RYZEN 3000 Boost Survey

>RYZEN 3000 Boost Survey

youtube.com/watch?v=DgSoZAdk_E8

Boost is not guaranteed.
What is he butthurting about?

Attached: 5gDNpG.jpg (680x485, 279K)

Over 80%+ (depends on the CPU model) are not hitting advertised clocks. As in, if the box says 4,5GHz clock, you have a 20% or less chance that you'll actually hit that frequency, even with PBO enabled.

As a AMD user, that's fucking horrible and I bet there will be a few lawsuits coming hot and heavy for false advertising.

Boost is not guaranteed

And one more thing I'd like to point out: 20% or less chance that you'll hit the advertised frequency on one core on a CCX.

My 3600 hits 4.2ghz
:^)

Yeah they really should have written a 100mhz lower number on the box and avoided this problem.

But have factors like the motherboard and cooling been factored in? I don't trust that more than 20% of people have the VRMs and cooling to actually boost to the extent of spec

>those comments from AMDrones giving AMD a pass
Yikes

Thermal density of the chiplets is too fucking high. Power hasn't decreased enough to match the decreases in area scaling.

>Boost is not guaranteed.
Why not? It says boost freq right on the box. This is like saying that a CPU is 8 cores when it's actually 4. Or a car has 200hp when it actually has 100hp.
>Oh well SOME of our cars can achieve 200hp, but it's not guaranteed
AMD is going to get fucked in the ass in court. Their marketing people should have printed lower boost clocks on the box, by like 100-150mhz.

Really makes you think

Attached: mmwaf1992nj31.png (2216x957, 587K)

They need to mention their motherboard since power delivery and firmware are the determining factors in boost at the moment.

My 3600 only hits 4 ghz no matter how much more voltage I pump into it, it won't do more.

Boost would be considered a modification to expected performance, the defense is iron clad in that "overclocking" culture for generations has known clocks aren't promised when running modified. t. tech law guy

Boost isn't guaranteed because its workload dependent, and individual mobos play a huge role in it as well.

>Why not? It says boost freq right on the box
It says maximum boost frequency. It does not say that you will achieve that frequency guaranteed. Same goes for Intel chips, which can easily be limited by poor motherboards and cooling since they're such fucking housefires.

You retard, boost is considered part of the normal operation of the processor. It works with or without PBO or any other special elective features and you would have to modify to stop the boost behavior.

Motherboards with 16 phase VRMs that can handle like 700W can't do it. The chips are broken, the fact that AMD reduces boost frequencies with the 1.003 AGESA microcodes is proof. An ASUS rep outright said they had reduced it so they chips don't burn out long term

>even with PBO enabled

>2nd thing he mentions in excluded data is PBO enabled devices

Your intel and nvidia boost is not guaranteed either.

my 3600 only hits 4.05 and i am kinda pissed

literally the fucking mobos
I guarantee all the people that aren't hitting max boost freq bought the cheapest compatible mobo (and I'm not saying it's a bad thing to do that, but being surprised by the fact that you can't overclock it to it's maximum freq is as brainlet as it gets)
i mean, do you expect them to advertise the lowest possible freq you will see with no cooling and the cheapest mobo you can get on a hot summer day? what the fuck

mine hit 4.2 on the older bios the newer one I haven't seen it go over 4.1

i planned to get a 3700x but def gonna wait for amd to fix that

Im on MSI's newest, havent used the older ones since I very recently got it

Not sure what I'm doing wrong, but I can't even get a stable 4.05GHz on my 2600X, it's prone to freezing up/crashing often. This is using my motherboard's built-in "game boost" setting though, so maybe I'd have better results trying to manually raise my clock speeds.

That's a retarded anal/o/gy. Cars of the same model line can produce widely different hp figures due to variances in qc. Furthermore, advertised hp figures are only peak figures. Many small displacement engines don't make anywhere near their advertised peak hp rating until the upper range of their powerband, near red line sometimes. The auto industry is filled with marketing bullshit and broscience.

>You retard, boost is considered part of the normal operation of the processor.
It's considered a normal part of the operation of the processor when conditions allow for it. This has literally always been the case with both Intel and AMD chips. Zen 2 chips are thermally limited when it comes to boost, so they don't boost as high as they could. Gaymans Nextass literally proved this by lowering them to sub-ambient temperatures and getting higher boost with nothing else changed. The only thing AMD changed in the new BIOS updates is lowering the max boost temperature from 85°C to 75°C, which has a corresponding effect on how high chips will boost in the average desktop.

They didn't actually prevent the chips from boosting higher, so long as you can meet the defined conditions for it to happen. This has literally always been how CPU boosts work. Good luck with your class action lawsuit against something working as defined by spec.

People didn't read the fine print and this is what happens. AMD's "turbo" for zen 2 is the maximum a single core on the processor with all other cores idle is guaranteed to reach. It is sadly a side effect of higher volume/higher yield binning to keep prices low. Any frequency you attain on all cores above the base frequency is already technically OC'ing.

If AMD did bin these chips to hit turbo on all cores you'd be paying like ~$600 for a 3800X. Granted I was hoping that AMD would have made a chip to at least match the i9-9900K OC'd to 5GHz on all cores just to rub it in intel's face but given AMD's approach to beating intel through lower cost, higher Nthread performance, more efficient processors it was never a realistic outcome in the first place. Still, to outperform a 5GHz i9-9900K in every other metric and get on average 95% the same performance in gaming for a lower cost is a feat in and of itself.

Attached: _id1565695743_343178.jpg (718x456, 53K)

This video is talking about single core OC, not all core

G-g-guys please don't bring this up, think of intel, they are the real jews, just please let amd do what ever they want, they need the money!
THINK OF MY AMD STOCK!

Chiplets BTFO - AGAIN!

Attached: confused.webm (450x472, 1.44M)

Still not a worthy sandy bridge replacement then

Even a 3600 wipes the floor with your poorfag chip in any workload there is, copelet.

The way AMD binned these chips means only 1 core out of all the other cores actually hit the boost clock frequencies. That means you have to isolate that 1 single core and put the others to idle to see this boost frequency. PBO + autoOC + Bios update + new windows scheduler is also needed to see this single core boost.

It's not AMD's fault nobody read the fine print. I don't like personally, it's an intel-grade tactic but still better than intel.

Depends how high you OC it to and wether you fully enable all the 40+ (50+?) security mitigations which cripple I/O and core performance. C2D is basically unusable because of that.

Attached: AMDMSIfirmwareupdateboostchanges.png (1496x950, 74K)

>i5-2500k $216
>Ryzen 5 3600 $199
I think we all know who the poorfag is

>It's not AMD's fault nobody read the fine print.
> it's an intel-grade tactic but still better than intel.
I don't understand you people.

>Boost is not guaranteed.
Then stop writing the max theoretical on the box

MY 3600x cannot boost to 4.4ghz. ever.

so it sucks.

what pos motherboard do you own

its not about Chiplets, its about AMD being tards again. Seriously they cant go a single launch without at least one fuckup.

>buys trash motherboard
>IT'S AMD'S FAULT
Motherboard shills and their piece of shit VRMs should be shot.

Attached: 1567346799073.png (712x614, 35K)

I really dont undestand, it's like a company is holding them as a hostage or something. "It doesnt do what was promised, b-b-but it's still great"

Fix your garbage VRMs faggot motherboard shill.

This never happen on Intel, the clocks are always kept and you can even oc further.

Attached: [000554].gif (1307x734, 23K)

Too bad every high end x570 board is like 200 dollars

literally fucking who

fuck dumb ass gif is busted.

Attached: [000554].gif (600x337, 252K)

Isn't this the guy who has a hooker GF? I'm sorry but I cannot take any kind of information from a man like that seriously.

Tried an expensive Gigabyte x570, Cheap B450 asus, Cheap B350 Asus.

the fucking B350 boosted the highest. So yeah its not about motherboards or at least the motherboards are not the sole reason.

that is true. an expensive intel works like it should, is easier to OC. But it costs twice as much for the same performance.

>Intentionally go out of my way to buy an Asus X570-Plus over an MSI X570-A PRO
>BRO MSI HAS SHIT VRMS BRO LOOK AT THIS CHART REDDIT MADE BRO t. pcbg
>Now this
I'm not mad that "my motherboard" boosts 75MHz less than the other motherboard, I'm even mad that I potentially got shilled something. No, I'm mad that there's so much fucking disinfo floating around that it's nearly impossible to make an informed buying decision about computer parts EVEN AFTER doing weeks of research, reading journo reviews, and reading user reviews on websites.

Attached: 1563035782732.jpg (720x873, 62K)

It fucking sucks and I feel somewhat let down like everybody else. But at the end of the day it's still a better option than intel for the time being.

Maybe AMD can redeem themselves with Zen 3 on 7nm+ EUV by hitting at least 4.5GHz on all cores in 2020 but all we can do now is wait.

Try lowering the voltage to achieve a more stable boost, just pushing voltage can be harmful to ryzen 3000 CPUs.

And why does asrock b350 boost higher than half of x570 boards with the same CPU? Explain.

Attached: 1541181531230.jpg (1280x361, 228K)

>It fucking sucks and I feel somewhat let down like everybody else. But at the end of the day it's still a better option than intel for the time being.
See this is the cope I don't get. You aren't forced to buy, if neither gives you a good deal and does anti-consumer shit, then neither is getting my money.
Why people like you go
>AHHH It's terrible, but I will still take it ,because I hate intel more, so I will support amds scummy practices

Imagine running 9900k on cheap biostar shit and then complain it doesn't go above 4.5GHz.

bcuz its v good :)

the absolute state of amdrones

They're still extremely well-performing CPUs for a much better price than the competition, you fucking retard. Trying to conflate the two exposes you as the Intel shill that you are. Real-world performance is LITERALLY ALL that matters. Enjoy arguing about 25MHz differences in boost clocks whilst I enjoy actually using my CPU.

It looks like Zen3 will be a completely different arch family from Zen2, so whatever issues are present here may not transfer over.
Papermaster has also said that Zen3's focus will be energy efficiency. One of the big issues here is the thermal density of the chiplets.

No matter how bad AMD fucks up they're still saints compared to Intel's malicious PR, lies and bribes.

I won't, in fact I'm going for a 2600 +MB combo soon. Especially since the thread scheduling thing isn't going to get on LTSC until 2021.

All I'm saying for anybody still willing it's still a better deal than intel. With all the crippling security mitigations required for basic security, high end Z motherboards with 100 chokes, VRMs, and phases to even hit 5GHz on all cores, and the high end 360mm custom AIOs required to keep the chip from fucking melting zen 2 is still worth it if you're willing to put up with mainstream windows and spend more than for a value zen+ system.

Attached: Screenshot_20190901-124320.png (720x1280, 250K)

remember when Jow Forums manually oc their cpu instead of relying on this auto oc shit?

>built-in "game boost"
Fucking stop using that shit right now jesus christ. All it does is pump a fuckload of voltage. If yolu're as unlucky as me then you are better off using a voltage offset and letting PBO do its thing.
I could never even get 4ghz stable on all cores without crashes or ridiculous voltages. Multicore barely gained anything but singlecore performance suffered.
Right now I'm converting some videos at its sitting at exactly 4ghz, 70C and on lower workloads I regularly see 4.25ghz single core boost and 4.15ghz all core boost for short bursts.

>hitting at least 4.5GHz on all cores in 2020
>2019
>still caring about muh gigguhurtz
As long as its faster who fucking cares

For anyone not in the know 4GHz zen 2 = 4.5GHz coffinlake due to 15% IPC uplift over zen+.

>that image

Attached: spurdo_sparde_vector_by_kevinino-d6ehtjk.png (1024x749, 117K)

>tell people not to get intel becuse it's a bad deal
>only praise
>tell people to also not get amd because it's a bad deal
>OMG INTEL SHILL DIE!
hm.

>People who probably don't even own a zen2 are out there damage controlling this right now
I bought a 3900X knowing I won't hit the boost and I don't care. The CPU is still good without that extra 75MHz. But AMD fucked up and this exaggeration isn't defendable as a practice.

You're probably right it just sucks that AMD didn't outright tell people, "hey you're probably never going to see boost clocks on all cores, k?"

AMD is undoubtedly on the bath to becoming intel unfortunately but going with intel right now is akin to shooting yourself in the foot. Going with AMD is like stubbing your toe. No company is ever going to be a fucking pure saint, they're after your money in the end.

Would you really take a bullet to the foot over stubbing your toe?

As far as ive been here almost no one on Jow Forums does this.

>don't support either

>Would you really take a bullet to the foot over stubbing your toe?
>no you must choose one, you MUST, YOU MUST, and clearly it should be AMD.
Again I don't understand this. If you have a working pc, just use that until they give you a good deal with no scummy pratices. If consumers actually did that, the companies would be forced, yes forced to deliver that, instead knowing they can get away with it.

Meanwhile my 2700x is almost as fast as a 3700x and uses the same power and it's at 4ghz all day Al core low temps
Dumb fucks who bought first run silicone are dumb fuck em

Attached: 1418684956425.jpg (337x361, 18K)

>nominal conditions
what does that even mean?

You read that right there some that do oc the large majority says they dont need to oc and end up buying new shit instead of overclocking.

You and I both know this is never going to happen, but at least we can get the scraps left over from all the cutthroat competition. See at least in my case it's going to be a massive jump from a 2GHz laptop APU.

Nope, that's a 15% jump in IPC from zen+ so it's at least that much more faster. But imho zen+ is still better value than zen 2 until those eventually drop in prices from zen 3 release.

Updated BIOS/windows thread scheduler, PBO+autoOC, AND you isolate the 1 single core that actually hit boost and put the others to idle.

Zen 3 is kinda useless for anyone happy with zen+ or even Zen og if I'm honest

Guess I might dump a 12 core 3900 in my x470 if I absolutely need to upgrade down the track but they don't even reliably clock that much faster so it's kinda pointless

>Right now I'm converting some videos at its sitting at exactly 4ghz, 70C and on lower workloads I regularly see 4.25ghz single core boost and 4.15ghz all core boost for short bursts.

AMD CLOCKLETS BTFO

Attached: clocklets.png (403x402, 28K)

Yeah, I'm manually incrementing my clock speeds up now, and it seems to be running a lot smoother, hell it's running cooler too, though that's probably because boost is disabled

manually incrementing my cpu ratio*

whoa, are you telling me that the voltage regulator is the determining factor in the sustained power draw of a processor?

what do you mean i can't just buy the cheapest micro atx board on newegg and get sick nasty clocks? wtf amd

Now how many security mitigations do you have fully enabled? Spectre and Meltdown tend to slice SSD performance in half especially on older intel processors.

Attached: 1558039242262.jpg (700x421, 33K)

>be amdpoorfag
>wait a decade for AMD to catch up to intel so you can save like $100 for equivalent processor
>turns out its still not fast enough
>spend the saved $100 on more expensive motherboard
>still get cucked by newer BIOS

>pic related, state of AMD boost clocks

Such is the life of an AMD poor fag.

Attached: Invisible-Boy-Mystery-Men-Kel-Mitchell.jpg (500x567, 63K)

Also
>base OC

in a duopoly, everybody loses. this is not an issue of one side being better it is an issue with there not being any true competition

amd is only competing with intel, and as long as intel keeps acting jewish, then there's no reason for AMD to be anything but the lesser of two evils

>be intard
>buy high end 100 choke/vrm/phase Z motherboard to attain 5GHz OC
>buy custom 360mm AIO to keep the chip from melting too
>get all that performance boost nullified by crippling 40+ security mitigations for the bare minimum security possible

Attached: 1540209531358.png (552x661, 288K)

the base clock on a 7600k is 4.3ghz single core turbo. and i got all 4 cores at 4.7 baby

clocklets still btfo

my 3570k can't go above 4.2GHz without gigantic voltage tunings, which is barely 10% above stock, so it's not like Intel is better.

And whatever is their current gen mainstream flagship (8700k? 9700k?) hits 100C unless you use watercooling. Or if you delid and use liquid metal, then a hyper 212 keeps it under 70C.

Do you guys get paid to do this? I can't imagine shilling for free.

Keeping the same clocks with 5% better IPC while lowering the voltages and temps drastically would be kinda great ngl

i have 2600 and 3600.
3600 is noticeably faster.
the 2600 bought 6 months after launch and is a good bin. oc'd to 4GHz 1.225V.

don't underestimate the advancements they made between ivy bridge and kaby lake. the 7600k can run a 4.5ghz overclock on a stock cooler without seeing crazy temps, and the stock cooler is dogshit

but at the same time, that was before they scaled up the core counts, and even though kaby lake might be pretty decent at smaller core counts i look at the 8, 12 core kaby lake-x processors and have to wonder if they're just hoping that you're never going to all core turbo that many cores

uh people don't have any self respect and they pretend to be a devil's advocate and argue in favor of their own opinion so that it looks better, despite the fact that doing that is straight up acting against your own best interests

5-10% in gayman is fuck all
I'll wait till they solve the bugs and get zen2 cheap eol

Dunno why dumbasses side grade so quick just fucking wait

>>turns out its still not fast enough
Why does it have to 100% just as fast? Why can't it be 90% just as fast and still be worth it as long as it's cheaper? Why is everyone so keen on burning money? My 4.15GHz all core 2700X cost me ~$200 in a $300 B450 CPU/motherboard bundle a couple of months ago, also got Cl 15 3200MHz 16GB RAM kit for just $100. An i7-8700K CPU ALONE is like what 350 fucking dollars?

Pic related: The 2700X is OC'd to 4.2GHz on all cores and the i7-8700K is OC'd to 5.0GHz on all cores btw.

>inb4 i5 stutterfest

Attached: 1080p_MIN.png (1324x1665, 101K)

>user opening up his pc case

Attached: stable temps.jpg (2160x600, 115K)

PUNKED
AND
CLOWNED

Attached: 1565541479593.jpg (1462x1462, 309K)

By putting your motherobard on life support, I give it 2 years max.

>it's an intel-grade tactic but still better than intel.
this. hypocrisy

Attached: 1565442767949.jpg (359x335, 8K)