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
Paul MaudDib
May 3, 2006

TEAM NVIDIA:
FORUM POLICE

MaxxBot posted:

Thank you, I'm a big CPU nerd so stuff like this is right up my alley.

Agner Fog is the best. If you haven't yet, also check out another of his works in this series, Microarchitecture.

Scott Wasson (formerly of TechReport) is another amazing guy to read/listen to. Unfortunately he's mostly gone silent since getting a job at AMD doing hardware-engineering-y things. He was way out in front on the whole frametime issue and all-around knows his poo poo. Is his old podcast still worth listening to anymore/any replacements?

David Kanter's great. I wish he did more podcasts or writing or whatever.

SemiEngineering also has some really interesting reads focused on litho tech/semiconductor fab stuff. I read an interesting article recently on some of the challenges they're facing with Extended UV (which is now beginning to push into X-ray territory)[/url]. I'm pretty sure there was another article recently that went into some of the stuff they're doing, where they're now instantly ionizing a little blob of metal into plasma order to get a perfectly round blob that behaves as close to ideally as possible, or some poo poo like that.

pyf interesting tech reads?

Do we have a SHSC or Cavern of Cobol book/article/what-i'm-reading thread for this stuff?

Paul MaudDib fucked around with this message at 04:39 on May 30, 2017

Adbot
ADBOT LOVES YOU

Theris
Oct 9, 2007

Inside the Machine by Jon "Hannibal from Ars" Stokes is great as an approachable introduction to in-depth CPU stuff. Unfortunately it's pretty outdated. (Only goes up to K8 and Core 2) He's supposedly working on a new revision but has been pretty quiet about it.

GRINDCORE MEGGIDO
Feb 28, 1985


It's a great book. I'm around 3/4 into it.
Just expect all your geeky friends to want to borrow it, because they will.

I miss the old Arstechnica.

VealCutlet
Dec 21, 2015

I am a marketing god, shave that shit

PerrineClostermann posted:

I look at my laptop with an Intel CPU and AMD GPU and laugh every time :smug:

I think the opposite is pretty common in new builds of the last few months. R5 w/ a 1060/1070 is a pretty good rig.

I upgraded my 1080 to a ti and chucked a Kraken G12 on it the other week. The R5 1600 isn't going anywhere though... maybe it will once threadripper gets here.

GRINDCORE MEGGIDO
Feb 28, 1985


Wasnt bitcoin mining the cause of 480 sellouts as well?

VealCutlet
Dec 21, 2015

I am a marketing god, shave that shit

GRINDCORE MEGGIDO posted:

Wasnt bitcoin mining the cause of 480 sellouts as well?
Yep

champagne posting
Apr 5, 2006

YOU ARE A BRAIN
IN A BUNKER

Which is a drat shame considering it would've been a nice card to go out and buy at the time.

GRINDCORE MEGGIDO
Feb 28, 1985


What causes miner purchases to bomb?

I wish I'd bought one when it hit a low.

GRINDCORE MEGGIDO fucked around with this message at 08:38 on May 30, 2017

Drakhoran
Oct 21, 2012

GRINDCORE MEGGIDO posted:

Is the demand constant, are miners buying them for the entire product life?

No, prices on all kinds of cryptocurrency are extremely volatile. One week miners will be buying up every card making them hard to get for actual gamers, next week they sell all their used cards on eBay bringing retail prices crashing down.



In unrelated news. Here is a picture of two Asrock X399 boards. Looks like the official name for the Threadripper socket will be TR4.

eames
May 9, 2009

GRINDCORE MEGGIDO posted:

Is the demand constant, are miners buying them for the entire product life?

From what I understand "mining" efficiency (coin per watt) is always on a steady decline, the price per coin is highly volatile and power costs are constant.

The spikes in GPU demand happen when the price goes up beyond the power costs of calculating a coin. Assuming a constant price, efficiency will catch up and inevitably make it unprofitable again, at which point people sell GPUs.
My understanding is that prices would need to rise forever to keep demand constant and a sudden crash leads to cheap used GPUs for all gamers.

(correct me if I'm wrong, seasoned cryptogoons)

BurritoJustice
Oct 9, 2012

Drakhoran posted:

In unrelated news. Here is a picture of two Asrock X399 boards. Looks like the official name for the Threadripper socket will be TR4.

AMD's parallel naming is getting pretty goofy, Intel has socket R4 on chipset X299 and AMD has socket TR4 on chipset X399.

Scarecow
May 20, 2008

3200mhz RAM is literally the Devil. Literally.
Lipstick Apathy
Look at how loving gigantic that socket is god damnnnnnnn

wargames
Mar 16, 2008

official yospos cat censor

Drakhoran posted:

No, prices on all kinds of cryptocurrency are extremely volatile. One week miners will be buying up every card making them hard to get for actual gamers, next week they sell all their used cards on eBay bringing retail prices crashing down.



In unrelated news. Here is a picture of two Asrock X399 boards. Looks like the official name for the Threadripper socket will be TR4.

The people who claim a single pea sized dab of thermal grease will have something break in their mind once that lands in front of them.

fishmech
Jul 16, 2006

by VideoGames
Salad Prong

eames posted:

From what I understand "mining" efficiency (coin per watt) is always on a steady decline, the price per coin is highly volatile and power costs are constant.

The spikes in GPU demand happen when the price goes up beyond the power costs of calculating a coin. Assuming a constant price, efficiency will catch up and inevitably make it unprofitable again, at which point people sell GPUs.
My understanding is that prices would need to rise forever to keep demand constant and a sudden crash leads to cheap used GPUs for all gamers.

(correct me if I'm wrong, seasoned cryptogoons)

The only thing I'd add that is there's constantly "new" coins being made and a certain kind of sucker buying GPUs for the new stuff without anyone having time to figure return vs electricity ratios, so that drives demand when prices aren't particularly good for the main established coins.

SwissArmyDruid
Feb 14, 2014

by sebmojo

Drakhoran posted:

In unrelated news. Here is a picture of two Asrock X399 boards. Looks like the official name for the Threadripper socket will be TR4.

How long before people start saying poo poo like, "Oh, finally AMD decided to get with the program and use LGA sockets."

I've got a tic in my left eye just thinking about it.

eames
May 9, 2009

64 lane PCIe, quadchannel ECC confirmed for Threadripper. :popeye:



also this is what happens when you bolt a AM4 cooler to a SP3/TR4 socket and yes those are torx bolts:

NewFatMike
Jun 11, 2015

Hot deets on that AMD RoG notebook.

quote:

The dimensions of this notebook are 41.2cm x 27.2cm x 3.4cm.

:stare:

New Zealand can eat me
Aug 29, 2008

:matters:


When do we find out how much THREADRIPPER costs? My wallet is whimpering in the corner after that $2k i9 announcement

PC LOAD LETTER
May 23, 2005
WTF?!
Maybe in a few days. I'm hoping they'll offer the 16C/32T version for very close to $1K. Those are the magic mindfuck psychological numbers they'd probably need to hit to move lots of those chips fast.

I dunno if I'd get one, really 8C/16T is more than enough for me right now but at that price I'd be tempted. It'll depend on what the platform costs too. That loving socket does not look cheap.

Rastor
Jun 2, 2001


Well, 80s nostalgia is big right now after all...

PC LOAD LETTER
May 23, 2005
WTF?!
Size seems typical for a high end 17" gaming laptop to me. The battery life is short on most of them yeah you do end up treating them like a 80's portable PC that you plug in everywhere.

Sininu
Jan 8, 2014


That's not much at all tbh.

Regrettable
Jan 5, 2010



That website will not load for me at all.

Seamonster
Apr 30, 2007

IMMER SIEGREICH
Seriously, a Threadripper mobo in the Aorus gaming line?

Anarchist Mae
Nov 5, 2009

by Reene
Lipstick Apathy

Seamonster posted:

Seriously, a Threadripper mobo in the Aorus gaming line?

It's for playing Ashes of the Benchmark. Or alternatively a his/hers two seat gaming computer.

eames
May 9, 2009

Seamonster posted:

Seriously, a Threadripper mobo in the Aorus gaming line?

I haven't seen a single X399 (or X299 for that matter) motherboard targeted at workstation users so far. :shrug:

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
It could be for gaming cafes now that gpu passthrough is a thing. 1 threadripper, 4 gpus, 4 seats :fap:

Anarchist Mae
Nov 5, 2009

by Reene
Lipstick Apathy
Rise of the Tomb Raider just got patched for Ryzen, now it performs 50% better than before.

Drakhoran
Oct 21, 2012

That's a significant improvement.

crazypenguin
Mar 9, 2005
nothing witty here, move along
I really want to know what they did to accomplish that.

Beautiful Ninja
Mar 26, 2009

Five time FCW Champion...of my heart.

crazypenguin posted:

I really want to know what they did to accomplish that.

A significant portion of that is bug fixing, Ryzen + Nvidia combos apparently performed extremely poorly in the game in DX12 pre-patch, far below what you'd expect compared to other games. There is some speculation that Nvidia's drivers don't handle Ryzen properly so I wonder if Nixxes worked around it or if it was a Rise of the Tomb Raider engine bug this whole time. There were some generic CPU related fixes as well that should have improved both Intel and AMD performance in this game with this patch as well.

http://www.tomshardware.com/reviews/amd-ryzen-7-1700x-review,4987-6.html - You can see Rise of the Tomb Raider DX12 results here, you can see that Ryzen was performing really out of whack here, the 1700X was barely faster than an FX-8350.

Beautiful Ninja fucked around with this message at 19:04 on May 30, 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
Did AMD drive the optimization effort or did the studio? I want to know if they will repeat/share this across other games

eames
May 9, 2009

Risky Bisquick posted:

It could be for gaming cafes now that gpu passthrough is a thing. 1 threadripper, 4 gpus, 4 seats :fap:

... and terrible performance because of virtualization bugs that AMD won't acknowledge. :(

https://community.amd.com/thread/215931

The promised fix (AGESA 1006) doesn't help with NPT performance, it only fixes groupings to enable GPU passthrough.

PerrineClostermann
Dec 15, 2012

by FactsAreUseless

Beautiful Ninja posted:

A significant portion of that is bug fixing, Ryzen + Nvidia combos apparently performed extremely poorly in the game in DX12 pre-patch, far below what you'd expect compared to other games. There is some speculation that Nvidia's drivers don't handle Ryzen properly so I wonder if Nixxes worked around it or if it was a Rise of the Tomb Raider engine bug this whole time. There were some generic CPU related fixes as well that should have improved both Intel and AMD performance in this game with this patch as well.

http://www.tomshardware.com/reviews/amd-ryzen-7-1700x-review,4987-6.html - You can see Rise of the Tomb Raider DX12 results here, you can see that Ryzen was performing really out of whack here, the 1700X was barely faster than an FX-8350.

Didn't Adored do a whole thing on this problem a month ago?

wargames
Mar 16, 2008

official yospos cat censor

eames posted:

... and terrible performance because of virtualization bugs that AMD won't acknowledge. :(

https://community.amd.com/thread/215931

The promised fix (AGESA 1006) doesn't help with NPT performance, it only fixes groupings to enable GPU passthrough.

They know of immou groups issues and are actively working on it.

SwissArmyDruid
Feb 14, 2014

by sebmojo
I'm hoping someone can tell us how many of the pins on those new Threadripper sockets are unused.

eames posted:

... and terrible performance because of virtualization bugs that AMD won't acknowledge. :(

https://community.amd.com/thread/215931

The promised fix (AGESA 1006) doesn't help with NPT performance, it only fixes groupings to enable GPU passthrough.

Won't acknowledge, or haven't had the time to address, with Computex and Threadripper taking a higher priority?

eames
May 9, 2009

SwissArmyDruid posted:

I'm hoping someone can tell us how many of the pins on those new Threadripper sockets are unused.


Won't acknowledge, or haven't had the time to address, with Computex and Threadripper taking a higher priority?

We'll find out soon enough I guess.

Speaking of Threadripper, Linus dropped some pretty heavy hints that Threadripper is two Ryzen 7 cores on a single package. I always knew that the architecture would be modular but having what appears to be two identical sockets merged into one was eye opening. Hopefully performance won't suffer too much if it's an SMP-like architecture.

https://www.youtube.com/watch?v=nynBEGCtg90

wargames
Mar 16, 2008

official yospos cat censor

eames posted:

We'll find out soon enough I guess.

Speaking of Threadripper, Linus dropped some pretty heavy hints that Threadripper is two Ryzen 7 cores on a single package. I always knew that the architecture would be modular but having what appears to be two identical sockets merged into one was eye opening. Hopefully performance won't suffer too much if it's an SMP-like architecture.

https://www.youtube.com/watch?v=nynBEGCtg90

If the interposer is pretty expandable it shouldn't be an issue, and amount to huge savings.

Malcolm XML
Aug 8, 2009

I always knew it would end like this.

eames posted:

We'll find out soon enough I guess.

Speaking of Threadripper, Linus dropped some pretty heavy hints that Threadripper is two Ryzen 7 cores on a single package. I always knew that the architecture would be modular but having what appears to be two identical sockets merged into one was eye opening. Hopefully performance won't suffer too much if it's an SMP-like architecture.

https://www.youtube.com/watch?v=nynBEGCtg90

Uh of course it is

Zen is built using 8c zeppelin modules connected by infinity fabric

The bigass 32c thing is effectively a 4 chip module


And yes NUMA will be an issue but just wait until poo poo gets patched and it'll be fine

Adbot
ADBOT LOVES YOU

Malcolm XML
Aug 8, 2009

I always knew it would end like this.

wargames posted:

If the interposer is pretty expandable it shouldn't be an issue, and amount to huge savings.

Bingo this allows for lots of cost savings on construction since big dies are crazy expensive

Intel is likely gonna do the same thing with EMIB

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