[ 3 / biz / cgl / ck / diy / fa / g / ic / jp / lit / sci / tg / vr / vt ] [ index / top / reports / report a bug ] [ 4plebs / archived.moe / rbt ]

Due to resource constraints, /g/ and /tg/ will no longer be archived or available. Other archivers continue to archive these boards.Become a Patron!

/g/ - Technology

View post   

[ Toggle deleted replies ]
>> No.77966145

So it's 40% faster than zen 2?

>> No.77966148

Yeah, about there.

>> No.77966311
File: 1.32 MB, 598x640, tenor.gif [View same] [iqdb] [saucenao] [google] [report]

Holy shit, this is insane. We need to stop Lisa before she becomes too strong...

>> No.77966492
File: 243 KB, 1195x1452, 1601363150973.png [View same] [iqdb] [saucenao] [google] [report]


>> No.77966756

Ye but how about the price?

>> No.77966872

>NEW 1080p doesn't matter

>> No.77966893

umm its a cpu test, and 1080p will put more strain on the cpu, not gpu....der

>> No.77966989

1080p hasn't mattered to me since years ago.

I know it's just for a test, I just don't care for because those are high end parts on a low end res

>> No.77967009

>the most popular resolution doesn't matter
Intcels on suicide watch.

>> No.77967119

>still loses to a 5900X that uses a fraction of the power and doesn't require an industrial chiller to safely use

>> No.77967355

>hehe cache go boom

>> No.77967392

So now 5xxx will be Zen 3 regardless of CPU/APU? Good news I suppose, it does seem silly that my laptop is a 4700U and my desktop is a 3700X when they are both Zen 2 with 8 cores.

>> No.77967421

>All this performance gain on the same socket
Intel could never.

>> No.77967425

AMD: sells you a new CPU for the same socket
Intel: sells you a new socket for the same CPU

>> No.77967630

Too bad it was a mistake to keep the same socket and won't happen again. It has been a shitshow to keep compatibility between the cpus and boards. They might as well have made new sockets with all the gymnastics they had to do for it to still not work like intended. New CPUs can't use old boards despite being the same socket and other boards need a bios which disables sections of the CPUs that fit the same socket. The fact that you can't simply get an AM4 socket CPU for your AM4 board is stupid. I bet they vowed to never do this again after they had to ship dummy CPUs to customers for them to flash AM4 boards to work with AM4 CPUs. That couldn't have been cheap.
>average amd framerate 72
>average intel framerate 73
You what? What is the difference between an average framerate and average CPU framerate? Does an average CPU framerate actually mean anything outside a benchmark? Is the generic average framerate number what you actually get in a system?

>> No.77967649

>You what? What is the difference between an average framerate and average CPU framerate? Does an average CPU framerate actually mean anything outside a benchmark? Is the generic average framerate number what you actually get in a system?
Average framerate is GPU-constrained because this is a GPU benchmark. Average CPU framerate is 1 divided by CPU frame times. It's basically a way to extrapolate CPU performance from a GPU benchmark. Unfortunately this is the best we have right now.

>> No.77967704

>Average framerate is GPU-constrained because this is a GPU benchmark
Isn't this a game benchmark? Wouldn't it being a 1080p benchmark make it CPU bound since the GPU isn't so stressed?

>> No.77967746

The intel deals i've been seeing the last week make a lot more sense now. IN SHAMBLES.

>> No.77967758

considering how obvious it was that buying an AMD socket was the good option with the upcoming annoucement, yeah

>> No.77967773

Reminder hairy fat fuck intelnexus shilled for kaby lake threadlet shit for games in 2017 that aged like dogshit.

>> No.77967774

If it's still under the 2 weeks of no-fault warranty or however long your country guarantees (I assume you don't live in a shithole without no-fault warranty), you can just return it and get your money back to buy Zen 3 in a couple weeks.

>> No.77967795

>poozen fags who plays at 1080p?!?
>also poozen fags heh look at these 1080p benchnarks intel btfo!!
Fucking Poozen Drones

>> No.77967803

Everything that's released in 2017 barring the 8700k is horribly outdated today anyways

>> No.77967807
File: 12 KB, 331x152, images.jpg [View same] [iqdb] [saucenao] [google] [report]

>inteljeet always making shit up out of asshurt

>> No.77967850
File: 18 KB, 267x297, 1537394779764.jpg [View same] [iqdb] [saucenao] [google] [report]

You are seething so hard your tranny neovag is dilating on its own.

>> No.77967852



>> No.77967910

Pretty sure the bios update to support it is out for a lot of them already as well, if not all.

>> No.77967947

I want more benchmarks not just one game, but this definitely looks promising even at 1080p

>> No.77967957

everyone shilled how 320p benchmarks would show how much more useful Kaby would be for future GPUs
here we are now.

>> No.77968012

No. There are renoir refreshes that use the 5xxx naming scheme. The u-series 5000 apus will be made of a mix of zen2 and zen3 parts.

>> No.77968099
File: 103 KB, 760x929, thatcope.png [View same] [iqdb] [saucenao] [google] [report]

The COPE from 10900k owners have begun

>> No.77968255

The intcel cope is already in full effect and it's hilarious. All these years pushing the 720p benchmarks where jewish cpu has 500 fps while AMD cpu has 450, now the first time amd pulls ahead and suddenly cpu bound benchmarks don't matter!

>> No.77968319

>intel fangoys praise 1080p benchmarks ad nauseum
>suddenly 1080p doesn't matter LOL
14nm++++++ corelets BTFO.

>> No.77968333

>that poozen is holding back RTX 3080

it was lol

13% slower at 4k
7% slower 1440p

>> No.77968410

>Joel Hruska
reviewers aleady have cpus

>> No.77968419
File: 845 KB, 800x2500, 1599536402729.jpg [View same] [iqdb] [saucenao] [google] [report]

>be minimum wage cuck
>save every bit of money you can flipping burgers for years
>buy the best consoomer tech to impress you minimum wage friends
>new better tech comes out
>no chance to afford an upgrade for the next several years
>the cope posting begins

>> No.77968561


>> No.77969411
File: 1.63 MB, 9272x3610, anime_amd_1.jpg [View same] [iqdb] [saucenao] [google] [report]

Based AMD. You just can't stop the power of anime and AMD.

>> No.77970868
File: 17 KB, 460x460, 1598825691595.jpg [View same] [iqdb] [saucenao] [google] [report]

As expected, this is going to be an absolute monster.

>Tfw have the 3900x but kinda want to max this board out with the 5950x

>> No.77970887

Does anyone actualy play that?

>> No.77970914

Normal framerate is GPU bound because that's a GPU stress test. CPU framerate is how many frames the CPU processed and sent to the GPU pipeline, regardless of whether the GPU actually ended up drawing those frames or dropped them.

>> No.77970927

So the normal framerate is actual gameplay and the cpu framerate is synthetic?

>> No.77970967

oy vey

>> No.77970969

The normal framerate is GPU bound and depends primarily on GPU performance. The slight 1-2% variation between the two benchmarks could be due to one of the setups having a slight overclock.
The CPU framerate is measured within the game engine by recording how many frames the CPU added to the queue.

>> No.77971019

The thing is though that in the current in built benchmark of ASOT Ryzen cpu's already do very well, like pretty much on par with Intel when you see DF use it in their video's.

But you don't see that performance parity in many games.

That makes me kind of suspicious of this.

I mean I'm hopeful about it and I certainly hope AMD beats Intel at the final thing they cling onto, gaming performance. Because then they will have to step up and make something better.

>> No.77971034

>people are surprised newer chips beat older chips
The absolute state of /g/

>> No.77971058

>intel now throwing in more cores than amd

When did the world turn upside down?

>> No.77971061

Here's a 3800X benchmark with the same hardware and the same settings for comparison:
The 5800X is literally 50% faster than 3800X. Both are AMD, so there's no AMD bias to factor in here. 50% generation-on-generation improvement is bloody insane.

Comet lake released 10 months after Zen 2. Zen 3 will release 5 months after Comet Lake. It's fair game to compare Zen 3 to Comet Lake if it was fair game to compare Comet Lake to Zen 2.

>> No.77971431

Anything that lets me push more games into 1080p/240hz+ territory for a lower price is more than welcomed.
I'm not some kind of retard that sticks to a weirdo resolution, so I'll make the next jump when 4k/240hz is available.

>> No.77971599

Just sell your 3800X before 8th of October crashes the used market and then buy a 5800X.
It's 2080Ti vs 3080 all over again.

>> No.77971633

Who plays in 1080p with all the settings to low while using a $400+ CPU?

>> No.77971643

Intcels. But now they don't have even that to fall back on when defending their shitty purchasing decisions.

>> No.77971704

um sweetie 720p is the new hotness

>> No.77971758

t. 2002

>> No.77971781

amd leapfrogging intel will just force intel to ditch their fabs and start buying up fab space at tsmc and samsung

>> No.77971857

How did you tune ram that tight and high? What was your starting point?

>> No.77971940

>Except amd already bought up tsmc

tsmc is magnitudes bigger than amd

>> No.77972045

>just don't have a CPU for a month+

>> No.77972160

If a cheap ass Athlon's performance is acceptable, then why not just keep the 3800x and be done with it?

>> No.77972390

Where's my 1600 bros at?

>> No.77972725

ayo dese whyte bois better hand ovr there womyn or else we gon riot

>> No.77973685

this is just one benchmark
intel tends to be more consistent, while zen has certain games it just underperforms in
I doubt that will have changed

>> No.77977194

A strong cpu IS useful for 1080p retardd

>> No.77979338

Who buys fast RAM and does not enable XMP or set timings manually? Apparently most people.

>> No.77979363


an overclocked intel cpu benchmark for comparison


>> No.77979396 [DELETED] 

Here's an overclocked 10900k with 4500mhz ram at 4k, better comparison. Seems like this oc'ed 10900k is about 25% faster than the leaked 5800x.

>> No.77980462

Damage control in full force.

Name (leave empty)
Comment (leave empty)
Password [?]Password used for file deletion.