Still winning

still winning
who /700fps/ club here?
get fucked lisa

Attached: oh nononoononoonononooohahahahahah.png (650x450, 42K)

Other urls found in this thread:

twitter.com/RyanSmithAT/status/1147867180282699782
xanxogaming.com/reviews/amd-ryzen-9-3900x-review-english-dethroning-the-intel-core-i9-9900k/
reddit.com/r/Amd/comments/cacwf9/psa_ryzen_3000_gaming_performance_is_being_gimped/
twitter.com/SFWRedditVideos

what the fuck is this? How did AnandTech go from a repetuable source into the worst shit that has ever been released? Not even fucking Intel themselves benchmark at 768p minimum rofl

>not even minimum or 1% minimum FPS

fuck AnandTech. Salon/BuzzFeed of computer hardware

Mr. Anand left for Apple. Cyclone blew his mind that much

Right? I can't believe those shills would hide Ryzen's massive stutter problem. What a load of shit.

Attached: 1562514639563.png (1679x835, 505K)

>768p

Attached: 2368922342156.jpg (1600x1150, 200K)

just mad because you lost LOL
if you arent playing in 1366x768 minimum you arent a real gamer

intel wins on minimums COPE amdronesDD

>Ubisoft

Attached: 1562512312835.jpg (300x300, 64K)

>AMD

Attached: 1561922150243.jpg (600x600, 85K)

no raytracing no buy

>768p
HAHAHAHA
Is this 2001 again?

i wish

Attached: 1525375698209.jpg (1095x1195, 287K)

Is there even 1 review where intel shills didnt cheat?

Attached: Untitled 1.jpg (661x34, 10K)

How does intel cheat reviews?

Attached: 1013423_164144617097928_2086563000_n.jpg (960x945, 55K)

>not benchmarking lost planet and quake 3 at 640x480 minimum with 2080 ti

it's like they are trying to bottleneck

Attached: outstanding-fansub-quality-018.png (479x360, 304K)

>gaming is everything
>I only care about gaming and at 1080p
Are we being invaded by /v/? There was a time when gamers were told to fuck off to /v/ and were told to install Gentoo. New-Jow Forums doesn't care that a budget CPU can compile the Linux kernel faster than the top Intel CPU.

What is wrong with these shitters?
I was playing in 1024x768 in 1995/1996 with mechwarrior 2 and descent 2 in DOS with VESA drivers.

They don't? Thanks for the confirmation!

That was epic, epic for the win!

Fuck off and kil yourself op shill

Attached: 2019-07-07 16_31_09-Here's the deal (Ryzen 3900X, 3700X, 9900K, 9700K benchmarks) - YouTube.png (1912x1050, 172K)

NOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

THIS CANT BE HAPPENING!!!!!!! INTEL BEATING US IN OUR OWN GAME!!!!!!! AMD FAILED!!!!!!!

Attached: 111257.png (650x450, 41K)

480p tests WHEN?

How fucking dumb are you?

Go look at the article, they go up to 4k ultra

Attached: SmartSelect_20190707-164042_Samsung Internet.jpg (1409x966, 348K)

i'm not trusting anyone who doesn't block ads in his browser.

Attached: jimbo.jpg (501x498, 18K)

These must be the "real world" scenarios Intel was talking about

hmmmm

Attached: GTA.jpg (1280x677, 83K)

Hey they're more real than cinebench

>spent $800 a few months ago on an already outdated piece of shit
>i-i won g-guys...
LMAO

>$329 vs $488
Hmmm.

Correction
>$339 @ 65w vs $374 @ 95w without cooler and no upgrade path.

>gtx 1080
>clearly GPU limited
>ashes of the benchmark

>wow, the gtx 1080 is capped at 62fps no matter what cpu we use, great test

They're trying to tell us something, you clauds. They're trying to tell us that 4K is a meme, a meme, a meme.
All games should be played at 800x600

then surely intlel would also suffer?

Cope harder brainlet

AMD Niggers cant even afford 4k let alone 1440p 144 so why bother

>$320 CPU
>Motherboards atleast $50-$100 more than their intel counterparts

GJ you waited and still lost and spent pretty much the same amount of money too

It does though. Are you blind?
Or do you really think the difference between 89.00 and 89.60 "average fps" makes skylake-X worth buying? Or that 87.34 vs 90.70 is a difference worth $50?

Ask me that when X570 motherboards dont cost $100 more than their Z390 counterparts.

>Playing a 786p on minimum at 712 FPS on any monitor
> It's almost like you want to be beaten by a player you can't see because you set resolution and view destance at their lowest setting.

Still won't get you more than 18FPS in ArmA3 multiplayer

>> It's almost like you want to be beaten by a player you can't see because you set resolution and view destance at their lowest setting.
crosshair pixels bigger than distant enemy

You guys OK? They bench on low resolution to ensure the GPU isn't the bottleneck.

What bizarro would do you live in?

>he doesnt know about scaling

OH NO NO NO NO NO NO
AHAHAHAHAHAHAHA
LOOK AT THE TOP OF HIS FRAMES! AHAHAHA
LOOK AT THOSE MS!
AHAHAHA

Attached: 2019-07-07 17_30_19-Navi Max! AMD RX 5700 XT Review_ 'Super' Price vs Performance_ - YouTu (1920x1080, 3.22M)

ITT: sperging intcels

Attached: ey.png (268x322, 133K)

Man everybody in this thread needs to HAVE SEX
H A V E S E X
A
V
E

S
E
X

That is the problem with benchmarks.
Nobody plays on those resolutions, and most people don't own a 2080ti or similarly powerful gpu.

including you, rajeet

Attached: indian.jpg (800x600, 90K)

i had sex TWICE this year I'll have you know!!!

>4k
driver issue

>what the fuck is this? How did AnandTech go from a repetuable source into the worst shit that has ever been released? Not even fucking Intel themselves benchmark at 768p minimum rofl
Not fucking kidding here but the game I play for money pretty much needs 240-300 fps and

doesn't count if they're indian

>he's a whore who plays csgo or r6s for """money"""

Jow Forums - technology

>he's a NEET incel who shitposts on Jow Forums for "free"

A worse game but thanks for ignoring the point of most of the post.
I'd be happier if Nvidia or AMD actually put some effort into supporting integer scaling at the driver level than just performance gains that require me to buy new hardware for.

I think you meant to type
Jow Forums - benchmark wars

No he meant Jow Forums - cope wars

Attached: b450.png (1920x1080, 435K)

BLUE BAR BIGGER
RED BAR BTFO

n

Attached: b450 no pbo.png (1920x1080, 282K)

i

Attached: b450 pbo.png (1920x1080, 371K)

ce

defend this

Attached: R6S.png (1124x630, 437K)

>he has to use an expensive AIO to beat AMD's included cooler.
kek

They benchmark it on low settings to ensure the CPU doesn't get bottlenecked you absolute and utter fucking mong.

THE KING REMAINS FIRMLY SEATED ON HIS THRONE

Attached: hpit-530_hpit_530_01_800x800[3].jpg (800x800, 139K)

Attached: zoomer.png (537x481, 78K)

That's the whole point of showing 768p, so you can see what the game engine is doing with the CPU when not bound to the GTX 1080.

It's like you're actually retarded or something.

Using a water cooler because you have to buy one if you want to hit those clock speeds.

And reaching 10 FPS? playable 1024x768 only started to be a thing with riva TNT/Voodo 3

You can hit 5Ghz with a Nh-d15, you don't HAVE to go water.

You DO need good cooling, but it's not 100% required to go water.

>$20 cooler included in the box vs $100 Noctua
Hmm. Something ain't right here.

>768p

What the fuck. Why stop at 768p? Let's see some 240p, or 144p benches. Everyone knows the real pros game on postage stamps.

That's 5Ghz all core OC, if you want just stock clocks, you can obviously just get a Hyper 212+ or similar.

You can't OC poozen on the stock cooler either, and you're gonna need an NH-D15 if you want it to hit max boost clocks

I want to see them compared under the same conditions. Not the ideal conditions. We all know Intel would look bad. But it would interesting nonetheless.

No manual OC, stock cooling, Fox only

Would be a fascinating video. Naturally the Intel bois could be tossed a bone, and offered to use an Intel stock cooler even if the processor doesn't come with one.

new architecture by amd literally works harder at higher resolutions though.

No mitigations either

it was AT LEAST 20 probably cinematic 24.

>3,75% difference between 3700K and i7 9700K
at that point, it's just error margin. Go for the cheapest one in your area.

> with no Fallout/ZombieLoad fix applied.
twitter.com/RyanSmithAT/status/1147867180282699782

>Intel stock cooler.
kek. No need to be THAT cruel to Intards.

Until agesa 1003 bios comes out we will never see true ryzen 3000 potential.

V1.0 non-beta BIOS comes out soon. Everything out there is beta right now.

Genuine question: are the recent influx of benchmarks at retarded resolutions just straw-grabbing/shilling? I really can't think of a reason why someone would benchmark CPUs in games and not use the relevant resolutions people actually play at other than because an irrelevant test gives them the outcome they want i.e. their team wins.
I get the arguement that "it shows what the CPU is doing when not in any way bound by the GPU" but if 99.99% of people are going to be playing at 1080p or higher, why would anyone even care?

Notice how many of them are not transparent with the specs and testing methodology? If they don't tell you this they are a SHILL!

look at the actual review, they cover 768p, 1080p low, 1080p high, and 4k ultra.

>1% lows still above 180fps
>somehow this is a problem with stutter
>average framerates only mattered when they were higher on intel, and we just ignored 1% lows
>now that AMD has higher averages all that matters is 1% lows
baka yeet senpai tbqh desu

Testing both CPU's on an AIO only tells you their maximum potential. It does not tell you their actual value. Test both on stock coolers (or the equivalent of what the Wraith coolers are for Intel on BOTH platforms). Then we will see where the performance to value proposition lies. Most people will stick with the included stock cooler on AMD.

Notice how (((Steve Burke))) never says benches like this are "misleading"

So why is the i7 beating the i9 that costs 100+ more

it's so you know that even though at 1080p/60fps all cpus are the same, at 1440p and 4k all cpus are the same, and at 144hz in 1080p both amd and intel can do it, you need to know that if it was 720p and you had a 700hz monitor that only intel can get you there...because their cpu is better...

>average
>a range
average literally exists to not be a fucking range, retards.
what the red is, is the 'normal fps' not the average. jesus christ

because the 9700k without hyperthreading can run cooler and potentially turbo for longer or slightly higher clocks than the 9900k under the same thermal conditions and game load.


It appears the game isn't using more than 8 threads, so the 9700k is capable of performing just as well as the 9900k as the extra threads are meaningless in many games.

They are going to have to retest gaming benchmarks.
>Scroll down to the important notes near the end.
xanxogaming.com/reviews/amd-ryzen-9-3900x-review-english-dethroning-the-intel-core-i9-9900k/
reddit.com/r/Amd/comments/cacwf9/psa_ryzen_3000_gaming_performance_is_being_gimped/

Attached: fffffffffffffffff.png (731x1259, 527K)

>all reviews are positive
>new drivers are coming that will make shit even BETTER
kek incels on suicide watch

seething amd poorfags

anand tech hq
>right how can we make ryzen to look bad?
>just test every possible resolution even create some that doesnt even exist in the first place and use the one that shows the greatest gap
>YOU ARE BRILLIANT