found that code from

>>61676384
Worth it.

>>61672010
212 Evo is a 140W TDP cooler that has to get VERY loud to reach that rated cooling.
Most other coolers like an H7 are underrated and use lower RPM fans. But I'd still recommend better than an H7 for a 7700k overclocked past 4.6-4.7.

The 212 Evo is really no better than the Ryzen 1500X stock cooler at a given noise level.

File: 0cCECjW.png (1MB, 1200x890px) Image search: [Google]

1MB, 1200x890px

>>61677403
On the same exact test, the D15 was measured at 49.8 dB and the 212 Evo at 47.1 dB, but you know lets just go and pull shit out of our asses that works too. Intel stock rates 61.1 dB.

Btw the pattern continues with 85w and 125w loads. 7.9C vs 9.3C vs 15.2C, then 10.9C vs 12.8C vs 23.2C; D15, 212E, Intel stock.

Your Noctua junk isn't worth the price.

File: 1487297768743.jpg (125KB, 960x878px) Image search: [Google]

125KB, 960x878px

>>61676097
I CAN'T TAKE THIS ANYMOOOOOORE

>>61672646
Yep, they needed a strawman to mock intel users so the BR thing come in handy

File: 1432661914532.jpg (14KB, 220x200px) Image search: [Google]

14KB, 220x200px

>>61677964
>intel users
>implying it's not just one hue

>>61677964
Read the entire thread from the beginning, fucknozzle.

>>61678069
No way I'm reading a wall of shitposting like that

File: 1433583472269.jpg (64KB, 361x691px) Image search: [Google]

64KB, 361x691px

>>61678193
I'm having trouble finding words that describe your level of retardation

File: Lolcat_FYU.jpg (32KB, 500x375px) Image search: [Google]

32KB, 500x375px

>>61678271
It's fine to be retarded anon, ask your mom.

File: 1436242449995.jpg (24KB, 670x549px) Image search: [Google]

24KB, 670x549px

>>61678314
>icanhascheezburger.com

>>61678379
hownewru.gov

>>61678379
can i haz broadwell?

[spoiler]taken from an old tech forum[/spoiler]

>>61672010
>CLC gets you to around 10c
Argument invalidated

>>61677567
Bias shill test.
That's the problem with comparison tests that are easy to obfuscate.
They could have had more ambient noise on one test, microphone further away, different angle. They could also outright lie and there is no way to get objective, verifiable data from them the same way you can see the framegraph of a game being played or something.

A 212 Evo is fine for 95W of cooling, yes. It's similar acoustics to the AMD stock coolers at that level.
But they advertise that cooler at 180W TDP. It gets very loud at 140, at least without a soldered/delided IHS.

>>61678957
Same test platform senpai, fixed microphone. FT doesn't fuck around.

File: 1479919678833.jpg (108KB, 640x742px) Image search: [Google]

108KB, 640x742px

>>61678569
10C over ambient

>>61670747
Yeah, that's why everyone got in trouble. Oh wait, they didn't because nothing happened and you're fucking brainwashed by the media.

>>61667937
7nm goes into production next year. Mom didn't say when it will be out.

>>61666876
"Journalists" are a fucking joke.

>>61667490
Really fishing for those (((You's))) aren't you buddy?

When we say intel has shills everywhere, including obscure taiwanese basket weavers boards, that's because it's true.

>>61666955
>HEDT
>Single-threaded workloads

oh shit nigger what are you doing

>>61680490
>basket weavers
Have you seen Taiwanese puppet shows? They are pretty technical. It applies to this board as well

>>61667418
>I would really like to watch it, but it's nails on a chalkboard.

I'm sorry for your crippling autism

Here is the thing
HEDT is a very small market so these i9 vs TR threads are pretty much useless but...
AMD epyc server CPUs are 100% raping the intel xeon chips which is where intel makes 60% of its revenue
This is why inbtfotel is scared because their cash cow is going away.

>>61680695
Here's another thing.
Most people didn't really care about the HEDT platform because it cost too much. Content creators won't build a 12k $ rig unless they make mad dosh. TR brought HEDT down tho almost mainstream segment. I expect every YouTube tard to use TR in the future.

>>61666968
Threadrippers have 4.2 GHz XFR. Check the article on The Tech Report, it has some good details on the CPUs, like the memory/core modes that will be selectable in BIOS.

>>61680774
Thanks for the information.

Will Threadripper enhance my 100+ porn tab experience?

>>61666766
I'm late to the thread... Are there benchmarks that show performance available yet? I am really curious how the single core performance and how the temperatures compared to the i9s are, but I haven't seen shit yet.

>>61666766
>same price
>same tdp
>30% better performance
>lolol Ryzen is shit look how it beats the absolute shit out of the 7900x by ONLY 30%!
>THREADRIPPER DEAD ON ARRIVAL ONLY WINS BY 30% THAT'S RIGHT IT WINS! BY 30%!

>>61680899
August 10th is when it's released and when we get reviews.

>>61680956
Intel shills lost all their handles to shill for Intel, give them some break.

>>61672010
the d15 will keep something either 350-480 watts under a 30 degree delta if the heat is properly transferring, it matches most 2 fan clc's and is only outdone by a significant (over 5degree difference) by some 3 fan clc's but depending on how they test, the clc's may not have had time to come up to temperature parity.

>>61667313
everything he said is what I was on a phenom II, zen was my deadline for a new cpu, and if amd was not going to be competitive, I was going to get a 6800 or 6850 and be done with it.

>>61681131
Its above most of the 80mm and 120mm radiators, below the 140's and up though. CLC's are shit compared to a custom loop though. The 212 will catch most 80mm's as well.

It's really not worth the premium.

>>61666766
Is this the part where you and your street shitting friends get laid off?

>>61668212
How about be more mad at intel, when amd knew they were in shit because of intel they offered to give 1 million cpus to hp but it would cost hp far more money to take the cpus then use intel because of intel bribes.

amd ran the fuck out of money, and bulldozer was slowed down and rushed at the same time, the earlier plan was to get the single core performance up and they had ways, just no more dev time could be taken, and after it was out, they switched from improving single core to improving power use.

so on so forth, intel fucked everyone they could over as hard as they could because intel has next to no talent working for them, how many times has intel been usurped by people significantly smaller then they were?

>>61668305
I buy based on which company is better + can I deal with any short comings.

I buy amd's gpus because I despise very business practice of nvidia, but amd is ballpark as good in my price/need range as its nvidia counterpart, if not slightly ahead.

amd until zen was not an option in the cpu, but post zen, I can't justify an intel cpu, even if I had infinate money I can't justify an intel cpu.

>>61668468
it was 104% of dells total profit is what intel was bribing them at their worst.

>>61668805
um... I had a prescott on 95nm, I beleive 3.2 ghz with ht.

My friend had a mid 2000 clocked athlon xp, and it shit on my cpu so fucking hard there was no comparison.

Funny thing was, I had to buy a prebuilt as to young for a job but old enough to have a say in what we bought hardware wise for the computer, no one offered amd, and I can't hate intel enough for me getting that piece of shit cpu and having to live with it for as long as I did.

>>61681788
2GHz Athlon 64 was better than 3,2GHz Prescott in most things, yeah. The HT could help Intel, but little software could use it, or at least use it well.

>>61667767
I could see the 6 core coffe lake taking some sales from amd, as 6 cores and high clocks will be hard to pass up especially if they can get it on the same socket, but the moment amd drops zen2 intel is fucked in every way imaginable.

the current process amd is on is targeting 3ghz
the process they will move to is targeting 5ghz and as far as I heard it's completely on track.

>>61670585
overclocking a hedt up till ryzen was a real thing because 4 cores were not enough, but there were no mainstream 6 cores.

I would still oc a hedt cpu to a reasonable clock speed like 4.0 or something there around, the socket can take it, the cpu can take it, and its not so much of an oc that its unreasonable to cool.

>>61681026
>>61680956
to be completely fair, amd wins 30% with 60% more cores, that is an issue if you have tasks that dont scale as well and need to use less.

and no fucking shit, if a task uses less why are you getting a 16 core cpu, well I have tasks that use 16 cores and I have tasks that use 2 at most.

That said its very hard to argue that those smaller tasks are worth an intel for when I could have the small tasks doing shit, and then run a large task right next to it that will eat 14 cores and ask for more.

>>61679903
>>61679903
>"Journalists" are a fucking joke.
>>61667227
>https://www.cjr.org/q_and_a/jared-yates-sexton-trump-jr-tweets-russia-qa.php
Read that. This guy did real investigative journalism.

>>61681826
If Coffee Lake 6 core can legitimately work on the majority of Z170 boards, let alone any, I could see that yes.
But I don't see that happening.

Hell with fucking X299 many of the boards don't support Kabylake-X, and won't support the 12-18 cores either, despite being the X299 chipset that's supposed to be compatible with them.

File: wizard.jpg (17KB, 281x240px) Image search: [Google]

17KB, 281x240px

http://www.phoronix.com/scan.php?page=news_item&px=Linux-NUMA-Sched-Fair

Improved performance for NUMA systems on Linux. Just in time.

>>61682219
because intel decided 'we want the boards to support quad core dual channel and quad channel, not have all the lanes and have all the lanes' and then at zero hour when they realized how much of a thread threadripper was, good you are supporting 4-10 cores... heres 12 to 18, 3 days before gdc if I remember right.

pretty much since sandy bridge how many of us asked 'did I really need to upgrade the fucking motherboard for one more fucking pin?' if intel is going to do anything to counter amd, this would probably be the best area for them to attack as it doesn't really eat into their margins to much.

>>61668212
Are you retarded? Intel's anticompetitive practices are well documented and the Jews in the government didn't do a thing as Intel tried to kill AMD by isolating it from almost every. Intel didn't hold any thing back

>>61683133
I know why X299 support is a mess.

But Intel has always and only done 2 "generations" per chipset for a LONG time.
That, and, presumably the 6 core is going to be like a 140w TDP part. The 7800X is 140W TDP, and it's 3.5GHz. Isn't the 8700k supposed to be 3.7GHz base?...

While the 7800X is slightly underated TDP compared to the very overrated of the 7900X which is really a 185W TDP chip, there's just no way that the 8700k is 95W TDP.
Intel claims the 7700k is 95W TDP, but really the 7700 non-k is (which they claim is 65) and the 7700k is more around 115.

So these boards that are made to support at max a 115W TDP CPU aren't going to seat a 140-150W TDP one even if it lies and says it's 95 on the spec sheet.
And that's just assuming it uses the same socket, PCIe lanes, and a number of other features.

I wouldn't be surprised if what ends up happening is that it is the same LGA1151, same PCIe lanes, and it'll be backwards compatible with Skylake and Kabylake but that Z170 and Z270 won't be *forwards* compatible except for maybe a few high end boards that might get BIOS updates.

>>61683992
if you oc the 7700k it seems that there are two numbers from anandtech, at 4.8 they said it was 179 and 122 watt
at 4.6 it was 152 and 95
at 5.0 it was 210 with no other number given.

apparently the second number is the one they pay attention to not the first, so if they sold the 6 cores as a 3.6 to 4.0ghz chip and allowed people to oc it, then that would be perfectly ok with the socket.

I have a custom water cooling loop, so TDP or power usage is pointless to me.
If TR preforms like leaks show, I'll pick one up when the platform matures a little.

>>61681907
Yeah a joke

do the math again OP, this time properly.
Intel 140w/10 cores = 14w per core
AMD 140w/16 cores = 8.75w per core
Why would I pay more for a less power efficient processor with lower performance?

>>61685719
Now do a calculation of how shitty those cores are performing wise

>>61685719
You did it wrong. TR is 180 watts.

140/10=14
180/16=11.25
14/11.25=1.24

AMD's own calculations put it at 29% more efficient, though, so there's probably something missing from this bare bones napkin math anyway.

>>61685812
That's the point, they aren't since they are using less power.

>>61666766
I hate when new graphics cards/CPUs come out, /g/ is polluted with shilling for months.

File: intel fans enjoying skylake-x.jpg (263KB, 1600x851px) Image search: [Google]

263KB, 1600x851px

>>61685812
They perform so well, the CPU can't even keep itself cool.

>>61685812
Theoretically the i9 has cores running with 20% more performance. However to do this it basically hits 90c under stock settings, so it's pretty much a wash.

File: 1501600030116.jpg (208KB, 960x874px) Image search: [Google]

208KB, 960x874px

>>61681623
>Chosing inferior hardware due to feels
>Muh underdog

>>61685869
Let's wait toy ruppe releases and you will see how much real power this mess uses. Amd always like about tdp

>>61685912
You can barely even fucking type, you fucking troglodyte. Fuck off and never come back.

>>61666766
Who's "we"
The shill division?

Reminder: Brand loyalty is cuckoldry

>>61685818
7900X is not a 140W TDP chip.
Intel says it is but they're lying. It's at least 185.
I don't know how this still confuses idiots on /g/.

AMD's numbers are from actual power usage in benchmarks.

>>61666766
>Same price
>Lower power draw
>Higher performance
>More cores
>"""Confirmed underperformance"""

>le tdp is power draw meme
Fucking hell /g/ get a grip.

>>61681826
>the process they will move to is targeting 5ghz
source please

>>61686931
>7nm summit ridge is aimed at higher clocks
i wonder what comes after 7nm.. 4nm? 0.5nm? the long wait till graphene? better cooling tech and higher clocks?

>>61666955
>buy $1k CPU to play LoL and watch animu
Look at this bozo

>>61687206
we have to go deeper
subatomic particles arranged into logic gates

By the way, if i wanted a cpu
And i had only 200~250 miserable euros
Would a ryzen be worth it over an i5?

why does everyone on here treat hardware companies like sports teams

who cares lol

>>61687993
yes.

>>61666766
>same price
>15-55% performance boost
I'm sold.

File: 1500798033250.jpg (19KB, 366x380px) Image search: [Google]

19KB, 366x380px

>>61680546

>>61687262
Making new elements in the name of 1773 gaming?

>>61685884
t. poo in the joo

>>61687206
Just wait©