Intel will release Skylake revision B after several OEMs complained about power management bugs in Skylake for which Intel failed to deliver a microcode fix.
Revision B will fix most of the bugs in the Skylake errata in hardware.
Intel has promised the OEMs to fix the rest in microcode updates for Rev. B.
>>54055312
lel, intel failed hard at skylake. Those microcode """""fixes""""" actually SHUTS DOWN specific features of the CPU to stop the problem being reproduced. Same happened in prime95 bug, you literally pay for the features and improvements that are going to be shut down in the future. Enjoy your micro """"fixes""""
[citation needed]
>No source, just shitposting
OP IS A FAGGOT
Skylake power management fixes in Linux WHEN??
http://www.phoronix.com/scan.php?page=news_item&px=Skylake-Concerns-MJG
>Intel's own documentation suggests all low-power idle states be enabled for long-term reliability, yet Linux isn't doing so.
LINUX, NOT EVEN ONCE
>>54055312
Source: ass of anon
>>54056096
>>54056140
>>54056185
Why are you trading your honor, just to shill a company that doesn't care about you?
>>54056206
>after several OEMs complained about power management bugs in Skylake
cite the source for your claims and everybody's happy.
no need to invoke the OMGZ-LE-SHILLZ-maymay.
>>54056171
the forum-postings contradict the article, though.
>Phoronix
>>54055312
The Skylake architecture as a whole has been less buggy than the previous one. I don't know why you are trying to spread misinformation about a problem that is largely irrelevant.
There is no Skylake revision B because Intel is launching Kaby Lake in Q3
OP is a lying faggot that doesn't even know the Intel roadmap
>>54056259
>Calculate prime numbers.
>Crash.
Yeah this architecture sure is free of bugs.
>>54055312
>google skylake revision b
>no relevant results
I got rused.
>>54056259
>try heavy computation
>unexpected behaviour
>next day intel releases a microcode to fix the problem, which is eventually found out that it closed certain features to stop the bug from happening
>>54056367
>40 billion dollar research budget
>still releases buggy chips
lel, I thought jews were supposed to be smart.
>>54056367
>>54056285
They fixed the bug. Whats the problem? Haswell was 50% more buggy.
>>54056456Every CPU in the history of mankind had bugs (aka errata). Each. and. every. one. And they will continue to have bugs.
>>54056460
They "hid" the bug, not fixed it. They closed certain features that may lead to that bug.
>>54056497
What, specifically, were these features you speak of so often. Prime95 works fine with skylake now. Provide your source, also (unless it's your ass, which I have already assumed.)
>>54056545
)))
>>54055312
>Still using my hexacore phenom II 1090t
>Laugh at intelcucks
>>54056626
>Using AMD HOUSEFIRES slower than Intel dual-cores
>any year
>>54056456
K. Well, find all the bugs.
Good luck. You release every 2 years.
>>54056626
http://support.amd.com/TechDocs/41322_10h_Rev_Gd.pdf
enjoy your amd bugs :^)
>>54056768
Was there a final number on that? It got to 700, but it was skipping around alot.
>>54056768
Never had any freezes in 5 years of extensive usage. Don't care about errata, there is enough of it for every complex chip.
>6 more months till zen
>>54056707
Still using a 3770K and its good
>>54056840
I want to sniff Kanye-samas head
>>54056626
Also using a 1090T. It's still pretty based overclocked to 4GHz.
>>54056639
My 1090T actually runs a lot cooler than the 4790K that it replaced. Only gets up to ~45 during muh gaymes and 55 during stress tests, despite a 25% overclock and a decent voltage bump. The 4790K got up to the high 70s during stress tests using the same Noctua NH-U14S.
>>54056882
i have the 4790k, it is a housefire starter. hottest cpu i ever owned
>>54056882
>>54056984
The FX-8350 is the undisputed king of heat
>>54056984
One thing I noted with mine was that the motherboard was setting the voltage way too high out of the box. Almost 1.3V for "Enhanced Turbo"/4.4GHz on all cores. I managed to cut 0.05V off and bump it to 4.6GHz when I got around to doing it manually, although it still ran hot even then.
>>54057040
It's actually pretty impressive how cool Thuban runs considering the housefires to follow. Still uses a lot of power, but it just doesn't get very hot at all.
>>54057040
never owned that. obviously there are cpus that run hotter (fx-8390 for example)
>>54056244
Yeah, the Phoronix comment section is very reliable. It's not like it's full of raving german lunatics or anything.
>>54056707
>You release every 2 years.
>processors literally getting slower now
>>OY VEY, RELEASE THE SKYKRAKEN
https://bugs.freedesktop.org/show_bug.cgi?id=88012
https://bugzilla.kernel.org/show_bug.cgi?id=109051
Intel in charge of anything, nope.
>>54056143
skylake power management is shit in both windows and linux. intel fucked up.
>>54056840
I'll probably consider Zen if AMD really delivers something that can compete Haswell-E
Unsure but I would support it
>>54055312
>made by the jews
>>54056367
classic chicken bits.
>>54057040
It isn't, the 8150 was hotter, and the 9590 is the actual CPU housefire King, outside of that marketing CPU the P4's are unbeaten, chip wise Fermi is still the king
>>54057040
You know the Phenom IIs have more IPC than FX
Whatever happened to AMD
>>54056333
This.
Tried last 24 hours, last week, last month, year etc
tried B, 2 and multiple combinations - absolutely 0 relevant answer except this exact thread already being archived.
>>54056171
"Yes, it's not an issue on desktop parts. I've updated the post to say that."
No problem if you use skylake on pc only laptops have problems.
Disinformation and shilling for other companies intensifies on /g/ day after day...
>>54060867
>Phenom IIs have more IPC than FX
Only the Bulldozer ones. Piledriver generally trades blows with the Phenom IIs, beating them in some workloads and losing out in others. They also overclock much better than K10 chips do, so the gap is widened a little there too. It's trivial to get an FX chip up to 4.5GHz or so, whereas most Phenom IIs top out between 3.8-4GHz unless you start upping voltage beyond 1.5V.
>mfw Zen will have more problems than Skylake ever did and Jim left early because he knew it couldn't be saved
Say what you will about Skylake, but Intel is still better staffed and better funded when it comes to CPU development.
>>54061389
Shit Wrecker left because his job was complete. He wouldn't just bail on a project because it was going badly. You don't get to be the most in-demand man in the industry doing things like that. AMD simply couldn't afford to keep him hanging around not doing much, and I'm sure he wanted to move on again anyway, as he always does after a few years.
>>54061429
Jim Keller isn't an idiot either. He knows whether or not a microprocessor design could be salvaged. The fact that he left well before Zen was taped out (when he left Apple and AMD the first time around, he stuck around until after the design was finalized) shows that he either lacked confidence in the final design to put his full influence behind it, or he looked for the earliest opportunity to bail out.
>>54061457
Desktop Zen was taped out before he left. The quote from AMD's CFO back in December which confused matters was relating to server parts. I don't agree with your assessment of why he left either. You're just seeing what you want to see, and history isn't on your side. Jim Keller doesn't fail, and he spent three years working on Zen. If it was a spectacular failure, his reputation would take a massive hit regardless of when he left.
>>54061528
Believe in what you want
But don't come crying to me when Zen lacks support for new features like XPoint and NVMe, has worse single-threaded performance than Haswell, and when an 8-core Zen can't even match a 6-core Kaby Lake in multithreaded benchmarks.
>fix most of the bugs
THAT'LL BE THE DAY
http://www.intel.com/content/dam/www/public/us/en/documents/specification-updates/desktop-6th-gen-core-family-spec-update.pdf
Are you fucking kidding me, I just bought a 6700k a week ago
>>54062487
Desktop parts don't have a lot of bugs.
It's the mobile parts that are plagued with power management bugs.
