What in the fuck How https://www.techpowerup.com/reviews

File: overclocking.jpg (85KB, 786x727px) Image search: [Google]

85KB, 786x727px

>>59446062
Some more neat stuff

>>59446062

>CPU that gets the most work done also uses the most power.

Looks pretty normal to me.

>>59446074

remember to subtract 20c to get real temps.

imo the 1800x should be left at default. it cant oc for shit.

File: Screenshot_20170317_151340.png (234KB, 1373x422px) Image search: [Google]

234KB, 1373x422px

>>59446082
It's using 28% more power for 10% more performance at best.

>>59446095
wtf? 400 mhz more and zero performance gain?

>>59446082
this

the i7 is probably sitting at 99% load and the 1800X at ~70%

>>59446109
In gaming, XFR boosts a core to 4.11GHz and it's disabled when OCing.

File: power_multi_thread[1].png (11KB, 500x250px) Image search: [Google]

11KB, 500x250px

>>59446082
>>59446112
actually this confirms it

>Corsair 16 GB Vengeance LPX DDR4
>@ 2133 MHz 15-16-16-35
>@ 2666 MHz 16-16-16-36

>>59446062
WOW A FUCKING I7 WAY BETTER THAN RYZEN NEED 20 MORE WATTS TO RUN

INTEL IS FINISHED

File: Screenshot_20170315_113004.png (84KB, 726x790px) Image search: [Google]

84KB, 726x790px

>>59446149
Yeah, way better.

File: 1488728804432.png (45KB, 960x540px) Image search: [Google]

45KB, 960x540px

>>59446062
>>59446062
>testing the power consumption of the least efficient AMD 14nm processor on the market instead of its way more efficient little brother that's 5% slower

>>59446095
How did the performance gap drop from 20%+ to 10% in 2 weeks?

>>59446206
Early reviews likely didn't disable stuff like core parking which seriously fucks with Ryzen's performance.

>>59446206
TPU wasn't using ultra buggy ASUS motherboards, also they disabled HPET and disabled balanced mode that parked cores nilly willy.

If only they managed to get higher clocked memory working, it would look even better.

>>59446176
>b-b-but total war...
>w-what about fallout 4...
>it's not fair

>>59446095
>only 11% faster in reslet resolutions with 10% higher single core clockspeed and 20% higher all core clockspeed on top of 10% higher IPC in applications that use 4 cores

Wow Intel, whatever the fuck are you planning next year?

Would have liked to see one of Intel's 8 cores in the review for comparison.

>>59446275
There's shitloads of 6900k comparisons out there.

http://www.hardwarecanucks.com/forum/hardware-canucks-reviews/74814-amd-ryzen-7-1800x-performance-review.html

>>59446272
>Wow Intel, whatever the fuck are you planning next year?

MOAR CORZ on skylake v3
And I guess dual core cannonlakes for laptops lmao

>>59446091
That's why anyone smart will get a 1700 and OC it.
Or a 1700X if you want to squeze another 200Mhz more from the OC

Dunno if that's worth $100

>>59446285
Yeah but it only makes sense to compare them with TPU's results if TPU is the one doing the tests. We don't know if TPU or if the other reviewer's methodology is flawed so the result may vary.

'This is epic. We're assuming you've sifted through our game-test results before seeing this page, and so you'll find that the gaming power draw of the 8-core Ryzen makes Intel's quad-core i7-7700K look bad. The power-draw is as much as 30W lesser!

Ryzen is hands down the most energy-efficient performance CPU AMD ever made, and easily outclasses Intel's 14 nm "leadership." Good show. '

Once memory speeds are sorted out Ryzen will match and sometimes beat the 7700K hands down (including games which will only make better use of Ryzens architecture) whilst using less power.

We told you it was Windows, motherboard BIOS's and development of current games and apps holding it back

>>59446329
Ryzen seems to heavily favor high speed memory. The memory speeds may be almost as important as CPU clock speed for Ryzen performance wise.

TPU's admin hates AMD so quite surprised seeing him give it a positive review.

>>59446409
Yeah that's a real shame since memory is really expensive right now as it is. But maybe there is more to be had from BIOS optimizations.

>>59446422
Which is why this is probably the best review so far and to be trusted. He notes how the ecosystem is shit though and how AMD should have fucking got their act together with mobo manu's.

>>59446206
Don't worry, you'll still see the same shills and retards post the day one benchmarks months from now.

>>59446261
Day1 benchmarks on crashy asus boards are all that matter.

>>59446139
Holy shit that's insane, or are the 7700k's just overclocked housefires?

>>59446206
>>59446221
>>59446456
every entertainment tech website rushed out their Ryzen "review" so they can cash in on the hype

still waiting on anandtech's ryzen gaming benches...

>>59446310
There have to be more differences, the 1700x TDP is 30 higher. I'm still uncertain which would be the best option.

File: software bottleneck.png (849KB, 1024x574px) Image search: [Google]

849KB, 1024x574px

>>59446062
>one cpu is being used 80-100%
>the other cpu is being used 40-50%
Seems to make sense that the cpu under less stress is using less power.
Ryzen's poor gaming performance is down to the current state of optimisation and will get better with time as more of the cpu is being used.

>>59446544
link to this video?

File: Ryzen power.jpg (75KB, 711x533px) Image search: [Google]

75KB, 711x533px

>>59446062
This is likely measured from the wall.
TPU testing methodology is really slipping.

The 1800X stock will pull 110w under a 100% AVX load.
The i7 7700K will pull about 95w.

In a gaming workload the 1800X will pull just under 60w.
The i7 7700k will pull about 75w.

>>59446222
Yeah they only tested with 2133 and 2666mhz RAM.
LegitReviews showed a nearly 8% increase in game performance going from 2400mhz to 3200mhz RAM.
AMD also has a similar result up on their own site.

>con
> Lacks integrated graphics

Oh come on W1zard

>>59446544
More like why the fuck is there so much load on the SMT threads instead of the physical cores? Scheduler working fine my ass.

File: 02-Power-Consumption-Gaming_w_711.png (23KB, 711x534px) Image search: [Google]

23KB, 711x534px

>>59446568
Figures for the i7 7700k here:
http://www.tomshardware.com/reviews/intel-kaby-lake-core-i7-7700k-i7-7700-i5-7600k-i5-7600,4870-8.html

Really shows the efficiency of AMD's new chip.

>>59446544
>Ryzen 134.5 FPS
>I7 130.5 FPS
>Ryzen's poor gaming performance
What.

>>59446543

SenseMI. We haven't seen everything out of it yet.

So, AMD shills. Can PSP be disabled yet?

>>59446445
> AMD should have fucking got their act together with mobo manu's.
Nothing but minor teething issues of a brand new platform.
Intel had full on recall of some things when they first launched Sandy Bridge due to bad chipsets. AMD's only problem in regards to motherboards has been vendors not fully sorting out their BIOS. Thats such a tiny thing to fix its barely a road bump. Though it did hurt them somewhat in reviews, they'll fully recover from it.

>>59446685
Yeah my brother had to return his mobo for a new one to fix the SATA 3 issue. Shit happens. Intel cucks have conveniently vacant memories when it comes to that kinda thing (or were too underage to know at the time)

>>59446594
Maybe it's reading it as 16 cores and simply throws threads wherever it can.

>>59446675
Not yet. AMD is a big company there's probably a lot of people behind the scenes that they have to consult about the feasibility of it.

I don't expect them to allow it but they might ship chips without it altogether which is a good compromise I think.

woah, TPU also known as The ultimate Taiwan shilling site gave AMD decent review

>>59446594
Microsoft hasn't fixed their scheduler.

Same deal as with Bulldozer's early problems.

>>59446848
Its not all that great. The conclusion includes lack of integrated graphics as a negative. The whole write up has minor quips worked into it, and the power consumption test being metered at the wall is pretty shit.

>>59446808
they will open part of code that is theirs
they did it with older versions I think

>>59446865
scheduler is not main issue, HPET apparently gave the biggest performance gain when disabled meaning it's not that hard to fix

>>59446878
nobody reads those, all people will see is graphs where 1800x~=1700 is only 5% away from i7 and costs 20% less

>>59446914
>1800x=1600x(for gaming)=~1700

fixed, in this case 1600x is 70% cheaper
8c will stay niche, anyone sane will buy 1600x instead of i5

>>59446878
Integrated graphics might be a positive for some, I'd say the lack of it is more of a neutral point though. People that buy these chips don't give a fuck about integrated graphics so it doesn't even affect them.

>>59446848
Without 10k$ worth of testing hardware like Tom's has it's pretty difficult to test power consumption

>>59446929
>anyone sane
so not a lot of people

>>59446987
Proper high refresh osciliscopes are way more expensive than $10k. USD

>>59446987
honestly out of the wall is more useful way of testing, you want to know what system would work like with this CPU not just the CPU

>>59446197
You can set the 1800X to work in it's efficient frequency sweetspot, just like any R7

>>59446726
You don't need to be underage to forget X99 and first stepping Skylake problems

>>59447126
Or you can overclock a 1700 or keep it at stock(zen sweetpoint) and save $200

>>59446503
it seriously is. 8 cores at 100% load, only 40 watts more.

Ryzen isn't the shit wrecker we hoped it would be but jesus fuck it's efficient

>>59446569
why would a circuit draw any substantial amount of power if it's doing literally nothing at all

>>59447169
I'd say its a pretty outstanding.
Teething issues aside, its super efficient, and it can trade blows with Broadwell clock for clock, only being slightly behind Skylake/Kaby Lake.

People were expecting it to only offer Sandy Bridge performance on average.

>>59446476
Said boards also brick themselves. Classical ASUS launch, i love how they cooked my 5820k.

>>59447197
It's not? It's just wasteful silicon

>>59446409
>Ryzen seems to heavily favor high speed memory.
It's internal data fabric (interconnect between CCXs) works at half the effective clock of RAM.

>>59447202
It was only expected to have a 40% increase over faildozer but it managed 52% which is pretty outstanding when you think about it. Only time will tell how this will pan out. The release could have been better no doubt but at last we have some stiff competition in an otherwise stagnant tech market.

>>59447229
It's half the effective clock right? So it's the "true" clock speed of the memory then?

>>59446675
It uses ARM TrustZone so i highly doubt that.

>>59447209
ANUS has really turned to shit, everything up to Nehalem and ASUS is recommended, most shit after just got worse.

Holy shit I'm about a few more ASUS bugs away from denouncing them as ABIT-tier

>>59447229
Rumor has it that AMD will address this issue. Probably with better chips further down the line (early adopters always get screwed anyhow). I doubnt we'll see any microcode improvements to this but who knows.

>>59447248
Kinda, yes.

>>59447260
other rumor has it AMD will do huge microcode revision in May that will add better RAM support

>>59447209
My Prime B350 Plus is stable, everything working fine more or less, but holy shit overclocking on this board is terrible. Having to enter multiple different sub menus to effectively alter one thing. Setting memory clocks and timings on it was a chore, and it wasn't because of poor compatibility, it was just finding where the fuck everything was located. Their UEFI is really shit.

In retrospect I should have waited for something else to come back in stock.

>>59447248
Yeah. A 2400mhz DDR4 kit = 1200mhz data fabric.
The actual nuance of how DDR4 signals is so complicated its not worth going into though.

>>59447260
That isn't an issue, its just how the clock ratio works.
Using faster memory increases the speed of the data fabric. Its how everything maintains coherency.

>>59447299
did you sue random "not validated" RAM sticks?
I'm curious how ryzen treats any non validated RAM if you manually set up timings

It's pretty amazing what AMD is doing with such a high latency memory controller, if they can get DRAM calls down to 50-60ns in a year or two these things will be monsters allround

File: amd-ryzen-ddr4-scaling-latency.jpg (77KB, 849x531px) Image search: [Google]

77KB, 849x531px

>>59447389
I used a 2933mhz HyperX kit. The DIMMs are good, worked in another system, booted fine in my Ryzen build, but they defaulted to the default 2400mhz profile.
The ASUS auto overclocking tool flat out doesn't work at all. Selecting 2933 or even 2666 from the drop down autoOC menu causes the system to do an endless boo tloop. It apparently doesn't try to adjust the voltage of the memory at all, but tries to increase the clocks while leaving it at 1.2v.

Unless you go through a handful of different menus and set everything by hand it won't boot. I don't think I'd ever buy another ASUS board after this one.

>>59447427
3200mhz RAM with C16 timings yields a latency reading in the latest AIDA64 of 76ns.
Faster memory with tighter timings gets you down to mid 60ns range. Any improvement to the IMC in Zen2 will effectively make system memory so fast that they're already at the point of diminishing returns. Very high effective utilization, extrapolating over 95% of theoretical bandwidth, and low latency at the same time.

What was AMD thinking putting the ASUS in the review kit instead of the Taichi or Aorus? At least these two don't crash the system or lose performance.

>>59447563
Some dudes got Aorus boards.

>>59447563
ASUS probably forked over some cash to be the main partner, but then totally dropped the ball.

>>59447563
>What was AMD thinking putting the ASUS in the review kit instead of the Taichi or Aorus? At least these two don't crash the system or lose performance.

There's memory bugs on the Aorus as well, but that seems to be its only glaring problem at the moment.

Remember asus with x99? Took them 2 months to fix it

>>59446062
And this is why the ccx exist, bigger, monolithic core complexes like Intel has are harder to control when power is in question.

>>59446149
>20 watts
150 -120 != 20
>intel fanbois btfo by simple maths
Its no wonder you guys can't pick the better value cpu.

>>59447718
They also suck some gigantic dicks to scale. I wonder how fucking BIG 32coarz Skylake-EX will be.

>>59447767
Lol yields

File: GAME COMPARE.jpg (2MB, 1004x5792px) Image search: [Google]

2MB, 1004x5792px

>>59446062
Since the other shill thread died before I could make my post to comment on the matter here.
Here.
>Games comparison: 33 for intel and 7 for ryzen.
>Workload comparison:15 for intel and 8 for ryzen
>Over all: Ryzen costs 200 more and only surpasses the 7700k in "H.265, h264, Wprime, blender, cinebench, veracrypt, 7zip" so if you use those a lot and the performance really matters to you get ryzen 1800x
>for everything else it's both cheaper and better to get the 7700k.

>>59446062
The intel chip has more shit on the chip, and requires higher voltages

File: WORK LOADS COMPARE.jpg (1MB, 1004x4936px) Image search: [Google]

1MB, 1004x4936px

>>59447860

File: OVER ALL COMPARE.jpg (303KB, 1114x1062px) Image search: [Google]

303KB, 1114x1062px

>>59447860
>>59447881

>>59447869
Eh, no? 7700k is smaller.
>>59447860
It was nuked, Ramesh.

>>59447860
I'd say that's fair but the chip wasn't intended to compete with the i7 7700K. It's meant to be a lower cost alternative to the 8 core i7s but it also has a performance advantage on them in a few cases and lower power usage as well

>>59447869
More shit? The Ryzen is a fucking SoC, has a full southbridge integrated on top of that has GMI links for multi-socket support as well, which are useless to this market segment, there's a reason it has more transistors.

File: Jim Keller.png (400KB, 816x1020px) Image search: [Google]

400KB, 816x1020px

>>59446062
"That's not even my final form."

- C.S.W.

>>59447901
Naw dude, Ryzen is smaller, AMD made a big deal about it, it's about 10% smaller

https://www.techpowerup.com/230446/amds-ryzen-chips-10-smaller-than-comparable-intel-skylake-dies

>>59447925
Meant like less FPU and stuff

>>59446562
https://www.youtube.com/watch?v=ylvdSnEbL50

>>59447860
Jesus christ

Ryzen is fucking garbage

What the fuck was AMD thinking

>>59447940
Ryzen cores are smaller, the die itself is bigger than 7700k.

>>59446897
>>59446865
>>59446544
PCper already went over the scheduler issues, AMD has stated there is no issue

https://www.youtube.com/watch?v=6laL-_hiAK0

>>59447940
The CCX is smaller than 4 of intel's cores, that is not die size. Learn the difference.
The Zeppelin die is somewhere around 200mm2~ desktop Kaby Lake is 126mm2.

>>59447954
>what the fuck was AMD thinking
Selling server dies is totally okay, Ramesh. Now stop samefagging.

>>59447981
Zeppelin is 195mm2.

>>59447987
Are you trying to say Ryzen is a server CPU? You realize they marketed it to desktop users.

>>59447860
so what are you trying to say is that it is 5% difference?
wow! 500Mhz is 5% performance!

>can't go beyond 4.2 GHz
this is an awfully nice server CPU... but that's it!

>>59447860
>Over all: Ryzen costs 200 more and only surpasses the 7700k
>>59447954
No one is supposed to be buying the 1800X or even 1700X, 99% of the R7 1700s will clock to 3.9 or 4ghz, making the higher end chips worthless.

From there the R7 1700 literally wrecks the 7700K in anything that fully utilizes it's threads.

>>59448005
Yes, the same way Intel's HEDT line is shitty leaky Xeons. It's business, Ramesh.

>>59447974
>taking pcper seriously

come on

>>59448001
I haven't seen anyone officially measure it once delided, and AMD hasn't stated it anywhere in their press material that I'm aware of.
Sweclockers has it at 192mm2 in their launch review though.

>>59448027
And what exactly is wrong with PCper?

>>59448014
>500 cpu is 5% worse while costing 200 more.
>THIS IS FINE YOU SEE, because it's only 5%
Amd shill, please...

>>59448017
>From there the R7 1700 literally wrecks the 7700K in anything that fully utilizes it's threads.
Okay post proof. Because the pictures I posted was the "proof" an amd shill posted in the last thread.

So

Is Intel slashing prices? Since I hear Ryzen is a disappointment from some Anons that maybe Intel might change their minds and all.

>>59448050
>comparing quadcores to octacores
And that's why you'll never be rich, dumb Ramesh.

>>59446543
>I don't understand TDP: the post

>>59448039
>1060 will be 60% more powerful than 480

>>59448019
It's really bad business and destroying consumer confidence in AMD's brand.

>>59446329
Number of people who care about power draw during gaming: 0

>>59448069
>And that's why you'll never be rich, dumb Ramesh.
What exactly is your point?

>>59448050
CPUs:500-350=200
motherboards:200-150=100

according to you.

File: Premiere.png (55KB, 822x546px) Image search: [Google]

55KB, 822x546px

>>59448050
Are you trying to shill for intel or are you just stupid?

It has twice the cores/threads of course it's going to wreck it in anything that fully utilizes them

http://www.gamersnexus.net/hwreviews/2827-amd-r7-1700-review-amd-competes-with-its-1800x

>>59448072
Where did they say that?

>>59448016
The memory clock speed is also a huge factor in its performance though.

>>59446865
>Same deal as with Bulldozer's early problems.
>bulldozer is "broken" for 5+ years
>finally "scheduler patch" comes out even though the chip is literally EOL
>less than 1% difference in most cases
top kek

>>59448019
>poonigger subhuman AMDead retard calling anyone Ramesh
Is this a new meme or something? AMD is a failed Indian company with a poojeet CEO.

File: cores.jpg (97KB, 558x695px) Image search: [Google]

97KB, 558x695px

>>59448069

>>59448050
Look at you how hard you're fighting for Intel it's funny as fuck

>>59448077
They give zero (0) shits, consumer market is unstable and full of fuckretarded normies. They are clearly aiming for server&mobile markets.

>>59448080
It's actually a pretty big deal for ITX PCs, laptops, and Lan Centers

>>59448080
really? nvidia shills care very much about it

It looks like the spammers from the other thread found us. This was a nice thread for a while at least.

>>59448115
>Lan Centers
How's Brazil this time of year?

>>59448077
Nobody ever praised AMD for its marketing.

>>59448017
You realize the 1700 loses even worse to the 7700k

They are the same price and the 7700k is like 20% faster

Horrible deal to buy a 1700 desu senpai

File: 1452368119479.jpg (52KB, 419x480px) Image search: [Google]

52KB, 419x480px

>>59448095
>double the cores and threads
>only 30% faster in heavily multithreaded workload

>>59448095
>Are you trying to shill for intel or are you just stupid?
>how dare you disagree with the AMD narrative

>It has twice the cores/threads of course it's going to wreck it in anything that fully utilizes them
Yeah, yeah and if you look at the benchmarks that is happening in only a few select applications. Refer to the proof >>59447860
>>59447881
>>59447893
And please refrain form making bold claims with out actual proof next time.

>>59448105
Rajeesh, you're shittalking fellow Indians. How could you, filthy bastard?
>>59448122
Hey, you can always go to something like Anandtech forums. /g/ needs IDs and flags.

>>59448080
Yep Intelfags didn't care about it at all during the Bulldozer series

>>59448153
>/g/ needs IDs and flags.
I think flags alone will suffice.

>>59448156
>Yep Intelfags didn't care about it at all during the Bulldozer series
It was something to point and laugh about but nothing that really mattered in practice

>>59448153
>filthy bastard?
Poonigger detected.

>>59448165
>I think flags alone will suffice.
Why, to validate your racist delusions? Keep your racist propaganda on /pol/, Nazi!

>>59448109
I am not fighting for intel, I am just sick, SICK of amd shills shitting up /g/ and /v/ nonstop with their fucking lies.
Proof and benchmarks set what is true in the real world.
Your endless damage control and shilling doesn't Accept the fact, get fired. Go get hired in something more respectable like cleaning those shiting streets.

>>59448165
I need IDs to filter the fuck out of local shitposters.

>>59448178
Yes-yes, Ramesh, I'll give you (you)s, you need to feed your family.

Hehehehhe

Games

>>59448178
>A concern troll is a person who participates in a debate posing as an actual or potential ally who simply has some concerns they need answered before they will ally themselves with a cause.

>>59448144
Premiere's multi-threading actually works better in older verisons from what I've heard

>>59448143
So I should spend the same amount of money for half the cores, slightly better gaming performance, and a chip I have to delid if I want to really OC it?

>>59448146
The 7700K is only going to beat the R7 1700 in software that isn't 100% multi-threaded.

It also means the 7700K is going to beat the 6850K and 6900K in those workloads, sans a well overclocking chip.

>>59448177
At least fucking try to properly shitpost, it's getting fucking boring.

gonna be fun to see what you guys are saying 5 years down the line when Ryzen owners don't have to upgrade but 7700Kfags are left in the dust playing ancient games that use 4 thread, heh

>>59448144
That would be an argument against the $1100 i7-6900k
It's not an argument against the $330 R7 1700

>>59448146
>if you look at the benchmarks that is happening in only a few select applications
same could be said for intel

File: AMD SHILLS new narrative.jpg (86KB, 1142x397px) Image search: [Google]

86KB, 1142x397px

>>59448210
>trolling
Ah yes here we go again Pajet
> "everybody who disagrees with me, wants to see proof, or posts proof is a troll"
Let's see how long you can keep this shitty damage control up.

>>59448265
(you), Ramesh. Get a trip so I can filter you.

>>59448213
>The 7700K is only going to beat the R7 1700 in software that isn't 100% multi-threaded.
It's going to beat the 1700 harder than it beat the 1800x. You can look at the benchmarks and estimate for yourself.

>It also means the 7700K is going to beat the 6850K and 6900K in those workloads, sans a well overclocking chip.
I am not telling you to buy an overpriced 6900k, dumbass.

File: samefag not samefag.gif (283KB, 652x349px) Image search: [Google]

283KB, 652x349px

>>59448265
nice try shill

>>59448178
This.

I hate AMD shills so fucking much.

Mods need to ban AMD shills on sight.

>>59448281
>I am not telling you to buy an overpriced 6900k, dumbass.
but you're telling him to buy an overpriced jewby lake? lol

>>59448213
7700k kills the 1700 in basically everything, it's not even close. It even kills the 1800X.

>>59448287
>I hate goyim as much as I hate competition
Nice try mr. Shekelstein.

>>59448282
I mean it doesn't even have to be that difficult, he could just edit it in mspaint

>>59448303
>the 7700k that beats the 1800x in terms of price and performance is overpriced.
What does that make AMDs product then?
Trash and even more overpriced?

>>59448311
AMD is not competition you dumbfuck, if they were I would actually consider buying them.

AMD is like the retarded red-headed step child of the tech world. They need to be put down for their own good.

>>59448315
>mspaint
doesn't run on Pajeetnux 2005 edition

Honestly it feels like I'm talking to bot.

There is literally no reason to buy a 1800X or a 1700X unless you are into heavily threaded applications. For general purpose and gaming a 1700 should suffice plus you get the added advantage of extra performance in highly threaded applications. It's a toss up really. Gamer? Get a 7700K. Gamer + productivity and a little future proofing? Get the 1700. They are the same price (except you get a free $30 cooler included with the 1700).

>>59448287
Truth be told, they used to have excuses like "optimization" or "muh smoothness" and I have worked through all of them with benchmarks and proof.
Now they just spam "lol same fagg trolling"
I am not sure if its actual shills any more or kids that think shitposting is cool or something.

>>59448322
>What does that make AMDs product then?
>Trash and even more overpriced?

Yes sadly that is exactly what it is.

Ryzen is so fucking bad it's actually hilarious at this point.

>>59448331
>y-yes goyim AMD are not competition, buy Itanium
Guess how it ended.

>>59448306
>intel shills actually believe this
a 1700 has ZERO dropped frames in obs. would you like to know how many gtames the 6700k and the 7700k dropped?

File: cpu-test.png (26KB, 500x450px) Image search: [Google]

26KB, 500x450px

>>59448342
100% wrong

the 7700k absolutely annihilates the 1700 in the vast majority of applications.

It even beats the fucking 1800X

>>59448369
Go ahead, show me the benchmarks on that.

Boт этo тeaтp oднoгo ceмeнa.

>>59448370
It seems to be losing to it there. Is there something you're trying to tell me?

File: ryzen-streaming.png (129KB, 1440x1080px) Image search: [Google]

129KB, 1440x1080px

>>59448369
Holy fuck AMD shills are so delusional

The 7700k kills a 1800X in streaming, I don't even want to know how bad it is against a 1700

>>59448370
I only see 1880X on that graph. All indications are the 177 OC'd uses a little more power than the 1800X and gives similar performance.

>all these intel shills panicking AMD managed to close so much of the gap on their first try

>>59448399
>1800X, 1700. Fuck my typing.

>>59448416
t. AMD shill

Where did Intel find a guy this dedicated? He's been at it nonstop since launch.

>>59448416
The RX480 was like this at first. Now it matches and even beats the GTX1060 in a lot of games.

>>59448399
The 1700 is definitively slower than the 1800X of course.

When the 7700k beats the 1800X, it's a horrible sign for the 1700

>>59448388
>can't read
7700K OC is leading the pack at $300
enjoy you're $500 paperweight

It's ONE dude replying to everyone including himself. I want more of that. I also want flags and IDs so I can filter retards when the actual discussion starts. Do it gookmoot.

>>59448416
>AMD on damage control for weeks, shitposting and shutting down anything with lies until proven otherwise with facts and bencmarks
>ha-ha guys l-look at intel shills panic, s-stop posting proof and proving me wrong

>>59448390
holy fuck indeed, I think at this point even rare /g/ visitors can recognize (you).

>>59448423
You know being so anti-AMD only proves you are a shill, right? A regular person would be glad intel has some competition.

>>59448462
You know being so pro-AMD only proves you are a shill, right? A regular person would wish AMD actually gave Intel some competition.

>>59446594
How do you know which core is SMT and which isn't, you assume it's ordered by real cores first?

>>59448281
Um no the 1700 is the same chip as the 1800X, you should be overclocking your Ryzen chip to 4ghz or as close to it as you can get

>>59448475
Intel has already dropped prices. They are feeling the competition.

I wonder if they'll test the 32 core Naples in games, it's only fair, not like a $10000 CPU can lose to a $300 Intel

>>59448497
Intel hasn't dropped prices a fucking cent, go look at new egg and amazon

It's fucking pathetic, AMD needs to get their shit together

>>59448499
Test it in 480p. With NUMA disabled ofc.

>>59448487
>Um no the 1700 is the same chip as the 1800X, you should be overclocking your Ryzen chip to 4ghz or as close to it as you can get
It's the same chip that overclocks worse.
But as soon as you have benchmarks that prove your point in overclocked 1700 beating 7700k, I am all for seeing it.

>>59448483
Physical core followed by SMT thread.

Cpu1 - core
Cpu2 - smt
Cpu3 - core

Etcetc

>>59448512
http://www.microcenter.com/product/472529/Core_i7-7700K_Kaby_Lake_42_GHz_LGA_1151_Boxed_Processor

>>59448527
Nearly every 1700 seen review is hitting 3.9ghz or 4ghz

the 1800X can barely get past 4ghz at times

>>59448541
>microcenter
they don't count their CPUs are always cheaper

$349 MSRP as normal at newegg

https://www.newegg.com/Product/Product.aspx?Item=N82E16819117726

Fuck the thread started so nice.

>>59448541

microcenter has always sold i7s for $299 and i5s for $200...

people actually used to offer services because microcenter has deals with cpu + mobo combos.

>>59448542
Post proof. Benchmarks and graphs with a link to the sources.
With the level of AMD shilling and damage control I don't want to read a post and accept it as fact anymore.

>>59448541
That's a microcenter sale retard, Intel didn't drop their prices a cent:

https://www.newegg.com/Product/Product.aspx?Item=N82E16819117726

Intel would rather take a marketshare hit than lower margins, holy shit people they said so themselves in their investor meet.

>>59448584
Not him but SiliconLottery tests came up with 70% of 1700s can hit 3.9 with reasonable voltages.

File: intel-israel.jpg (145KB, 1500x1054px) Image search: [Google]

145KB, 1500x1054px

>>59448613

>>59448584
Did you even look up R7 1700 benchmarks before posting?

https://www.pcper.com/reviews/Processors/Overclocking-AMD-Ryzen-7-1700-Real-Winner

By simply enabling that option in the ASUS UEFI and rebooting, our Ryzen 1700 processor was running at 4.0 GHz on all cores! For this piece, I won’t be going into the drudge and debate on what settings ASUS changed to get to this setting or if the voltages are overly aggressive – the point is that it just works out of the box.

-
-
http://www.gamersnexus.net/hwreviews/2827-amd-r7-1700-review-amd-competes-with-its-1800x

Assuming you’re willing to put in the five minutes to OC and run a little higher temperature, the 1700 is able to complete the scene in the same time as the overclocked 1800X, as both land at around 3.9 to 4.0GHz in our production tests. We could sustain a 4.0GHz clock in gaming, but production workloads stressed stability and forced us to drop to 3.9GHz.
-
-
https://www.bit-tech.net/hardware/2017/03/08/amd-ryzen-7-1700-review/7

Push the cores up to 4.05GHz, though, and the tables are turned and graphs flip on their head. In the same game, the Ryzen 7 1700 is now top of the graph and a mid-table result in Cinebench R15's multi-threaded test leaps up to the top spot, bettering the overclocked Core i7-6900K and Ryzen 7 1800X.

10 R7 1700s OC'd
http://oc.jagatreview.com/2017/03/overclocking-binning-10-prosesor-amd-ryzen-7-1700/

>tfw invaded by /v/ermin and actual discussion will be impossible until next year

File: 1488400157075.jpg (114KB, 960x874px) Image search: [Google]

114KB, 960x874px

>>59448652
Kill yourself AMD shill

>>59448613
Some pretty deep financial calculation must have gotten into this decision.

>>59448702
It'll be forever impossible unless flags and IDs.

>>59448702
>workload comparrisons get posted>
>ignore those
>gaming performance is also posted
>OH NOOO NOT THE VERMIN! and it was such a good amd shill threat too...

>>59448390
>>59448390
Holy fuck intel shills are so delusional
theres not a single streaming site that goes over 60fps
so the important part is DROPPED FRAMES
the cpu has to SEND that info and 4 cores is just not enough

>>59448752
>benchmarks show ryzen is a very solid workstation cpu
>LOL RYZEN IS TERRIBLE LOOK AT THESE GAMING BENCHMARKS!
Fuck off /v/ermin.

File: OVER ALL COMPARE2.jpg (338KB, 1603x1032px) Image search: [Google]

338KB, 1603x1032px

>>59448791
>benchmarks show ryzen is a very solid workstation cpu
This
>>59447881
and pic related.

>m-m-m-m-muh vermin
Shill please go and stay go.

>>59446074
How does it draw less power when running at the last two frequencies?

File: Capture.png (24KB, 426x477px) Image search: [Google]

24KB, 426x477px

>>59448837
As mentioned in the other thread the memory speeds are not equal so the test isn't exactly fair to one side.

>>59448845
Probably throttling from tjmax, put it on a water loop or sand it.

>>59448859
The source used was posted by an amd shill in another thread as proof.
I didn't pick this site. I just went over the "proof" amd shills posted.
Don't claim Ryzen is superior to me, post "proof and then when I point out it still lost in a lot of applications you go
>nuhuhu it doesn't count b-b-b-because of something

Go ahead, post a benchmark that actually proves ryzen is superior. I will wait.

Seems someone's getting increasingly more nervous how scores are much different now than on day one.

>>59448915
And the guy tries his hardest to samefag.

>>59448837
Can't I just OC a 1700? I'll get longer life out of AM4 as well, plus ucode updates n stuff.

>>59448879
>throttling at 70C

Seems a bit conservative.

>>59448910
>even when gimped the ryzen is competitive with the i7
But go ahead and keep crying shill.

>>59448859
But surely if each platform is beanched using the highest speed RAM it can use, the test is equitable?

>>59448969
Rypoo is shit OC!

File: Skärmavbild 2017-03-17 kl. 19.53.44.png (85KB, 609x774px) Image search: [Google]

85KB, 609x774px

>>59448910
>In b4 "but only games matter!!1"

>>59449029
At least post kernel compilation chart.

>>59448998
That mobo supports up to 3200.

>>59448998
It's kinda questionable. I don't like it because it means there's two things being tested at once. If they tested each chip with the 2133 and 2666 memory then it would be fair. Apparently the explanation for the 3000 memory not being used in Ryzen is that the board may not support it but that still gives the i7 an advantage outside of raw performance.

File: Ryzen is FIXED, not really.jpg (152KB, 1082x802px) Image search: [Google]

152KB, 1082x802px

>>59448982
>even when gimped the ryzen is competitive with the i7
Actually thats the new and "fixed" benchmarks.
In the other thread I poseted old benchmarks and ryzen fans cried those were old and that the new Ryzen is all fixed and shit.
I asked for proof and got the exact site and benchmarks I posted.
It's funny to be honest that I post the NEW and "fixed" benchmarks and you still go
>even when gimped the ryzen is competitive with the i7
Shitty shitty, shills.

One dude is keeping this thread alive. That's some quality autism and insane dedication.

File: 7700k.jpg (442KB, 1920x1080px) Image search: [Google]

442KB, 1920x1080px

>>59446329

How can the 1700 compete with a 7700K on a custom loop?

>>59449029
>posts one, ONE benchmark
>see ryzen is PERFECTION
And AMD SHILLS blame me for cherry picking....
Reffer to this >>59447881

>>59447299
I stopped by microcenter they had 4 prime b350s on the clearence rack for $40. I was so damn tempted to pick one up for my upcoming r5 build.

But the fact they were on the clearence rack to begin with turned me away.

>>59448998
the architecture of kaby lake is not new. the market has been optimizing it for 4 years now!

>>59449104
>4
Try 9. It's all the same shit since Nehalem.

>>59449067
>Actually thats the new and "fixed" benchmarks
The benchmark clearly show the 1800x using slower RAM.

>>59446062
That's odd mine a 4.7ghz barely cracks into the high 80Ws at any point during use.

>>59449164
Don't blame me, I only used the "proof" that amd is fixed and better provided by AMD Shills.

If you want to give me a source and proof that ryzen isn't shit, I will be happy to take a look at it.

>>59448167
I saw LOTS of people recommending Intel over the Bulldozer on /v/ just for the energy consumption, even when the both allowed you to play at 60 fps and the amd chip was cheaper and had better multithread performance

>>59446675
I don't know, I saw an option in my BIOS to disable it, but with the way people act, it's probably doesn't do what I think you guys are talking about.

>>59449010
A 30%+ OC on a 1700 is fantastic, you practically get a 1800X for $200, this is even doable on some B boards and on stock cooler if you have some luck with voltage bins, this is beyond fantastic value.

>>59449237
I think this is mainly an issue on compatibility with coreboot/libreboot. I think the code that runs on the PSP is encrypted and you cannot make the system work without the PSP or the code. That option might just disable the PSP after the boot phase I guess.

>>59449220
Intelfags only care about power consumption when their chip is drawing less power than the closest and equivalent. When Intel draws more the argument is that it's a high performance part, are you poor or something etc.

I couldn't give less of a shit if the difference is 20W or 60W, it would never be a driving factor when considering a desktop cpu for me.

>>59449343
It does spell out a rather nice future for ITX boards, mobile and server chips.

https://www.newegg.com/Product/Product.aspx?Item=N82E16813132927
Z270 Hero 229

https://www.newegg.com/Product/Product.aspx?Item=N82E16813132963
X370 Hero 254

25 dollars

>>59448390
Now post the unbugged results

>>59449377
>ASUS

Lmao

Both board quality(components AND firmware) and supply has been abysmal this launch.

Now that vendors saw that people are actually interested in this the Pinnacle Ridge launch should be much tamer.

>>59449361
Obviously that's the only time i care about power draw. I'm pretty excited to see how Zen scales down and what the ULV shit brings to the table. Thinking about cheap 4c/8t laptops makes me moist.

>>59449404
>512 Vega cores + 4/8 CPU at 3.0GHz with 3.7 turbo

>>59449404
we know how Zen scales down, and the answer is really fucking well. Keep in mind Zen was built with the server market first and foremost in mind where perf/watt is king, and on that point AMD hit a home fucking run.

I would not be surprised if AMD's APUs and mobile chips end up beating intel's offerings. We just gotta hope the fucking OEMs stop sucking intel's dick and actually sell some decently kitted out AMD laptops.

File: 1455122603501.jpg (32KB, 375x344px) Image search: [Google]

32KB, 375x344px

>>59449435
AMD DO IT

blow the market apart

>>59449435
The 1 CCX mobile APUs won't have any reason to have high clocked memory support that's not DDR4L, no need for the GMI links, that big a southbridge isn't needed either, it's gonna be a lot smaller die on the uncore and CPU part so they will clock higher, on top of that the chips will be cherrypicked like server ones for maximum Fmax/Vmin efficiency, on top of that I heard it will not be fabbed at 14nm LPP but something more tuned, but this is rumor at this point.

>>59449461
Hahah you don't know shit stop pretending you do.

>Huudrrrr we know it's a server chip first and foremost


Stop pulling shit out of your ass you are not helping AMD mate. Feel free to infer and postulate but get out of here with that unsubstantiated

>we know

Shit

>>59449537
Oh "you heard" that shit did you. Cite your sources child and quit smearing your poo on the walls with your rumourmongering.

>>59449568
Every CPU's got multi-socket interconnects and scales down linearly on voltage and clocks to 2.2GHz, of course it's a server design.

Buying a 1800X or 1700X is dumb for desktop use unless you are a cuck. OC'd 1700 or 6700K/7700K is where it's at. Either are a good choice. You pays your money...

>>59449402
It's like you never seen any real abysmal launch.

>>59450078
I've been posting on this board since Barcelona launched, I'm well aware of shitty platform launches.

>>59446074
this is the coffin in which Ryzen will be buried

Ryzen needs those clock speeds to be at least 400mhz faster for every spot on that voltage:temperature axis.

>>59450123
>server die needs high clocks

Lol

>>59448702
Next year Zen+/zenver2 drops, your best bet at decent discission is digging through random forums

>>59447296
Just wait (r)

>>59450123
You know that beats the shit out of Intel, right?

>>59447209
despite MSI's usually disgusting gaymer look I buy them
luckily the aesthetic market is heavy into whites grays and simplicity now :^)

>>59450684
I don't buy MSI because of Killer junk.

>>59450123
>i want to burn data centers down
Dude what.

>>59447237
>expected to have a 40% increase over faildozer but it managed 52%
compared to something released 7 years ago
that's not impressive at all. really sad to see how stagnent the market has become.

you cucks actually believe intels 1-5% gains (which are just clock gains) each year to be the norm... tsk tsk too many millenials.

>>59451365
>Excavator
>7 years ago

You little kids really need to try harder when you shitpost. Its just sad.

File: fuck this gay earth.jpg (167KB, 600x414px) Image search: [Google]

167KB, 600x414px

is this why its been having less performance? power? why havent both sides shown this? fanboys and shills are retarded. good thing i am only here for shitposting otherwise i'd be pissed.

>>59451479
The CPU is underutilized in games, you can see by the overclocked gaming benches that score the same as stock.

Game-specific optimizations will fix this.

>>59451534
>Game-specific optimizations will fix this
that should be up to amd and motherboard manufacturers.

>>59451628
AMD and motherboard manufacturers don't optimize games.

>>59451628
They already said at least 300 gamedevs are working on Ryzen-specific optimizaition.

>>59451628
Mobo vendors and AMD have nothing to do with game engines.

>>59446609
OYYYYYY VEYYYYY GOYIM SHUDDIT DOWNNNNNN

When will we reach the point when CPUs are gonna integrate fixed function logic for DX11/12/Vulkan drawcalls?

>>59446594
The reason more cores are getting loaded at half speed is because Windows scheduler prioritized power saving, 1 high clocked core at full tilt uses more power than 3 cores at 50%

>>59446609
>>59451724
Yeah lets ignore alll the other benchmarks.
>>59447860
>>59447881
>>59447893

>>59448118
Only when it suits them

>>59448105
fuck off Ramesh

>>59448370
>benchmark from intel marketing outlet no. 56
POO IN THE LOO PAJEET

>>59451755
But in this case it impacts performance badly and Windows doesn't know it, it throws work at SMT thread while a physical core is unused, it traverses the data fabric to put other cores to work instead of filling up the first 4 cores and THEN moving to the other 4.

>>59448423
t. cuck

>>59446082
Nice meme. 7700k at 4.8ghz has identical performance to a 4-3.8ghz 1800x while using more power >>59448209

>>59447299
>My Prime B350 Plus is stable, everything working fine more or less, but holy shit overclocking on this board is terrible. Having to enter multiple different sub menus to effectively alter one thing. Setting memory clocks and timings on it was a chore, and it wasn't because of poor compatibility, it was just finding where the fuck everything was located. Their UEFI is really shit.In retrospect I should have waited for something else to come back in stock.
Did you notice the underclocked memory and the CPU seeing 1.5v even though the default is 1.3? Fuck anon it is not worth it. I'm not touching the settings on mine. My Ryzen can play music and idle until the replacement MSI board shows up.

>>59451757
See >>59448859

>>59451757
see >>59451794

File: 7770 doesnt support ram high than 2400.jpg (42KB, 740x195px) Image search: [Google]

42KB, 740x195px

>>59451820
>>59451831
KEK!
I like how in one ryzen shill thread people complain the 3000MHZ ram is giving 7700k an unfair advantage.
And then in another thread the very same shills are saying 7700k is shit because it can only run ram at 2400MHZ

>>59451814
It actually has better performance than the 1800x and it costs 200$ less.
But sure if you are building a server then the extra power cost might be something to consider.
BUT OH FUCKING WAIT, MOST PEOPLE AREN'T BUYING 7700k or 1800x FOR A SERVER.

>>59448005
amd made one cpu to do it all, and in that case, it effectively hits every market, it's not number 1 in desktop, but it killed intel's E line, while offering up enough performance to justify it over a 7700k on the 1700

>>59448077
did not lose confidence here.

>>59451881
Are you actually unironically braindead or you're merely pretending, Ramesh?

>>59451881
It's almost like people can be wrong.

>>59451881
>stock memory

Aka nobody but people buying lowend patriot/kingston stuff, all decent memory kits are overclocked OOTB and run fine on both platforms, though for AMD it depends on the motherboard until the microcode updates.

>>59451881
Nice post pajeet

>>59451928
fuck off, amd is poo in loo shit just buy intel please

>>59451905
Performance wise they're practically the same while the AMD chip's clock is significantly lower. However it's literally double the core/thread count which makes it future-proof for the near future, at least the next 5 years easily. You can bet games will utilize more cores/threads more efficiently and god knows what you'll be using your pc for by that time. That's a good investment long-term. Besides, the fucking 8 core consumes less power than intel's shitty i7 4 core pos. The whole POINT of stagnating their ipc was to make their chips more efficient and they couldn't even manage that after 3-4 years of producing incremental non-upgrades. AMD is the better choice, and I'm saying this whilst having running a 6700k.

>>59452057
you dont own a superior intel product, i know of this that you dont have intel because you would know how poo in the loo amd is, clearly retared, please buy intel

File: coolio.png (56KB, 1012x666px) Image search: [Google]

56KB, 1012x666px

>>59452109
not an argument. I regret it, but I bought it last february when I needed to upgrade. Couldn't wait sadly.

>>59452133
fake, please buy intel for real this time

>>59446685
The interesting part is that the AM4 boards don't really have a chipset, there's only some I/O functionality that's handled by the boards. What's traditionally known as north and south bridge is fully integrated into the chip. Less potential problems with mainboards, more potential to fuck up the main chip and having to do expensive respins.

So as long as the chip itself isn't broken many things should be fixable via BIOS and microcode updates.

File: 2.png (51KB, 525x318px) Image search: [Google]

51KB, 525x318px

>>59452155
Just noticed your name. Wellmeme'd. But seriously, intel fucking sucks compared to the AMD chips.

>>59447114
No. Its a CPU test, you want to know what the part in question sucks up, how much the VRMs have to deliver. No one cares about the rest in a scenario like this.

>>59447767
Well, their current server chips use an interconnect between two ringbus systems. AMDs infinity fabric really should be much easier to scale up. But the more important part is that it'll be integrated into future GPUs arches and I/O stuff. They'll be able to puzzle together their semi custom stuff really easy like that.

>>59452183
no it doesnt fuck you poojeet

>>59452326
>Well, their current server chips use an interconnect between two ringbus systems.
Which is pretty retarded idead because it imposes latency penalty due to switches. I guess Intel's new uarch will have as much modularity as possible because they are basically jury rigging interconnects for larger dies now.

>>59452376
LOL AMD IS SHIT, BTFO AMD ON SUICIDE WATCH

>>59452420
Intel is shit too. Long live VIA!

>>59452420
I used to buy only intel but the new Ryzen release has me switching sides, I only use the best of the best, Thanks AMD!

>>59452376
The ringbus approach offers advantages too I assume, it's just a nightmare to scale up. They had it developed and just stuck with it as core count grew. AMD started from scratch, having high core count and the importance of a highly scalable on and off die interconnect for their semi custom parts in mind.

It'll be really interesting to see how that developes in the next half decade or so.

>>59452462
>The ringbus approach offers advantages too I assume
Yes, of course it does, at lower core counts. It's a fucking nightmare to scale up though. I wonder how big and fuckretarded 32c Skylake-EX die will be.

>>59452426
>>59452439
AMD shills, buy intel please

>>59452512
But i shill only VIA, and i do it for free. Long live VIA!

>>59452512
I would if they were still #1, they arent.

Plz make the next gen more then a 1% IPC improvement and ill come back.

>>59448337
The thread started off good, with actual discussion, and then at this point: >>59447860 the intel shill(s) arrived and it's just been corporate sanctioned shitposting since then.

>>59452728
>guy posts benchmarks
>this triggers the AMD shills
>shitposting goes to maximum levels
>somehow it the fault of the guy, not he shills
wew lad

>>59447860
>Civ 6 fps at 1080 is nearly equal to that at 1440p
How?

>>59446569
Lack of integrated graphics is a big minus. I use an old athlon ii as a external HDD type server. Had to buy a new graphics card for the two times a year I need to screen output for bios or something.

I don't care if it can just output text or 2D 1024x768 as long as there is some kind of IGP.

>>59448064
My opinion is that it's AMD who needs to slash prices or deliver something with Ryzen R5 in April.

I don't really care about i7700k vs Ryzen 1800X. $120-200 is enough for a CPU.

>>59448095
Yes, Ryzen beats everything Intel on the few workloads that max all cores. Sadly it bumbs out in everything that doesn't.

>>59448100
Too bad you can't run Ryzen om high memory speeds.

>>59448080
>>59448112
>>59448115
Power draw is a huge deal. Bills aside, you can slam a massive 140W TDP rated heatsink on a 65W TDP CPU and have the fan set to kick in at 48C and have a fanless system unless it's under full load for 20 minutes. Yes, this is my setup and the fan only moves if I encode a kpop concert to x265 so it takes less space on my phone or something like that.

>>59448370
>2% difference
>literally killed it