Oof!

Oof!

Attached: poos on jew.png (2560x1440, 1.04M)

Other urls found in this thread:

linustechtips.com/main/blogs/entry/1567-the-chiplet-problem-with-gpus/
youtu.be/jzaXvEPyKd0
imgur.com/a/evCRbq9
twitter.com/NSFWRedditVideo

Attached: poos on jew 2.png (2560x1440, 1.63M)

does this consider power usage?

Coming up

Attached: poos on jew 3.png (2560x1440, 1.73M)

came on Jow Forums to post these exact 2 screenshots

Attached: poos on jew 4.png (2560x1440, 1.65M)

Why such disparity?

this whole thread is making me consider the 1660ti very seriously

As Ngreedia

Attached: poos on jews 5.png (2560x1440, 1.71M)

What are you trying to show us here? Despite your AMD shillery this is making 1660 Ti look very good.

what's with the price tags?
so for half the price you can get 90% of the performance?

>Not buying the better value Vega 56 instead
retards

>not paying a premium to get the best and buying a RTX Titan

Looks like 5700, Vega 56, and 1660Ti are the good cards, and Vega only counts if the high power consumption is okay with you.

>no CS:GO
>no Overwatch
>no Apex
>no fortnite
>no CoD

Feels good using a based Vega 56.

the 5700 should've been 325$

here in my 3rd world poopsock the vega56 is over 100 dollaroos more expensive than the 1660ti

I feel for you bro

Navi is actually pretty good desu sempai I was expecting a huge flop
Looking forward to 5800/5900, if we get a 64 or 80 CU Navi that'll be breddy gud

>undervolt

Attached: 20190507_175454.jpg (2592x4608, 2.66M)

Paying more to save more mustard race

linustechtips.com/main/blogs/entry/1567-the-chiplet-problem-with-gpus/

I'm waiting for a multi chiplet GPU. Even though I KNOW there's several issues with it, I'm hoping AMD will be able to come up with a solution that acts as a single GPU.

Is it surprising that different games work differently well on different microarchitectures?

Attached: mpv-shot0004.jpg (1280x720, 110K)

Attached: mpv-shot0005.jpg (1280x720, 113K)

So it's either the 2070-Super or the 5700XT which are the best deals for use & longevity.

Thanks user.

Attached: mpv-shot0006.jpg (1280x720, 115K)

Fucking based

Attached: 20190705_053555.jpg (890x720, 255K)

I'm really curious to see where they'll go with Navi from here. Given that the 5700 cards are 36 and 40 CUs, there seems to be lots of room to scale in both directions.

If they could scale it up, I'd expect the largest GPUs to be the ones to be released first, though, so there has to be some kind of technical difficulty there. Since the Radeon VII exists, though, it obviously can't be related to 7 nm yields.

If that's your plan then you should just buy a 590, then and save $70.

It can, fabs can only produce so much 7nm wafers.

That's one of the reasons graphics cards and CPUs are released in a staggered manner in the first place.

this
1. 1660ti
2. rx5700
3. vega56

And the 5700 series is still better value. Wait for Sapphire though.

>2 percent faster for a hundred dollars

Wow amd truly btfo!

Don't forget the industrial chiller to get to 5.2Ghz

>Barely any mention of shitty XT thermals, 5700's minimal OC capacity vs 2060S, 43 dba noise from blowers at load
>Larger benchmark sample shows performance deficits for Navi compared to HWU review
Review was a little too positive on what looks like a middling card.
If they can get non-blower coolers at their respective base prices it might be a reasonable buy.

Attached: relative-performance_2560-1440.png (500x970, 52K)

100 fps average on 1440p? That's pretty fucking based for $400. Calling it right now, whatever bigger die surfaces from this cute 40CU card will be the GOAT card.

GN's review is good. Based on power consumption the thermals should be fairly decent when board partners design new coolers. I really don't understand AMD's decision to revert to the stupid ass blower design.

>watching gn

Taking suggestions for better channels that cover thermals, noise normalized thermals, power consumption, frame times, and compare with a decent number of other cards.

Looks quite darn good for a potential console GPU.

Holy fucking shit

>RX 590 is good value all of a sudden
When the fuck did this shit start

Oh no no

Attached: Screenshot_2019-07-07-20-12-20.png (1280x720, 402K)

shhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh

Attached: Screenshot_2019-07-07-20-15-20.png (1280x720, 328K)

I implied no such thing. I just posted the carts for reference.

AMD shouldnt even have given this fat shill a card.

I know you didnt I was just saying

Well, look at the price. The MSRP was fucking insulting, now it looks much better.

>AMD shouldnt even have given this fat shill a card.
Why? All he criticized was that the blower design sucks, which is totally correct.

After his disastrous shill video where he compared amd using a preset he didnt like to principled technologies. Everyone should boycott that fat fuck.

>no shit games & csgo
good

Do they also take into account the value of how much power they draw from the wall? These charts are so stupid.

What I find interesting is that the 5700XT boosts to over 2000MHz, of course with the blower fan at 100%, just imagine what will be possible with the AIB in august that will probably have two eight pin and better power options.

>After his disastrous shill video where he compared amd using a preset he didnt like to principled technologies.
Sauce?

165W for 5700
180W for 5700 XT

>2% slower
>20% cheaper
jesus christ

How have you not seen it dude?

youtu.be/jzaXvEPyKd0

Hes so egotistical too I dont think he ever acknowledged how stupid this was.

He was correct though. AMD attempted to show an "AMD high number, Intel low number" comparison that had no impact on actual streaming quality.

>How have you not seen it dude?
Forgive me for not watching every vidya that gets put on Jewtube by e-celebs.

That being said though, how was that """benchmark""" by AMD not stupid and retarded? I'd get it if it were presented as a cheap funny jab at Intel, but when they seemingly unironically make it part of their official marketing statements, it's only fair to lambast them for it.

His results contradict every other review out there. I'm not surprised the fat retard can't even measure thermals correctly.

>testing on taxing preset bad!

I bet you see no problem benching CPUs on games at 720p or 1080p with a 2080ti either do you? Stupid fag.

The test wasnt about which preset you like to use it was about comparing two CPUs doing a difficult task. Furthermore the fat shill himself did the same thing he accused amd of. Kill yourself apologist.

>how is it not stupid

How is testing CPUs on a taxing preset stupid? Should a 2080ti be testing on 1080p because most people dont have 4k?

And again, GN themselves have tested CPUs using 10mbps stream settings before. So if you attack amd for it the fat egoist shill is guilty too.

Since this shitty nu4cuck loves range banning networks for everything and I cant upload files apparently anymore.

imgur.com/a/evCRbq9

People are too stupid to realize this. Especially Steve at Gay Mans Next Ass. It's like that PCIE-4 test they did. Nobody will use that either but it shows that it is faster under certain loads and conditions.

Intel shills are dumb as a rock.

>How is testing CPUs on a taxing preset stupid?
Testing CPUs with taxing workloads is fair, but this was obviously not just any benchmark, but cherry-picked to find a specific choke-point making Intel look totally unusable despite it being a totally unrealistic scenario.

They are so desperate to prove that amd and Intel are just as bad as each other (tm) they will go to any lengths.

Sad fact is amd is actually better than Intel in most every single way and now were seeing that.

How the fuck is picking a taxing PRESET cherry picking? Thats far more realistic than some artificial bench yet no one whines about that.

>cpu x is capable of y
>cpu z isnt capable of y
>MISLEADING!

AMD's CPU's should be tested against Intek K series CPU's using an equivalent air cooler that fits both platforms. Anything else is unfair.

Wat. It wasn't the bitrate he criticized, but the quality setting.

>Thats far more realistic than some artificial bench
Noone ever uses the "slow" setting.
>>cpu z isnt capable of y
>y being something that nobody cares about

Nah, it's pretty fair when you buy a CPU that whatever comes in the box is tested vs whatever comes in a competing product's box. Benchmark aftermarket coolers later.

>I care about 720p 500+fps benches with a 2080ti

Intcel hypocrites

>I bet you see no problem benching CPUs on games at 720p or 1080p with a 2080ti either do you?
Nice strawman.
>it was about comparing two CPUs doing a difficult task
They specifically chose streaming to indicate that in streaming, AMD does better. But they did it in such a way to suggest Intel CPU's can't be used for streaming effectively. Hence, the misleading numbers.
>Nobody will use that either but it shows that it is faster under certain loads and conditions.
Another meaningless test to show differentiation where there isn't any to be found.
I don't understand AMD's need to choose bizarre, unrealistic workloads to show off their product. It doesn't need it to look good.

So you should test Intel k series CPUs with no cooler?

Strawman? Its literally not. Its a completely pointless test used to compere and isolate the cpu. The stream test is a taxing preset some say isnt necessary in the real world used to tax and compare two CPUs.

take off the jew blinders

>>I care about 720p 500+fps benches with a 2080ti
That would be your position, retard, in that it's just as unrealistic and unimportant as streaming on "slow".

amd always looks good on benchmarks but when it comes to actual performance it's always subpar

it's like they choose the best binned cards for reviewers and everyone else gets the garbage leftovers

Yet everyone, including fat fuck Steve, use tests like that to ISOLATE AND COMPARE CPUs. What is the point in constantly testing things with tasks that dont properly tax them?

>What is the point in constantly testing things with tasks that dont properly tax them?
At least game testing at 720p isolates the CPU and might potentially work as a proxy for demonstrating smooth scaling to potential future GPUs. Just picking a particular data point with non-smooth scaling and laughing at "it doesn't work" isn't useful. It might have been potentially useful if they had smoothly adjusted the quality settings and demonstrated where the respective CPUs fall off. But even then, I don't see what kind of real-world scenario that might have been a proxy for.

upset?
i am, i was going to get a 3700x, but it just doesnt give the framerates i want

>testing streaming with a taxing preset is bad and isnt indicative of streaming
>testing on a setting no one will play on low res with the fastest GPUs is indicative of performance

Don't forget your $100 AIO bro

Again, one of those can work as a proxy for a useful metric, while the other doesn't.

>streaming isnt a proxy for streaming

*NEW streaming doesnt matter

"Slow" streaming isn't a proxy for anything.

Oh shut up. We all know AMD wins at reasonable streaming levels too, the point is that there's no reason to do useless demonstrations.

Forgot pic.

Attached: mpv-shot0007.jpg (1280x640, 94K)

Yes it is. Its literally streaming.

Yes there is. Just like playing on ultra or 4k ultra. Widen the gap and tax the cpu harder. Show multitasking capability more

fuck AIO's either go full custom watercooling, open loop, or get a tower cooler

I chose the dark rock 4 pro, cools within 4c of the best AIO's at 10dB less

Attached: 1561762585201.gif (264x320, 929K)

>fast

Enjoy your grass artifacts

Aios are a meme scam for losers

>Yes it is. Its literally streaming.
>streaming is just streaming, there are no differentating factors
>Just like playing on ultra or 4k ultra.
Again, those are metrics that display smooth scaling.

thats what i just said, fuck AIO's

>Enjoy your grass artifacts
Fix your buggy encoder.

Most people will use the stock cooler though. I want to see an equivalent cooler to the one that comes in the AMD box that fits both platforms tested. How is that unreasonable?

no the 5700 should have been $200

its not

im not denying the 3700x is good price/performance

but that doesnt matter to me if its off by 20+ fps causing dips below what i want to stay above

I even bought some b-die ram too, guess that doesnt matter as theyll work fine on TelavivTel

Youre saying streaming on a taxing preset doesnt indicate better multi tasking or better streaming performance. Listen to yourself.

Like it won't transfer over to streaming in 4k or anything else.

>I want to see an equivalent cooler to the one that comes in the AMD box that fits both platforms tested. How is that unreasonable?
It's kind of unreasonable if you consider that most people who buy Intel wouldn't buy a comparable cooler to that. If anything, it would be useful with a comparison between the performance on the stock cooler vs the performance on a comparable cooler to what people would use with Intel CPUs, to see if it makes a meaningful difference.

I know I was agreeing. AIOs are a meme

Most people buying AMD CPU's stick with the included stock cooler though. Using an AIO or fancy Noctua cooler and comparing it to a $20 stock box cooler is just silly. I want to see how they compare under similar conditions. Not the ideal conditions.