Ryzen review thread

File: 01165650911l.jpg (199KB, 752x1002px) Image search: [Google]

199KB, 752x1002px

oc3d:
https://www.overclock3d.net/reviews/cpu_mainboard/amd_ryzen_7_1800x_cpu_review/19

File: passmark.png (26KB, 609x323px) Image search: [Google]

26KB, 609x323px

>ryzen
>barely beating the 7700k in MT
>b-but muh bang for the buck
>turns out it also has worse cost benefit


HAHAHAHAHAHHAHAHAHAHHAHAHHAHAHAHAHAHHAHAHA

This is just too good. I don't think my sides can take it anymore.
My ribcage is literally in danger of bursting of laughter.

So, has anyone deactivated half of the cores, then go for OC and then do gaming benchmarks?

I am interested in that.
If it is the same as with all cores activated, then it's actually quite good.

It should further lower the powerconsumption.

Or did I miss something and the CPU automatically puts Core into Idle mode or even deactivating them?

File: Ashes-Ryzen.png (20KB, 637x521px) Image search: [Google]

20KB, 637x521px

>losing in Ashes of the Memegularity, the only game designed to work with 8 cores
>to the fucking 7700k
>to the fucking 7700k underclocked to 3.8ghz

Where were you when AMD is kill

keep in mind that many benchmarks are against unclocked intel K cpus which every sane person overclocks 500-1000mhz for 10-20% more performance AT LEAST

>>59200851
>comparing a mid-range chip to AMD's highest end chip in price/performance
not deceptively misleading at all

>>59200777
DELETE THIS

File: 1488470011746.png (140KB, 600x450px) Image search: [Google]

140KB, 600x450px

>>59200877

File: flat_800x800_075_f.jpg (18KB, 248x189px) Image search: [Google]

18KB, 248x189px

Gonna short AMD pretty soon.

Ryzen DoA. AMD is fucking finished now.

>>59200915
Absolutely hilarious. Ryzen is literally bulldozer 2.0.

>>59200873
from what ive read i dont think anyone has disabled cores in ryzen master yet, just gone for bios/mobo OC.

File: image.jpg (163KB, 640x480px) Image search: [Google]

163KB, 640x480px

Where's the 8 machops?

Fuck Ryzen and Vega, 7700k and 1080 ti here I come

>>59200988
why on earth would you need such expensive hardware?

>>59200777

HAHAHA AMD FAG BTFO

All that money and waiting on your pre-order when you could of bought a 9590.

>>59200873
Turbo downclocks them I believe.

File: 1488369363003.jpg (312KB, 1208x869px) Image search: [Google]

312KB, 1208x869px

AMD customers, don't open this.
It's for your own mental safety.

Reading those benchmarks and I just think that whether I should buy an i7 7700k instead. Like, even an Ryzen 1700 costs USD355.54 while one 7700k costs USD328.23 in my hometown.
And even in 7zip compression/decompression and also Chrome HTML5 tests 7700k does better than an Ryzen 1800X, what the hell?
Don't know if the so-called SMT thing in Ryzen is problematic or what, but as I mainly game and onto Internet on PC maybe 7700k is a better choice. I quite regret spending a month waiting to see how Ryzen does, to be honest.

>>59201020
personally I would wait for coffeelake, intel is making 6-core desktop CPUs at regular consumer price points then

or wait a long time for 10nm desktop chips

File: bench issues.png (147KB, 984x792px) Image search: [Google]

147KB, 984x792px

>>59200915
>>59200946
>>59200877

>>59200996
Why on earth do you think everyone is a thrifty poorfag like you?

>>59201039
You mean to tell me that next year's product will be better than this year's?

Oh God, this is looking bad.

>>59200892
>comparing things that cost the same is unreasonable REEEEEEE

>>59200935
just shorted 100 shares

>>59201018
delete this

Ryzen IPC > Intel IPC

https://www.youtube.com/watch?v=V5RP1CPpFVE

>>59201101
R7 1800x costs ~$500
i7 7700K costs ~$350

You're retarded

wait im confused, is ryzen shit or good?

>>59200777
For any french fags out there:
http://www.hardware.fr/articles/956-22/retour-sous-systeme-memoire.html
Explanation of the shitty game benchmarks for ryzen.
TL;DR:
L3 cache behave in a new way and windows is not YET able to correctly handle that kind of cache

>>59201087
>trusting a retarded youtuber over many other legit sites that produced similar results

OK

>>59201144
The 7700K costs about $300 in my area and blows the 1800X the fuck out. What is your point?

>yeah well this is more expensive so it's okay that it's worse, stupid!
jesus christ you are dumb

>>59201094
it good in everything besides gaming though

the heck is wrong with SMT? disabling it apparently increases FPS drastically
and not everyone could OC their memory, guru3d says 3200 works fine, others can't go beyond 2666

>>59201018
>hey guys look at my overclock XD
>whats prime95 XDDDDDDDDDD

>>59201152
blunder of the century

>>59201161
i dont trust french people

>>59201089
>Why on earth do you think everyone is a thrifty poorfag like you?
I'm not poor, I just don't see the point in spending thousands on a computer when something less than a thousand will suffice

oh wait you're one of those gamers aren't you
grow up

>>59201174
Way completely miss my original point.

>>59201152
it good for work, it's same as 4790K for gaming

>>59201176
>Value 1.296V
>Min 0.408V

Means it's on offset-voltage, meaning that it was under load at the time that photo was taken.

>>59201152
It's as good as Intel's 8 core chips basically. Shills are shitting on it by comparing it to 4 core chips.

>>59201198
>Way completely miss my original point.
Then please articulate it instead of just throwing around petty insults and being a jackass in general.

Bottom line:
price
performance in various applications

In this case, the 7700k beats the 1800X for almost half the price. Deal with it.

The problem is clearly the memory controller.

It's the bottleneck.

>>59201184
But you trust broken benchmarks?

File: 8072_47_amd-ryzen-7-1800x-cpu-review.png (59KB, 567x1117px) Image search: [Google]

59KB, 567x1117px

>>59201152
Depends on your perspective.

>https://www.reddit.com/r/Amd/comments/5x3y81/amd_suggested_that_we_test_the_cpus_with_a_gpu/
LOL
>inb4 reddit

File: 1700 v 7700.png (147KB, 1024x519px) Image search: [Google]

147KB, 1024x519px

3.9ghz Ryzn 1700 vs 5ghz Kaby Lake 7700k.
Ryzen does admirably in well threaded titles.

Nothing but launch teething issues going on right now, so shills are desperately shitposting.


>>59201161
A victim cache isn't a new thing.
Whats new is that DRAM is the LLC in Zen. It greatly benefits from high bandwidth and tight timings.
The system memory latency was addressed in a BIOS update last night, for ASUS at least, but they messed something up and auto voltage levels increased. Everyone will be adding an addendum, or a follow up review to their original articles in a few days or following weeks. Board vendors still have to sort out BIOS.

>>59201228
Compare the i7 7700K and the R7 1700 dumbass. They're the same price. That's a more fair comparison, though they still have completely different intended workloads.

>>59200851
Literally the only benchmark showing such a result, and probably due to Intel doing some shady shit with the program.

File: 1472680126877.jpg (71KB, 800x534px) Image search: [Google]

71KB, 800x534px

>>59201161
again? and apparently if you put it in high power mode it runs better, also better fps without SMT
everyone reporting different stats all over

the heck is going on?
I do not want to buy 7700K, i'd rather wait when my 2600 dies.

>>59201249

seems like everything else on that thing is better than the ryzen including the cheaper 7700K

>>59201174
6900k also loses to 7700k while costing twice as much as r7

File: Untitled-1.jpg (62KB, 1024x1024px) Image search: [Google]

62KB, 1024x1024px

>>59201228

but it isn't competing with the i7700k. It's an 8 core prosumer chip. It isn't a gaming chip.

>>59201309
And?

>>59201238
doesn't look that way, bandwidth is 50gb/s, latency? it's not that much bigger to cause that worse performance in games

File: Screen Shot 2017-03-02 at 12.12.05 PM.png (65KB, 816x144px) Image search: [Google]

65KB, 816x144px

>>59201277
>Compare the i7 7700K and the R7 1700 dumbass. They're the same price

>>59201277
>Compare the i7 7700K and the R7 1700 dumbass. They're the same price
>the 1800X is clocked higher but loses to the 7700K
>so you should look at the 1700, which is even slower
really activated my almonds

>>59201039
>Coffeelake
>6-core desktop Intel CPUs at regular consumer price points
Nah, Intel is known for that enormous Shlomoric spirit deep inside its soul. And I doubt whether a so-called "15% perf improvement" over Kaby Lake can be achieved, with referring to many Skylake vs. Kaby Lake comparisons.
Maybe Skylake-X could bring us some surprise, but with this time's AMD's colossal failure... I shouldn't have high expectations of Intel's HEDT platform becoming cheaper by even just one cent and go straight to Z270 instead.
>10nm desktop chips
I won't wait for another long period. My Phenom II x6 can't quite keep up and I need an up-to-date chip to let me game harder on my PC.

>>59201175
>disabling SMT apparently increases FPS drastically
Sauce, please? Because in some game benchmarks I've read (I forget the name of the reviewer) and the increase is quite close to nil.

>>59201200
>same as 4790K for gaming
>what AMDrones actually believe
Most game benchmarks show it's doing worse than the 4790k, which is a fucking 3 year old i7. Ryzen is literally fucking nothing.

>>59201338
https://www.amazon.com/gp/offer-listing/B06WP5YCX6/ref=dp_olp_new_mbc?ie=UTF8&condition=new

SAY IT WITH ME /G/:

B L U N D E R
O F
T H E
D E C A D E

>>59201349
gamers nexus, min fps jumped 30-35%

hardware unboxed says that gta5 sometimes stutters with i7, but doesn't with 1800x

reviews are all over the palce

>>59201347
The chart is about Price/Performance. The higher up you go the more money you have to pay to increase your performance numbers. This is not hard.

Are people on /g/ really so retarded to compare R7 to Intel's 4 core CPU instead of their 6 and 8 core X99 processors? I expected /v/ to be retarded but I didn't expect it from here.

File: 1488150996590.jpg (79KB, 1280x720px) Image search: [Google]

79KB, 1280x720px

>>59201373
B L U N D E R
O F
T H E
D E C A D E

File: 1488470541190.jpg (2MB, 2000x2000px) Image search: [Google]

2MB, 2000x2000px

AHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHA

>>59201195
Look: if you spend an extra $800 on a machine you use for four years, it comes out to around 50¢ per day. That's less than what it would cost to feed an African child and as a bonus you get a bunch of dead nigger kids

>>59201372
see >>59201347
>>59201378
And?

do you have a point to make or not

File: 2.jpg (69KB, 440x531px) Image search: [Google]

69KB, 440x531px

>>59201380

>Implying it's not shills Intel has sent fearing for their now proven retarded profit margins.

Gotta keep that Jewish milk flowing

>>59201380
it's about price ball park now, it's retarded yes, but the dream was to have 6700k gaming perf. and 6900k work perf. for $400.
we almost got it.

File: C0Ej4xMW8AAk0w9.jpg large.jpg (15KB, 336x368px) Image search: [Google]

15KB, 336x368px

>>59201419

give it a few BIOS/Windows patches and I think it'll get there. Most are reporting weird results relating to the new uArch.

>>59201419
this remind me that 1700 not x has exactly the same gaming performance as 1800x
and 6800 work performance

but I've read only 5 reviews might be wrong

>>59201446
reputation is soiled though, gaming crowd is stupid they can't adapt to new reality in few weeks
it's forever launch day for them as demonstrated by 480

>>59201207

So you've run prime95 for at least 12hrs to prove the overclock is stable?

File: bTYwsyCv4cVLAWPSxtLpUL-650-80.png (17KB, 650x365px) Image search: [Google]

17KB, 650x365px

>>59201299
Yes the worst case people as posting. Other times you will be gpu bottnecked. 1800x and 1700x are not good gaymen CPUs.

>>59201380
>Are people on /g/ really so retarded to compare R7 to Intel's 4 core CPU instead of their 6 and 8 core X99 processors? I expected /v/ to be retarded but I didn't expect it from here.

>If you buy a $300 7700K that beats a $500 1800X, you're retarded
If you kill your enemies, they win

>>59200777
>>59200840
>>59200851
So, it's just another bulldozer then? Fucking really?

>>59201511
>$300 7700K that beats a $500 1800X, you're retarded
7700K is 100% behind in work loads

File: CsoG7HVWgAAohLo.jpg large.jpg (13KB, 480x480px) Image search: [Google]

13KB, 480x480px

>>59201490

but it isn't a gaming chip.

It's a prosumer chip. Who gives a fuck what gamers think?

If all you do is play video games get an i7 7700K.

If you do stuff like video production, encoding, compiling, or whatever other workstation tasks there are then R7s are a great deal. They can run games they just don't perform ideally.

>>59201539
welcome to new reality when gaymers matter

>>59201531
less than that actually, not disagreing with your point, just your numbers

>>59201528
Buldozer couldn't keep up against a locked i5. Those low clocked ryzen 7 does pretty well against unlocked i7s.

http://hwbot.org/submission/3473875_der8auer_cpu_frequency_ryzen_7_1800x_5802.93_mhz

5.8ghz on LN2.

>>59201549
>If all you do is play video games get an i7 7700K.
may as well stay on 2500k or 2600 then

>>59201377
So I read throught Gamers Nexus' Ryzen reviews a bit. And yeah, 1%/0.1% FPS got improved when SMT disabled. But still i7 7700k does better than it by an obvious perf margin.
Maybe we can be a bit optimistic/objective to say that SMT in Ryzen is quite buggy, at least in real games.

>>59201549
>If you do stuff like video production, encoding, compiling, or whatever other workstation tasks there are then R7s are a great deal. They can run games they just don't perform ideally.
so it's the AMD FX-8350 all over again?

>>59201549
but you dont seem to grasp the severity
the 1800x gets obliterated by garbage tier i3s

https://www.purepc.pl/procesory/premiera_i_test_procesora_amd_ryzen_r7_1800x_dobra_zmiana?page=0,15


the 7700k will win in real world multitasking and gaming, these benchmarks even compare the 7770k unclocked which does 5ghz+ on air easily

I love how /g/ goes from 95% gamers to 50% work station users whenever AMD releases a new chip

>>59201276
>A victim cache isn't a new thing.
That's not the point.
A splitted L3 is a "new" thing or atleast not the traditional way of doing it.
Each core can access the differents L3 caches which result in higher latencies.
It's a kinda hybrid L3 cache.
There is NO software atm that handle that kind of thing correctly.

>>59201290
>high power mode it runs better
An higher frenquency is generally better
>also better fps without SMT
more L3 cache for each core

>the heck is going on?
AMD fucked up hard on the driver part as usual.

>>59201560

I'm a developer, after reading benchies I ordered my 1800x. I don't wanna spend 1300 on CPU + mobo for 8c/16t but I spent 650 on the combo. Was a no brainer.

>>59201614
Not really. The 8350 has garbage single thread performance.
the Ryzen chips merely have 'good' single thread performance

>>59201175
>the heck is wrong with SMT? disabling it apparently increases FPS drastically
could it be a windows thing? for example game using two threads for bulk of the processing and windows pushes those two to thread1 and thread2 without taking account that those two are on the same physical core.

without setting the windows10 "high perforormance" mode it fucks up the boosting at least

File: received_10210586577914398.png (17KB, 1671x83px) Image search: [Google]

17KB, 1671x83px

I really wanted it to work for AMD. To be honest it has worked. They exceeded their 40% IPC goal and once again have relevant cpus. However the price now is an issue. That will become better though if we have $130 r3 4c chips that can compete with $250-$350 Intel chips at gaming.

For now though outside of primarily work station use they aren't a great value. Think I'm going to grab a 7700k to retire my fx8350

>>59201670
Bulldozer had scheduler problem initially, I'm thinking this is probably the same issue. Difference is with Bulldozer the extra cores were actual cores so there wasn't a huge performance difference before and after. With Zen the SMT threads are simulated, there's no extra hardware core to pick up the slack so there's a performance hit instead.

>>59201172
The youtuber actually had ingame footage of FPS and core usage. And his results are similar to some others.

>>59201703

>not grabbing an R5 1600x which will clock a bit higher than the 1800x but have 2 less cores

any fallout 4 benchmarks?

>>59201758
https://www.purepc.pl/procesory/premiera_i_test_procesora_amd_ryzen_r7_1800x_dobra_zmiana?page=0,15

its not looking good atm

>>59201491
Prime is largely irrelevant for Intel CPUs post-Haswell. You can pass your vaunted 12h p95 test and crash in a game in less than 20m.

>>59201419
I think AMD may have actually made a mistake pricing it that low. X99 chips are exorbitantly expensive, putting them out of the range of retarded gaymers and the like. No one really cares about X99 gayming benchmarks. But R7 is cheap enough to attract people who aren't professionals or prosumers, so when it can't compete with th 7700K which is the gayming king because of high clockspeed/IPC, people go on a shitposting spree.

>>59201648
It's almost like the kids that invade to /v/ fuck off back to /v/ when it turns out that a CPU has 5 fewer FPS than the competition

>>59201755
Not available til summer and I promised to sell my current rig to a guy for $750 March 10th and can't pass that up

File: 1403979994916.gif (981KB, 342x239px) Image search: [Google]

981KB, 342x239px

>>59200777
Marketing statement from AMD regarding poor gaming benchmarks in some scenarios:

>“As we presented at Ryzen Tech Day, we are supporting 300+ developer kits with game development studios to optimize current and future game releases for the all-new Ryzen CPU. We are on track for 1000+ developer systems in 2017. For example, Bethesda at GDC yesterday announced its strategic relationship with AMD to optimize for Ryzen CPUs, primarily through Vulkan low-level API optimizations, for a new generation of games, DLC and VR experiences.

>Oxide Games also provided a public statement today on the significant performance uplift observed when optimizing for the 8-core, 16-thread Ryzen 7 CPU design – optimizations not yet reflected in Ashes of the Singularity benchmarking. Creative Assembly, developers of the Total War series, made a similar statement today related to upcoming Ryzen optimizations.

>CPU benchmarking deficits to the competition in certain games at 1080p resolution can be attributed to the development and optimization of the game uniquely to Intel platforms – until now. Even without optimizations in place, Ryzen delivers high, smooth frame rates on all “CPU-bound” games, as well as overall smooth frame rates and great experiences in GPU-bound gaming and VR. With developers taking advantage of Ryzen architecture and the extra cores and threads, we expect benchmarks to only get better, and enable Ryzen excel at next generation gaming experiences as well.

>Game performance will be optimized for Ryzen and continue to improve from at-launch frame rate scores.” John Taylor, AMD

>>59201835

>summer

It's "the coming weeks" anon, not summer. Begin Q2.

File: compile.png (61KB, 645x690px) Image search: [Google]

61KB, 645x690px

>>59201648
Install /g/entoo.

>>59201794
>Prime is largely irrelevant for Intel CPUs post-Haswell. You can pass your vaunted 12h p95 test and crash in a game in less than 20m.

Prime95 was irrelevant in 2006 already. IntelBurnTest found you problems in 5 minutes that Prime95 could not find in 12 hours.

>>59201648
I work in AutoCAD, and I do video editing on the side for a youtube channel. A prosumer CPU is exactly what I need, with an SSG to boot.
If gamers actually cared about performance that much they'd be playing games instead of shitposting on 4chan.

>>59201410
You really are /v/ meets /pol/ aren't you.

>>59201783
hmm thanks

>>59201842
DAMAGE CONTROL

>>59200777
AMD delivered, performance are fucking good.

File: 1396272573542.jpg (317KB, 1067x1264px) Image search: [Google]

317KB, 1067x1264px

>Disabling SMT improves performance
???

>>59201877
IBT (i.e. Linpack) isn't relevant either, I'd go as far as to say that nothing involving heavy AVX torture isn't relevant to actual day to day stability. It's not on my Haslel at least, my OC repeatedly passed 2h in IBT and would crash in less than 10m in BF4 every single time. It's fine if you want to catch grossly unstable settings, but isn't very good for relatively "subtle" issues.

>>59201324
If you have at least two brain cells, you could probably make out the point he's making

File: kern compl.png (39KB, 600x750px) Image search: [Google]

39KB, 600x750px

>>59201864
Kernel compilation times are really good too.

>>59201842
That's actually pretty cool if Bethesda shit becomes optimized for moar corez, FO4 seems to run like shit no matter what hardware you have.

>>59201755
>R5 1600x which will clock a bit higher than the 1800x
source?

>>59201967
>isn't relevant
Brain fart, I mean
>nothing [...] is relevant

>>59201955
Yeah there's definitely some kind of fix needed here. I wonder if AMD will be able to make Microsoft to issue a fix for Windows 7 though. If they can't get that then Ryzen will be crippled on one of the most popular OSes

>>59201955
I think there's a bug or something. Remeber reading the notes on the Linux kernal how 4.10 had to be optimized for SMT on Ryzen CPU's because it wasn't working properly. Imaging the next W10 update will target that as well.

>>59201967
If you want to see if an OC is stable play two 4K youtube videos at once, run a virus scan, and start a skype video call while running Dolphin.
It kills everything if the system isn't stable.

>>59201996
Isn't W7 explicitly unsupported? Good luck.

>>59201976
You're right, but I want him to say it explicitly so I can greentext it and mock him with images

>>59201612
It's literally on the OS now for it to schedule the correct tasks to the correct threads.

as a user of gentoo linux all i care about is compile time.

most workloads these shitty benchmarks test against are likely better suited for GPU accelerated shit anyhow, be it gaymen or other high floating point heavy math.

the globalists jews at intel don't want you to do productive things with their chips

http://www.infowars.com/watch-alex-jones-show/

Looks like yet another shit fest from AMD:

Fucking every time.

>>59202033
www.phoronix.com/scan.php?page=article&item=ryzen-1800x-linux&num=1

Ryzen shows only a couple significant performance deficits in the Phoronix test suite. Compiling times are very good

File: 01164548136l.jpg (69KB, 402x401px) Image search: [Google]

69KB, 402x401px

Overclocking results look really meh, most reviewers could do 3.9 and 4.0. Don't waste your money on 1800x and 1700x. Get 1700 and do whatever clock it can get if you are into linux.

Do you fags literally not do anything but gaymen shit

Isn't this tech board

>>59200777
I don't get how the 1800X seems to match the 6900K in productivity tasks and even in single threaded synthetics, but somehow manages to fall so far behind in gaming? Both single and multi threaded performance seems high. Maybe buggy PCI-E implementation?

>>59202061
maybe if i can fit it in my budget I might get one and kill this i7-3930k retarded waste of money then.

>>59201648
What? This thread is nothing but MUH GAYMS shit.

>>59201622
I do like how it's beaten by the i3 in fallout 4 and then suddenly jumps to the top in TW:WH

>>59201622
>purepc.pl
Yeah, no.

>>59202075
I clop to furry porn

>>59201858
Their "Q2 release" is expected in June. 3 months

>>59201842
>intel optimized code
>runs bad on amd
really makes ya think

>MUH GAYMES

Is this board solely frequented by degenerate /v/ermin shit?

Can we get some deep learning/neural network coding numbas

File: awd414444.png (31KB, 693x413px) Image search: [Google]

31KB, 693x413px

So performance is great except gaming being 10%-15%~ off the pace. I'll take that. At least we know AMD will continue to work on performance for the next 5 years as opposed to Nvidia and Intel who abandon after a year so they can sell new shit with a 1% performance gain for 500 shmekels.

I have a 4690k and won't be upgrading till late 2018 at the earliest btw, in case anyone accuses me of being a shill.

So it's shit for gaming, fuark.

Guess I'll jump onto some 2nd hand 6700k

>>59202076
Games are an entirely different beast, unpredictable, and inconsistent in their demands depending on what happens in a given scene.
With games you have a lot of sporadic IO access, huge chunks of data stream in all at once, pop in can cause pronounced dips in min frames, and huge variability in frame times. The way caches are utilized is entirely different. Even the best threaded games aren't as parallel as rendering workloads. If a game can utilize 8 cores, a few of them might only be loaded 10-20% where as rendering a video will load everything pretty evenly.

>>59201990
By virtue of having less cores to deal with

>>59202076
1800x probably doesn't have the single core IPC of intel.

probably doesn't help that intel likely spends more time enhancing floating point math on their CPUs which likely helps with "cpu bottleneck" gaymen benchmarks.

Overall I don't think it's an issue unless you like running at low detail and low res. Otherwise I imagine the GPU more than readily does virtually all of the work that isn't logic based.

>>59202136
It performs about as well as you can expect an 8 core to perform in gaming. The R3s and R5s will probably be much better at gaming in general whenever they come out.

>>59201290
2 PCs here, 2500K and 6600K; no perceived difference. Literally stick with it until it dies.

>>59202076
>>>
Games are shit pandering to whatever brand gibsthemdat

>>59201373
It's not though

I'm definitely getting one. Dem cores and threads man, at half the price

>>59202149
>By virtue of having less cores to deal with
>source: my ass

>>59202029
>>59202061
Well, if Ryzen does good on Linux gaming then I will consider buying an Ryzen 1700/1700X. If SMT can be done right on Linux that's a good news to me.

File: laugh spoky.gif (2MB, 360x270px) Image search: [Google]

2MB, 360x270px

>>59202200
>Ryzen
>gaming
>Linux
>gaming

>>59202170
can't you turn the cores off and clock higher? didn't any reviewers try this?

>>59202170
>It performs about as well as you can expect an 8 core to perform in gaming.
It performs worse than the 6900k in gaming, which is an 8 core too.

there are a few problems regarding ryzen

the smt and l3 will be probably fixed on the creators update(fast ring now)

on linux they dont have a kernel for their smt (4.10 will be the one)
and the new sound chip they are using on the mobos dont have drivers for linux yet

I just ordered a 1700. Fite me.

>>59202170
Even an i7 6900k wins BOTH i7 7700k and Ryzen 1800X. Your statement doesn't make any sense.

>>59202095
me too (except I don't "clop")

>>59202068
Yeah, OC really is pretty disappointing. I really expected more out of the 1800X considering it can do 4.1GHz boost at stock and 95W TDP. Not even being able to get all cores up to its maximum XFR boost with water cooling is pretty shitty.

Really what is the point of this single core turbo meme when the chips can handle all cores being at the turbo speed or higher 24/7? Makes no sense to me.

>>59202218
It should be possible to disable cores but I haven't seen benchmarks of it yet.

>>59202223
That's fair but I mean it is still way cheaper and it's usually very similar if not on par with it.

>>59202271
It's cheaper than a 6900k yes, but more expensive than a 7700k which beats it by a lot in gaming

>>59202264
I haven't seen reviewers trying 6/4-core only overclocks though. If anybody saw some post link

>>59202276
by a lot? the worst i saw was a -12% and that considering all the problems ryzen has...

the cpu is what amd said it will be nothing less nothing more

Anyone know if the ASUS X370 Pro has AM3 cooler support?

File: 1.jpg (185KB, 1001x595px) Image search: [Google]

185KB, 1001x595px

>>59202337
>the worst I saw was -12%
Well you didn't look very well then

why the fuck would you buy 8 core processor and THEN disable its cores wtf

>>59202303

There are too many things that all the reviews that I've seen didn't bother trying. They really should have planned this shit better. OC3D's video review was a fucking hour long and covered basically nothing.. they just read a few numbers from their benchmarks

>>59202033
Same and I already haave an 1700x on the mail. I'll likely install some bleh distro first to get some baseline feel on how the hardware handles and then go full on gentoo mode on it

>>59202370
It's probably because these benchmarks are rushed as fuck. The last minute MB updates didn't help in the situation.

>>59202362
>Overcrap.
>Unironically using this as an argument for why 1800X is crap.

Jesus wept this board is just /v/ isn't it

I bet the chink motherboard companies are taking bribes to sabotage them somehow

>>59202363
Trying to gauge the performance of an R3 or R5 quad core. Having 8 cores limits the amount of room in the TDP so overclocking and turbo doesn't go as far.

>>59202403
>hurr that benchmark doesn't count because it doesn't fit my narrative
Are you 12?

File: Untitled.png (148KB, 1213x727px) Image search: [Google]

148KB, 1213x727px

for $500 should i get a 7700k/mobo bundle or a 1800X alone?

>>59202363

for the memes duh

Because chips with fewer cores are outpacing it due to higher clocks. It makes sense that if higher clocks can be achieved with fewer cores then it could be a viable "profile" for gaymers

>>59202426
You are the underage /v/ermin posting about video games on a technology board. What does this have to do with kernel compiling, prime95, video encoding, virtual machines and such

fucking nothing

>>59202362
Is known that Overwatch only uses 2 cores.

>>59201195
It's a matter of perspective. You're not spending money on stuff. You're spending it on yourself.

>>59202303
I haven't either, would be interesting.

File: hän on ok sen kanssa.png (187KB, 600x339px) Image search: [Google]

187KB, 600x339px

>>59202362
>mfw my display is 60Hz 1080p

>>59202438
wait for the prices to stabalise. Ryzen litterally just came out.

>>59200877
Oxide games is incredibly excited with what we are seeing from the Ryzen CPU. Using our Nitrous game engine, we are working to scale our existing and future game title performance to take full advantage of Ryzen and its 8-core, 16-thread architecture, and the results thus far are impressive. These optimizations are not yet available for Ryzen benchmarking. However, expect updates soon to enhance the performance of games like Ashes of the Singularity on Ryzen CPUs, as well as our future game releases.” - Brad Wardell, CEO Stardock and Oxide

Creative Assembly is committed to reviewing and optimizing its games on the all-new Ryzen CPU. While current third-party testing doesn’t reflect this yet, our joint optimization program with AMD means that we are looking at options to deliver performance optimization updates in the future to provide better performance on Ryzen CPUs moving forward. " – Creative Assembly, Developers of the Multi-award Winning Total War Series

As we presented at Ryzen Tech Day, we are supporting 300+ developer kits with game development studios to optimize current and future game releases for the all-new Ryzen CPU. We are on track for 1000+ developer systems in 2017. For example, Bethesda at GDC yesterday announced its strategic relationship with AMD to optimize for Ryzen CPUs, primarily through Vulkan low-level API optimizations, for a new generation of games, DLC and VR experiences.
Oxide Games also provided a public statement today on the significant performance uplift observed when optimizing for the 8-core, 16-thread Ryzen 7 CPU design – optimizations not yet reflected in Ashes of the Singularity benchmarking. Creative Assembly, developers of the Total War series, made a similar statement today related to upcoming Ryzen optimizations.
CPU benchmarking deficits to the competition in certain games at 1080p resolution can be attributed to the development and optimization of the game uniquely to Intel platforms – until now.

>>59201955
Gotta be like when Windows needed patching to make Bulldozer go from "completely awful" to "not a total disaster".

>People thought Ryzan was gonna curbstomp Intel

Really? Did they really think this? It was my belief this whole time that Ryzan would get just close enough to force competition on Intel which would in turn force them to lower their prices.

>>59202012
Yeah, I wonder if we might see differences in phoronix tests.

>>59202460
Look at this mad motherfucker

The reality is, it sucks for gaming, stay mad bitch

File: 1310429820535.jpg (13KB, 300x300px) Image search: [Google]

13KB, 300x300px

>>59202472
Really? Then why is the 6900k, a much lower clocked cpu higher on that list than the 7700k?

>>59202575
In most cases it did, only thing is iffy ram support and gaming performance, everything else is top notch

>>59202575
>>People thought Ryzan was gonna curbstomp Intel

well in single perf is quite close despite being gimped since windows doesnt have support yet

in MT is light years ahead

in perf/watt and perf/dollar doesnt have a competition

>>59202575
Well it was aggressively priced. It's half the price of their flagship 8 core chips with competitive performance. The biggest difference between R7 and the 8 core i7s aside from the price is that R7 runs at 95W or less. There's even a 65W part which I still think is really impressive.

>>59202622
because the internal bandwidth of the 6900k is triple compared with the 7700k same goes for ryzen

thats why they top out on 299 its not the game its the engine tailored to the uarch

is ryzen good with /3/ stuff?

>>59202634

I would delete this

>>59202634
I just don't get why people bought into the hype and preordered like fucking idiots, even with Jewtel being pure fucking AIDS.

Stuff like the RAM timing or the L3 Cache not being supported are exactly the reason why one should first wait if any problems would pop up, aside from benchmark results.

>>59201955
That happens on various workloads on Intel too, and can also be seen when going from SMT4 to SMT8 on POWER8.

SMT is far from an uncontroversial technology, since it makes the threads running on the same CPU share various resources like cache that may degrade their performance more than linearly.

>>59202684
Yep that's one kind of workload that it excels at. It's good at anything that can take advantage of its cores and threads. Plus it's quite cheap relatively speaking.

>>59202684
GN's 16 monkey-heads rendering test puts it very close to 6900k levels of performance. It's literally a 1-2 sec difference in render time.
http://www.gamersnexus.net/hwreviews/2822-amd-ryzen-r7-1800x-review-premiere-blender-fps-benchmarks/page-6

File: [Asenshi] Flying Witch - 09 [D339EE49]_0001_00:11:48_[2016_09_30_04].jpg (162KB, 1600x900px) Image search: [Google]

162KB, 1600x900px

Where are my confirmations?
Does it support ECC?
Does it have instructions intact so I can passthrough?
How does it fare on emulation?
What's up with the RAM?
How does SMT fare on Linux 4.10?

I compile everyday, so grabbing one looks very tempting considering I'm still on a 2500k. I don't play newer games, but I still do play games. It looks like the R7 would be a good buy.

>>59202721
because its not a problem with the cpu..

the imc is purely on the mobo aibs they said they will need 1 month but they released it yesterday

the l3 is a windows 10 problem (its fixed on the creators update currently running on the fast ring) the same goes with the smt (because the memory controller is very efficient )intel had the same problem when they launched HT

this isnt a case of bulldozer that amd waited for ms to make it a bit "better" its literally new technology without support even the sound chip doesnt have drivers yet

>spend 350€ for an i7 + some for a high end mobo
>still play games at 1080p @60Hz
I don't get these people. At that resolution and and framerate ryzen isn't going to bottleneck you at all and you get the benefit of incredible multi threaded performance.
A very small percentage games with 122-144 Hz monitors.
I got my 1440p monitor 4 years ago and I love it. Certainly wouldn't trade it for 2 1080p no matter what refresh rate.
At 1440p or higher, this amazing is amazing

>>59202068
It does seem clockspeeds are going to be Ryzen's greatest problem. Especially if the details regarding the upcoming R5/R3 parts are true, they hardly offer higher clockspeeds than the R7 parts, desprite having as few as four cores.

I wonder what the deal is. If it's just a manufacturing thing, we could be looking at higher clocks in the future as Glofo tunes their process, but if Zen has some intrinsically bad critical paths, that may spell mediocre performance in the future too.

>>59202721
>RAM timing
Wait, what's wrong with the RAM implementation?

>>59202068
4.0ghz is what the xfr does moron LOL
3.9 is the xfr range for the 1700x

do you even try?

LOL /g/ mods are fucking faggots and AMD fanboys.

They are deleting every negative Ryzen thread right now.

>>59202842
they had a huge cas latency for some reason with some mobos this has been fixed since yesterday

those that didnt mention this on the reviews are those that shilled the most

>>59202819
Retard who has no clue about CPUs.

Going to 4K resolution, the GPU is clearly the bottleneck.

It's no wonder that the CPUs gets closer in the benchmarks.

Get fucked

>>59202664
>the internal bandwidth

Top fucking kek, what the fuck are you even talking about

>>59202160
>1800x probably doesn't have the single core IPC of intel.
It did pretty much on par with the similarly clocked i7 6900K on Anandtech's single-threaded benchmarks, so to me it looks more like clockspeed being an advantage for Intel on some benchmarks.

>>59202858

Actually try reading the reviews fucking retard. 90% of reviewers haven't managed to push X1800 to over 4.1 GHz.

This is case of AMD doing super aggressive binning yet again and releasing products with literally no OC headroom.

>>59202886
>board flooded with capslock spam and memes with "ryzen" in the OP
>15+ threads
>hmm I wonder why they're being deleted
almonds: activated

Any folding@home or BOINC benchmarks?
Is this a good crunching machine?

Idk how you convince someone who does a lot of Adobe CC rendering to switch to Ryzen over Intel though...

>>59202892
Most reviews published by big sites were written weeks ago, and not updated last minute.
Sites like Tomshardware and Anandtech will be publishing follow up articles once everything is mostly sorted out.

File: 2.jpg (152KB, 1270x685px) Image search: [Google]

152KB, 1270x685px

>>59202664
>it's never AMD's fault!!!

>>59202076
Games are literally coded by monkeys and lazy ludites.

>>59202939

>5 years old 3770K beats Ryzen 7 1800X in avg fps

KEK

File: david-face.jpg (261KB, 900x1167px) Image search: [Google]

261KB, 900x1167px

>>59200877
>>59200915
>>59201249
>>59201276

Stop playing games.

>>59202913
i think you need to read a bit more about how a cpu works instead of saying "TOP FUCKING KEK"

>>59202909
>ryzen isn't going to bottleneck you
Retard. I'm saying that for people who have moved on to 1440p@60Hz or higher it's a good deal. Certainly for people who game on 1080p@60Hz, the majority.

>>59202808
>How does it fare on emulation?
This wouldn't be the chip you use for emulation. Most emulators are lightly threaded, dolphin is 2 cores max I believe and even that isn't strictly correct because some games require that you disable that I think.

The R3s will probably be the best in the Ryzen lineup as far as emulation.

>>59202975
I think you need stop talking out of your ass and wake up to the real performance.

It's shit for gaming, just accept it

Man I totally care how well the CPU performs under games coded by drooling retards.

File: 90f7eade0b9f420cb51df4d2817df76b.png (59KB, 631x659px) Image search: [Google]

59KB, 631x659px

>>59202939
meanhile on a more serious site
guess what

>>59202984
It's not winning either against the i7-7700K.

With a stronger GPU, the difference will become bigger again.

>>59202995
The only really performance demanding emulators are for the PS2, Wii, and WiiU.
Ryzen runs just fine in Dolphin and PCSX2.

>>59203014
6900k shits all over it, nice

>>59202999
>It's a shit CPU because some idiot code monkeys made messy code and can't utilize 8 cores and all other tech aboard
ok

who cares other than gamers?
and if you are a "gamer"
fucking kill yourself

>>59203029
at only a thousand dollars

File: 1463305343924.jpg (125KB, 1375x749px) Image search: [Google]

125KB, 1375x749px

>>59203007
>games coded by drooling retards

Ahahaha, surely butthurt is real.
I bet those people coding those games that sell million times, will have a higher earning and a better life than you will ever have.

>>59203029
>shits all over it, nice
Yeah at costing 1800
mean while 1800 sits at 600

File: 1488463564529.png (47KB, 680x628px) Image search: [Google]

47KB, 680x628px

>>59203014
That precious side of yours shows 1800x being 12% behind 7700k on average so I don't know what's the point you're trying to make.

>>59203029
thats not the point moron you cant have a 4c cpu shitting on 8c while the game is known to be using up to 20 threads

File: 12717956_596838670464823_1454185656159592850_n.jpg (14KB, 480x268px) Image search: [Google]

14KB, 480x268px

I was really hoping ryzen would make jewtel step up their game or lower prices

but it looks like they won't need to now

>>59202995
>This wouldn't be the chip you use for emulation.
Of course. If I wanted a dedicated emulation machine, I would just grab the best single threaded CPU out there and overclock it to oblivion. An unlocked Pentium, i3, or a 6700k/7700k would do. Thing is that it's not the only thing I do since I mentioned compilation. I do encode every now and then and the 3500k does the job, but I'm sure twice the cores will help me with that.

>people are suddenly starting to compare the 1800X to an i7-7700K for half the price

Kek'd

The 1700 costs about the same as the i7-7700K and stands no chance atm.

File: ryzen_smt_issue.jpg (92KB, 422x400px) Image search: [Google]

92KB, 422x400px

>>59201612
>SMT

One of the reviews "disabled SMT" which on their board means that they disabled entire CCX. They got no performance change on a halved chip.

Apparently Windows ( surprise ) fucks with thread scheduling and drops them to SMT instead of preferring full cores first.

>>59203054
Game development jobs are a meme. Long hours for shit pay, and your customers send you death threats if you don't cater to their bullshit.

>>59203097
>Windows is too much of a hot garbage OS to use modern CPU.
predictable

>>59201656
CPU Caches are transparent to OSes.

>>59203097
It was a problem with Intels HT early on. Seeing as SMT is relatively new, I don't see why SMT wouldn't be facing the same problems.

>>59203068
the point is whoever re do the benchmarks with the current bios will get a lot of different results mostly because there will be no memory latency to fuck up the game jayishillforintellalltimeztwopesos already hinted at that..

Anyone seen their Amazon preorders updated yet?

>>59200873
So far there are only gaming benchmarks with SMT off, and it shows no big difference ...

File: Intel.jpg (67KB, 368x540px) Image search: [Google]

67KB, 368x540px

>Intel CPU performance still superior to AMD's Ryzen
>Ryzen only made Intel CPU even cheaper

>>59203021
Sure but what I'm saying is you will get your 60 fps on your 1080p monitor with rizen for years to come and have great multi threaded performance outside of games.

what happened to the "Jim Killer can do no wrong" thing?

File: smt1vssmt0.png (120KB, 806x580px) Image search: [Google]

120KB, 806x580px

>>59203135
GN's benchmarks show that it hurts the 1% lows and 0.1% lows big time

>>59203110
I would expect it to be more of a compiler problem than an OS one.

>>59201528
So, it's just another retard then? Fucking really?

File: Utter cuckery.png (38KB, 874x495px) Image search: [Google]

38KB, 874x495px

It's sad, really.
AMD pisses all over them, and they can't get enough.

>>59203147
did he failed? NO there isnt proper support for ryzen yet neither on windows or on linux

he just created yet another thoroughbred b(for those that actually remembers )

>>59203149
Yes, the 1% lows and 0.1% lows actually determine how you perceive the game is running.

Sudden dips of frames are really noticeable if the average FPS is high.

>>59203183
he's so smart that he created a product that is underwhelming because no OS on earth can use its amazing power

really makes u think

>>59202920
Reviews are fucking idiots when it comes to overclocking. Every time there is a new platform they do the basics and are surprised when they get shit results. Wait for the real OC dudes in the community to work with it and understand how to OC it. It may well be a very shit OC'ing part, but I will wait for some dudes who are willing to spend a week tweaking to write some guides.

>>59203176
(You) have to go back.

>>59203114
Because the problem and the solution is exactly the same as it was for Intel and should already be solved?

>>59203147
That's why he left to Tesla in 2015. he knew how disappointing it is going to be.

>>59203027
All emulators are very reliant on single core performance, and since Ryzen's shit at it, it's just going to get mollywhopped in that field.

>>59203209
But they can use it. The new Linux kernel and the Windows 10 Creator's update should both increase it's SMT/cache performance massively.

>>59203246
Keep your shit posting in /v/

>>59203243
Depends how the front end of the cpu looks like to the os. If the chip reports "16 cores fo realz" then the os can merely shit itsself.

>>59203243
AMD might be doing their multithreading a bit differently so whatever makes Windows run well on HT might not be applicable to SMT right away. However, considering that they're very similar functions, it won't be long that we might get a fix available.

>>59203268
>stop making AMD look bad!!
Make me, pajeet.

>>59203265
even if it does, the damage is already done

most sites wouldn't even bother rebenching and Ryzen will always be viewed as a failure

>>59203243
It's a little different for every implementation. OS support for modern CPUs isn't at all seamless, and AMD never fails to drop the ball at launch. Bulldozer had similar problems. They were just overshadowed by much, much worse design defects.

>>59203296
You're forgetting they still have the R5 and R3 launch which will be much more attractive to gaymers and normies.

>>59201987
hahaha implying bethesda the developer will ever optimize their archaic engine.

that is most likely referring to idtech games.

>>59203276
It's not like the OS is "finely tuned" for Intel's HT or anything like that. It's just a matter of preferring real CPUs to virtual ones.

>>59203296
failure? how label it as a failure? only shills on /g/ that started to compare it with 7700k while amd clearly said that they are going against 6900k...

to that regard is quite close to it even beating it some times for half the price

and having lower tdp

>>59201377
Half the time jumped 30-35 %, the other half 1 %.

>>59203325
Yes, and AMD might be doing its addressing differently from Intels.

>>59203330
>lower (((tdp)))
Actual thermals fly out of control if you try to break 4.1GHz. These will never challenge the 7700K in games. Not even remotely.

R5 vs. Core i5 is going to be a more interesting question.

>>59203357
The CPU topology is presented in the ACPI tables, so in that case they're doing their "addressing" wrong, quite simply.

>>59203330
no bought the fucking 69's dude. they are pure meme's starting at $1000. For gamers there is no point in buying anything past the i7 7700k.

>>59203381
Which is a clock that AMD does not sell at. Chips currently tested are most likely earliest production units and/or rejects from dies selected for Naples chips.

>>59203389
Can't you just accept the fact that the implementation is not exactly the same?

>>59203137
>implying these aren't good news.

Are you retarded? Because you seem to be retarded.

gaymes gaymes gaymes gaymes gaymes gaymes gaymes gaymes gaymes gaymes gaymes gaymes gaymes gaymes gaymes gaymes
There, a gayme for every Zen thread.

>>59203414
Of course the implementation is different, but that doesn't mean that it doesn't fit the protocol. What would you even suggest that the OS *could* do differently that it doesn't already do for HT?

>>59203397
a casual gamer most probably will buy something up to 6c at most

but the 8c arent for them they are targeting all the streamers that need 2 setups to stream properly because the next best 8c is at 1000...

plus its not like the 1800x is horrible on games its doing quite well actually but its not a 7700k....

>>59203397
(((gaymers)))

For content creators 69 chips are steal. I compile shitload of stuff at work and at that job 20 minutes your highly paid developer spent on browsing 4chan are wasted. If you cut that to 5 minutes then that may shit a free 6950x every month or so.

>>59200851
>>barely beating the 7700k in MT

PassMark is not exclusively multithreaded, idiot.

https://www.tek.no/artikler/test-amd-ryzen-7-1800x/377125

Norwegian test of the 1800X. They actually like it.

>>59203434
Let's just say that because of the different implementation, whatever works on HT might not work immediately on SMT.

>>59203450
Now you are into the enterprise realm. For the person working out of their bedroom its much more likely their workload can be hardware accelerated and in that case better off getting a better GPU anyway.

>>59203014
why are their results that different? consumer won't be smart enough to tweak their stuff working proper

>>59203512
Problems with Windows and early motherboard BIOS. It all fixes itself more or less transparently.

>>59202088
It should be a surprise to no-one that Bethesda's games are not optimized in the slightestl.

File: 1428451278185.png (1017KB, 848x900px) Image search: [Google]

1017KB, 848x900px

GN:
> For now, though, the 1800X is a disappointment, and is not a processor we recommend for gamers when considering the price-point. If you were to buy it, disable SMT for gaming. It’s mostly detrimental, likely due to resource contention among threads in gaming environments. Even mixed workload users should consider when and where software acceleration is better than GPU acceleration, if ever for their needs, before purchasing the 1800X.

what a poc

>>59203548
and SMT is purposed to better the performance while now shutting it down gives worse till same performance?

File: K9N5Aev.png (45KB, 960x540px) Image search: [Google]

45KB, 960x540px

TheStilt wrote up a very in depth technical analysis of Ryzen. The best on the web by far.
https://forums.anandtech.com/threads/ryzen-strictly-technical.2500572/

His measured power numbers match what Tomshardware saw with their oscilloscopes. The Ryzen 7 1800X at stock pulls 110w max when you really hammer all the cores. Pushing it to 3.9ghz on all cores static increases power to just under 120w. All core 4ghz would be right around 130w~ if your chip got there and was stable.
Pinning it to 3.3ghz its just under 59w.

>>59201528
It's halfway between bulldozer and the hype

https://translate.google.com/translate?hl=en&sl=de&tl=en&u=https%3A%2F%2Fwww.golem.de%2Fnews%2Fryzen-7-1800x-im-test-amd-ist-endlich-zurueck-1703-125996-6.html
ryzens IMC is shit that's why games are so slow

>>59203640
I'd say it met the hype it just wasn't what /g/ expected. I think /g/ expected Ryzen to just BTFO of every Intel chip under $1000 and it ended up besting a lot of them but still lost to a few in certain areas.

>>59203512
lots of reasons..

smt
being more tailored to intel ht
L3 cache problems
general miss handling on the windows as usual

guys this clown posted raw result footage

https://www.youtube.com/watch?v=BXVIPo_qbc4

why is there such a difference between reviewers results?

This is about what I expected. I still am getting an 1800X because I can afford it and I have some brand loyalty to AMD. Reasoning for that is that their CPUs have aged well for me so far and I agree with their design philosophy more closely. I'm not planning on upgrading my internals for a very long time after this unless a breakthrough comes about. I have an 8350 currently so damn near anything will be a major upgrade in single core performance. Whilst I could also buy a 7700K and go Intel, I don't necessarily think it's a better CPU. I expect the 1800X to age very well. What really baffles me is my friend who does TONS of video rendering like that's his job, a professional video editor. He goes and buys a 6700K on last gen prices knowing full well AMD with their history of moar corez is dropping a vastly improved uarch soon. He and I should swap CPUs for the workload we do if I had a 1800X

>>59203714
Mobo/BIOS issues

>>59203717
get the 1700 and oc save some cash

>>59203485
If that's what you want to convince yourself, don't let me stop you.

>>59203735
I did that, just came today. Still waiting on the mobo and ram...

File: d5r8sd9qt0jy.jpg (112KB, 746x962px) Image search: [Google]

112KB, 746x962px

>>59203714
because most mobos had fucked up bios
then there is the smt problem clearly the OS isnt handling it well

the chip is quite good but till the smt bug gets fixed by ms(and you know how well ms fixes bugs) we wont really be sure about its true potential

>>59203601
CPUs rarely use all resources at the same time. The idea behind SMT ( and HT ) is that if the CPU is using FPU, and has ALU free, then you can schedule another thread that uses ALU. However if you have another thread that uses FPU, than it needs to wait.

Here is the issue:

2 core / 4 thread cpu.

You have a program that uses 2 threads and does something mundane, say rendering, heavy FPU work.

If you schedule threads to physical cores, you get fast results. If you schedule both threads to one core, then you get results slower.

The bug with SMT scheduling we seem to have here is that, say 8 thread games ( witcher ) seem to run on four physical cores on the Ryzen instead being schedule to all eight cores.

This is either a bug in the game, OS on the CPU itsself.

>>59203735
Eh, I have never had the top SKU of anything, I want to go all out this gen and not upgrade for a while. I'm putting it with an Aorus AX370 Gaming 7 and probably 16 or 32 gigs of DDR4 (whatever is cost effective and can do dual channel)

CPU is going under the best air-cooler on the market with improvements. Cryorig R1 ultimate with Noctua Industrial PPC 3000 RPM fans. Got my 5 year old 8350 up to 4.5 on my first overclock so we'll see what it does with RyZen.

>>59201087
Joker is a fat ass and uses too much thermal paste

I have a 4770k

Do I upgrade to 7700k or just wait for Coffee Lake or X299/X399/etc?

File: 9oVGc83.png (64KB, 960x540px) Image search: [Google]

64KB, 960x540px

>>59203635
>850 points in Cinebench 15 at 30W is quite telling. Or not telling, but absolutely massive. Zeppelin can reach absolutely monstrous and unseen levels of efficiency, as long as it operates within its ideal frequency range.

Intel is going to lose some enterprise marketshare.

>>59203714
what's interesting is that intels cpu seems to have usage at 70-80% while amds cpu shows usage at 30-40%

looks like alot of reviews didn't out to much time into their reviews, they just wanted to cash in on meme views.

>>59202197

source: every processor ever made with lower core counts on the same arch

File: [Bakemono] Nanatsu no Taizai - Seisen no Shirushi - 01 [720p x264] [69C752D0]_0001_00:12:07_[2017_02_24_09].jpg (97KB, 1280x720px) Image search: [Google]

97KB, 1280x720px

>no one talking about anandtech's review
You're all fags.

File: su bae.png (50KB, 927x544px) Image search: [Google]

50KB, 927x544px

>inb4 leddit

>>59203814
>4 days given to reviewers for a processor
>random enthusiast mobos
>shitty generic Noctua cooler
>only ONE pair of certified supported RAM sticks

AYYMD fucked up, they should have been given a couple of weeks.

>>59203821
Because it was written over a week ago, and they're going to redo everything in their upcoming part2 article.

>>59203812
>implying either Kaby Lake or Coffee Lake will be real upgrades from Haswell
Either get a Ryzen octo or hexacore, or wait for Skylake-E.

>>59203848
Most big reviewers have had their samples for 2 weeks. Only smaller outlets got chips sent to them this week.

>>59203817
>source: every processor ever made with lower core counts on the same arch
but that's completely wrong in every way?

>>59203859
I'd wait for Skylake-E for moar corez but since Intel won it would probably cost me an arm and a leg.

>>59203868
linusmemetips review was a fucking joke

>>59203744
You seem to be convinced that what worked for intel would, 100% out of the box, work right away on a different CPU brand made from an entirely new architecture with a different implementation of multithreading.

>>59203868
and within that two weeks is a last-minute mobo update that can drastically change performance

>>59201983
>>59201864
more benchmarks lik dis

>>59203714
am i going insane or is there less stuttering/clearer image on the 1700 side?

I got BIOS update today.
Is it related to 1700-1800 being released

Anyone with some review/graph of 4k gaming?
mostly interested in fallout4 but others are interesting aswell


im really concered about the 1080p numbers atm, in fallout 4 ryzen was lacking severly at sub 100fps getting beaten by i3s and my current i5 4690k.....
I was expecting the 150fps+ gaming to get hit and thats fine but we are going down in fps and its worrysome.

>>59203879
>implying Linus does actual reviews at this point

Purposedly breaking the test rig was unprofessional and a publicity stunt done in poor taste.

File: 8Rch6JF.png (55KB, 1509x905px) Image search: [Google]

55KB, 1509x905px

>>59203813
As indicated by the Vmin-Fmax curve, Zeppelin's voltage scaling is perfectly linear until 3.3GHz (25mV per 100MHz). The first deviation ("Critical 1") from this linear behavior can be seen at 3.3GHz. The second and the final deviation ("Critical 2") can be seen at 3.5GHz. Beyond this point the voltage scaling is neither linear or recovers even temporarily, and the CPU is requiring higher voltage in increasingly larger steps to scale further.

>The ideal frequency range for the process or the design (as a whole) appears to be 2.1 - 3.3GHz (25mV per 100MHz). Above this region (>= 3.3GHz) the voltage scaling gradually deteriorates to 40 - 100mV+ per 100MHz.

Wew. A good undervolting die is going to be incredible. 75w 3.7ghz here I come.

>>59203912
Some other people who got chips shipped early mentioned the smoother game play in BF1 as well. Less stuttering while shit is loading. Makes sense given no core is getting near 100% load.

>>59203714
>fucking moonlight sonata in background with that guitar crap

wtf

>>59203959
Probably because Fallout is betehesda shit game

File: frametimes.png (82KB, 647x409px) Image search: [Google]

82KB, 647x409px

>>59203912
>>59203967

Ryzen has consistent frametimes in BF1.

>>59203987
Frametimes is all I care about considering I use mid/high range tier gpu's. Guess I'll wait for the 1600 reviews.

>>59203782
so what we see now is their 4core/8thread performance? kek, that would be fun

>>59203967
On the high-end models the actual (effective) voltage for the base frequency (e.g. 3.6GHz on 1800X SKU) can be anything between 1.200 - 1.300V. Meanwhile the actual (effective) voltage for the highest single core boosted PState (XFR, e.g. 4.1GHz) can be as high as 1.47500V.

>(XFR, e.g. 4.1GHz) can be as high as 1.47500V.

This is extremely interesting on the technical side of things. Here we have a processor on a 14nm FinFET process with its ideal range being 2.1ghz to 3.3ghz with voltages of 700mv to slightly over 1v.
Yet they are pushing over 1.45v through the chip in peak boost states. Global Foundries had to have created a transistor library specifically for AMD. Oxide insulation has to be THICC. On another interesting data point we have seen LN2 over clockers hit 5.3-5.8ghz on LN2 with voltages between 1.7v and 1.97v thus far, most all having chips return to usable conditions after their bench runs.
The process that these chips are fabbed on is incredibly robust.

>>59203782
>The bug with SMT scheduling we seem to have here is that, say 8 thread games ( witcher ) seem to run on four physical cores on the Ryzen instead being schedule to all eight cores.
Highly doubt that. Windows scheduler knows if 2 logical cores belong to the same physical core. Bulldozer bug was fixed like that, instead of acting like a real 8 core windows was patched to take bulldozer cores belonging to same module act like 2 logical HT threads, so if there was 4 threads it was distributed to each module.

"SMT bug" seems to be AMD optimizing SMT for server workloads rather than games. Intel's first HT iterations degraded performance on some cases too.

File: Power.png (20KB, 600x400px) Image search: [Google]

20KB, 600x400px

>>59203635
>Ryzen at 3.9ghz actual CPU power consumption is under 125w
>Broadwell-E actual CPU power consumption is 135-149w stock

Ryzen's perf/watt is very high compared to intel's.

So I guess I should get a 1700x or 1800x and then wait for Zen+? I want to upgrade from the sandy vagina.

>>59204188
Intel's HyperThreading *still* has some performance degradation in gaming workloads. It is an OS level and game engine issue, not a hardware level one.

>>59204223
Too bad that doesn't extend to good OC scaling with extra voltage. You can barely get above 4GHz.

>>59204237
Wait for R5 results unless you need moar corez.

>>59204243
There isn't much to do on OS side at this point if Windows is treating Intel and AMD logical cores the same, MS should have optimized it since more than a decade

>>59204188
Which still can be worked around by OS.

>>59204291
The moar cores the merrier. Encoding on this is a pain in the ass.

>>59204266
OC headroom does seem really low, but power consumption isn't the limiting factor. Power delivery on the boards seems to be a big culprit, so are thermals.
The 14nm LPP Vt Ryzen is fabbed on has some incredible characteristics, but super high frequency isn't one of them. Not yet at least.

>>59204334

if you're encoding anything then yes. The 1800x and 1700x have no competition in price/performance and will still game perfectly fine.

>>59203987
might have to do with just more cores and its more stable, even in the benchmarks there isnt much of a gap from min to avg

>>59204355
Sure, but pretty much every review I've seen used fairly high-end motherboards, I don't think they'd have extremely shit power delivery. Thermals in the end are a function of power consumption, or are you trying to say you think there's some other issue like with Intel's mainstream CPUs which aren't soldered to the IHS?

I remember reading somewhere that Ryzen apparently has some sensors in the IHS too, so there probably won't be any delidding.

There definitely seems to be a RAM bottleneck bug. 3200 or 2666 Mhz doesn't seem to make any difference except with Doom. I remember the digital foundry review where there was a noticeable boost with higher speed RAM.

>On the Asus Crosshair VI Hero used by ComputerBase in this test, DDR4-3200 was also possible. Additional performance does not bring the average of the fourteen games tested, however: DDR4-2400, DDR4-2666 and DDR4-3200 are about the same. DDR4-1866, on the other hand, is clearly not enough, on average the performance in the games decreases by 15 per cent. In the top (Shadow Warrior 2), there is a 23 percent loss of performance.
translated from https://www.computerbase.de/2017-03/amd-ryzen-1800x-1700x-1700-test/4/#diagramm-watch-dogs-2-speicherskalierung

itt literal children complaining about 100 vs 90 frames on 1080@60hz when for literally anything else it's the same or (2x) better