Think this is amazing? What if I told you Ryzen actually h

Make it stop.. Please make it stop..

t. Big Blue

>>59619035
>Little

Guess you can't start a thread without IT'S OVER XXX IS FINISHED

>>59619003
wow is intel shit

>>59619003
IT'S OVER INTEL IS FINISHED

POwer usage is irrelivant to most people fag

why would they waste money to put something no one can use?

>>59623485
>POwer usage is irrelivant
>compiling is irrelevant
>video encoding is irrelevant
>smooth gaming is irrelevant
>streaming is irrelevant
>OC without premium price is irrelevant

what else is irrelevant?

>>59623485
It's only irrelevant when Intel gets stomped into the ground.

>>59623550
Because it's not ready to enable yet? These dies are used for both server and mobile parts, which this would help immensely with power draw as integrated voltage regulators can switch voltages in smaller increments AND around 10 times faster than motherboard ones.

>>59623726
you think they will for desktop parts? don't think it's that easy, would be nice

>>59623769
Maybe some future revision or maybe with a future firmware update, but it's way more important to get it running on server and mobile parts so I don't think AMD is spending too much time on this at the moment for desktop.

>>59623726
Watch it be a Zen2 feature on desktop and bury Intel into the ground.
A FIVR has a similar effect as a half node shrink on power consumption, lol cannonyields

>>59623591

Cinebench, blender, IPC (only megahurtz count) and cores are all irrelevant last time I checked.

>>59623550
They didn't waste money, this is a very good futureproof move, Intel did the same thing with Haswell and Broadwell but decided to not use them anymore because it made a massive hotspot on their CPU and increased TDP, so now you got Skylake and newer without the integrate voltage regulator.

It's there, it will be enabled sooner or latter, and that's good to know.

File: 55329.png (30KB, 650x470px)

30KB, 650x470px

>>59623550
Here's an example what happened after Intel put the FIVR on their Haswell chips compared to Ivy Bridge.

>>59623821
>people are already "next time for sure"ing ryzen
>it's not even out for most people yet
K
E
C
K

>>59623591
anything that isn't photoshop and dota.

>>59623992
Don't fanboy, disabling certain features either due to profit, time or thermal constraints and using it in the next generation is something that has been used in IC design for a while, a great example is the first Gen Fermi, which had 512 CUDA cores on the die but only 480 enabled, and the 2nd Gen Fermi was simply the 480 with the enabled cores and a unfucked TSMC process.

File: ..png (1KB, 50x52px)

1KB, 50x52px

>>59624022
>RYZEN IS THE NEW FERMI

File: Intel-Core-i7-7700K-vs-Core-i7-6700K_Power-Consumption-Normal-840x630.png (100KB, 840x630px)

100KB, 840x630px

>>59624041
Intel can only wish.
If anything's the fermi, it's Delidlake

>>59619003
>What if I told you Ryzen actually has a integrated voltage regulator (like Broadwell/Haswell) that's NOT even enabled for desktops?

I'd say such an action on Intel's side would trigger you AMD shills to shitpost for months, yet it is quite okay when AMD does it.

>>59624106
If Intel had it maybe their CPUs wouldn't be housefires, too bad they don't (anymore)

Stupid question; is the proper way to pronounce the acronym "dLDO", dildo?

>>59624286
Why not

>>59623966
Can we get s temperature graph now?

>>59624141
That's why they took them off, genius

>>59624547
They all run slightly above room temperature.

>>59624581
Whatever Devil's Canyon is a known house fire

>>59624706
Devil's Canyon is using solder, not semen for its interface material so no it's not a housefire like the other mainstream crap

>>59619003
Core M like CPU with 4 cores when?

Imagine that shit with Vega fanless.

>>59619003
>It's ignored and the motherboard regulator is used.
I can bet half the reason is because mobo vendors could price mobos higher, AMD and Intel just keeps removing stuff off the mobo PCB onto the CPU/chipset and it's driving mobo vendor profits nuts

>>59624722
And you are now a certified ignoramus

File: the final proof.jpg (395KB, 2400x2000px)

395KB, 2400x2000px

>>59624017

>>59624763

Q6600 a few years ago

>>59626473
Waht the fuck

Remember when /g/ wasn't filled with advertising shills?

>>59624722
>Devil's Canyon is using solder
Why do you dumb little kids just say things?
The Haswell refresh Devil's Canyon does not have a soldered IHS, it uses intel's proprietary TIM, and temps are just as bad as Haswell because of it. It greatly benefits from deliding.

http://www.kitguru.net/gaming/uncategorized/anton-shilov/intels-devils-canyon-chips-ngptim-is-still-not-efficient-research/

Don't spread your lies here, kike.

>>59619003
not sure if you watched what amd did, but they took the retarded mess that was bulldozer, and made it VERY power efficient by the end of its development...

Amd has FAR more experience with power issues than intel does, now that they have a core that is worth a damn, everyone thinks this low power stuff came out of nowhere.

>>59626995
did lisa give you a bj to make you shill out like that

>>59627029
dude, they took a power hungry low ipc piece of shit, and turned it into a something good, relitivly good, they just never put any of the newer parts on full pc parts, only mobile or apu

>>59627029
He's not wrong. Excavator is so efficient compared to Steamroller almost entirely down to power delivery optimisations. AMD spent a lot of time polishing a turd, but it was worth it because they gained a lot of experience on how to squeeze the most of a given process node.

>>59623550
To not cannibalize their single socket power efficient server offerings and work out firmware kinks with motherboard vendors
>>59627029
AMD kept reducing power usage of their chips in their 28nm process, maintaining them competitive with Intel's 14nm parts on certain workloads
Specifically the mobile APU's
On the same 28nm process compared to Intel's 14nm, that's 3 nodes behind
The only way to do this is by improving everything else, all of this expertise made it into Zen

>>59623966
that surely can't be wholly attributable to FIVR though. It would be nice if it can help Ryzen too, but people shouldn't be expecting miracles from one feature alone.

>>59626995
They did?

>>59628847
bulldozer started out slower than phenom II, and the last desktop version is better than it by significant numbers, bulldozer was also a power hungry monster, with a 225 watt cpu variant. Part of that was due to the insane stock oc, but a lot of it was architecture itself, and they managed to get variants down to reasonable power levels, though performance was still lacking.
they had years of working on this one issue exclusively, where intel had separated themselves on many fronts.
Make no mistake current bulldozer derivatives are not good for how powerful they are, but are reasonable in power they demand.

ITT: paid AMD shills

Fuck off with your marketing and kill yourselves

>>59629343
You aren't going to get paid Pajeet, specially with such low quality posts

>>59628847

They took a 125w 8350 down to a 95w 8370e.

They have 4 core APUs pushing 1.4v Vcore from the factory down to 65w.

Mobile Bristol Ridge APUs have a 35w TDP and it's just an underclocked, undervolted desktop part.

Throw in Lisa Su's emphasis on performance per watt and IPC for all new semiconductors and AMD is a brand reborn.

>>59629639
The APUs really fascinate me. I have an A8 7650k and it's really fantastic for the price.

>>59625292
So, Ryzen provides a smoother gaming experience. Nice to know. Thank you.

File: 1460135885539.jpg (811KB, 2560x1440px) Image search: [Google]

811KB, 2560x1440px

>>59631665

>>59632479
That's been patched in windows 10 though, issue was that too many tasks were being put on the virtual threads when there was still idle cores available.

When did you last update your shitposting repository?

>>59628136
That's far more than any minor architectural update could bring, it's all on the same 22nm node too.

Switching voltages in the 100 of ns is a pretty big deal compared to switching voltages every 10ms

>>59632694
are you blind? ryzen has flat line, i7 goes all over
OCed 1700/1700x will get same flat line

>>59633741
Fucking look at it again, it's ryzen that has the heavy dips

File: 1468611411932.jpg (790KB, 2560x1440px)

790KB, 2560x1440px

>>59633799
see that red line? that red line is 1800x