Ryzen freeze bug found and fixed

>The issue with Flops was found and fixed in the beginning of february.
>The current µcode version dates to 01/27/2017, so the fix is obviously not included yet (due to the time required for validation).
>Flops is only affected when the SMT is enabled, so disabling the SMT can be used as a temporary work-around (until the actual fix arrives).

forum.hwbot.org/showpost.php?p=480922&postcount=30

Other urls found in this thread:

www-ssl.intel.com/content/www/us/en/processors/atom/atom-z8000-spec-update.html
danluu.com/cpu-bugs/
theregister.co.uk/2017/02/06/cisco_intel_decline_to_link_product_warning_to_faulty_chip/
twitter.com/NSFWRedditGif

So is it not shit yet? Should I just wait for the next generation after Ryzen for AMD to work out the bugs?

What's Flops?

It's like Skylake, also had freeze bug with numeric calculations (Prime95). Also got fixed by ucode.

Should be name of the benchmark that tripped the bug.

this sort of thing isn't uncommon with new instruction sets/uarches. with any luck they can fix it with a ucode update. skylake had a similar issue, and before that haswell's TSX was bugged and had to be disabled, and even before that the phenom had some issue similar to this (could be misremembering).

Phenom B2 revision, it decreased performance on quadcores. Linux has some workaround that limited the slowdown a lot, but you had to disable the ucode workaround in the motherboard BIOS.

GO AMD GO

YOU CAN DO IT

was that the TLB bug?

> also had freeze bug with numeric calculations (Prime95).
So that's why my screen was just going black

all cpu launches have bugs lol.

flops is a very specific bench that is tailored for each cpu its not generic it needs to be built to the specific cpu at the time

it doesnt have even haswell support yet let alone zen

Probably not. You had to set up p95 in a certain way, with specific parameters to trigger the bug.

>tfw all 14nm intel iGPUs are unusable (and not fixable) due to hardware bug (with exception of newest generation) and nobody cares, no major site mentions this
>amd hardware bug found, immediately fixed
>all review sites go into full ape mode

>tfw all 14nm intel iGPUs are unusable (and not fixable) due to hardware bug
what

>14nm intel iGPUs are unusable


I'm guessing you are talking about broadwell?because that one works fine,

T&L bug (aka. alpha blend bug) which results in all dark textures going into full FFFFFF nigger mode
>works fine
It does, if the game developer decided to include a WORKAROUND for this issue. 90% of gaymes do not include this workaround.

>FFFFFF
kek, meant 000000.
all textures are just black, so you can't see shit. You can try left4dead2 on dark maps to see it for yourself
Also problem exists in all pre-z8*50 atoms (ie. x5-z8500 and z8700)

Linux only? because this doesn't happens on Dolphin or any other emulator even shit GlideN64

I'm talking about windows. Never tried to install linux on modern computers. I'm not a virgin

>tries igpu bugs to see what causes them
>not a virgin

Are you sure this is a Intel hardware probleman d not just a driver problem? AMD also has problem with alpha blend in OGL but this is the first time i've heard of Intel.

>d not just a driver problem?
no, intel officially released a documentation about this (spec update).
Problem is NOT fixable via microcode or driver

can you link to it?

I've seen the document for Atoms and i7-5600u serie CPUs
www-ssl.intel.com/content/www/us/en/processors/atom/atom-z8000-spec-update.html - atom
page 23

CHT30
SoC May Experience an Incorrect Pixel Alpha Component in The Render Target
Problem:
Under certain complex 3D Render pipeline conditions, the graphics subsystem may
experience an incorrect pixel alpha component in the render target.
Implication:
Due to this erratum the graphics subsystem may experience an incorrect pixel alpha
component in the render target
This erratum has not been observed with the commercial applications tested.
Workaround:
Applications can be written to avoid the conditions necessary for this erratum to occur.
Status
For the steppings affected, see the Summary Tables of Changes

But it looks same in other spec updates

The onomatopoetic sound my dick makes when I take it out

You do know that all of Intels products have tons of bugs and some have no fixes or workarounds, they just exist.

same is most certainly true to AMD

You're a dumb faggot

Those are using Broadwell Gen8 GPU

Skylake Gen9(Apollo Lake shares the same Gen9 GPU from Skylake) & Kaby Lake Gen9.5 GPUs are very different microarchitecture

>intel releases unusable product
>nobody gives a shit
You're dumb. Faggot.

except intel officially announced they're gonna stop caring about quality control to gain advantage on the market
also read danluu.com/cpu-bugs/

AMD ironically is much better in this matter

>Phenom TLB bug
>better

TOP KEK

>intel TSX bug
TOP MASTER KEK

Intel is the x86 standard, everyone codes to Intel's 99% marketshare

You don't like it, too fucking bad

>everyone codes to Intel's 99% marketshare

>fixable via bios patch
>comparing it to a hardware, non-fixable error that affects all users
You're so full of shit. Kill yourself

>Killing performance by patching the BIOS
>good

TOP KEK

at least it works, unlike intel cpus
theregister.co.uk/2017/02/06/cisco_intel_decline_to_link_product_warning_to_faulty_chip/
hahahaha, rekt jewtel paid shill

>Intel indicated in a January 2017 revision of its Atom C2000 family documentation that the chip line contains a clock flaw. Errata note AVR.54, titled "System May Experience Inability to Boot or May Cease Operation," explains that the Atom C2000 Low Pin Count bus clock outputs (LPC_CLKOUT0 and LPC_CLKOUT1) may stop functioning. Permanently.

Intel status:
[X] FINISHED & BANKROUPT
[ ] NOT FINISHED & NOT BANKROUPT