Ryzen Discussion Thread 2

File: tmp_2421714949578718601591885308.gif (280KB, 384x216px) Image search: [Google]

280KB, 384x216px

R5 1600
Aurous x370 Gaming K5
16 GB Corsair lpx 3000
Running at 3.7 with a wee bit more voltage
Memory at 2933
This Mobo works but it's shit for enthusiasts or high oc seekers, grab the regular version (5) instead
Good thing this isn't for any too intensive

Is there any news yet on what improvements zen2 will get? Beyond just "better performance" and being 7nm?

I want to make a new computer with a ryzen CPU, but I'd really like to have better single thread performance first.

>>61484042
Higher IPC
Higher clocks

>>61484094
Are we talking small improvements or big enough improvements to close the gap with intel?

So Tech Deals recommends a 6C/12T R5 over a 4C/8T. I know it's not a huge margin in price, but once I'm in 4C/8T territory anyway do those two extra cores really matter? Or is that only if I want to do some major multitasking? I don't even run two monitors, let alone secondary processes to have on a second screen while doing intensive stuff like gaming.

>>61484136
It should close the gap at least.

>>61483782
1700x
crosshair hero 5
32gb 3200mhz g skill, can only hit 2933
getting 3.8ghz at 1.33v
>>61484136
The "gap" is already set to be closed
The only thing holding Ryzen back is clockspeed and zen2 is going to fix that. Ryzen has a higher IPC that anything Intel has right now and if that is increased even further intel is in for some shit
t. someone who sold their 6700k for Ryzen

>>61484182
Yes, with Ryzen you want those extra cores for sure.

>>61484216
Even though my usage is almost always "single concurrent application Facebook machine with 5+ year old games?" How would the two extra cores benefit that?

>>61484193
>>61484211
Sounds good.
I guess now I just need wait.

Also, I've heard that the x370 chipset is better than b350 for overclocking.
Is that true? If so, why?

>>61484238
Yeah, you probably won't see much use on those 2 extra cores in that case, but having them there will ensures that your build won't become obsolete as quick.

>>61484182
It's one of those "might as well because I can" type of situations. Go ahead and do so unless your budget is incredibly tight.

>>61483782
R7 1700 at 3.6GHz 1.2V(for stability)
Asrock AB350m Pro4
Corsair LED 3000 running at 2666, probably could run at 2933 if I messed with the timings and such.

Paired with a RX 570 Sapphire Nitro+

Just great in every way, no complaints whatsoever. It works wonders for 3D modelling and rendering, and a side of gaming, can playThe Withcer 3 on Ultra at 60fps, if i turn down the memeworks and people density to "very high" instead of ultra.

>>61484257
They can be due to higher quality components used
I won't go into huge details about it but basically yes x370 is better for overclocking

>>61484293
Those Sapphire Nitro+ cards are good shit.

I wish I could get ahold of another one, but miners fucking drove the price through the roof.

>>61484267
>>61484290
I guess what really didn't have me sold, beyond "would I even need the cores," is the out-of-box close speed difference. The 1500X 3.5/7 gave me more confidence in single core performance than 1600's 3.2/6, even though Tech Deals is recommending a comfortable OC to 3.7 on the 1600.

Do you know the (confident) limit on the 1500X with stock cooler?

>>61484391
With the stock cooler?
Probably 3.7 or 3.8 all cores.

>>61484257
Honestly they really aren't aside from maybe allowing you slightly more control in the bios.

The VRM's have basically been proven to be the same on the b350 and x370 boards for the most part, I'm using an MSI Tomahawk Arctic and got my ram to 3466mhz and r7 1700 to 3.9ghz @ 1.3375v, if i try to go to 3.95ghz i need to pump far more volts into the CPU, i could, my noctua d14 could handle the heat, but 3.9ghz seems to be my CPUs highest reasonable clock and I accept that.

>>61484431
Yeah, that's what I suspected would bug me. I can break even on OC, but my fallback to stock settings is even better by 100-300MHz. I get the whole "why not" thing but it's a really hard sell when I think about it. Maybe too hard for only $25, I dunno.

Thank you for all of the feedback though. It's been very informative.

R7 1700
ASUS Crosshair VI Hero
G.Skill 32GB 3200MHz CL14

I've gotten the CPU stable at 3.9GHz at 1.35v and the memory won't even post as anything above 2933, however at 2933 it sometimes has issues booting. But just hitting the reset button typically gets it to boot no problem. And I can easily attest to the performance gains on the ram for sure, gaming does seem a REALLY nice uplift in performance (From 2133.) 144 FPS at 1440p with a 1080ti is no problem for the 1700 now, and that's at stock clock speeds. Overclocking it makes it almost a non-issue in any game I throw at the combo.

Probably get memed asking here, but fuck it.
I bought i5 7600k & Z270 motherboard on sale for about $400 USD, and am now considering trying to sell it for about $350 since it's never been used, and is still relatively new.
Would that be realistic to expect? Newegg has a combo at about $160 more than I'd get from selling the Skylake, which shouldn't be too hard to get.

File: 1467843122829.jpg (60KB, 640x425px) Image search: [Google]

60KB, 640x425px

>>61483782
Dual rank IS quite a bit faster than single rank
2933Mhz dual rank is superior to 3200 single rank
And dual rank can be stable at 3200, but often needs really high SOC voltages
In addition, of you're on a gigabyte gaming 5 and want to get the highest memory speeds, the older F4 UEFI is your friend
But you wont have the latest AGESA update with that, so keep that in mind

And does anyone know why it takes 30 minutes to reset the UEFI on that board when the automatic reset after boot failure fails? Really annoying

R7 1700
MSI B350M Gaming Pro
2x4GB DDR4 Patriot Viper Elite

Still overclocking/testing the R7, but it runs really hot: 70C at stock everything. Undervolted it runs fine, but further stability testing is needed; at 3.4Ghz/1.075V with the stock cooler and 30 minutes of Prime95 small FFTs, I'm getting 73.5C. What the hell is going on? Most people are getting temps way lower than mine.

My RAM runs fine at 2400Mhz though.

1600x stock
x370 gaming 5
Sapphire 390 Nitro
3200 LPX modules, running them at 2133 for now

I really don't see a point increasing memory clock yet, ran titanfall 2 yesterday( first "new" game in a few months) it chewed the thing at 70 fps on 1440p
honestly was impressed, my 390 got a new life
was ready to tweak CPU/RAM at this point, no need.

>>61484968
are you looking at tdie temperature?

>>61484981
at least run it at 2666, 2133 to low man

>>61485027
What does Ryzen Master report as the "Temperature" on the main page? Is that tdie?

Should I wait for zen2? I want to upgrade to high core master race but I can wait.

What's the stock R7 1700 voltage? Also, what's the usual undervoltage success rate? I know it varies from chip to chip, and the motherboard would have a hand to play, but I'm curious as to what a general undervolt will land at while being stable.
I have no plans on overclocking, and I'd imagine 3GHz on Ryzen is just fine for video editing/rendering and some VM meddling, while keeping a manageable power consumption and temperature.

>>61485058
yeah, it's high then
it's normal for stock cooler though under stress load, what cooler you got?

>>61485028
what for? it works more than fine for everything I throw at it

>>61485126
Just use the offset to undervolt
About -0.11V seems to be possible on most 1700

>>61485126
some people report lowering up to 0.1v for stock clock
more lucky ones even report slight OC with undervolting

>>61484786
untweaked dual rank often beats untweak single ram, but after subtiming optimization tables tends to turn

>>61485147
Oh, I forgot to mention stock cooler in my post. If that's normal for stock cooler we're all good. It took me a few months to save for the R7 1700 (I had an AM3+ motherboard before then, $450 minimum investment to upgrade), but I'll upgrade to a non-stock cooler ASAP. I'm thinking of either going for the Noctua NH-L9x65 or a liquid cooler (I only have room for a 120mm radiator), but it'll take a few months to save up for the upgrade (college bills are a bitch). What would you recommend for the 1700 @ 3.5-3.7Ghz? Not looking to go up to or past 4.0.

File: ram.jpg (61KB, 476x452px) Image search: [Google]

61KB, 476x452px

Running this RAM at 2933Mhz on a gigabyte gaming 5, F6 UEFI
F4 UEFI allowed me to run it at 3200Mhz without increasing the SOC voltage too much, but F6 has less problems overall

>>61484136
current process is geared for 3ghz operation, and most cpus will hit 3.8 without getting danger volts with many able to hit 4.0

the 7nm process is apparently geared for 5ghz operation, which if you can oc the same amount, would mean the cpus could hit 6.5ghz

however, i'm just taking it at face value as every cpu will be 200mhz around 5ghz minimum and likely xfr or whatever will deal with the oc values.

>>61483782
1700
taichi
32gb 3200 ripjaw i believe

haven't oc'ed yet as I was waiting on a further down the line bios so I don't have to re dick around for hours each time something better comes. haven't gotten around to upgrading the bios as we have had fairly stormy weather and don't want to risk bricking the fucker.

my ram is at 2400 as that was all that was stable without dicking around.

>>61485217
I'd stick with stock(unless it's very loud for you), stress testing temps are irrelevant, normal use you probably get around 60C?

>>61485327
Idling I get between 35-50C. So I think I'm ok temperature wise. Thank you, anon!

File: 4qsfhkk.png (129KB, 1562x332px) Image search: [Google]

129KB, 1562x332px

>>61484136
5% IPC increase
20% clock increase
moar cores

>>61483782
1600 (3.9Ghz 1.35V)
Gigabyte GA-AB350-GAMING 3
G.Skill Ripjaws V 2x8GB F4-3200C14D-16GVK (3200Mhz)

Upgraded from 2500k. Really satisfied that I don't need to close some programs just to play games.

>>61483782
r5 1600 @ 3.9 @ 1.34, currently working my way down in voltage
asrock taichi, launch bios
8gb EVGA 3200, running at 3200mhz/ CAS 15
honestly been too lazy to change bios, evga is what came with this, i bought it all second hand. The difference in single and dual is large, my buddy at 3.5 dual is at the same performance as me at 3.9 with single. 3200 was a breeze, just set and forget in XMP, probably due to the single dimm used.
Mobo is nice, easy settings, sometimes resets OC, although that may be due to the no mobo battery.
as for the memory scaling there is a really good thread on OCN, after 3200 it appears latency matters more, as in 3466 CAS 15 < 3200 CAS 14, so on.

File: 1479579958400.jpg (313KB, 623x1200px) Image search: [Google]

313KB, 623x1200px

Are the ITX boards good yet? (ASRock, Gigabyte)

Looking to overclock a 1600x on a custom loop and switch to Zen2 or Zen 3 in the future.
Would also settle for mATX. Main reason for going ITX would be reusability of the board in a HTPC/Portable/Home Server build.

>>61485762
yes actually, they very recently came out, so they're out of stock almost everywhere rn, give it a week or three and you should be fine though

Since this is a Ryzen thread, I'll post here.
Should I sell my i5 7600k & Z270 and my old laptop to buy Ryzen 7 1700 as well as a motherboard/ram for it? It isn't a full system like I left out in a previous post, it's just the CPU/Mobo. All I'd need is memory for it. For reference, my current system is an Athlon 860K at 4.2GHz, 8GB DDR3 1600MHz on an ASrock A78 motherboard.
The Ryzen combo is about $550 USD, so assuming I can sell the laptop for about $200, and the CPU/Mobo for about $350, should I take this route?
I can move the video card, power supply, etc. but for now I was mainly considering the CPU, motherboard, and memory.

Ideally, I'd like to hop on Ryzen over Kabylake because higher core count (6/12 or 8/16 vs 4/4), better stock cooler (or none for 7600k), lower power draw (65W 1600/1700 vs 91W 7600k), and as Ryzen is pushing for more cores, I'd imagine developers will probably slowly start rolling out optimizations for higher core counts. I do some video editing/rendering, do some VM work, and would like to not worry about closing tabs on Chrome to be able to do anything. I'd also like to buy the system to just upgrade soon(tm) and not have to worry about upgrading again for ages (probably AM4's end of life).

I'm indecisive as fuck, and would love some guidance from someone that knows what their stuff, because I sure as fuck don't. Thank you to anyone that can help, I know my post is messy to read, but again, I'm indecisive about everything.

So for all of you that have been semi-following the mysterious segfaulting issue on Ryzen, it seems like something has finally surfaced.

https://community.amd.com/thread/215773?start=270&tstart=0

>CPU 14 COR ICACHE L1 IRD error

Oh dear, this could be pretty bad. Not as bad as the TLB bug but this could easily impact AMD's finances. No microcode update is going to fix a malfunctioning L1 instruction cache.

>>61484042
>7nm
HOW

>>61486217

Does that mean they need a new stepping? Oh god, and since their lineup is all out basically besides TR, they would need to at least respin all the Epyc processors at least and try and sell off the older steppings or liquidate them.

That seems pretty bad. But it only seems to be a productivity issue so far with software development. It doesn't seem like it would kill games or something like video editing.

>>61486333
For the plebs on Windows this won't matter much.

>>61486289
https://www.globalfoundries.com/technology-solutions/cmos/performance/7nm-finfet

>>61486362
shit
I was gonna jump in on Ryzen this fall, but now I kinda want to wait for this

Windows 7 or 8.1?

>>61484182
A top of the line 6-core CPU from 2010 is as capable for most uses as a top of the line 4-core CPU from 2014

The landscape is a little different now, but take from it what you will.

>>61486401
Windows 7. 8.1 is really not getting any love from AMD.

File: misato laugh.gif (1021KB, 500x484px) Image search: [Google]

1021KB, 500x484px

>>61486401
>Windows 8

R5 1600
Asrock AB350M Pro4
Corsair vengeance LPX 3000MHz 2x4GB

RAM is running at 2933MHz, haven't tried overclocking the CPU yet.

>>61486430
>>61486436
And if I use Windows 7 Enterprise and only install Security-Only updates I'll be safe from telemetry?

File: you gonna get blankly stared at.jpg (84KB, 720x400px) Image search: [Google]

84KB, 720x400px

>>61486448

>>61486448
Security updates are shit.
More important are the OS stability updates, security updates have jack shit to do with that.
Like the update that uncripples your boot times if you have more than 2TB of storage.

>>61486476
Do they have telemetry? I only have a 1.1GB SSD anyway.

>>61486483
>1.1GB SSD
nani

>>61486509
1.1TB, sorry.

>>61486526
The individual kb installations don't have telemetry bundled in them.

R7 1700 @3.65 1.185V
Asus prime x370-pro
2x8gb Corsair vengeance lpx 3000 Mhz - runs @ 2933Mhz , no crashes

I feel that it is quite wierd that it runs it at 3.65 all cores at 1.185V.

So far ryzen feels awesome.

>>61486542
So I'm safe with Enterprise and only taking Security-Only and Stability updates?

If I use my 1800x up until Zen2 releases, do you think I can resell/trade it in for a discount on that sweet 7nm?

File: 1482696210676.png (127KB, 1456x873px) Image search: [Google]

127KB, 1456x873px

>>61485528
Either this process node is next-level magic or there's a bit of marketing fluff going on.

Zeppelin B1 (Ryzen series) operates at 3.045Ghz at 14LPPs 'overdrive' voltage
> >3Ghz operation
And at 'nominal' voltage operates at 2.46Ghz

For the 5Ghz target to apply to Zen2/Zen3, whichever is designed for 7nm,
>binned 1800X capable of 2.675Ghz at 0.85 drive voltage
>people now rumor Zen 7nm to reach an 87% clock increase from a full node shrink

>>61486217
The most common fix seems to be increasing voltages or a replacement CPU. Could be a problem with binning and/or voltage regulation, not an unfixable hardware bug.

>>61485985
Shameful bump for my dumb ass.

>>61486570
pls respond

R5 1600 [email protected]
B350 tomahawk
Corsair LPX 3200@c16 only working at 2933@c14
Rx480 8g
I like my ryzen system

apu when

>>61486766
Probably?
You are asking people to look into the future, nobody knows the answer better than you.

>>61486840
Guess I'll just be hopeful.
>tfw full AMD systems are finally good again

>1800x
>Gigabyte Gaming 3 (cheapest B350 MOBO)
>HD7850 (waiting for Vega)

>>61486862
>HD7850
Only thing that this GPU lacks is vram.

>>61486884
It's doing okay on my 1440p monitor, I'm deliberately staying away from software/gaymes with higher requirements until I can get an aftermarket Vega.

>>61486619
>Either this process node is next-level magic or there's a bit of marketing fluff going on.
It's next level magic.
It's also denser than fucking Intel's 10nm.

File: img_cbb70957ulu2v.jpg (158KB, 1024x768px) Image search: [Google]

158KB, 1024x768px

>>61485893
apparnetly the VRM coolers are shit on both the ASRock and the Gigabyte.

http://forum.asrock.com/forum_posts.asp?TID=5650&PN=1&title=ab350-gamingitx-ac-extremly-hot-vrm

https://www.reddit.com/r/Amd/comments/6nmg8a/gigabyte_gaab350n_owners_how_are_your_vrm_temps/

>>61487110
first of all
>Buying Fatality when Pro4 exists

Now, that this is out of the way, what is thermal foams decimation levels? It's really easy to say "oh that foam lololo" but knowing what you are talking about, takes some actual knowledge of physics.

>>61487157
It's the only ITX board ASRock sells and the reported tempereratures are high even at stock settings.
Doesn't take much physics to know that the cooler can't be good.

>>61486393
buy infrastructure now(before RAM jumps in cost again)
and get either 1600 or 1400 for AM4
insert zen2 into same AM4 next year

>>61486570
probably.

>>61487303
Is RAM really expected to increase in price again? Why?

>>61487354
rumors, ssds gonna jump as well
even though more and more fabs start to pop up and production increases even more phones will use it
it will plummet do the dirt eventually, but not soon

So when is Zen 2 supposed to come out? Waiting here with a decade old C2D computer, or should I just buy now?

>[email protected] stable
>gigabyte x370 k7
>2666mhz ram

Intel is fucked

>>61487388
Now that sucks.

>>61487392
just go R51600 now, if it won't be enough by the zen2 time switch it, same socket is a nice thing to have

>>61487392
Fucking years dude. Mid 2019 if they're quick with it.
Upgrade your c2d.

File: Ryzen.png (190KB, 1024x576px) Image search: [Google]

190KB, 1024x576px

>>61483782
CPU: R7 [email protected] (1.4v)
MB: ASUS Crosshair Hero 6
RAM: 2x8 Gb Ripjaws V @ 3000 MHz

>>61487416
>>61487485
Will try and do.

>>61486619
14LPP is a low power node, while their 7nm process seems to be geared towards high clock operations.
Even within the same feature size there are HUGE differences in clock speeds due to different process characteristics.

>1600 @ stock
>Gigabyte AB350M Gaming 3
>TridentZ 3200 CL16 @ 2933 MHz

I'm not planning on overclocking at the moment, my RX 480 is bottlenecking everything regardless. Besides my current case is very narrow and I'm not sure if I should replace it or just get a smaller cooler that fits inside it.

>>61486444
What's your RAM model number?

>>61487741
https://www.amazon.de/gp/product/B0123ZBPDA/ref=oh_aui_detailpage_o00_s01?ie=UTF8&psc=1

>>61483782
>CPU
1700 @ 3.9Ghz 1.325 vcore, 4Ghz requires 1.45 vcore which i'm not keen on.
>Motherboard
Asrock Taichi using the latest AGESA 1.0.0.6a
>Ram
Reused from previous intel build. Corsair lpx 3000Mhz CL15 (Dual-rank, samsung B-die) Running at 2933Mhz 16-17-17-35 1.365V

Installing the Ryzen balanced windows power profile improved performance quite a bit in games and synthetic benchmarks. My previous build used an OC'd 6600k @ 4.6Ghz and it completely pales in comparison.

tfw pc retard with a 1600 and cant figure out how to clock my corsair vengenance 3000 ram
its running at 2133 now and when i go to bios and set it to 29xx it fails to boot and prompts me to change the overclock in bios

mobo is asus prime b350m -a

>>61487752
Danke

>>61486217
Well, finally some progress here.
If the cache itself is malfunctioning, it could spell trouble. Do we know if it's actually caused by the cache or another component?

File: 1484297239974.png (574KB, 605x558px) Image search: [Google]

574KB, 605x558px

I really hope they do a 7nm threadripper update. I'm considering going for the 1920X in the gamble that they'll do a revision on the same socket down the line.

Hope the gamble pays off. Anyone else planning on the goofy route?

1700 with Corsair led 3200 ram, can't for the life of me getting 3200. Only can go as high as 2933. Single rank :(

File: Ryzen1700.png (150KB, 962x420px) Image search: [Google]

150KB, 962x420px

Praise Ryzen

Am I the only one who bought the 1400?

What the correct procedure for ram overclock? Any guides?

>getting anything higher than 2666mhz

Don't fall for the ram speed shills, anything higher than that is just wasting your money

>>61487773
so it not boot at 2400 ?

>>61487773
Use the XMP preset and set like 2666

>>61488422
Did you have any trouble installing Windows 7?

>>61488659
Not with Ryzen, it's great for Ryzen.

>>61488784
Nah, it was fairly simple and easy.

Just a standard Windows 7 Installation with tweaking the boot drive with some USB drivers to enable it to work. Easy.

>>61488883
How exactly did you do it? I ordered a PS/2 keyboard and mouse, that works, right?

>>61483782
Will Ryzen ever got IGPU?

Next AGESA update when?

>>61489673

AMD are retarded for not including an iGPU on at least the R3.

Why the fuck would I buy an R3 for basic tasks if you need a dedicated GPU to go along with it?

With Intel all you need is a G4560 or i3 to do simple shit, and they'll also play 4k 10bit 60fps which is good enough for at least the next 5 years.

>>61489704
Because then why would anyone buy their upcoming APU line? Although I agree. I'm itching to build an AMD server to replace my current pentium/H97 itx build in my server now.

Should I use Winows' built in high performance power plan or AMD's balanced power plan when overclocking? And why?

>>61489704
I actually thought at first I could get away with buying the r5 1600 and then adding GPU when I have the money to immediately buy a better GPU.
Now I must include the VGA cost in my build if I ever want to use Ryzen.

>>61489774
I got a Nvidia GT210 for my little sisters build for $25 new. Can do 1440p displays and plays back 1080p content no problem. You can use it as a placeholder. And 90% of those cards are passively cooled too, so no noise. Plays her Minecraft and shovel Knight at 30-40-ish fps at 1080p.

>>61489704
You are going to have a rude awakening when Ryzen APUs release with 11 Vega CUs, which means at least 3 times the performance of current Intel iGPUs.

>>61489899
>which means at least 3 times the performance of current Intel iGPUs.

Nigga, I just need to play videos. If i wanted to play vidya I'd get a dedicated gpu.

>>61483782
Fuck off Pajeet marketer

>>61489911
Have you ever tried to use a laptop with a dedicated GPU?
Their batteries run out in less than an hour because of the massive power draw that a dedicated GPU causes

File: 3.9 4c8t 3066 14-16-16-16-34.png (129KB, 563x560px) Image search: [Google]

129KB, 563x560px

Anyone have any AIDA64 benchmarks?

Also post yours.

>>61489970
>Have you ever tried to use a laptop with a dedicated GPU?
No, that's for faggots.

Anyone running an ITX Ryzen build? I'm not massively bothered about cutting edge gaming so:

Best motherboard?
Best R5 bang for buck?
Best smaller ITX case?

>>61490664
Just get a mATX.

I have a ASRock AB350M(1x PCIe/2 RAM slots) and I haven't had any issues with it and it was only $63.

AB350M Pro4(2x PCIe/4 RAM slots) also has really good reviews for only 10-15 more buckeroos, although I think it's actually an oversized mATC and won't fit smaller mATX cases.

File: mpxr3fy.png (51KB, 518x794px) Image search: [Google]

51KB, 518x794px

>>61490730
can vouch for this motherboard, really good for the price

>>61485985
I'm pretty much you and just bought a 1700 with gigabyte k3 x370 and corsair 3200 16gb last night. I am still on a 1055t and 8by ram til it arrives but I've been waiting for 2 years to make the leap. Pull the trigger.

>>61490730
I hear ya. Specs are good but I don't really want a huge (relatively speaking) mATX case. What's the best compact mATX case? I don't need drive bays, not looking to dual SLI...

File: 11-208-059-01.jpg (20KB, 640x480px) Image search: [Google]

20KB, 640x480px

>>61491104
I'm enjoying my XION-310.

Have GPU prices been coming down yet? I was thinking of upgrading my 2500K but I'd also like to replace my R9 390.

>>61486217
oh boy this is going to be interesting

I deliberately hold off buying new archs for precisely that reason.
X299 turned out to be a disaster and Ryzen is still unsafe until this is fixed.
Glad I got a 6800K

File: 301.jpg (194KB, 1920x1100px) Image search: [Google]

194KB, 1920x1100px

>>61491104
fractal for no window, picrelated for window

>>61483915
I bought the aurous x370 gaming k5 on amazon prime day because I am stupid. Is this a good board? I have the r7 1700 and I am not going to overclock.

>>61489911
This. IGPUs are grossly overpowered for everyday tasks, but anemic for gayman and GPGPU jobs. The best solution would be to have an MCM solution with one option delivering 1 (one) CU for servers and workstations to just show something on the onboard video ports. This would free up a PCIe slot on entry-level servers while also removing the need for extra video chips.

>>61483782
Wait, so *ALL* mitx boards for AM4 are shit? What a bummer.

File: build.png (63KB, 1146x749px) Image search: [Google]

63KB, 1146x749px

>>61490664
im going with the biostar x370 board. asrock and gigabyte also released some boards as well. only thing that i'd watch out for is low profile cooler compatibility. Noctua is offering free am4 bracket upgrade kits for the l9i and l9x65, not sure about other coolers.

if you have money:
-ncase m1
-dancase a4 sfx
-evga hadron air/hadron hydro
-in win 901

if you don't have money:
-define nano s
-rvz01e
-node 202
-silverstone sg13
-coolermaster elite 140/110
-phantex enthoo evolv itx
-thermaltake core v1

>>61494625
>Fractal for no windows
You mean Fractal Design Define C mini? That thing is barely smaller than their ATX towers.

R1700
Asrock Ab350m Pro4
Corsair vengeance 3200, running @ 3192

I love Ryzen, lads

I probably will update when the third version is out

I want a small cheap ryzen homeserver, probably with 1600
What I need is
> ECC support
> at least three pci-e
> amd-v, virtualization support

I could finally retire my x5650 virtualization server if I'd get a good setup with ryzen

File: 1498671079654.jpg (84KB, 1280x720px) Image search: [Google]

84KB, 1280x720px

>>61495272
>I probably will update when the third version is out

>>61495176
>barely smaller than atx towers
that's what you get with matx. If you want a small machine you have to go mitx.

>>61495353
I mean Zen 3 (three)

>>61495176
>MEGA
>surprised when it's fucking big

Why should I get an X370 over a B350?

>>61495424
So like in 2019?

>>61495464
better sli support. That's literally it.

>>61495580
Great.

>>61495248
That's my setup too, had 3200 since agesa 4 as well.

Don't tell me you're running hawai as well?

>>61495464
https://rog.asus.com/articles/technologies/your-guide-to-the-ryzen-am4-platform-and-its-x370-b350-and-a320-chipsets/

It's worth pointing out that board manufacturers tend to put higher quality components and better VRMs on X370 boards. Also things like voltage measurement points, Q-code displays, on-board switches for various functions, etc.

>>61495540
Any problem you cunt?

>>61495346
I really think amd missed out on not having all 28 pcie lanes exposed on x370, would compete so well against Intel low end hedt and in small server spaces. There's enough for 24lanes in any configuration you want, an nvme m2 and the pch.

>>61495891
If i only intend to overclock a reasonable amount and nothing else special, is B350 fine?

>>61495949
Yep

>>61495453
?

>>61495959
Oh, I misread, thought it was MATX instead of mATX

>>61495952
How high should I be overclocking a 1800x?

>>61496045
1.4v

>>61496050
Isn't that awfully high? What clock speed?

>>61495899
ILL FUKKEN WRECK U KIDDO

Bet you'll wait for Ryzen 4 in 2019, fuck off

File: 1498069816639.jpg (110KB, 1199x900px) Image search: [Google]

110KB, 1199x900px

How would a R5 1600 with a reasonable overclock compare in single threaded performance to an i3 6100 overclocked to 4.6GHz? I want to upgrade to Ryzen for the multibread performance but don't want single thread regression.

File: 1484112145568.jpg (1MB, 3146x2358px) Image search: [Google]

1MB, 3146x2358px

>>61496045
AMD says anything higher than 1.425V for long-term use may cause premature failure. Personally I wouldn't even go higher 1.415V and that's if you measure that value from the board with a multimeter. The software readings from the sensor are fairly inaccurate and tend to under-report the actual voltage by around 60 mV in my experience.

The frequency you can achieve will depend on your chip's capability and when you hit that voltage wall. At some point it will start to take a lot (relatively) more voltage to keep an additional 100Mhz stable at 100% load. You'll have to determine where this is and how far you want to push it.

For testing CPU OC stability I recommend Prime95 Small FFT torture tests. I like to use in-place tests, 15 minutes per size, 1344 min and max. Run that for an hour and see what voltages you hit. FWIW I do basically the same thing when OC'ing on Intel platforms.

File: IMG_20170720_191204.jpg (3MB, 3024x4032px) Image search: [Google]

3MB, 3024x4032px

wish I had a computer to put this in, just gotta save up

>>61484042
AVX512, maybe?

>>61496754
Sleep with it at night like a good boy.

File: AMDramspeed.png (1MB, 1266x669px) Image search: [Google]

1MB, 1266x669px

>>61484042
Guessing better memory support
Higher base clocks and OC wall its nearly gone.
Better driver support out of the box
MOAR CORES
Single thread performance per clock is already very VERY good when paired with 3200mhz + ram
Problem is clocks are hardwalled at 4.0-4.1ghz and ram controller can be a bitch above 3200mhz ram.
Quad channel memory on TR might help this issue but I think the higher speed ram helps improve the infinity fabric. Ive read some people assume the ram mhz speed is the speed at which the design operates at. Slower ram = slower performance in single core applications as the cpu has to wait on the other cpus to catch up to what it just computed out.

>>61496754
give it lots of kisses

>>61496220
Whatever you can get out of 1.4v and your cooler.

>>61484182
yes they do for the following reasons
>games that can leverage the full 6/12 threads will push 50% more FPS this puts you ahead of an i7 7700k

and just as important

>games that can only use 4/8 threads will run faster and smoother because the last 2/4 threads will exclusively handle background system tasks and other open programs while those 4 cores will run the game exclusively without interruption.

People often noticed that even though the intel chips are technically faster in raw FPS the 6-8 core ryzens rendering games smoother than the i7 quad cores with less hitching/stuttering and this was because the 4 core intel chip had to run background tasks along with the game.

just installed "AMD Ryzen Master". Temp on low use is 41-45C . When i used AIDA64 i get temps from 50-70C swings. Not sure whats going on here.

1600X at stock speed. Noctua NH-U12S-AM4
Corsair Vengeance 16GB ram/ 3200 Mhz
MSI B350 Gaming Pro Carbo

>>61484094
Infinity glue©

>>61497701

>letting a utility full of malware do your overclocking for you

>>61497701
PSA for you and everyone else
Don't use Ryzen master. It is shit.
Do overclocking in bios and use HWiNFO for monitoring temps and voltages (for LLC)

>>61496045
>>61496660
Would not recommend going over 1.4V for any length of time

>>61497701
if you have a 1700X or 1800X your reported temps are deliberately shown to be higher than they actually are

Fairly new to CPU doo-hickeys. Got a build with a R5 1500x in it, gigabyte b350 mb. Tried overclock but honestly saw no difference in performance, did it from bios. Pretty sure I did it right also. Anyone else with this CPU and OC use? Maybe some numbers of what you have that works for OC and is stable for long periods would be super cool. Also, would getting higher speed RAM make that big of a difference? I have 8gb 2400mhz right now. Thanks!

>>61497939
depends on what you were using to test it
try a benchmarking program or some distributed computing like F@H or Prime95

>>61497939
and yeah, that RAM is going to hold you back a bit on overclocking

>>61498000
I'll have to give those a try. DESU I was just eyeballing my fps in games to see if there was really any difference. Does OC really put much of a performance boost for gaming on? If it's like 5 fps I'd honestly rather just run it stock than risk OC.

>>61498039
Would 3200 be a good area to hit? Or higher?

File: cputemps.png (38KB, 566x459px) Image search: [Google]

38KB, 566x459px

>>61497901
dont know if that also apply for the 1600X
>>61497806
Thanks for that info just looking at the readings and it dont look good i have been running the system for 3 months now with no issues with sudden shutdowns or freezing. But the temps seem way to high under low end use. Maybe i failed when applying thermal paste or something.

>>61497939
>8GB
>2400mhz

should have gotten at least 16 and 2667

>>61498098
That should say desu, i duno why it went to desu lmfao

>>61498114
Fuck, I've been bamboozled

>>61495844
Nah, a 1080 I got cheap after the EVGA explosion scare. It werks.

>>61498113
It was a first build, just need something to keep me under budget at the time. Planning to get 16 at 3200 soon

>>61498159
Unless they fix it in the meantime, be sure to get RAM that will actually hit those speeds with Ryzen

http://www.overclock.net/t/1627555/ryzen-memory-ic-collection-thread

File: am i cool.png (26KB, 400x400px) Image search: [Google]

26KB, 400x400px

Am i cool?

Gigabyte AB350m

>>61498110
If you subtracted 20 from your readings they would be right in line with what I would expect. I'm thinking your mobo does not perform the temperature offset so it's higher than the actual temperature. I would check the settings/manual and see if you can adjust it. I really wish AMD didn't do this bullshit with the temps.

>>61498217
OC that sucker a little more

>>61498234
Gets a little bit too hot, stock heat sink is pretty cheezy.

2933 Mhz Ram by the way.

>>61498256
Wraith Spire is really good for a stock cooler
How many case fans do you have

>>61498217
I seriously doubt that voltage reading is accurate, have you checked the voltage with a multimeter? I've had boards where CPU-Z reports 1.25V and the actual voltage is 1.36V.

>>61498213
Ohh this a neat list. So pretty I just pick up two of the 3200 8gb ones listed there and I'm good?

File: Speccy64_2017-07-10_18-18-08.png (49KB, 1074x761px) Image search: [Google]

49KB, 1074x761px

r8 my setup lads

>>61498384
Hold off for at least a little while
The mobo makers will probably push out a fresh BIOS update for Ryzen 3 chips, see what works after that

File: Capture.png (130KB, 971x681px) Image search: [Google]

130KB, 971x681px

>>61498266
4 Case Fans, it does well, but it will get hot under full load.

>>61498369
Yeah it does seem kind of low.

>>61498489
Will do, I'll keep an eye out. Thanks for the help

>>61498510
4 isn't nearly enough
Make sure you've got a fan on every available mount in your case

1800x @ 4.0ghz (1.4v)
Asus Crosshair VI Hero (AC-Wifi)
G skill trident Z 3200mhz x2 (16D) @ 2933mhz

Kind of pissed about the OC voltage. I'm using a corsair h110i and it keeps it idle at about 40 and load at around 68ish, but I've seen other people with 1800x do 4.0 and 4.1 for much much less.

>>61498546
#SiliconLottery
still, you got it to 4.0, which is good

Is Zen2 still schedule to hit the market by the end of this year?

>>61498615
>AMD undercutting their shiny new flagship product in less than a year
Summer 2018 at the earliest

>>61483782
Gigabyte Gaming original AB350
R7 1700@ 3.6GHz stock vcore and settings
Corsair LPX Vengeance 3000 @ 2933 XMP set

BIOS is v5 and it tok latest update or XMP to "Just Werk"

Noctura AM4 is in the mail but I got the AM4 kit form Noctura so I have to wait 2 weeks prob for international shipping and my Noctura D15 is going to be here tomorrow. So sad. My goal is 3.8GHz stable cool temps.

>>61498546
Don't sweat it too much, anon. 100MHz won't make a substantial difference either way.

>>61498663
I bought the Noctua AM4 kit from Amazon and they sent me the wrong one, it's for the UxS models or something. Sending it back and getting (hopefully) the right one from Newegg.

>>61498229
The Tdie reading is the cpu temp with the offset subtracted.

bump or else Intel wins

File: Screenshot_20170722-162440.png (368KB, 1080x1920px) Image search: [Google]

368KB, 1080x1920px

Thinking of doing a 1700 build with a Taichi. Will this Corsair RAM hit 3200 with the latest BIOS?

>>61501914
http://www.overclock.net/t/1627555/ryzen-memory-ic-collection-thread

>>61486217
>>61486217
>but this could easily impact AMD's finances
nobody cares about linux for the desktop you unemployed autist

>>61490026
what the fuck those speeds are terrible, like the L1 is half the speed it should be running at

File: 1500705458[1].png (297KB, 829x797px) Image search: [Google]

297KB, 829x797px

>>61501964
for reference 1800X @ 4.025GHz/3066MHz memory

>>61493270
>I'd also like to replace my R9 390.
why? its perfectly good for now even at 1440
the only thing i'd replace it for is vega if any good or straight to volta

Any 1700 users here?
How high did you OC at 1.35V; AMD says not to run it higher than that for the long-term

The RAM I got is Hynix M-Die. That's good, right? Will I be able to get the full 3200 on a B350 MOBO?

>>61502182
>AMD says not to run it higher than that for the long-term
wrong they say that for 1.425V

>>61502200
no Hynix M-die is like worst case scenario

>>61501928
Cheers mate

File: comedy gold xd.jpg (39KB, 576x432px) Image search: [Google]

39KB, 576x432px

>>61502200

>>61502221
>>61502241
The RAM guides say it's good.
>>61501928
>Hynix A/M-Die are mostly working @ 3200 MHz with AGESA 1.0.0.5+.

If I'm using an unsupported OS (Windows 7/8.1) how do I access the "boost clock"?

>>61502257
Maybe Single Rank but I've seen plenty of people struggling to get 3200 even on SR Hynix dies on AGESA 1.0.0.6 BIOSes

File: 1499836471639.jpg (1MB, 2048x1536px) Image search: [Google]

1MB, 2048x1536px

I'm looking forward to Zen 2 but it's making it difficult for to pick which CPU to buy for now.

>>61502262
magic

>>61502268
Guess I'll just have to try it and see. If it really doesn't work, can I just wait for an update to fix it?

do not get this combo, there is a high chance that it will cause many problems with booting, no higher than 2133mhz clocks of ram and bluescreens:
>MSI B350 PC MATE
>Ryzen 1600
>G.Skill Aegis (F4-3000C16D-16GISB)
the funniest thing is that these are supported rams on MSIs list

>>61502285
You can hope but with the amount of tweaking you can do with 1.0.0.6 I can't imagine they can do much else with BIOS updates to fix it but I would hope it does, I have Dual Rank M-die which is by far the worst case on the platform and I have it running at 3066MHz, so I'd love to get it at the rated 3200MHz and stop there.

>>61502272
if you're already thinking upgrade now, you really can't wait for Zen2
Jump in now, and then look for Zen3

>>61502304
>Dual Rank M-Die at 3066MHz
That's good enough for me desu
Thanks for posting this, you've made me feel better.

>>61502317
Can't guarantee it on a B350 mobo, I have a Crosshair VI Hero which right now it best case for memory overclocking.

>>61502325
Honestly I'm coming from a 2010 Inspiron 570, I doubt it's even possible for my new build to disappoint me.

>>61502341
Sure just don't go with high expectations for memory overclocking, X99 is still garbage at this and getting over 3500MHz is a stroke of luck so the platform definitely needs to mature.

>>61502352
The RAM is meant to be 3200, I don't think I have to OC for that.

>>61502433
Anything over 2133MHz is an overclock. Memory rating is the same as something you'd get from here: https://siliconlottery.com/collections/all/products/1700a39g

It's 'rated' at 3.9GHz @ 1.4V but that's with their testing.

>>61502442
Alright then, thanks for explaining. I think I'll just live with whatever I get and maybe upgrade RAM/MOBO when Zen3 comes out.

>>61490787
>1.424v
>only 3.8ghz
you lost silicon lottery

>>61490787
>1463MHz
>1.424v 3.8ghz
wew

File: le beckenbauer face.jpg (27KB, 514x297px) Image search: [Google]

27KB, 514x297px

>>61490787
>over 1.4V
>not even 3.9GHz

So for gaming, is the 1600 the best bang for buck?

More Ryzen ITX boards are finally on the way, i'm leaning towards the Gigabyte personally.

>>61502639
>ITX
why

>>61502607
yes

>>61502744
Because small PCs are cute~

Is it worth upgrading from a i5-4670 (non k) to Ryzen 5 1600?

>>61502607
No, a Pentium G4560 has a higher performance/price ratio.

>>61503016
you mean that cpu intel restricted sales on because it was killing i3 sales so it's harder to come by and its price went up now?

File: 9ce97019a934d8348504e058dba0be67.jpg (64KB, 564x465px) Image search: [Google]

64KB, 564x465px

>>61483782
I've been out of loop with cpu's for a while, can someone give me a redpill why the i9 series suck and why is Ryzen so good?

>>61502607
1600 or 1700
But basically, just don't even bother with any of the X-numbered processors
Unless you lose hard at the Silicon Lottery, you can OC most of the gap for much less money

>>61503073
The i9 series is much faster but in video games the CPU performance is much less important than the video card so it will give the same FPS as a much cheaper Ryzen in video games.

>>61503134
>The i9 series is much faster
not really it's a regression vs kaby lake

anyone with Corsair LPX 3200 16Gb ver 5.39?
how does it do? scared to try XMP

File: intel_core_i9_7900X_33-665x460.jpg (47KB, 665x460px) Image search: [Google]

47KB, 665x460px

>>61503168
The slowest i9 is only slightly more than twice as fast as the fastest kaby lake.

>>61503208
oh i thought you were talking about gaming since you claimed they were better than tryzen

>>61483782
Need to build a new PM current one is dying

Is threadripper better for gaymen/faster core for core? Ie. disregarding the fact that it has more cores

Been out of the loop

R7 1700 at 3.6GHz stock voltages
Gigabyte AB350N-GAMING WIFI
32GB Corsair LED 3000 at 2933MHz

I'd mess with the voltage but a cheap cooler and a Node 304 means it wouldn't last very long.
Absolutely flies through Revit nonetheless. Coming from a 3770 it's night and day.

Board:
AsRock Fatal1ty X370 Gaming X
CPU:
R7 1700 @ 3.8Ghz (AiO cooled - max 60°C under heavy load)
RAM:
16GiB Corsair Vengeance 3200mhz CL16 @ 2993Mhz

Ram is not going over 2993 - at 3066 it boots but bluescreens on torture test, on 3200mhz it doesn't post.
Performance is super solid at 2993mhz, so I don't really mind - and I guess a future BIOS update may deal with this. Admittedly, I was uninformed about the fact that the RAM was hynix and not samsung.

The system is incredibly fast and stable and at idle it is almost completely silent. If a video is running, you can absolutely not hear the system at all.

I came from an i7-2600k system from 2011 and I definitely notice that the ryzen system is way faster in most tasks.

>>61498546
Most of the fags running high frequency combined with very low voltage didn't properly stress tested and are also probably running some old AGESA.
New AGESA and prime95 version 29.2 are much much harder to pass.
4 GHz 1.4 volt is a very nice result, not many chip can do that.

>>61503269
Threadripper is beyond unnecessary if you're just going to game with it

>>61503541
>using prime 95
You want your chip to die?

>>61503280
Undervolt it

>>61483782
AMD Ryzen 7 1700
GIGABYTE GA-AX370-Gaming
16GB of Corsair Vengance LPX DDR4 @ 2400MHz 16-16-16-39

Verified safe

>>61503600
No, I just want it to not crash when running 4 or 5 hours encoding or compiling big stuff (LineageOS).

>>61503627
You shouldn't use prime 95 for stability testing, I heard newer versions can damage your chip, even Linus said it.

Use AIDA64 or realbench

>>61503639
That does not make any sense at all, not that I expected anything more from that linus faggot.
How can a program damage a CPU, or anything else for that matter?

File: ryzen overclock.png (473KB, 1289x1079px) Image search: [Google]

473KB, 1289x1079px

1700 overclocked to 3.8ghz via pstates
Crosshair hero
Corsair vengeance 3200mhz currently running at that despite not being on QVL

Just need vega to be good so I can go full-on with a 4k build.
Still trying to dial in the final settings for my overclock(s).

>>61503695

Prime95 can fuck cpus generally through the motherboard - namely aggressive LLC can easily push motherboards to pull way more juice than they (and the cpu) can handle.

>>61503754
>Vega
>4k
hahahahahahhaha

File: assrock.jpg (200KB, 1280x960px) Image search: [Google]

200KB, 1280x960px

ASUS Prime B350-Plus or ASRock AB350-Pro4? On a pure quality perspective.

I've always been an ASUS guy but I heard they lost the lead in Ryzen, and ASRock always seemed like poormans ASUS.

>>61503622
Why skimp on the RAM, brah

File: 1432506504306.png (128KB, 1280x892px) Image search: [Google]

128KB, 1280x892px

I have to run 1.4V to hit 3.9 FEEL MY PAIN.

how do you guys determine OC stability?

some OCs i cant even run cinebench without crashing but i can get months of uptime with various realworld tasks and other benchmarks, it's just CB that crashes it

>>61503831

1.4v is still under banic amounts of voltage so as long as you keep the chip cool its not a problem.

File: IBT high.png (291KB, 1465x889px) Image search: [Google]

291KB, 1465x889px

>>61503840

A proper stress test that hammers the cpu until it looks like a victim of a /d/-related doujin. The likes of realbench, AIDA64, Prime95, Intel burn test, OCCT and whatnot are all viable cpu stress tests.

>>61503754
>Prime95 can fuck cpus generally through the motherboard - namely aggressive LLC can easily push motherboards to pull way more juice than they (and the cpu) can handle.
Motherboard manufacturers using cheap components and making design errors is a thing, prime95 killing perfectly fine and good engineered hardware is a whole different story (that did not happen).
Following the same reasoning it is also risky to run folding@home, x264 encoding, compiling large programs, doing rendering, manipulating camera RAW files, etc.
Luckily I have a decent motherboard who can sustain pumping 200+ watt into the CPU without blowing up.

>>61503831
1.4 volt prime95 stable is not a bad result at all.
I guess a big chunk of ryzen CPUs can't do that.

>>61503771
Cons
- 3 + 2 Phases
- Bios chip is soldered
Pros
- Bios chip has UART header
- Heatsinks look and feel decent

everything else is pretty much standard for all motherboards.

File: 66f.png (581KB, 1145x635px) Image search: [Google]

581KB, 1145x635px

>set Performance mode in windows setting as AMD recommend
>CPU now barely underclock when idle/shitposting on the internet, but also doesn't overclock itself to 3.79 anymore

Wait what the fuck?
Should I put the balanced mode back up? Or is it gonna mess with my future overclocking? Even when/if I overclock, I still want the dynamic boosting and underclock (how do AMD call that again?)

I really like my cpu going under 1.5Ghz when I'm not doing anything, and nearly 3.8 on 2 cores is really all I need.

>>61503906
YOU HAVE TO GO BACK

>>61503906

Make sure the minimum and maximum processor states in the power plan are set correctly.

>>61503906
power gating still works, difference is less than 10w
temps should also stay the same

>>61503935
>>61503947
Oh, thank you. Seems to work now.

>>61503969
https://www.youtube.com/watch?v=bzxn1hS7Nq4

more extensive explanation
tldr you can run it 1.4v 4Ghz constantly and it almost won't draw more power than 1.2Ghz 0.8v at idle

>>61503985
because C-States are still working

>1700x@stock
>C6H
>3200MHz CL14
In single threaded operations it loses about 6% to
>[email protected] (1.05V vcore; stable for 48h prime small fft test using avx)
>P8P67(B3) Deluxe
>1600MHz CL9
And it loses horribly to
>[email protected] (0.97V vcore; stable for 48h prime small fft test using avx)
>P9X79-E WS
>1600MHz CL9 (quad channel)

>>61484042
>7nm?
Where'd you get the idea that zenver2 will be on 7nm manufacture?

>>61503985
Damn. Well this convinced me to overclock after all. I'll pay attention to the temps at idle myself to be sure (running in an mATX case), but yeah.

>>61502883
>I want to pay more for less features, less expandability and lower life components.

>>61498395
>2133 MHz RAM speed

If I'm going to use Win7/Win8.1/Linux can I use the boost clock setting?

>>61504197
>Win7/Win8.1
They still have telemetry m8

>>61504231
Not unupdated Enterprise versions. Also they're better OSs regardless of telemetry.

Can you answer my question or not?

When is Threadripper coming out? I'm genuinely interested to see how well it does in game. Of course I'm not expecting it to do better than say the 1800x but if it does as good as it it'd already be super impressive in itself.
The thing has 16 cores 36 threads at 3.5ghz and 4ghz turbo, if it works well there's no way you'd ever have framedrops or stutters, and imagine how well that beast could stream.

I've been having problems with my current machine (phenom ii which i bought some time in early 2011) overheating and shutting itself off the last couple weeks, will i have the same problems during summer with a ryzen machine

>>61503761
>Trades blows with full gp102 in geometry workloads
>Not capable of 4k
?

File: woooooooow.png (63KB, 321x357px) Image search: [Google]

63KB, 321x357px

>old PC shat the bed a few months before official Ryzen announcement
>got a super deal for a used PC with an i5-4570s
>want to upgrade to a 1600x
>it alone costs only ~35 € less than what I payed for the whole computer (sans GPU)
>mfw I don't have an excuse to upgrade until at least Zen2

Guess I'm stuck being a corelet for over a year longer.
Should I at least stock up on RAM while waiting? What are the chances that prices of RAM will rise faster than their quality? (I wanted to buy Goodram, but they don't even sell DDR4 any higher than 2400Mhz)

>>61504778
No.

>>61501914

Did you not read the OP?

That is single rank ram, which performs worse than dual rank. A dual rank at 2933MHz is faster than that one you listed at 3200mhz. The difference is slight, but if you're asking anyways go get dual rank.

>>61505234
How do I check what rank my ram is? I bought my stuff without even knowing about this

>>61504038
What will it take to get this fundamental concept through your head.
Ryzen has more cores at lower single threat IPC.
This was never a secret.
If you only use single threaded applications then by all means buy a cpu with stronger singke core IPC.
Some of us have work to do and don't exclusively use our pcs for muh childrens gayms.

>>61505313

Google, there are a few Reddit links which are legitimate and associate model numbers to dye type.

>inb4 leredditmemes

File: 3.7 8c16 14.png (130KB, 579x574px) Image search: [Google]

130KB, 579x574px

>>61501999
I was meming with 4+0 downcore.

File: 1496944759661.webm (3MB, 360x202px) Image search: [Google]

3MB, 360x202px

>>61503639
Amd engineers recommend Prime95 for stability testing. It does stress your CPU more than normal workloads but that is the point of the test

>>61503754
>LLC
There's your problem.

>>61504360
Wait until you see Zen2's Threadraper

>>61503985
It does make some difference
Not much, but its there
Like 14W at idle with p-states, 19W without

>>61505345
I just googled my ram instead.
It's dual rank! I'm so lucky, I honestly just purchased whatever DDR4 was cheaper when I bought all the parts

>>61503208
>threadripper breaks 3k

>>61503627
then test it under the worst loads it would be subjected to in real use

>>61502574
>>1463MHz
It means it's 2900
lrn2RAM

I keep hearing that if you go ryzen you will need atyleast 3200mhz RAM otherwise it will lack performance is this true?
If this is true the price difference between a full intel or amd system will be a lot smaller.

>>61507191
2x16 GB crucial 2133 cas15 currently running at 3200 cas16 without issue here, and this shit is even dual rank.
Paid 220 € for the kit.

>>61507269
What the hell, it can go that high?
Is there ANY ram OC guides for ryzen? I never oc'd RAM before.

>>61507511
Yes they do, at least my sample.
The trick is VDIMM 1.4 volt, Proc ODT 60 ohm, tRFC ~512 and tCWL=CAS (or CAS-2).
I don't think there is a single decent overclocking guide for ryzen, like there is not a single forum thread with a user overclocking results chart.
All I see are faggots posting half an hour of realbench, aida or some other bullshit and call it stable.
I have the feeling most of the overclock you see are not really stable anymore with new AGESA (and maybe also summer temperature).
Even Chew* (who is an entitled cunt BTW) stated more than one time that new AGESA made overclocking much more difficult.

>>61507662
Thank you. I also have a crutial kit (2444) so that's reassuring

I didn't update the bios at all so I doubt I have the latest AGESA.

>>61507269
>>61507511
>>61507662
>>61507912

Keep the silicon lottery factor in mind

>>61488901
Windows 7 has PS/2 drivers so it should work.

>>61508735
Perfect, now I just have to wait for my SSD.

I was thinking should I sell my 4590 to my brother and try out my luck with silicon lottery with the 1600?

I'd be interested in vt-d passthrough and virtualization: this would be a combined workstation (linux) and gaming (virtualized windows) setup. I have the 4590 setup currently, but would like to get more cores for guests.