Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us money per month for bills, and since we don't believe in showing ads to our users, we try to make the money back through forum registrations.
 
  • Post
  • Reply
PerrineClostermann
Dec 15, 2012

by FactsAreUseless
Two computer semi-illiterate friends have Zen systems and have reported no problems.

Adbot
ADBOT LOVES YOU

GRINDCORE MEGGIDO
Feb 28, 1985


A few people are in there reporting it happening at stock, maybe it's a specific mobo bios.

Kazinsal
Dec 13, 2011



I wonder if there are some chips that really lost the silicon lottery and just don't have enough volts at stock clocks to keep up under a really intense long-running 100% load situation.

eames
May 9, 2009

Kazinsal posted:

I wonder if there are some chips that really lost the silicon lottery and just don't have enough volts at stock clocks to keep up under a really intense long-running 100% load situation.

Some people in the Gentoo thread reported that higher load line calibration settings (= higher Vcore under load) fixed the compiler bug so you might be right

wargames
Mar 16, 2008

official yospos cat censor

Kazinsal posted:

I wonder if there are some chips that really lost the silicon lottery and just don't have enough volts at stock clocks to keep up under a really intense long-running 100% load situation.

Would zen2 with just a slight bump to volts all around be a lazy fix for such things or is zen 2 going to be on 7nm?

Also what happened to intels 10nm process?

Kazinsal
Dec 13, 2011



wargames posted:

Would zen2 with just a slight bump to volts all around be a lazy fix for such things or is zen 2 going to be on 7nm?

Also what happened to intels 10nm process?

This is entirely a wild-rear end guess based solely on my observations in both the CPU threads on SA, some threads on a few other boards, and conversations I've had about the CPU R&D process with some other folks in the industry, but I suspect the following will happen:

Zen 1.0 (current Zen) will continue to receive stability and performance improvements through microcode built into AGESA. Zen 1.5 (some folks are calling it Zen+) will happen early next year; it will likely be a process refinement of the Zen architecture with all the microcode-based improvements intact, and then some. It'll be when AMD is comfortable with the process they're working with and designing for.

I suspect a die shrink would be in the cards at some point if AMD can end friendship with GloFo, now Samsung is my best friend. There's no way they could do a competitive 10nm part on the GloFo process.

Cannonlake (Intel's Kaby Lake die shrink) has been pushed back a bit last I checked.

SwissArmyDruid
Feb 14, 2014

by sebmojo

Kazinsal posted:

This is entirely a wild-rear end guess based solely on my observations in both the CPU threads on SA, some threads on a few other boards, and conversations I've had about the CPU R&D process with some other folks in the industry, but I suspect the following will happen:

Zen 1.0 (current Zen) will continue to receive stability and performance improvements through microcode built into AGESA. Zen 1.5 (some folks are calling it Zen+) will happen early next year; it will likely be a process refinement of the Zen architecture with all the microcode-based improvements intact, and then some. It'll be when AMD is comfortable with the process they're working with and designing for.

I suspect a die shrink would be in the cards at some point if AMD can end friendship with GloFo, now Samsung is my best friend. There's no way they could do a competitive 10nm part on the GloFo process.

Cannonlake (Intel's Kaby Lake die shrink) has been pushed back a bit last I checked.

Maybe AMD won't be cutting ties just yet. But at least it may be because GloFo has a product worth offering, rather than being a goddamn albatross:

SwissArmyDruid posted:

It *appears* that Samsung, working with GloFo and IBM, may beat Intel to market with EUV.
IBM is touting 40% perf increase or 75% reduced power consumption from the new GAAFET transistor design vs commercial 10nm FinFET.

https://www.theverge.com/circuitbreaker/2017/6/5/15740500/ibm-5nm-gaafet-transistor-faster-more-efficient
https://arstechnica.com/gadgets/2017/06/ibm-5nm-chip/
http://wccftech.com/ibm-gf-5nm/

This is more bad news for Intel, I think. I can't remember if they had arrived at their next step after FinFET yet. Last I heard, they were going to switch to InGaAs at 7nm and ride FinFET all the way down to 5nm, or something like that.

edit: Now the question remains: Can AMD GPU-side survive long enough to get whatever comes after Navi onto this new 5nm process? XD

wargames
Mar 16, 2008

official yospos cat censor

Kazinsal posted:

I suspect a die shrink would be in the cards at some point if AMD can end friendship with GloFo, now Samsung is my best friend. There's no way they could do a competitive 10nm part on the GloFo process.

Cannonlake (Intel's Kaby Lake die shrink) has been pushed back a bit last I checked.

AMD is planning on skipping 10nm and using ibm/samsung/gloflo 7nm in 2018 for zen2, and zen3 will use 7nm refined.

Scarecow
May 20, 2008

3200mhz RAM is literally the Devil. Literally.
Lipstick Apathy
So is threadripper looking like a late july release?

Dante80
Mar 23, 2015

For the last month, I've been a happy mother of a 1600X based system. Here is what I got:


Fractal Define C
Ryzen 1600X (Corsair H110i Extreme)
Asrock X370 Taichi
G.SKILL Flare X 16GB 3200Mhz
XFX RX480 8GB GTR
Samsung 960 EVO 1Tb
Seasonic Prime 650W

Very satisfied so far. Was originally going for the 8 core, but 1600X had such a good price that I decided to go for it. With the money difference I chose to double the PCI-E SSD size and refrain from getting a HDD altogether.

Someone handed me some industrial PWM Noctua fans too (the black ones), so I promptly put 5 on them in the rig (2x140 on the rad, 2x140 on the top and a 120 on the back). Was seriously doubting whether that was a good choice as soon as I posted, those 2000rpm fans are pretty loud!. I was psychologically ready to dabble on speedfan for a day or two, but I found out that simply running the Asrock BIOS option for auto fan control made the whole rig whisper quiet. The fact that both GPU and PSU run fanless on idle/browsing/movies etc helps a lot too.

Have not tried OCing yet, other than setting the memory to its 3200Mhz 14/14/14 profile. The whole thing is reporting pretty good temperatures though, to the point that I don't hear the fans unless I'm stressing the system.

And running something like Furmark/Prime95 for an hour or so provides still glorious temps (that 280 AIO cooler is worth it)..albeit with very increased sound.

So far, so good.

Dante80 fucked around with this message at 11:46 on Jun 9, 2017

Combat Pretzel
Jun 23, 2004

No, seriously... what kurds?!

Anime Schoolgirl posted:

similar problems happened with sandy bridge and didn't really even get fixed until ivy bridge. especially with the P67 :shepicide:
Wait what? My old P67-based system was rock-stable (was rev. 3 tho).

JnnyThndrs
May 29, 2001

HERE ARE THE FUCKING TOWELS
Yeah, my main box is still a 3770/P67@ 4.3 setup and it's absolutely rock-stable no matter what abuse I throw at it. IIRC, the P67 bug that fried SATA controllers was caught before significant numbers of mobos were released to the public.

Anarchist Mae
Nov 5, 2009

by Reene
Lipstick Apathy
Ryzen Threadripper 1920 oh and also one of the supposed AMD Radeon engineering sample IDs matches the ID of an Intel HD Graphics Gen9 engineering sample?..

Anarchist Mae fucked around with this message at 13:36 on Jun 9, 2017

NewFatMike
Jun 11, 2015

Alienware was a little surprising on that.

As much as I love AdoredTV, Level1 is a little more grounded than our friendly neighborhood Scotsman, Pretty good roundup of AMD stuff from Computex:

https://www.youtube.com/watch?v=pI2F8K3Agjs

Anime Schoolgirl
Nov 28, 2002

Combat Pretzel posted:

Wait what? My old P67-based system was rock-stable (was rev. 3 tho).
early firmware for sandy bridge was really, really bad and had numerous microcode issues

they were fixed pretty quickly (2 months) at least

Malloc Voidstar
May 7, 2007

Fuck the cowboys. Unf. Fuck em hard.
lmao at the rumor being 1998 and the reality being 1920

Mr.Radar
Nov 5, 2005

You guys aren't going to believe this, but that guy is our games teacher.

Measly Twerp posted:

Ryzen Threadripper 1920 oh and also one of the supposed AMD Radeon engineering sample IDs matches the ID of an Intel HD Graphics Gen9 engineering sample?..

Rumor is that Apple is ordering custom Intel CPUs with AMD integrated graphics because Intel's graphics just suck so much, even at the high end.

repiv
Aug 13, 2009

Measly Twerp posted:

Ryzen Threadripper 1920 oh and also one of the supposed AMD Radeon engineering sample IDs matches the ID of an Intel HD Graphics Gen9 engineering sample?..

"Intel(R) HD Graphics Gen9; 694C:C0 (1720SP 47C 1GHz, 528kB L2, 10.4GB 800MHz)"

That can't be right, 1720 SPs doesn't divide evenly into the 64SP CU/NCUs that GCN is built from.

Paul MaudDib
May 3, 2006

TEAM NVIDIA:
FORUM POLICE

wargames posted:

Would zen2 with just a slight bump to volts all around be a lazy fix for such things or is zen 2 going to be on 7nm?

Yeah but it'd blow their power efficiency even worse than it already is. They are already practically tied with Intel's chips rather than being 30-50% ahead like their TDP numbers imply.

Anime Schoolgirl
Nov 28, 2002

repiv posted:

"Intel(R) HD Graphics Gen9; 694C:C0 (1720SP 47C 1GHz, 528kB L2, 10.4GB 800MHz)"

That can't be right, 1720 SPs doesn't divide evenly into the 64SP CU/NCUs that GCN is built from.
likely a type error, 1728 SPs would mean 27 CUs, which must mean it's a MCM design with an intel chip and a rtg gpu

or it could be the GPU on-die along with the intel CPU, in which case :psyduck: on how that will work

Anime Schoolgirl fucked around with this message at 18:39 on Jun 9, 2017

GRINDCORE MEGGIDO
Feb 28, 1985


Didn't Intel confirm this was false a little while ago?

SwissArmyDruid
Feb 14, 2014

by sebmojo
Intel said nothing, and Lisa Su said that they were not in the business of "enabling a competitor to compete with our products".

But Apple being its own, internal, walled-off garden in almost all ways that it is... I can see why an exception might be made there. It would be a product that would be sold nowhere else.

SwissArmyDruid fucked around with this message at 19:04 on Jun 9, 2017

Risky Bisquick
Jan 18, 2008

PLEASE LET ME WRITE YOUR VICTIM IMPACT STATEMENT SO I CAN FURTHER DEMONSTRATE THE CALAMITY THAT IS OUR JUSTICE SYSTEM.



Buglord
Conspiracy theory: Mac Pros are being held back to plop in Threadrippers. They could keep their margins and lower the price to the consumer :monocle:

BangersInMyKnickers
Nov 3, 2004

I have a thing for courageous dongles

There is zero chance Apple is rolling the dice on a new AMD cpu/chipset platform, they'll let problems settle out for a year or two before even considering it.

ArgumentatumE.C.T.
Nov 5, 2016

by Jeffrey of YOSPOS

Risky Bisquick posted:

Conspiracy theory: Mac Pros are being held back to plop in Threadrippers. They could keep their margins and lower the price to the consumer :monocle:

All Mac Pros will come with a free pony

Yaoi Gagarin
Feb 20, 2014

Risky Bisquick posted:

Conspiracy theory: Mac Pros are being held back to plop in Threadrippers. They could increase their margins and increase the price to the consumer :monocle:

I agree

EmpyreanFlux
Mar 1, 2013

The AUDACITY! The IMPUDENCE! The unabated NERVE!

repiv posted:

"Intel(R) HD Graphics Gen9; 694C:C0 (1720SP 47C 1GHz, 528kB L2, 10.4GB 800MHz)"

That can't be right, 1720 SPs doesn't divide evenly into the 64SP CU/NCUs that GCN is built from.

Maybe it's a fusion of Intel and AMD design, possibly Intel layout and command processor but AMD shaders and geometry processor? Intel works in threes, so a slight typography error could mean this is a scaled up and enhanced Iris 580 with GDDR5 memory and 384 bit bus, so 1728:216:27, but I can't for the life of me figure out that weirdass 10.4GB memory config. Everything else is divisible by three though.

eames
May 9, 2009

FaustianQ posted:

Maybe it's a fusion of Intel and AMD design, possibly Intel layout and command processor but AMD shaders and geometry processor? Intel works in threes, so a slight typography error could mean this is a scaled up and enhanced Iris 580 with GDDR5 memory and 384 bit bus, so 1728:216:27, but I can't for the life of me figure out that weirdass 10.4GB memory config. Everything else is divisible by three though.

10.4 GB shared HBM2 :getin:

GRINDCORE MEGGIDO
Feb 28, 1985


eames posted:

10.4 GB shared HBM2 :getin:

drat son

EmpyreanFlux
Mar 1, 2013

The AUDACITY! The IMPUDENCE! The unabated NERVE!

eames posted:

10.4 GB shared HBM2 :getin:

That's really only feasible with HBM1 though.

So for maximum WTF, it's the shell of a Intel iGPU scaled up to a desktop part, MCM'ed to a CPU, with most of it's guts replaced with AMD tech, uses AMD's HBCC tech, on an interposer to support 6 3-hi stacks of HBM1 and using half of the systems 16GB of DDR4 memory as shared memory (and thus reporting 10.4GB) for also having a stupid large 6144 bit bus. I cannot discern this theoretical parts purpose.

SwissArmyDruid
Feb 14, 2014

by sebmojo
One word: Apple.

Watermelon Daiquiri
Jul 10, 2010
I TRIED TO BAIT THE TXPOL THREAD WITH THE WORLD'S WORST POSSIBLE TAKE AND ALL I GOT WAS THIS STUPID AVATAR.
im kinda surprised (in a researcher sense) that they haven't gone with gaa sooner, since itd 'only' be what, an extra 2-3 masks to get that? theyd have to mess with the mandrels, but it shouldnt be too much.

PC LOAD LETTER
May 23, 2005
WTF?!
RWT did a great article on GAA for anyone wanting read more on it: http://www.realworldtech.com/intel-10nm-qwfet/

Its a couple years old now but much of the information still holds up.

Seems like DK was expecting it sooner too but maybe it just wasn't practical yet.

SwissArmyDruid
Feb 14, 2014

by sebmojo

Comatoast posted:

What in the world? It's like a 23 minute long MacRumors video.

Oh, trust me, he's actually pretty good at what he does.

If you want full-on bed-making GBS threads fanboyism videos, that's what RedGamingTech is for.

Watermelon Daiquiri
Jul 10, 2010
I TRIED TO BAIT THE TXPOL THREAD WITH THE WORLD'S WORST POSSIBLE TAKE AND ALL I GOT WAS THIS STUPID AVATAR.

PC LOAD LETTER posted:

RWT did a great article on GAA for anyone wanting read more on it: http://www.realworldtech.com/intel-10nm-qwfet/

Its a couple years old now but much of the information still holds up.

Seems like DK was expecting it sooner too but maybe it just wasn't practical yet.

Gah good loving lord that 'quantum well' gate :stonk: growing a layer of si in the middle of all that? granted, you really only need one mask for all of that metal, but still.

Maxwell Adams
Oct 21, 2000

T E E F S
Wait, hang on, I want to have the craziest theory.

At E3, Sony will announce that the hybrid Intel/AMD chip is going into their next portable system. It's a laptop. The Playstation Laptop, coming soon, with a price twice as high as the unreasonable number you're thinking of.

Anarchist Mae
Nov 5, 2009

by Reene
Lipstick Apathy


I can finally afford to upgrade. In three days I'm going to put my i3-6100 in a box and forget it ever existed.

wargames
Mar 16, 2008

official yospos cat censor

Measly Twerp posted:



I can finally afford to upgrade. In three days I'm going to put my i3-6100 in a box and forget it ever existed.

What does 134,699 feet convert to real money?

Malcolm XML
Aug 8, 2009

I always knew it would end like this.

Watermelon Daiquiri posted:

Gah good loving lord that 'quantum well' gate :stonk: growing a layer of si in the middle of all that? granted, you really only need one mask for all of that metal, but still.

Yeah a combo of iii-v and gaa seems a bit too much at the same time

I suspect Intel is going to do one or the other and then the next step will be to.do both


However Samsung TSMC and the relentless pace of mobile have basically brought a second and third fab competitor to the game so who knows

Adbot
ADBOT LOVES YOU

Malcolm XML
Aug 8, 2009

I always knew it would end like this.
Actually yeah that qw fet is a iii-v finfet

  • 1
  • 2
  • 3
  • 4
  • 5
  • Post
  • Reply