http://www.phoronix.com/scan.php?pa ge=news_item&px=Ryze

File: 1496431921862.jpg (2MB, 1920x3103px) Image search: [Google]

2MB, 1920x3103px

>no gentoo
wtf I hate amd now

BHAAHAHAHAHAHAHAHAHAHAHA

File: 1489690271931.png (3MB, 1920x1080px) Image search: [Google]

3MB, 1920x1080px

>Linux
>gaming
good stuff OP

File: vey.gif (1MB, 444x711px) Image search: [Google]

1MB, 444x711px

http://phoronix.com/scan.php?page=news_item&px=Ryzen-Price-Cuts

>1700X now for $349

>>60721585
Intel is shit, news at 11.

What else do you think would happen when you use a garbage OS?

>>60721471
Why is AMD incapable of properly launching a new architecture since like Athlon64?
They ALWAYS spill the spaghetti.
JUST WAIT 10.0 by now.
Zen+, here I come/wait, smhtqbw.

File: feels_amd_skele.png (49KB, 653x726px) Image search: [Google]

49KB, 653x726px

just wait for zen+ guys

>>60721471
>The issue is happening on multiple versions of GCC but I haven't seen any reports when using LLVM/Clang or alternative compilers.

>using dog shit GNU software

hahahahaHAHAHAHAHAHAH

>>60722537
Enough of this shit. I don't care what Intel/AMD shills do, but DO NOT SHIT ON A GOOD COMPILER FOR THE SAKE OF ANY COMPANY FOR FUCK SAKE, YOU JUST CALL IT SHIT BECAUSE YOUR FAVORITE CPU DOESN'T WORK WELL WITH IT, THERE ISN'T A SINGLE OTHER CPU THAT HAS PROBLEMS WITH YOUR CPU
fuck AMD and Intel for bringing retarded people like you

>>60722728
>totally new architecture, different from literally anything to date
>shit happens to not work with said architecture
>GCC wasn't updated for that
>DURR YOUR CPU IS SHIT THE COMPILER IS PERFECTLY FINE

That sounds next level retarded desu, the other compilers work perfectly fine and the processors were released sufficient time ago. This is a shortcoming, and it qualifies as shit software.

It happens, it even happened to Irfanpoo, which doesn't support .webp/m without an external, time wasting plugin.

>>60723234
so you're telling me Ryzen is not x86 or ARM, it's literally different™ architecture eh?

The few gentoo users also on the distro forums posting on that thread I trust over 9000 times more than damage controlling radeon/amdgpu -developer bridgman or some unpaid shills.

There is obviously some problem with the cpu, because this happens without using any of the specific, taxing instructions (avx/avx2).
1700x@stock, C6H, gskill 2x8GB rated @3200Mhz CL16, running at 2113MHz CL16. With gcc-4.8.5, 5.*.* and 6.*, these segfaults crop up randomnly.
It is impossible that the problem lies with compiler when using generic instructions only.

LOL, there is no such a thing as 'totally new' x86(amd64) arch and there is nothing 'revolutionary' in Zen design either when compared to Nehalem and up Intel designs.

Pretty strange that it's only happening with GCC.
>impossible that the problem lies with compiler
Could really be just an obscure bug. It's not impossible.

>>60721471
I love how retarded shills make the same threads about this shit over and over. >>60714396

>I'm going to back the unstable overclocks theory. I had been overclocking my R7 1700 to 3.95 GHz, but after an hour or so, it would crash while doing compile jobs. Now I am running at 3.9 GHz and it runs all day everyday no problem, no instability.

tl;dr gentoo nerds lost the silicon lottery, forced to dial back the OC a bit.

>>60723740
DELED

>>60721471
>AYYMD users on update watch
Everything has flaws, it will be fixed in a few day, nothing to worry about

>FMA bug
AMD is finished!
>got fixed

>VME bug
AMD is finished!
>got fixed

>this
AMD is finished!
>????

>>60721471
One of the top Ryzen shill memes btfo
Literally Bulldozer 2.0

I'm holding buying Ryzen because of this. This was exactly the kind of load I was going to put on it.

>>60723740
>>60723371
I have an ryzen 1700 at stock, linux will randomly crash, usually within a few days of uptime, while I'm using it, so about 36 hours. Yesterday I didn't suspend it over night because I needed a few programs to do shit, and one of them got a seg fault a few hours into the night, randomly. I hope they find the problem soon.

>>60723892
Read the thread. And the previous thread. It is likely a hardware-being-pushed-too-hard issue.

I got a 1800X, it constantly crashes every few hours when I try to run any heavy on it

From reading the various posts in the Gentoo thread

https://forums.gentoo.org/viewtopic-t-1061546-postdays-0-postorder-asc-start-25.html

..I'd say there are some serious issues with Ryzen.

One thing that seems common is that this happens most frequently with 4 sticks of RAM but it's also a problem with just 2 sticks, though it happens less frequently with 2 than 4.

>>60722537
GCC is probably the best, most mature and versatile compiler in existence.

It's important to note that the Glasgow Haskell Compiler's testsuite also exhibits the same kind of segmentation fault failures.

I don't know what is going on but I am guessing this is a factor: Not that many things will load all the cores and threads to the brink. It would not be that shocking if something that tries to feed 18 threads on a 8core/16thread CPU brings up issues that aren't noticed by more average usage.

>>60723371
>There is obviously some problem with the cpu

That's obvious. The simple fact that numerous people are reporting the exact same issues clearly demonstrate that there are some real problems with Ryzen. We wouldn't even have this discussion if it had worked as expected.

>>60723371
>there is no such a thing as 'totally new' x86(amd64) arch

This. It's supposed to be a x86-64 CPU. It should run x86-64 code just fine. If it can't then that's a shortcoming in the CPU. GCC, for example, works just fine on every other CPU. It compiles just fine on older AMD CPUs. It compiles just fine on Intel CPUs. It segfaults on Ryzen.

>>60723477
Please explain why this supposed GCC bug is only triggered by a Ryzen CPU.

>>60723954
So everybody's supposed to just downlclock and not get the performance that AMD shills are meming with their benchmarks? Ok lol

>>60723954
So, running 1700x@stock with Crosshair VI Hero, 8xmake operation is hardware-being-pushed-too-hard issue. Or just a single make operation. Or opening firefox.
Wow, I must tell that to FX-8350@stock too, with which such problems do not happen or to 2500k@stock, which from day one have not cropped a single segfault either. Or K6@500Mhz. Or 1090t.

There is literally dozen users telling that segfaults happen with STOCK-fucking-CLOCKS.

Someone on AMD payroll, like bridgman, spouting this kind of nonsense too is very unprofessional.

>>60724069
The CPU itself or the MOBO/chipset? Because the hexacore R5 looks perfect for my needs and a transcoding PLEX server.

I'm reading the gentoo thread and some reported that the new Agesa fixes it.
Clearly a microcode issue, will be fixed like the last 2 microcode bugs.

>>60724118
There are some posts where the issue was fixed just by using the most recent version of binutils. https://forums.gentoo.org/viewtopic-p-8057818.html#8057818

>>60724118
>This. It's supposed to be a x86-64 CPU
That's just the ISA, it doesn't mean every CPU will handle every instruction exactly identically due to different architectures.

It could be an errata in Zen, it could be a bug in GCC, or it could be both in how they interact.

>>60724161
> https://forums.gentoo.org/viewtopic-p-8057818.html#8057818
> Posted: Tue Apr 18, 2017 3:18 pm Post subject:
> Apr 18, 2017
> April

$ date
Sat Jun 3 15:06:22 CEST 2017

makes me wonder why this is news

There's no doubt new architecture teething issues suck, but running around claiming this is TEH END OF AMD is just retarded. It'll get fixed like any other issue if it is a microcode issue.

>>60724247
It's called a microcode update, dumbass. http://www.game-debate.com/news/22985/amd-ryzen-agesa-1-0-0-6-update-adds-support-for-memory-clocks-up-to-ddr-4000mhz

>>60724233
>>60724247
>>60724254
Jesus christ the FUD spreading samefag just doesn't give up.

>>60724233
>Source: My rage filled tears

>wahh waaah FMA and VME WILL NEVER BE FIXED RYZEN IS BUGGY
>get fixed in a month via microcode

Lmao, so desperate

>>60724270
I recognize his posting style. He spends all day on /g/ shitting up every CPU and GPU thread. What gives it away is that he samefags compulsively and spams the shit out of his responses. He really needs to be range banned for shitting up /g/

Does someone know how they fix this kind of stuff via microcode updates? I thought CPU Arch's have instruction matrixes and stuff like that hardcoded in literal hardware.

>>60724373
Firmware is a shitshow regardless of hardware, configurable microcode was precisely invented after the FDIV shitshow so a total recall wasn't needed.

>>60724373
AMD uses something called AGESA to update. http://www.anandtech.com/show/11447/amd-announces-ryzen-agesa-1006-update

What a shitty thread.

Intel desperation is so good

>>60724314
Kek what a desperate shitposter.

>>60721471
>>>>>gaming
>>/v/

It's already fixed. INTEL BTFO

File: c7xhOav.jpg (100KB, 533x400px) Image search: [Google]

100KB, 533x400px

>>60721471
>broken binutils, nothing to do with Ryzen at all

Corelets are getting desperate

Intel is fucking terrified by the sheer amount of shilling over this.

My god.

>>60721471
>AMD is shit on Linux
No shit. I fell for the Intel CPU + AMD GPU. Never again. Intel+nVidia is the only working combination on Linux.

>>60725338
your shilling is really transparent

I've been using AMD CPUs and GPUs on Linux for ages and never had problems.
Intel, on the other hand, absolutely needs microcode updates configured or it flips shit several times a day.

>>60725412
There's also this one time I bought and Intel NUC. It didn't even last till the end of the week because Intel fucked up the voltage regulator and it fried the CPU.

Asserting that Intel is a golden benchmark by which everything should be judged is retarded. They are ridiculously sloppy for a company that sells supposedly premium products for actually premium prices.

That being said, my Sandy Bridge EliteBook is still going strong, so there *was* a time when Intel was good.

>>60725538
Don't forget the atoms bricking themselves. https://www.extremetech.com/computing/244074-intel-atom-c2000-bug-killing-products-multiple-manufacturers

>>60725412
>shilling
I'm not. I've always liked AMD more than jewtel, I'm just disappointed in their GPU support for Linux.
AMD completely abandoned driver support on Linux for a majority of their cards. They literally supported my GPU for less than a month after it's release. The open source driver is fine, but not as efficient as official AMD drivers would be if they actually worked on them. It has a much worse performance than the windows driver while most nVidia cards run as good on Linux as they do on windows.

>>60723827
>>60721471
>>60724118
>>60724151
>>60724247
Already fixed, use GCC 6.3.

>poozen doesnt work
>expect to work properly
>pootards never gonna learn

>>60725604
What GPU do you have anyway?

The legacy cards that stopped being supported relatively recently are already several years old and Catalyst was trash for everything but 3D anyway.
Almost everything released under the new "Rx xxx" naming convention (and some high end cards from HD 7000 and HD 8000 series) is supported by the new AMDGPU-PRO driver.

>>60725703
Reminder to report shitposters.

>>60725703
see
>>60725678
>Already fixed, use GCC 6.3.

>>60725719
>GPU made in mid 2015
>old
It's R7 360.

>>60726203
Bonaire is from 2012.

>>60722728
>THERE ISN'T A SINGLE OTHER CPU THAT HAS PROBLEMS
You clearly did not ever see issue tracker of GCC. GCC supports miriads of architectures and there is no way there are zero bugs in architecture support.

stfu
>>60724058
>GCC is probably the best,
By what definition? Clang 4 output program outpaces GCC output in several situations involving parallelism.
http://www.phoronix.com/scan.php?page=article&item=gcc7-clang4-jan&num=1
>most mature
Ok, check.
>and versatile
Clang is drop-in replacement even if it does not suport that big number of architectures.

Hilarious seeing this intel shilling go so hard, glad I bought a ryzen.

>>60726696
Moving forward clang is the way to go. LTO is still broken as fuck and my work also needs a Fortran compiler though so I'm sticking with GCC at the time being but I plan fully to migrate when those things get fixed.

>>60723234
>hundreds of bottles
>competitors bottle opener can open all of them
>make a new bottle opener that can't
>blame the bottles

>>60721471

kek. why would we amd owners who run windows even give a shit?

File: 1392100729337.png (73KB, 744x1340px) Image search: [Google]

73KB, 744x1340px

>>60728568

>>60722343
>beats Intel to a pulp with Ryzen
>incapable

I wonder what that makes of Intel's arch.

>>60730314
I heard X299 is a rushed mess. And X99 wasn't perfect, either. Can't wait for that shitshow to get on the road.

>>60730190
>can't even post a relevant image
man you gotta be retarded huh?

>>60730314
>beats intel to a pulp
>intel still has 80% marketshare
delusion?

File: 1w9jrnoh.vichan-1.jpg (144KB, 768x960px) Image search: [Google]

144KB, 768x960px

>>60721471
Anyone one on xen? How does xen work on aymd risen and if i could use pci passtrought.
t xen /qubes fanboy

>>60721471
>using old compiler and expect it will always work

>>60731385
you don't have any other choice because intels HEDT stuff doesn't even have VT-d, it's either unlocked CPU or VT-d. iirc AMD has never disabled IOMMU virtualization in their products, even my old Phenom II had it.

New arch needs errata? News at 11!

>>60731580
>stuff doesn't even have VT-d, it's either unlocked CPU or VT-d
Not true since skylake

>>60731364
Does it look like to you that a CPU's performance is based on market share?

File: INTEL-AMD-CPU-Market-share.png (136KB, 1118x862px) Image search: [Google]

136KB, 1118x862px

>>60731364
It only launched 3 months ago, and they're already taking a bite out of Intel's marketshare. Hopefully that red line will keep going up.

>>60731831
Hopefully it doesnt since AYYYMD is and has been shit

>>60731831
>http://store.steampowered.com/hwsurvey/processormfg/
> -0.42%

Wondering, why it's still declining on steam? Yeah, I know that not everybody use it but still, amd cpu keep declining in share from start of the year.

>>60731867
Steam survey is not indicative of overall marketshare because there's a lot more people with craptops running Intel mobile CPUs, and more people with Intel CPUs that are better for gaming in general.

>>60731866
Reminder to report shitposters.

>>60731866
If you're an Intel fan why would you want that? Competition means lower prices and better products. You should want AMD to compete regardless of if you use it or not

>>60731866
>I don't want competition!!! It needs to be one sided!!!
I will never understand these people's mindset

File: 1426356128820.jpg (48KB, 387x357px) Image search: [Google]

48KB, 387x357px

>>60721471
Why is everyone so concentrated into shaming AMD for an honest mistake in a limited set of setups that has already been fixed while ignoring the elephant in the room that is intel with terrible business and anti-consumer practices? How is more important a mistake than being screwed on purpose?

>>60721471
People take too much pride in processors

>>60731831

You can notice the same bump for previous releases... Watch Piledriver do better than Ryzen.

>>60731962
Zen is not Bulldozer.

>>60731932
because amd fanboys insist that their favourite company is literally Jesus

File: file.png (708KB, 640x640px) Image search: [Google]

708KB, 640x640px

>>60721471
>only happens with GCC, not Clang
>blame Ryzen