AMD cpus can't ove

But, even after meltdowns and shite rehashes of the same CPUs, getting 5-10% performance increases over the past 10 years, there isn't a single AMD chip that can even come close to core to core performance of an Intel chip. Years and years into development they release zen, which is good for the price-point but still can't outperform any intel chip. A fucking 32 core processor gets beat out by a fucking 22 core from last generation? What the fuck is AMD even doing? A 16 core that can't go toe to toe with a 12 core? Or even a fucking 10 core?

The companies making the prebuilts the normies buy will.
But need to be enough price/performance to convince fucking over the jewish contracts.

Kek what ?

TDP isn't measured like that.

How do you think it's measured?

False Have you seen the clock for clock charts ?

AMD has an equation to calculate TDP its calculated by subtracting 42 from 61.8 and dividing by 0.189
see under power testing in this review
gamersnexus.net/hwreviews/3287-amd-r7-2700-and-2700x-review-game-streaming-cpu-benchmarks-memory/page-3

Weird, those power charts don't have the 2700.

U wot m8?

Attached: 1524501808777.png (500x613, 81K)

Attached: 1524501744405.png (600x805, 91K)

wait, what? how? is this true?

overclocking doesn't matter

you just know this will be said when zen 2 launches and intel doesn't have the clockspeed advantage anymore

i have a 2700x it's great, i don't plan on overclocking which is why i went with the x version.
fully compatible with my x370 mobo , fast as shit

Attached: 1462903937182.jpg (638x792, 61K)

>+4 - +7% performance
>"very rewarding"

I see what you did there

>I dont plan on overlooking
>which is why I went with the x versjkn
Is this bait?

He probably meant manual overclock, which is true, 2700x is better with automatic XFR2 overclock.

>LN2
>6ghz

Attached: amd_fx_8370_overclock.png (415x411, 137K)

Attached: intel_collage.jpg (9999x9999, 3.87M)

Still inferior to Intel's Pentium 4 @ 10GHz* :)

*10GHz could have been possible but the laws of physics are anti-semitic

I didn't think we'd come to a time where not overclocking to 5GHz on 6+ cores is considered shit so soon

I was impressed with my g3248 at 4.8GHz

>2.004v
Sweet jesus

It would be faster than Ryzen too.
AMD BTFO.

Don't do this to me user, I'm in public

Ill get charged with public indecency

>mfw FX9590+Dual R9 295X2

Attached: best-death-scenes-alien-movies-ash-e1494423890219.jpg (700x298, 40K)

iirc they didn't actually scale that well with frequency for some reason

Because they're shit. The 9590 was getting outclassed by i3s and it ran at 5 GHz.

Because his post is full of bullshit.
TDP works exactly like watts in = 95+% heat out
Intel shill is thinking of Intel's SDP

Sounds like a fun build.

It saddens me that dual GPU cards are no longer a thing.

With PBO + XFR2 you can get to 4.5Ghz when it is needed. Games do not typically load the CPU at all times. Games are bursty. You can get those single core boosts to 4.5Ghz with 1.5mV safely (temporarily) and a lot cooler overall than Intel. You end up at 8700K performance levels in gaming at least as seen by benchmarks. It ends up cheaper to buy overall too.

Yea indeed

If this is true then why hasn't Linus Jow Forums Tips made a video about it with benchmarks and all?

>i still don t believe this . anandtech already said they are retesting
That's quite normal, but as they said, they just had all updates for both intel and AMD if I understood the story so I wouldn't be surprised those bench are legit

They also benched using memory speeds that the CPUs officially support.

>what is an APU
pleb

I got a 1700x and looking that this I feel shat on by AMD because I bought what was essentially a half baked product for beta testing about an year ago and now the 2000 series is what Zen actually should have been.

How that ?

2933 for AMD, 2666 for Intel

That has always been the way of technology user.

The way I see is that 3 years ago, having a 8/16 thread cpu would be reserved for workstations and enthusiasts. Now I can render and game/stream with no bottlenecks on my old 4790k like before.

Wait for the 7nm 3000 ryzen, praying for 16 core desktop.

intel doesn't support a better one ? That's quite stupid considering the difference of performances and the small price (it's $5 or $10 ? max)

Fuck hoe are they so good? Just a refresh too

>still posting anandmeme botched shit

The fucking power draw though.

Why? Single stronger gpu or scalable like vega is a better solution

>use the proper patches
>b-b-botched! it's shit!
The eternal refrain of the Intlel goy.

Attached: 1522723396996.gif (200x150, 1.51M)

>we're investigating the discrepancies with other reviewers
>week later no word from Ian Cutress
the absolute state of amdrones

>Wait for the 7nm 3000 ryzen, praying for 16 core desktop.
6 cores/CCX -> 12c/24t desktop cores
24c/48t Kikeripper cores
48c/96t Starship server cores

The retards updated their benchmarks after being called out on thier bullshit.

Anandtech already updated parts of the gaming benches, it just wasn't announced yet, mainly because its limited/incomplete.

The proper benches has 8700K winning over 2700x.

>Dual R9 295X2
How's that housefire?

They're like the same thing, just ryzen 2 has few hundred mhz extra headroom, what are you talking about you retard?
>1700x
>not 1700
Confirmed retard

>six million kilohertz
Shut it down!

>Kikeripper

Attached: 1523586375590.png (2560x1440, 1010K)

>nitrous
>oxide
>cooling

Ebin

heh, first thought was "That can't be real" but after seeing this
>Der8auer
it's obviously real, it's happening, Intel's bankrupt and finished.

Not that anyone not trying to break records run their system on NO2 but hey, 6 GHz is a high number.

At this point compiler optimization can make the difference in a benchmark. There was one guy who ran a 7zip benchmark with a GCC compiled version vs a clang compiled version. And got appreciably different results. Also keep in mind that a lot of this shit that gets benched is compiled with intel’s own compiler.

>Run benchmarks for comparing amd vs intel
>Compile benchmark with intel optimised compiler
Who would even do this and claim to be independent?

The person doing the benchmark doesn’t do it themselves, but there are a lot of windows applications that use intel’s compilers in the distributed binaries, and I can’t blame them up until last year, AMD chips were worthless. On the other hand, I doubt intel’s compiler is as much about better optimization on intel chips as it is about crippling competitor chips.
The guy in the article I read was just proving a point.

The jews rule the world, user. Not saying this as a meme. It's the actual truth. Intel does everything they can to cripple the competition, EVERYTHING. From compilers to bribing companies to not use mobile ryzen, CTS labs bullshit, forcing people to apply meltdown patch on ryzen too even though it's useless but they do it simply out of spite and the list basically never ends at this point

But people want the blue pill, not the red one. They want the beautiful lie, not the ugly truth.

pretty much everyone recommends to leave 2700x at stock because auto overclock is faster and more efficient than all core ~4.2

it's not bait, it makes sense to buy the 2600 and overclock it like a 2600X but you don't get much out of overclocking the 2600X or the 2700X.

Whilst the truth might be ugly at first, would realising it not help in changing the current state of affairs into a pro-consumer one?

There's a reason I failed chemistry.

Attached: 1514944925659.jpg (720x736, 32K)

>current: i5-4570
>considering: R5 2600x
>purpose: gaymen
I want to last at least 5 years with my next cpu upgrade. Will the 2600x give me that?

>changing the current state of affairs
People don't want that change. They want the comfort the blue pill provides. Ignorance is bliss.

Nobody got forced into applying Meltdown patches for Ryzen, not on the Linux side of things, anyway (no idea about Windows). Yes, they did try but they were very politely told no by Linus and others.
>“I think we need something better than this garbage.”

>10,240 CUDA cores 32GB HBM2 Titan W with a tdp of 600w

Attached: 1522414507327.gif (400x300, 715K)

Why? Dual GPU cards are shit
Two independent GPUs are faster and SLI problems will occur anyway

Attached: lonely_hearth__Rabbit_-_Mozilla_Firefox_2018-04-25_23-38-13.png (145x137, 44K)

>comparing one company's 12nm to another's 14nm

Lol wew lad.

Attached: 1524530841669.png (856x846, 85K)

They're still being made. Fiji and Polaris based Radeon Pro Duos can still be bought now.

>20% higher performance compared to the titan V
lol, no thanks

>tfw not planning on upgrading my PC for another 2 years
>by that time the cryptomeme will be dead so GPUs and RAM will be affordable again and it looks like we'll have beastly 16 - 32 core hyperthreaded desktop processors available

I'm so glad I don't need a new computer right now, things will be so much better in another year or two.

Attached: 184847473838.jpg (572x321, 16K)

we're going to the moon goy.

>mfw I need a workstation but I'm making do with a laptop and try to toughen it out until TR2

>cryptomeme will be dead
do nocoiners actually believe this? Your best hope on the GPU front is proof of stake overtaking proof of work

currently running Ryzen 1700 stock. waiting for 4th gen 10 Core Ryzen

>10 core
ryzen can't do 5+5, only even numbers

It can if it has 6c CCXes (which we dont know if it will or not). Each CCX must have the same number of active cores as all other CCXes in the system. If AMD chooses to go with 6c CCXes with Gen 3 Ryzen, 12, 10, 8, 6, 4, and 2 core options are possible, although its likely anything below 3 cores/CCX is simply going to be labeled bad silicon and shunted over to Threadripper dummy dies.

considering 12C it is but that's TR territory, unless TR will become 24C and EPYC became 48C. DAMN