Gamer Nexus back at Ryzen again

> Sup Forums was implying BIOS updates would DRASTICALLY improve Ryzen performance
> see this
BASED KEK NEXUS
youtube.com/watch?v=8cHJ7FDZKg8

Seems like BIOS updates can't into magic guys, what really makes the difference is RAM clock speeds + Game Updates.

Other urls found in this thread:

gamersnexus.net/hwreviews/2865-ryzen-revisit-ram-overclock-windows-update-efi-updates
youtu.be/nLRCK7RfbUg
youtube.com/watch?v=kmLtww8SMpE
youtube.com/watch?v=fTJi_gLyXQQ
youtube.com/watch?v=QBf2lvfKkxA
youtube.com/watch?v=oi14tOZ_-kg
twitter.com/NSFWRedditVideo

Let me know when someone competent runs their tests.

Depends on what boards they tested on, some people did test on multiple motherboards, and some had significant increases in performance, and i think gigabyte's actually had less performance.

A video is a pretty inefficient way of displaying benchmark results. Just give us an image with the graphs on it you retard.

...

amd's microcode update for bios is out in may.. and was announced a month ago

...

>Gamer Nexus

A la basura

AMD retards don't care about facts.

You should know it by now.

Who's more competent? They're the only one that do a well documented research.

AdoredTV

08/15 Pajeet YT-techchannel

t.Red team

There is no alternative facts.

A BIOS update let me get from CL 15 to CL11 on my RAM

>Gamer Nexus
Isn't this that crybaby from 2 weeks ago?


>also seriously watching any YT channel with the word 'GAMER' in the name

from the GN article

You fucktard have no clue.

It's a legit site which also features YT videos.

Instead you redtards think AdoredTV is seriously neutral.

Braindamage is real amongst your kind.

>crybaby
wut?

It's Gamers Nexus retard

How about somebody who knows how to act like a fucking professional, instead of playing recorded conversations?

How about someone who doesn't feel his ego is more important then the data and has to add his pointless color commentary about shit instead of letting the viewer examine the results?

I can be the most thorough junk collector too, but I'm still collecting junk.

His testing is now officially garbage in, garbage out.

They're both shit, this being a whinny hippie while the other is a fat Scotsman .

Youtube celebs are fucking worthless, and you can shove them up your easily impressed ass.

Wouldn't it be nice of him to throw in the 7700K and 6900K data from Ryzens launch too?

Hi i am an eater, i eat everyday.

No one thinks adored is neutral. But then no one is disproving him or his points either.

His results are the closest to the real world, no bias or whatsoever.
> inb4 HOLY FUCK I GOT 300FPS BOOST WITH A BIOS UPDATE GIVE ME VIEWS PLEASE chanel.

Pic related is GTA V, this game benefits more from RAM speeds apparently.

Something something he got "threatened" and "nearly emotionally raped" by some guy on the internet and threw a massive fit.

He did
> pic related
read the article, it's pretty good.
gamersnexus.net/hwreviews/2865-ryzen-revisit-ram-overclock-windows-update-efi-updates

Wow, you should sic /jp/ on him, he'd probably kill himself.

INTEL ON SUICIDE WATCH
> SHUT IT DOWN
> DON'T REVIEW THE 1800X!!!!!!

>testing while BIOSes are still in flux and updates come every few days

What a waste of time, the result now will be worthless in a week.
And then worthless a week after that BIOS update because AMD is releasing their AGESA update.

>see comment by GN in the OP

Then see
From his very own shitty review.

Is this guy legit fucking retarded? Is someone else doing this for him and he doesn't know? Split personality?

that improvement came from a game update, you huge faggot.

What game besides DOTA and Ashes had an update for Ryzen?

makes sense that drake listeners and nvidia owners would both be jewish.

yea the bios updates fix the fucking overclocking and huge improvements on ram.. coming soon in may too
and that'll be the only performance microcode update that we're getting (may)

but bios just fixes all of what we can change.. it doesn't magically add performance yet

officially? none.

It's going to be a very gradual performance gain over time

better a jew than getting cucked by low fps

MSI and ASUS have been releasing on launch day BIOS updates daily, there's wildly different performance results on them, especially on ASUS ones.

The ones impacted less by performance are the ASrock mobos.

That is from the TWW getting a patch as well.

Nice 20% increase from the combination of win update and game update, though.

How does that compare to the 7700k and 6900k?

>i7-7700K @ 5.1 GHz

You either need an absurd cooling solution or win the silicone lottery to hit this kind of number. Maybe both.

nvm I see it now.

So it beats the 7700k in minimum frames now. Nice.
Thanks for posting.

you must be from the special snowflake class.
fyi most mainboards dont let you clock ram higher then 2133mhz without bios update.
the last updates i got for my mobo allowed me to clock from 2133 to 2933mhz.

the bios literly doesnt work with higher speeds.
LITERLY DOESNT WORK.

now you shit here that bios updates dont improve performance.

please have a sphincter infarct you brain dead retard. if you want to shill at least find real problems and point them out.

When is Pinnacle Ridge/zen2? Most of these bugs should be fixed by then, and I'll have enough money for a whole new system around September.

>it's not the bios
>it's the thing the bios update allowed
>gaymursnexus is never wrong

>I have one shitty board
>but I know how most AM4 boards work

This is the first I've heard this. I've instead of heard almost everyone with a B350 or X370 getting at least 2800, though 3200 less common.

Explains why 80% of launch day reviewers were done with 2133/2400MHz memory.

>that improvement came from a game update
Except that's unconfirmed, and the Warhammer patch mentioned nothing about improving Ryzen performance.

>shitty board
asus prime x370 pro

thats how they rolled out the x370 chipset bios and the ram compatibility was bad to say it mildy in the begining the last several bios updates made some improvements there but still it doesnt "just work".

Yes. day 1.

Now that's not the case. B350 boards did take longer to get updates, but most of them seem to have them.

The situation sure improved over day one.

Too bad all these faggots are testing gaming with the updates, not any multithreaded workloads like handbrake.

the b350 boards get the same firmware updates from upstream amd and these are the same that go into the x370 bios's.

Yes. They basically test them out first on X370 ones so they get updates quicker, then when they're more stable they get released for B350 ones as well.

I've heard of many people with many different B350 boards whether it's Gigabyte, ASRock, MSI, or Asus, all getting 2800 or 3200 fine.

>Tomb Raider

Wasn't that like 40% slower on Ryzen on day 1?

HUB has blown adored tv the fuck out twice in a row now. First with the 2500k vs fx 8370 benches and also the one where adored tv claims Nvidia are humping ryzen and that a fury x will make ryzen look better.

Adored tv is a paid shill amd affiliate and even does giveaways for amd on Twitter.

>AMD subhumans before release of Ryzen
>Gamernexus is most reliable reviewer of AMD products
>AMD subhumans after the release of Ryzen
>LOL STOP LINKING THIS SHILL

AMD niggers should put down.

...

You're welcome to try.

With DX12 it still mostly is.

DX12 on Nvidia with Ryzen drops performance like 30-40%.
It's a problem with Nvidia's DX12 driver.

The Fury X roughly matches the Titan X Pascal in DX12 games with Ryzen, that's how bad Nvidia's DX12 driver is especially on Ryzen.

>HUB has blown adored tv the fuck out twice in a row now.
Lmfao no he didn't. If you believe that, you know as little about what you're talking about as HUB does.
AdoredTv is bias as shit and gets quiet a bit wrong, but not as completely full of shit and wrong as HUB is.

HUB dude is such an idiot he had no idea what was going on in those other videos nor how to reproduce them like tons of others have.

>IntelNexus

kek

>He still tests using Nvidia GPU's and shitty DX12 API drivers

dismissed

Fucking this. Adored once again was proven to be a fraud and liar.

youtu.be/nLRCK7RfbUg

Nope. I'd believe video evidence over some random spreadsheet someone made up on the net.

You all better stop talking shit on Steve, or else.

>I believe what I want, thanks

Ok.

youtube.com/watch?v=kmLtww8SMpE
youtube.com/watch?v=fTJi_gLyXQQ
youtube.com/watch?v=QBf2lvfKkxA

Great evidence of Nvidia's driver being shit and their GPU sitting on its ass doing nothing?

youtube.com/watch?v=oi14tOZ_-kg
GamersNexus was talking about testing multitasking while gaming, too.

But it kind of blows my mind that Steve seems so wishy-washy about the idea that some people would, wow, actually have two monitors. Especially people spending $1200+ on a PC.
I thought that shit was common? Am I just one of the 1%?

>I'd believe video evidence!!!

How come both companies have an Asian CEO but only one has awful drivers anymore?

What has this got to do with the fact adored tv claimed a fury x would be better than a 1070 on ryzen because of apparent "Nvidia ryzen gimping"?

Nvidia's DX11 drivers are fine, in fact they're great.


Their DX12 ones are fucking garbage, and actually lower performance than their DX11 ones in most titles.

>Overclocks a 'golden' sample 7700K to 5.1Ghz.
>Does not use the most common 4.8Ghz a 7700K every 7700K owner can expect.
>Uses Nvidia hardware only and completely ignores AMD GPU's on Ryzen because 'muh bottlencks' when Nvidia's API is the biggest fucking bottleneck out there.

I've had enough of these asswipe shill 'reviewers'.

>a RX 470 beating a GTX 1060 on DX12
Yeah, nvidia's drivers are just fine. Ryzen is still shit. Nothing to see here.

What am I looking at here? I can't see anything irregular

every graph has a stock 7700K you bastard

You're obviously not following the topic.

Nvidia's DX12 drivers can't use more than 4 cores, this gimps both Ryzen and the 6900k

While their DX11 drivers can use more than that, due to a very interesting piece of technology they have implemented.

I have nothing to say about Adored, I don't watch his shit.

that doesnt refute anything he greentexted

In a perfect world without shitty APIs and shitty pajeet developers, the 470 would be slightly more powerful than a 1060 from a pure hardware resource perspective.

>Nvidia's DX12 drivers can't use more than 4 cores

How do you know this? Has anyone reputable like anandtech confirmed this or is it the regular speculation based on some random scenario in a game babble?

1. It's not hard to hit 5GHz on any 7700K, no need for cpu lottery
2. Who even cares? Overclocked Ryzen can't even com close to the stock 7700K anyways
3. It's AMD's fault for not being able to get higher clock speeds

AMD DX11 drivers are fine, too.

They're actually overall lower CPU overhead than Nvidia's DX11 drivers, and the difference in performance comes from that the scheduling is done on the GPU itself after the instructions are sent.

AMD GPUs favor instructions sent as a stream constantly, not in chunks every few MS, which is what happens when multiple cores are sending instructions that get sent at different times.

AMD could split it up in their DX11 driver but... it'd be higher CPU usage overall to shave off fractions of a MS due to the overhead of so many threads doing little work.
It works for Nvidia because their DX11 drivers does A LOT more work that the hardware scheduler itself does for AMD.

You don't see how shit DX12 with an Nvidia GPU is running on Ryzen while the AMD GPU outperforms that in both DX11 and DX12

>1. It's not hard to hit 5GHz on any 7700K, no need for cpu lottery
5.0 isn't 5.1.

5.1 is something around a 20% chance at reasonable voltage.
A little over 50% hit 5.0

It's pretty shilly to include a 7700k at 5.1 as what most people can expect of their overclocked results. 4.8, maybe 4.9, is what everyone can hit.

What the fuck is this pic?

See 1700X OC 3,9 b693 vs b970: it's 101 vs 115 1% low and 137 vs 152 AVG. It means that there is almost no reason to fuck around with SMT.

OP-pic is misleading: of course there might be only mild improvements if you are willing to cut MT performance or rebott each time before running a game but most people would not want that.

>They're actually overall lower CPU overhead than Nvidia's DX11 drivers

Digital foundry proved this to be false when they tested the rx 480 vs GTX 1060 on the new Pentium. In games where the rx 480 would lead with an i7 it would crumble with the Pentium whereas the 1060 barely dropped any fps. They even said it was the amd dx11 driver straight up.

I get your point, but why complain about 100MHz? That CPU is probably less than 1% faster than s 5GHz 7700K, which is pretty common.

All this goes to show is that most tech reviewers are incompetent and only interested in click bait.

They're already arguing 5 FPS blows Ryzen TFO, of course they're going to try a completely unrealistic OC for most people to try and get that extra edge so the teeny little bar graph looks that much longer.

How common is it without an AIO and/or delid job?

>So it beats the 7700k in minimum frames now. Nice.
The overclocked 1700x and fancy ram beats the STOCK 7700k in min frames only.
>nice
>nice
?????

> fucking around SMT
> ever
The human eye can't see those extra 2 fps

Its pretty shilly to include a 1700 at 4ghz when most people can't get anything more than 3.8 on all cores and anything more requires a truckload of power with a lot of instability. Most people a far as I've seen keep the 1700 at 3.8 on all cores.

the bios updates allow for higher clocks on memory
you fucking moron

It's not the driver it's the way AMD works vs Nvidia in loading tasks onto each core. There was a video from a developer explaining it posted yesterday but I can't be bothered finding it.

Vulkan and to a degree DX12 removes this issue but Nvidia are in no hurry to allow DX12 not Vulkan to take over for obvious reasons.

No. They did not prove it false. It's due to the game engines in most games.

AMD's DX11 driver makes the overall load on one thread higher, which means a game needs to separate its render thread from the main thread which most don't but they SHOULD do.
Nvidia's will however use more TOTAL CPU usage by a very large margin, but it's distributed over multiple threads.

what did you expect from an AMD subhuman?

Because if you have a min and max frame too far apart you get a fucking stutterfest due to latency between each frame. I'd rather play a game at a stead 60-80 fps than a gam at 45-110 fps.

Nothing really. All AMDrones need to fucking die.

They didn't include the 1700 at 4ghz you dumb faggot.
It's at 3.9 which everyone can achieve. That's the exact reason the 7700k should be at 4.8 or 4.9 as well.

So Ryzen went from 30% slower in games to 5% slower in a month?


Yikes, I can see why some are getting nervous

Lets be frank here, you can achieve any clock as long as you're comfortable with the voltage required.
Both 7700k and 1700

Amdrones are like the feminists of the gpu/cpu industry. They claim everything is trying to oppress them and keep them down and that everything is a conspiracy against them because they think they're superior.

The only community more cancerous is the sony fanboys.

>well documented research