ITT we agree that Ryzen is great

tru dat

hoping to build a 2600 rig next mo for my bday

Currently rocking a 6core Athlon. Looking forward to being the proud owner of an 8 core Ryzen in a couple weeks.

"""6 core"""

what? my cpu currently has 6 cores

WHEN THE FUCK IS ZEN 2 COMING OUT? I WANT TO UPDATE MY PC SO FUCKING BAD AND DITCH INTEL ONCE AND FOR ALL

This

>mfw I built 4 ryzen PC in the last 6 months and all are fine.

Attached: 1472592191749.jpg (1273x1640, 651K)

Just bought a new 2700X for $275, first AMD chip since Athlon 64...I expect great things.

I'm still using an old Core 2 Duo system and I'm thinking of building a Ryzen APU rig. Should I wait for the next iteration or buy now?

Samples late this year. Retail probably the first quarter next year.

But how well can it run the best os (macOS)?

Based 8400. Best purchase I've ever made.

The EPYC2 parts were set to sample at the end of this year, but apparently some of the big OEMs already have samples in hand.
No word when the consumer chips will launch, but its likely AMD will hold a massive press event about it early into 2019.

Su is really big on press. Not many people know but as soon as she took over as CEO she personally set up a couple of AMD's larger media gatherings.

then you clearly haven't been using it on GNU/Linux because Ryzen has had some really serious issues like random hangs when idle. You can still reproduce this by buying an old board with an old BIOS (I guess you need a 1000 series CPU too) & installing Linux and waiting 10 minutes. It'll hang. There's a pretty long kernel bug thread about this, it's very well documented. I had to use some kernel parameters to work around this for half a year or so until AMD finally pushed a BIOS fix.

no denying Intel usually wins in man-child benchmarks.

you're confused. A six-core Athlon's got six cores. You're thinking of the bulldozer chips

just buy now if the current one is too slow. there's always going to be some new thing around the corner

Are you referring to Intel's market share and stock?

Attached: 1536757163436.png (672x794, 422K)

The sad part is these are cherry picked benchmarks at 1080p or 720p. In real world performance the 2700X at stock settings is only ~5% behind the i7 8700K. Wow champ you got 2-4 more FPS than AMD and all it took with the most expensive AIO cooler you could find. lmao.

Attached: perfrel_2560_1440.png (500x970, 49K)

It is, I'm working on replacing two other PC's with Ryzen and next year after tax return gonna scrap my xeon home server for Epyc.
If you disregard dumb fanboy kids this is the best thing that happened in cpu market for years. And I do hope Intel makes something better, and I do hope AMD will return the blow. I will just buy the best thing considering performance / value / power draw.
Funny but since this is "technology board" shouldn't people be happy there's competition so performance will go up and prices down?
Seriously motherfuckers, all of you saying I have Intel what are you poor?
Can't you like... You know... Buy better thing after that?
And better thing after that, no matter the company?
Or... In the end... You are so poor to afford status symbol of the month and everything that's left for you is shilling so hard to kinda drown that buyers remorse? Cos mama ain't gonna buy you a new one?
Faggots this is great timeline, technology is moving forward. King is dead, long live the (new) King.

>intel 14nm shortage till mid 2019

Attached: 1530715628293.jpg (653x477, 88K)

prices for 8700K have already increased 100 bucks in Sweden. It costs 510 USD now (25% VAT included)

9900K is gonna cost like 800 USD here rofl

Here's a graph from big german retailer. Looks pretty much the same everywhere else

Attached: rqaGg9e.png (1500x650, 118K)

Based 8400 did it again. BTFO Rypoo garbage twice the price.

since you bring up the topic of benchmarks.. if the game runs just fine at 60 fps with stock settings then the CPU is probably good enough. The GPU too. Yet benchmarks are like oh 170 vs 173 fps at ultra high detail. all that doesn't really matter. besides, very few are going to buy a 1080 anyway

you seem a bit ignorant of what is going on with Intel right now. Intel CPU prices in Europe has increased 50-60% since early September. We're talking 1600 SEK for the Ryzen 1600 vs 2800 SEK for the Intel i5-8400. Intel actually nearing twice the price for the same performance right now. Also, see

Yea... no.

Attached: Capture.jpg (1369x390, 95K)

Mine has been great

Attached: battlestation_for_4chan.jpg (1632x1224, 473K)

This kills the Ryzen.

Attached: adobe_65275574_premiere_pro_cc_download_1293105.jpg (1000x1000, 41K)

>1.2% advantage
Yeah man, AMD is discussing bankruptcy as we speak.

Idgaf I bought the 8400 on sale for roughly 160$ 8 months ago. This should keep me up until 7nm.

1.2% less performance
100% more expensive

that's a really nice setup, user, nice desk and the chair looks pretty comfy. is it comfy? I need a new chair

Absolutely comfy. I weigh 250lbs but I bought a Big and Tall chair rated for 500. I feel like a king whenever I sit down

I've literally never had that issue and I've had Ryzen since launch.

>Adobe
Enjoy having 9 processes runninng all the time even if you disable all autostart adobe processes

real men use DaVinci Resolve

You are objectively wrong. There have been multiple issues with Ryzen's first generation and crashing. There is of course the infamous segmentation fault problems under high load affecting almost all early processors in batches of Ryzen, but there was also long-running problems with random freezes and lockups on Linux and even some Windows machines caused by deep sleep states after idling.

Not all CPUs were affected. I had to RMA two Ryzen 1700X due to the segfault problem, and both of them locked up completely when idle unless I disabled C6 state in the BIOS (for many people, even this fix did not work).

The ones that AMD gave me as replacement do not have the segfault problem, and they also don't seem to have any problems with C6 state. I do still occasionally get random unknown NMI from the processors, but that might be HPET.

0.05 shekels have been transferred to your bank account

We are not the same people, and why would a shill have a Ryzen?

$ lscpu
Architecture: x86_64
CPU op-mode(s): 32-bit, 64-bit
Byte Order: Little Endian
CPU(s): 16
On-line CPU(s) list: 0-15
Thread(s) per core: 2
Core(s) per socket: 8
Socket(s): 1
NUMA node(s): 1
Vendor ID: AuthenticAMD
CPU family: 23
Model: 1
Model name: AMD Ryzen 7 1700X Eight-Core Processor
Stepping: 1
CPU MHz: 2011.427
CPU max MHz: 3400.0000
CPU min MHz: 2200.0000
BogoMIPS: 6786.64
Virtualization: AMD-V
L1d cache: 32K
L1i cache: 64K
L2 cache: 512K
L3 cache: 8192K
NUMA node0 CPU(s): 0-15
Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl nonstop_tsc cpuid extd_apicid aperfmperf pni pclmulqdq monitor ssse3 fma cx16 sse4_1 sse4_2 movbe popcnt aes xsave avx f16c rdrand lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw skinit wdt tce topoext perfctr_core perfctr_nb bpext perfctr_llc mwaitx cpb hw_pstate sme ssbd sev vmmcall fsgsbase bmi1 avx2 smep bmi2 rdseed adx smap clflushopt sha_ni xsaveopt xsavec xgetbv1 xsaves clzero irperf xsaveerptr arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold avic v_vmsave_vmload vgif overflow_recov succor smca

Attached: 2018-09-28-201531_450x109_scrot.png (450x109, 12K)

I love Ryzen, even though I only have an R3 1200.

See

>why would shills try to find faults in their competitors product by testing it themselves

I mean, it's a working CPU for cheap. What more do you want?

Attached: 0.jpg (1500x650, 67K)

>This should keep me up until 7nm.
Glad you're going AMD, nobody should be forced to use Intel SecutryHoles™

Attached: 1538123630423.jpg (740x416, 37K)

I love how desperate Intel fanboys crash every thread even related to CPUs nowadays.

ryzen is bad goyim

Attached: 1514987452922.jpg (700x700, 58K)

Thrown in the dumpster again

Attached: DELID.png (1064x594, 256K)

It really isn't shocking that two different people with similar experiences running Ryzen on GNU/Linux show up in a thread about Ryzen when there are big common issues. As I said in there's a really long kernel bug thread about the idle bug since it was initially assumed it was a kernel problem (not a Ryzen problem) so people kept on discussing it there.

The segfault problem with early chips ain't a secret either and AMD did agree to refund when people complained.

It's not exactly shocking that people who've had a Ryzen since launch know what the problems early on were

Attached: neofetch-taeyeon-2018-09.jpg (1326x620, 205K)

>This should keep me up until 7nm.
Glad you're going AMD, nobody should be forced to use Intel SecutryHoles™

Attached: 1538159177414.jpg (740x416, 37K)

So you admit that the faults are real kek. What's the problem with pointing them out? They affect real people, you know.

GNU/Linux is a free and open source kernel and operating system.
It's the users who should fix the problems they have with it. GNU/Linux comes with ABSOLUTELY NO WARRANTY.

>no mention of the better motherboard, liquid cooler for the OC and the expensive B-die RAM
Every time. Every single time, this picture is posted the shill hopes nobody will bring this up so they can just fool somebody who just doesn't know any better.
Amd fags are as bad as intel shills, except they do it for free.

>Host: Gaming
>B-But I *need* to run Linux for muh gaymes
Absolute state of freetards.

>I prefer using this translation layer/wrapper or running Windows in a VM on Linux for a computer I use for gaming with performance penalties and unneeded problems because muh freedoms
yes :^)

>reeeeeeee why didnt you point out that intel garbage cant be overclocked

Attached: 1499960238314.jpg (960x878, 125K)

Because it doesn't cost users extra to get the stock performance.

Exactly a 3fps difference isn't even worth mentioning, especially when it very easily be a slightly biased test, and with video games, the benchmarks are always so inconsistent, you could run it once and get 40, and again and get 80. People need to buy for their use case, and price to performance. And when I say price to performance I mean the cost of motherboard and coolers and shit too. Comparing cpus for gaming is a joke, buy a better gpu.

I use Adobe specifically for certain proprietary features and plug-ins. I run Linux too so I have to dual boot just because of Adobe trash. I'd love to use DaVinci resolve, but it's Linux version is garbage, and it's missing features that Adobe isn't. If black magic put more work into adding support for other niches other than studios I think they'd get a far larger audience. I really hate using Adobe products, but until there are actual viable competitors to the level that blender competes with autodesk, Adobe will continue to be the leader, especially in video.

>It's the users who should fix the problems they have with it
no, not when it's a hardware bug or a BIOS bug. In case of the Ryzen idle bug we had to work around that in software until AMD finally, half a year later, came with a proper firmware BIOS fix. Hardware/firmware problems are on them and it's their responsibility to fix that.

yes that is the GAMING motherboard. With LEDs. It also happens to have dual 1220 audio which is useful and dual gigabit ethernet which is also quite useful (bonding).

don't do much gaming beyond Warzone and CS:GO from time totime, never been that interested in entertainment for children (and men who never grew up). Don't actually have a wintoddler WM, can't say I need it.

Making a side build
Not going for crazy OCs or anything but it may be running as a Plex server mainly but gaming on the weekends.
2700x vs 8500k?

That other guy with a Ryzen 1700X. Do you have ECC RAM?

Like I said before, the best we could hope for is that it made Intel step up their game and start trying again.

I must admit I didn't expect Intel to have become so complacent while AMD were asleep at the wheel that they would actually shit the bed when waking up again, rush some half-baked SKUs out, botch an entire process node, be revealed to have cut so many corners that it had a security impact beyond the corners everyone else cut, and wipe this much off their stock price.

My new server does not have a Xeon in it, because Epyc has more PCIe lanes. It's not quite like seeing the Athlon 64 vs the Pentium 4 again, but it's definitely interesting.

>then you clearly haven't been using it on GNU/Linux

Attached: yasunasmug.jpg (447x400, 23K)

have you checked prices recently?
its about the same price as a 2700(nonx)

>no problems on Windows
>It's a hardware problem that only appears on Linux so it's their fault!!!11

I have 1600 and it hangs when surfing/reading pdfs in Ubuntu... But never while gaming.
How do I find out if I'm affected?..

Attached: 1502261452790.png (1920x1200, 655K)

fools!! consumers get leftover products while big bro is fostering all the S class tech.. point is all these newer models of cpus are just repainted andrenamed to feed the very "needy" and inexperienced.. and that intel chip you just adopted? its a piece of leftovers from 2007

Attached: tumblr_lxagxpS2oJ1qg2ql3o1_540.jpg (540x405, 54K)

An x86 CPU should be able to execute valid code that works on literally any other x86 CPU without crashing. You are retarded.

The reason Windows did not show even nearly as many problems with it is because the ASLR implementation and CPU scheduler on Windows is different. People reported massive stability improvements after disabling ASLR on Linux, and it's not a fault of the ASLR implementation. Regardless of that, people were able to reproduce the issue on Windows:

github.com/corngood/kill-ryzen-win

It also crashes on Windows during some particular workloads, like H.264 and H.265 workloads.

The early Ryzen CPUs literally mangled memory addresses during load, causing ret instructions to jump to complete garbage addresses, and causing data loads from memory to access complete garbage addresses, resulting in invalid opcode errors and segmentation faults.

The problem is with the CPU, and that's why AMD instantly gives you an RMA if you're affected, because they know it is.

As for the lockups during idle, I can absolutely guarantee you that this happens under Windows, but it's much rarer because Windows wakes up the CPU cores a lot more than Linux does, and so the CPU rarely enters that deep of a sleep state.

If it's freezing when idling or not doing any heavy work, you probably have the C6 state problem. Try updating your motherboard's UEFI and if that doesn't work, disable C6 state. If that also does not work, set rcu_nocbs on the command line.

For your 6-core Ryzen:
rcu_nocbs=0-11 processor.max_cstate=1

And for a 8-core Ryzen:
rcu_nocbs=0-15 processor.max_cstate=1

You should be able to find a lot of information on it with just searching for "linux hangs ryzen idle" or something like that. Note that the C6 sleep state hangs are not at all related to the segfaulting.

As for the segfaulting, run this script and it should reproduce the issue within 10 minutes:
github.com/Oxalin/ryzen-test

And there's also: github.com/hayamdk/ryzen_segv_test

ryzen_segv_test takes much, much longer to reproduce the issue than kill-ryzen.sh in my experience, but is 100% an indicator for a problem with your CPU and requires far less RAM. If ryzen_segv_test crashes and dmesg looks something like:

[96962.978425] ryzen_segv_test[26723]: segfault at 3da7a ip 000055920e703d20 sp 00007ff30f318f10 error 6 in ryzen_segv_test[55920e703000+2000]
[97425.192645] ryzen_segv_test[27659]: segfault at 228041e8 ip 0000561a78c37d2c sp 00007fae0c9aaf10 error 6 in ryzen_segv_test[561a78c37000+2000]
[97603.109698] traps: ryzen_segv_test[13348] trap invalid opcode ip:55ce2f9f4d41 sp:7f31da654f10 error:0 in ryzen_segv_test[55ce2f9f4000+2000]
[97701.410617] ryzen_segv_test[8823]: segfault at 78f19d25 ip 000055bf2822dd2c sp 00007f171c915f10 error 6 in ryzen_segv_test[55bf2822d000+2000]
[99709.508601] ryzen_segv_test[26138]: segfault at 530db1b2 ip 00005653f233bd31 sp 00007f7b2bb42f10 error 4 in ryzen_segv_test[5653f233b000+2000]
[100423.185795] ryzen_segv_test[1611]: segfault at ae2164d4 ip 0000564dd460bd2c sp 00007f83d3fa2f10 error 6 in ryzen_segv_test[564dd460b000+2000]

Then you should probably RMA your CPU.

By "the command line" I mean as a command line option for the kernel, set in grub.conf.

Also, in my dmesg log there (from a long time ago, before I RMA'd) you can quite clearly see what happens. A faulty CPU completely mangles memory addresses:

>3da7a

This address makes absolutely no sense for any modern operating system.

2400g here
imagine buying a dedicated gpu in 2018

Arigatou user, gonna run the script and write back

I only have 8 geegs of ram though...

>And there's also: github.com/hayamdk/ryzen_segv_test
>ryzen_segv_test takes much, much longer to reproduce the issue than kill-ryzen.sh in my experience, but is 100% an indicator for a problem with your CPU and requires far less RAM. If ryzen_segv_test crashes and dmesg looks something like:

Run this overnight, while you're sleeping. If you come back ~8 hours later or something and there's no problems logged in your terminal nor in dmesg, you are fine.

If not, I would highly suggest RMAing your CPU, but obviously that will leave you without a computer for a while. That said, the RMA is quick. I am Norwegian, and the RMA address is in the Netherlands. I RMA'd two Ryzens and for both of them the RMA process took only 4-5 days before I got a replacement.

I hear good words about the process in the USA too, but for that I have no experience.

I'm Russian and I will have to RMA it in the hardware store I bought it. And probably won't be able to explain it to people that work here... But still I am curious if I'm affected.

Nvm, I googled and the RMA is global

Is it? I'll know in a few years. My 8350 and 6350 still suit my needs.

pls stahp

Attached: 488974844215.png (1093x754, 72K)

I see, I'm sorry I can't provide any advice for people living in Russia, but you should be able to RMA it directly to AMD regardless. You just have to find where to return it to.

They are going to ask you to mark the package you send to them with a ticket number for reference. I would personally suggest that you put one such ticket reference inside the CPU box itself, as well as somewhere outside on your packaging (clearly marked). That way the ticket number won't get lost because the transport company fucked up.

AMD knows that there's a widespread problem with their early batches of Ryzens and will provide you with an RMA if you can prove to them that you have the problem.

Show your nose

Had to create a swapfile, but I've got no errors, guess that's c6 state problem. Thanks for guiding me user. I'm going to bed but tomorrow I'll update my bios and live my life happily freeze-free (probably).

Attached: 1536821869544.jpg (800x631, 342K)

Oh, you're still around. Yeah, if an UEFI upgrade doesn't fix it, then you can add:

rcu_nocbs=0-11 processor.max_cstate=1

To your kernel command line. This will most certainly fix it. You can read about what it actually does here

utcc.utoronto.ca/~cks/space/blog/linux/KernelRcuNocbsMeaning

>no mention of the better motherboard
b450 isnt expensive

>liquid cooler
4ghz is possible on stock cooling

>the expensive B-die RAM
$30

you get btfo every single time you post this shit just stop

How can people be objectively wrong about their own experiences with a product? They've either experienced crashes or they haven't you retard

why is every fucking benchmark by intel faggots an FPS benchmark. Some of us don't give a shit about gayming.

Because replying "no crashes so far" to a post which is completely correct to point out that a huge amount of Ryzen processors DO have problems is an attempt to argue against it being a problem based on anecdotal evidence.

no, non ECC RGB RAM. :)

Exactly.

You much likely are, but even if it's something else you should follow
just to rule that out. Start by updating UEFI/BIOS and look for a new option in the BIOS named "CPU Idle Current" and set that to "Typical Idle". That will likely fix it. I used the rcu_nocbs option + a script on boot-up that disabled C6 on the package for half a year so that is also an option.

AMD Hackintosh is a thing now user

youtu.be/AUyF--fJaaM?t=69
Watch it.
>ASus crosshair 7 hero
>4.2 overclock using aggressively turned G.skill sniper X 3400Mhz with tightened sub-timings
>and the cooler has been upgrade to the Corsair H115i Pro

THAT is what was used to get the performance benchmarks you keep posting each time without context. Not only are yo downplaying the components used, but also their price.

Again if all of this escaped you, here is the man who made the benchmark, saying it CLEAR as cryztal

youtu.be/AUyF--fJaaM?t=1442
>Given that the intel CPU was over all faster this makes it better value choice for gamers.
>At the very least the OC Ryzen build will require an upgraded cooler, but for the results shown here you will need a high quality motherboard and some premium Samsung B-die.

THERE, now FUCK OFF you fucking slimy shill.

lol at the title "best $200 CPU"

right now it's more like $170 for the 5 2600 vs $300 for the i5-8400 so.. it's not hard to see what's got the way better value today.

I really hope Intel manages to get their shit together, AMD will quickly run into supply issues themselves if this keeps up and CPUs from both brands will become overly expensive.

kind of odd Intel shills aren't fired or given the week off or something, nobody's going to buy those extremely overpriced CPUs until/unless a) Intel gets their issues sorted out or b) AMD raises prices as much as Intel has the last month.

>better just divert attention from the fact that amd shills blatantly lie and keep posting benchmark results meant to fool people with no context
Nothing to see here people, GO BACK TO SLEEP! Plase don't acknowledge what amd fags are doing.

>muh IPC
>muh gayms
>muh FPS

Why aren't there any laptops with Ryzen cpus and RX cards that work well with Linux?