Intel: lol

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > Intel: lol

WhyteRyce: Dec 30, 2001

LRADIKAL posted:

This got started because someone said something to the effect of "...Keller realized Musk is a loving idiot..."

What's your point? Amazon ran a deficit for years, so maybe he's more of a Bezos, except, he's beating Bezos (and ULA, and Russia) with his rocket company. So I guess I'll make the claim that god head Elon will turn Tesla into the most important car company in the history of man. Do you agree that Musk is a loving idiot, or is it that not all of his businesses are profitable?

His rocket company being successful has nothing to do with his car company or how the car company is run

# ? Apr 27, 2018 22:02

Adbot: ADBOT LOVES YOU

# ? Apr 27, 2024 23:59

LRADIKAL: Jun 10, 2001; Fun Shoe

Nor does it have to do with his intelligence.

# ? Apr 27, 2018 22:36

Fame Douglas: Nov 20, 2013; by Fluffdaddy

Intel shouldn't have hired Musk to do their fab work. No wonder 10 nm is delayed again.

# ? Apr 27, 2018 22:39

mewse: May 2, 2006

Fame Douglas posted:

Intel shouldn't have hired Musk to do their fab work. No wonder 10 nm is delayed again.

Intel is a loving idiot

# ? Apr 27, 2018 22:42

Cygni: Nov 12, 2005; raring to post

LRADIKAL posted:

Nor does it have to do with his intelligence.

shuttup elon

# ? Apr 27, 2018 22:48

ufarn: May 30, 2009

How are Intel's new ML chips doing, speaking of? They sounded pretty cool, but I don't know whether they're seeing major adoption, assuming it's even released.

# ? Apr 28, 2018 00:11

Inept: Jul 8, 2003

LRADIKAL posted:

Nor does it have to do with his intelligence.

quote:

Musk countered that this was one reason we needed to colonize Mars�so that we�ll have a bolt-hole if A.I. goes rogue and turns on humanity.

quote:

He told Bloomberg�s Ashlee Vance, the author of the biography Elon Musk, that he was afraid that his friend Larry Page, a co-founder of Google and now the C.E.O. of its parent company, Alphabet, could have perfectly good intentions but still �produce something evil by accident��including, possibly, �a fleet of artificial intelligence-enhanced robots capable of destroying mankind.�

He is dumb and crazy.

# ? Apr 28, 2018 01:57

GRINDCORE MEGGIDO: Feb 28, 1985

Inept posted:

He is dumb and crazy.

He likes recycling theories and making them dumber.

# ? Apr 28, 2018 02:10

JawnV6: Jul 4, 2004; So hot ...

Inept posted:

He is dumb and crazy.

Are we still debating if the son of a guy who bought an emerald mine on a whim has some unspecified intelligence or is just lucky? Guess I made the dumb choice with non-emerald-mine-owning parents.

# ? Apr 28, 2018 02:43

PerrineClostermann: Dec 15, 2012; by FactsAreUseless

This sounds like the right wing press in here with all the Musk hate.

# ? Apr 28, 2018 03:17

fishmech: Jul 16, 2006; by VideoGames; Salad Prong

PerrineClostermann posted:

This sounds like the right wing press in here with all the Musk hate.

Ah yes, the right wing press, famous for despising racism, supporting workers rights, and wanting the drat fool to not massively gently caress up building a car.

:confused:

# ? Apr 28, 2018 03:39

PerrineClostermann: Dec 15, 2012; by FactsAreUseless

fishmech posted:

Ah yes, the right wing press, famous for despising racism, supporting workers rights, and wanting the drat fool to not massively gently caress up building a car.

Have you not seen all the crying over how Musk is a scamming conman talking about green energy and improving the world to steal tax payer money while his companies fail and produce zero results?

# ? Apr 28, 2018 03:56

fishmech: Jul 16, 2006; by VideoGames; Salad Prong

PerrineClostermann posted:

Have you not seen all the crying over how Musk is a scamming conman talking about green energy and improving the world to steal tax payer money while his companies fail and produce zero results?

No, I haven't seen your favorite dumb conservatives because I've been busy seeing this:

And uh yeah Tesla can't make their stupid "autopilot" work, they can't produce their Model 3 cars. And beyond that, Elon's hyperloop poo poo doesn't work and can't really work, his Boring Company is horrible and SpaceX mainly gets by because of the people running it having more freedom from his meddling than the other companies do.

All this bullshit is why it was totally the right move for the guy to ditch Tesla for an Intel offer.

# ? Apr 28, 2018 04:07

LRADIKAL: Jun 10, 2001; Fun Shoe

Ha! Now do the same style take down for Intel!

# ? Apr 28, 2018 04:38

Anarchist Mae: Nov 5, 2009; by Reene; Lipstick Apathy

fishmech posted:

TLDR; dude, bro, did you know capitalism is bad and exploits people???

# ? Apr 28, 2018 05:02

Canned Sunshine: Nov 20, 2005; CAUTION: POST QUALITY UNDER CONSTRUCTION

fishmech posted:

And beyond that, Elon's hyperloop poo poo doesn't work and can't really work

The hyper loop does work? :confused:

I think there's potential questions behind the passenger-related economics of it given the small capsule sizes, but it's scientifically feasible and it's really just a matter of getting the engineering balanced with the capital costs.

# ? Apr 28, 2018 05:07

fishmech: Jul 16, 2006; by VideoGames; Salad Prong

SourKraut posted:

The hyper loop does work? I think there's potential questions behind the passenger-related economics of it given the small capsule sizes, but it's scientifically feasible and it's really just a matter of getting the engineering balanced with the capital costs.

They've built scale models over short distances that don't carry passengers, and haven't worked out any of the issues like "a single puncture compromises the whole length's partial vacuum, which will then destroy at least a few hundred feet around the puncture if not for miles and miles around" or "it's difficult to fit air pressurization into the small pods to ensure safe cabin pressure unless you allow the pressure in the tunnel to be much higher than desired for the high speeds". That's before we get into things like the cost effectiveness or profitability or even whether the raw concept has any real advantage over simply building a normal high speed rail track which doesn't need hundreds of miles of airtight tubing to never ever get damaged or else the whole line's rendered useless.

In other words it's like the promised 12 GHz single core x86 chips of the early 2000s, but with transportation.

# ? Apr 28, 2018 05:23

PerrineClostermann: Dec 15, 2012; by FactsAreUseless

fishmech posted:

No, I haven't seen your favorite dumb conservatives because I've been busy seeing this:

And uh yeah Tesla can't make their stupid "autopilot" work, they can't produce their Model 3 cars. And beyond that, Elon's hyperloop poo poo doesn't work and can't really work, his Boring Company is horrible and SpaceX mainly gets by because of the people running it having more freedom from his meddling than the other companies do.

All this bullshit is why it was totally the right move for the guy to ditch Tesla for an Intel offer.

You think I read conservative news? It gets posted in every drat Elon-related thread everywhere, dumbass boomers whining about how horrible Musk is, how he has no actual achievements and uses others to suckle at the government teat.

This thread is full of the same.

# ? Apr 28, 2018 06:41

Canned Sunshine: Nov 20, 2005; CAUTION: POST QUALITY UNDER CONSTRUCTION

*** siren *** effort post *** siren ***

fishmech posted:

They've built scale models over short distances that don't carry passengers

Yes, that's how development and engineering works. I work in the water/wastewater treatment industry, and all new technologies gets small-scale pilot testing and go through years of validation before attempted at a large-scale, so I just have a different perspective and don't see the alarm yet. And fyi - they have conducted full-scale tests, though yes, over short distances.

fishmech posted:

and haven't worked out any of the issues like "a single puncture compromises the whole length's partial vacuum, which will then destroy at least a few hundred feet around the puncture if not for miles and miles around"

This is solvable too, with the same approach used for piping that might see vacuum: you bury or encase it. It won't be the "sexy" exterior appearance that Musk initially wanted/touted, but as demonstrated by recent permitting for underground installations, I think they've realized this is likely the way to go. Concrete encasement would be ideal, but obviously very expensive. But encasing it completely eliminates your points above.

The concerns about potential cascade failures due to a single puncture are reasonable, but again, that's solvable as well. As a reminder, the system is a near-vacuum, but technically it's still positive pressure. Of course the exterior atmospheric pressure is much higher, which is why most refer to it as a "vacuum". In the event of a puncture, exterior air will attempt to fill the vacuum, and the flowrate of air is limited by a maximum pressure differential across a given orifice size such that without adequate venting, the structural failure of the tube either at the puncture or at another weak point that you reference would occur. However, again, you could do the same thing you do with piping: provide vacuum relief, i.e. engineered locations to safely eliminate the vacuum at designed locations along the hyper loop path. This is actually fairly easily achievable with design approaches and equipment already available, so I wouldn't call this a major obstacle.

This would obviously result in the pods coming rapidly to a halt, so that's obviously the greater concern, but again, that can be addressed as well (such as extendable emergency brakes, which I believe they're already working on).

fishmech posted:

or "it's difficult to fit air pressurization into the small pods to ensure safe cabin pressure unless you allow the pressure in the tunnel to be much higher than desired for the high speeds".

I don't even understand what you're trying to say here. Scientists and engineers have already dealt with both extremes of this (the near-vacuum of space with pressurized space capsules, and deep sea submersibles seeing significantly high exterior pressures relative to the internal pressure). If you're referring to the use of higher pressure air behind the capsule for thrust, a dedicated air handling unit can probably accommodate this. Of course powering it becomes an issue and it obviously adds size and weight, but again, it's achievable.

fishmech posted:

That's before we get into things like the cost effectiveness or profitability or even whether the raw concept has any real advantage over simply building a normal high speed rail track which doesn't need hundreds of miles of airtight tubing to never ever get damaged or else the whole line's rendered useless.

The cost effectiveness and profitability are the biggest obstacles, I fully agree.

In terms of "any real advantage", the current belief that I've seen amongst the hyper loop designers is that it can hit several hundred miles per hour, while the fastest high speed trains using maglev technology are typically a third of that. So yes, that's a definitely advantage.

And maglev itself utilizes specialized "track", so it's not as if you can simply re-route the train if the track does get damaged. If you're referring to trains in the 100-150 mph range that utilized existing track, then yes, they do have that flexibility. But again, you're talking 1/5th or so the potential speed of the hyper loop, and that's on straight runs only, so even it has its own significant limitations. And even it often costs dozens of billions of dollars to implement.

fishmech posted:

In other words it's like the promised 12 GHz single core x86 chips of the early 2000s, but with transportation.

Except I don't think we ever saw a 12 GHz x86 chip, while we have seen full-scale pilot testing of the hyper loop, albeit for short distances.

So fishmech, I'm pretty sure you haven't really thought or researched a lot of your talking points. Care to revisit it after you have a chance to?

# ? Apr 28, 2018 06:45

Anime Schoolgirl: Nov 28, 2002

maybe we don't need the merged x86 thread after all lmao

# ? Apr 28, 2018 07:09

fishmech: Jul 16, 2006; by VideoGames; Salad Prong

SourKraut posted:

*** siren *** effort post *** siren ***

Yes, that's how development and engineering works. I work in the water/wastewater treatment industry, and all new technologies gets small-scale pilot testing and go through years of validation before attempted at a large-scale, so I just have a different perspective and don't see the alarm yet. And fyi - they have conducted full-scale tests, though yes, over short distances.

This is solvable too, with the same approach used for piping that might see vacuum: you bury or encase it. It won't be the "sexy" exterior appearance that Musk initially wanted/touted, but as demonstrated by recent permitting for underground installations, I think they've realized this is likely the way to go. Concrete encasement would be ideal, but obviously very expensive. But encasing it completely eliminates your points above.

The concerns about potential cascade failures due to a single puncture are reasonable, but again, that's solvable as well. As a reminder, the system is a near-vacuum, but technically it's still positive pressure. Of course the exterior atmospheric pressure is much higher, which is why most refer to it as a "vacuum". In the event of a puncture, exterior air will attempt to fill the vacuum, and the flowrate of air is limited by a maximum pressure differential across a given orifice size such that without adequate venting, the structural failure of the tube either at the puncture or at another weak point that you reference would occur. However, again, you could do the same thing you do with piping: provide vacuum relief, i.e. engineered locations to safely eliminate the vacuum at designed locations along the hyper loop path. This is actually fairly easily achievable with design approaches and equipment already available, so I wouldn't call this a major obstacle.

This would obviously result in the pods coming rapidly to a halt, so that's obviously the greater concern, but again, that can be addressed as well (such as extendable emergency brakes, which I believe they're already working on).

I don't even understand what you're trying to say here. Scientists and engineers have already dealt with both extremes of this (the near-vacuum of space with pressurized space capsules, and deep sea submersibles seeing significantly high exterior pressures relative to the internal pressure). If you're referring to the use of higher pressure air behind the capsule for thrust, a dedicated air handling unit can probably accommodate this. Of course powering it becomes an issue and it obviously adds size and weight, but again, it's achievable.

The cost effectiveness and profitability are the biggest obstacles, I fully agree.

In terms of "any real advantage", the current belief that I've seen amongst the hyper loop designers is that it can hit several hundred miles per hour, while the fastest high speed trains using maglev technology are typically a third of that. So yes, that's a definitely advantage.

And maglev itself utilizes specialized "track", so it's not as if you can simply re-route the train if the track does get damaged. If you're referring to trains in the 100-150 mph range that utilized existing track, then yes, they do have that flexibility. But again, you're talking 1/5th or so the potential speed of the hyper loop, and that's on straight runs only, so even it has its own significant limitations. And even it often costs dozens of billions of dollars to implement.

Except I don't think we ever saw a 12 GHz x86 chip, while we have seen full-scale pilot testing of the hyper loop, albeit for short distances.

So fishmech, I'm pretty sure you haven't really thought or researched a lot of your talking points. Care to revisit it after you have a chance to?

They're not conducting full scale tests when they can't even reach normal railed high speed train speeds, let alone the 600+ mile per hour claims they were making, especially when they don't even have a setup that can run between say two towns with prototype stations.

Just saying "bury it or encase it" doesn't mean you're going to successfully avoid the issues of the partial vacuum vessel having a hole blown through to the air, or an underground part shifted about by seismic activity breaking the vacuum etc.

Again the problems of the cascade failure aren't really solvable, just somewhat mitigatable.

I'm talking about the pods being able to stay comfortably pressurized for the planned multihour trips across long distances within the cramped pods and all that. Again, you don't even need to do that if you up the air pressure of the tunnel quite a ways, but then you get less of the speed benefit you built the thing for.

China already runs regular non-maglev high speed rail up to 300 miles per hour in real world testing on the public rails, and the maglevs they use there go up to 311. Japanese maglev models that aren't out yet but are in testing on actual track have hit 375 miles per hour, and the French were able to test on their conventional rail high speed trains at up to 357 miles per hour. Now I went and checked the testing with the hyperloop vehicles and they seemed to have a hard time getting to even 300 miles per hour right now, placing a good 57 miles per hour slower than the fastest train running on straight standard French high speed rail track, and 75 miles per hour slower than Japan's prototype maglev.

So, you could hire on some french engineers and equipment right now and go build a real conventional high speed rail line and run trains on it in revenue service at least up to 325 MPH for a safety margin, and probably finish building it before a practical hyperloop is even close to finished. Frankly I doubt that they'll ever acheived the touted speed of sound in a tunnel actually able to cover distance between cities, while that 325-350 mph conventional railway can be built today, hell it can be built with 3 year old technology.

I bring up 12 ghz precisely because it was a speed promised and never achieved -the last serious attempt was an AMD Bulldozer overclock test from 2011 where they held stable at 8.8 ghz with a rather elaborate cooling system for a short while. Plus again nah, we're not seeing a hyperloop delivering the speeds or passengers or anything that was promised.

# ? Apr 28, 2018 08:12

Lockback: Sep 3, 2006; All days are nights to see till I see thee; and nights bright days when dreams do show me thee.

This is a really stupid derail. Can you please take it to anywhere in GBS?

# ? Apr 28, 2018 14:30

zebez: Apr 27, 2008

Anime Schoolgirl posted:

maybe we don't need the merged x86 thread after all lmao

True.

# ? Apr 28, 2018 14:32

B-Mac: Apr 21, 2003; I'll never catch "the gay"!

Anime Schoolgirl posted:

maybe we don't need the merged x86 thread after all lmao

Best post here in recent memory.

# ? Apr 28, 2018 14:38

The Illusive Man: Mar 27, 2008; ~savior of yoomanity~

*Intel announced 10nm not shipping in volume until 2019 (if ever)*
*sees lots of posts in the Intel thread, assumes interesting discussion of Intel�s fortunes has ensued*
*�Oh...�*

I�m not looking forward to when Tesla actually reports their financials next week. :ohdear:

Has anything been confirmed about exactly *why* Intel�s 10nm process is so boned compared to the rest of the industry? Global Foundries and TSMC seem rather confident of hitting 7nm, while Intel is 3+ years past their target date and still sweating the ramp for 10nm.

E: I get that Intel�s 10nm is not directly equivalent to that of other fabs, but it�s still a pretty bad look.

# ? Apr 28, 2018 14:45

Fame Douglas: Nov 20, 2013; by Fluffdaddy

I'm excited to see the discussion about Elon Musk when Apple announces their switch to ARM processors.

# ? Apr 28, 2018 14:50

Khorne: May 1, 2002

Space Racist posted:

Has anything been confirmed about exactly *why* Intel�s 10nm process is so boned compared to the rest of the industry? Global Foundries and TSMC seem rather confident of hitting 7nm, while Intel is 3+ years past their target date and still sweating the ramp for 10nm.

7nm/10nm are marketing numbers. Intel's 10nm is equivalent to their 7nm.

I should put "roughly equivalent" because they are different processes, but the density and performance should be similar.

The biggest difference is probably that the 7nm coalition is borderline "everyone who isn't intel". It's a combination of IBM, Samsung, and Globalfoundries.

TSMC should have 7nm at production capacity around when Intel has 10nm up. If not sooner.

Khorne fucked around with this message at 15:03 on Apr 28, 2018

# ? Apr 28, 2018 14:52

ufarn: May 30, 2009

Someone hire Keller to fix this thread.

# ? Apr 28, 2018 15:14

crazypenguin: Mar 9, 2005; nothing witty here, move along

Space Racist posted:

Has anything been confirmed about exactly *why* Intel�s 10nm process is so boned compared to the rest of the industry? Global Foundries and TSMC seem rather confident of hitting 7nm, while Intel is 3+ years past their target date and still sweating the ramp for 10nm.

It's all rumors and guesswork at the moment, but the guesses I hear are:

Intel was pretty ambitious about 10nm (iirc, there's some important ways it's better than competitor's 7nm) and was trying to "prove Moore's law still held"
This had them focused a little too much on improving density, rather than improving costs, which is arguably more important.
Intel is pretty arrogant and huffed their own farts about their unassailable process leadership
Other foundries ran into trouble scaling down their processes, and---get this---it seems they worked together and shared a fair amount of work with each other to help each other figure out the problems. Intel probably considers that a weakness, lol.
Management gently caress ups.

# ? Apr 28, 2018 19:19

Deuce: Jun 18, 2004; Mile High Club

Lockback posted:

This is a really stupid derail.

# ? Apr 28, 2018 21:14

redeyes: Sep 14, 2002; by Fluffdaddy

the gently caress happened on this page of the thread?!

# ? Apr 28, 2018 21:57

Stanley Pain: Jun 16, 2001; by Fluffdaddy

Anime Schoolgirl posted:

maybe we don't need the merged x86 thread after all lmao

*nods sagely*

Things sure got interesting in here though.

# ? Apr 28, 2018 22:12

canyoneer: Sep 13, 2005; I only have canyoneyes for you

Fame Douglas posted:

Intel shouldn't have hired Musk to do their fab work. No wonder 10 nm is delayed again.

nanofab will double capacity every 12 months until orders are filled!

# ? Apr 28, 2018 22:28

GRINDCORE MEGGIDO: Feb 28, 1985

I thought thread got good. What is Intel most likely to get Keller working on?

# ? Apr 29, 2018 02:53

Malcolm XML: Aug 8, 2009; I always knew it would end like ｔｈｉｓ．

GRINDCORE MEGGIDO posted:

I thought thread got good. What is Intel most likely to get Keller working on?

Infinity fabric and modular socs

# ? Apr 29, 2018 03:34

WhyteRyce: Dec 30, 2001

He's going up revive that DVR project

# ? Apr 29, 2018 03:36

isndl: May 2, 2012; I WON A CONTEST IN TG AND ALL I GOT WAS THIS CUSTOM TITLE

Jim's gonna whip up an ARM blueprint from scratch before riding off into the sunset as Intel abandons x86 for their coming decade of ARMCORE technology.

# ? Apr 29, 2018 04:04

BIG HEADLINE: Jun 13, 2006; "Stand back, Ottawan ruffian, or face my lumens!"

Or they're just going to throw a lot of money at him in the hopes he can find a shortcut around the "Demon at 7nm." Or 10nm in their case.

# ? Apr 29, 2018 05:10

Laslow: Jul 18, 2007

Jot this little nugget down, wall street: It's obviously Itanium 3.

# ? Apr 29, 2018 06:20

Adbot: ADBOT LOVES YOU

# ? Apr 27, 2024 23:59

movax: Aug 30, 2008

That was an interesting (and honestly entertaining) derail to read waiting in line for the Victoria Peak tram but let�s say it�s run its course and cherish these page numbers forever.

Topic chat: after writing off say 10% of chip area for test/debug stuff, what percentage is actually still OG x86 stuff now that we�ve got area eaten for video decode/encode and other functions? Have we lost die area to ISPs yet? With all the SIMD extensions getting a dedicated DSP a la Hexagon is not likely but I�m wondering what the next IP core to get laid down is. Video will continue forever, mobile style ISP I doubt unless there�s a serious push to get into mobile (lol), audio isn�t that demanding and it�s on the PCH, PCIe RC is a first-class citizen in CPU land...

# ? Apr 29, 2018 10:43

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > Intel: lol

«‹›740 »