As game developers, we call our automated agents "AI" in a long tradition of overloading and bastardizing words from other fields, but we all understand it's not actually real "AI" of any sort. I mean, even pathfinding goes under the term "AI" for our purposes. So, yeah, this is deep learning, but no more "AI" than what we do inside the games. Very often, we actually have to work to make our opponents *less* effective, because computers have so many advantages over players, especially in any game at all where reflexes count, or broad analysis of lots of details is important.
But more to the point, Elon keeps talking about regulating AI to prevent it from destroying the world. Every time he talks about this, he sounds like an unhinged lunatic that has some irrational fears about something he doesn't deeply understand. I still haven't heard a realistic scenario about how AI is going to go about doing this. And let's be honest... the perception of his capacity for rational thought on matters outside his domain of expertise was NOT helped by his declaration that he's 99.9% certain we're living in a computer simulation.
Besides which, how exactly would one "regulate" this, short of simply banning AI development by private enterprises? Massive governmental oversight requiring a programmer to pinky swear or sign in blood that they'll use those neural networks for good instead of evil? I honestly don't get it.
There are plenty of home-grown cynics, now vastly expanded in numbers since the US started wholesale spying on its own citizens, and even now continues to fight to push backdoors into our encryption "for our own protection." I'd posit that our three-letter agencies and colluding leadership have done far more to alienate its own citizens' trust in their government than the collective effort of all the paid 50 Cent Army shills over the past decade.
Continuing skepticism in the face of overwhelming evidence doesn't make you look clever or thoughtful. It makes you look like a partisan zealot refusing to admit when you're wrong. Here, you can read it right on their own Twitter feed:
The scientists who authored the report are confused as to why this is a story, and you're going to quibble over "publicly available" semantics? No, it was put up on public sites for "public comments". That's hardly equivalent to a "leak", and you damn well know it. I can't believe you're seriously making that argument. Or else, you're doing an awesome job of trolling me right now.
And seriously, "fake news" is an alt-right conspiracy? Are you kidding me? That was a narrative started by the left/mainstream media to explain how Hillary could have possibly lost the election. But the left gets hoist by their own petard whenever something like this happens, as reporters occasionally demonstrate a complete disregard for the most basic fact-checking before they breathlessly run a story that just happens to show the current administration in a bad light. We saw it happen with the Washington Post and the "Russians hacking the electrical grid" nonsense (they didn't even bother contacting Burlington Electric for a statement), and we see it now again with the non-existent fact-checking of the New York Times before publishing a completely made-up story.
Would I call that "fake news"? No, not really. Just plain bad journalism. But in the end, it really amounts to the same thing - misinformation presented as news.
“It's not clear what the news is in this story,” Robert Kopp, a climate scientist at Rutgers University who is listed on the report as among the lead authors, said on Twitter.
Another scientist who authored the report, Katharine Hayhoe, a professor at Texas Tech who leads the school’s Climate Science Center, also emphasized that the report is already publicly available.
“Important to point out that this report was already accessible to anyone who cared to read it during public review & comment time,” she tweeted. “Few did.”
This is not some "leaked report". New York Times reporting at its finest, trying to drum up controversy where none exists. The report isn't fake. Just the reporting.
You forgot Cinemax. Yes, I covered that with "plus a number of smaller specialty channels", in case you missed it. And you make all those options sound like a bad thing. I disagree. I'm not interested in most of that, and I'm thankful I don't have to pay for the channels I don't want. On the other hand, the specialty channels are awesome at giving you lots of what you DO want for a reasonable price.
I subscribe, like you, to Netflix and Prime, which gives me plenty of "normal" TV. I also subscribe to a couple of anime channels, which again, gives me a smorgasbord of anime to watch at any time. All this for under $35 a month, which is far less than what I'd pay for basic cable, and all for a much better TV experience.
What's more, I can choose to pay any range I want, from $10 a month up to $150 (or more?) if I was a television glutton, or just had infinite money to spend on stuff I didn't need. This sure seems like "a la carte" to me. A perfect system? Of course not. But far better, in my opinion, and demonstrably cheaper than cable has been for the last decade or so.
Yes, a la carte was what we wanted. But we didn't want it to come with dozens of monthly bills, we wanted to still have just a single bill.
We don't have "dozens". At the moment, we have FOUR. There's Netflix, Hulu, Amazon Prime, and YouTube as the current big four, plus a number of smaller specialty channels. Now add Disney as well, if it succeeds. But most people are not planning to subscribe to ALL of them, just as few people subscribed to ALL the channels available on cable.
If you were getting a single bill, it would mean there wasn't any competition between these services. A single monolithic company that serves everything you want is a pipe dream. That's what the cable companies were, and that's exactly WHY a la carte programming was never offered. I'll take the lessor of the two evils, I think.
Really, all this fragmentation will kill the business. Without one stop shopping and reasonable prices, it's better to just go back to bittorrent.
I'm not sure I agree. Isn't "a la carte" what we wanted and were unable to get with the cable companies for many years? Well, now we have it. These individual services have to compete with the current price models and with each other, meaning that you can still subscribe to four services easily and still pay far less than the vast majority of cable packages.
I don't think we'll ever have more than half a dozen major services, and the rest will be highly specialized or also-rans. I'm totally fine with that, so long as the services don't collude to bring their prices up collectively.
A better title would be, "Researchers fool Google's TensorFlow library in laboratory tests".
As it turns out, they did NOT test this against actual self-driving vehicle image recognition, but a generic deep neural network library. This seemed obvious, as there are still no commercially available fully autonomous vehicles, but I skimmed the paper to confirm it.
There was another issue I noticed as well. They resized all their training images down to 32x32 pixels. I admit I'm no expert in neural networks, but this seems like it would greatly favor the ability to fool classification algorithms. Maybe someone more knowledgeable can correct me if I'm off base here. Still, my suspicion seems to be confirmed by this little gem:
"Our final classifier accuracy was 91% on the test dataset."
So, their baseline algorithm only worked properly slightly better than 9/10 times. Should we believe that this represents the state of the art that will be applied in actual self-driving vehicles? It seems like the researchers didn't even have a highly robust classifier from the start.
I believe the merits of the paper lie in demonstrating this as a theoretical concern, but this should in no way be construed to represent a definitive threat against actual vehicle systems. You can't necessarily blame the researchers for the crappy headline, of course, as the title is "Robust Physical-World Attacks on Machine Learning Models". But I wouldn't necessarily rate this as the most robust research I've ever seen either.
No, I'm talking about how there's a position called "VP of Diversity", and the dogma that gender imbalances MUST be caused somehow by latent sexism - nevermind that most of the workers are good, open-minded, progressive liberals. The notion that perhaps women, as a group, aren't as interested in careers as programmers can't possibly be entertained, because such a notion would undermine many careers built around promoting diversity.
I agree with some of the conclusions in the memo, but not the way he argued his point. This is a common failing with many men. They often argue points dispassionately and logically, but human beings are not completely dispassionate or logical (not just limited to women). The feelings of your co-workers matter. Companies that allow bad apples to get away with anything they want simply because they are "top performers" are beyond contemptible (see: Uber).
So, yes, he "dispassionately" argued his points, all backed up by carefully sourced scientific data, and ended up inflaming many passions who didn't care about those scientific sources, and it cost him his job. Personally, I felt that the firing was an over-reaction, not to mention the intolerance displayed by people who were so repulsed by a different viewpoint that they would have refused to work with him. But I wouldn't argue that Google didn't have the right to fire him over the brouhaha he caused.
Seriously, stop blaming conservatives for this. I've personally upmodded your comments even when I disagreed with them because I saw that happening to you. Ironically, I believe you also complained about conservatives modding you down in that same thread, which was a bit galling. I've had plenty of occasions when I've been modded down as well for my own views. It's not just you. It happens both ways, and it's equally annoying./end rant
Do you really think Google's VP of Diversity gives two figs about scientific studies? Pointing out an inconvenient fact will get you fired. That's the reality.
This is political correctness taken to it's (il)logical extreme. He called women names and made them feel bad about themselves. Therefore, he deserves execution, but unfortunately, there seem to be laws against that. The worst they can do is fire him, as it turns out.
Well, here's what Wikipedia had to say about this trait.
Neuroticism: (sensitive/nervous vs. secure/confident). The tendency to experience unpleasant emotions easily, such as anger, anxiety, depression, and vulnerability. Neuroticism also refers to the degree of emotional stability and impulse control and is sometimes referred to by its low pole, "emotional stability". A high need for stability manifests itself as a stable and calm personality, but can be seen as uninspiring and unconcerned. A low need for stability causes a reactive and excitable personality, often very dynamic individuals, but they can be perceived as unstable or insecure.
Okay, so he called women "emotionally unstable" or "insecure". I'm not certain how reading the Wikipedia definition makes that sound any better.
Oh damn, I can see why he was fired. I just read his memo in full, and while I don't necessarily agree with him on many of his points (even as a fairly conservative male), he made a lot of sense in some areas.
But "biological differences" between the sexes is a massive hot-button topic. I can't speak for gender diversity advocates, but I suspect they believe that's used as an excuse for perpetuating injustice and sexism. Many males, on the other hand, likely view it as a statistical or scientific issue, and don't see how we could possibly *ignore* biological differences when looking at root causes of gender inequality in many fields.
If I had to say what it was the put it over the top in terms of firing, I'd have to say when he claimed women suffered from a greater degree of neuroticism than men. Honestly, given most of the fucked-up personalities I've worked with have been men, it's hard for me to figure out how he came to that conclusion. Pointing out that women tend to be less assertive, leading to fewer promotions and leadership positions probably didn't do him much of a favor either.
Anyhow, after reading that paragraph, I concluded he'd have been less likely to be fired if he spit a loogie at the VP of Diversity.
It's not all that surprising that gaming benchmarks don't scale as well to large numbers of cores. Videogame programming isn't a field in which performance can simply scale nearly linearly based on the number of hardware threads available. That's because the CPU is performing a huge number of very diverse tasks among all it's engine components, and there's a great deal of global coordination that occurs on a central database. It's essentially a heterogeneous workload, and those just don't scale as well.
Thinks like a 3D renderer or video encoding benchmark, in which you can divide up portions of the screen or encode successive frames on different threads - those sorts of things will scale nearly linearly with the number of threads, because it's a largely homogeneous workload.
So, it's not necessarily about synthetic vs read world benchmarks. It's really about how well the application in general scales to multiple threads. For some things, it's relatively easy. Others... not so much.
Someone should ask the leader of those nutjobs (a retired firefighter) if he thinks it's nefarious that firefighters go around intentionally burning houses down. What? Yep, houses that are scheduled for demolition are occasionally burned down by the fire department. Because, you know... they need to actually practice their craft under as realistic circumstances as they can manage. It's sort of a win-win, as it saves on demolition costs for the owner in addition to helping train local community firefighters.
It's the same thing here. The military and firefighters are perhaps a bit similar in that they don't typically have enough real life "on the job" training, and when it happens, it's very dangerous. So, they need to do a lot of training and simulations to prepare professionally as best they can.
A moron navigating with a GPS is going to still be a moron navigating with a map and compass. It's not the tools that turn people into idiots. Plenty of people got lost or had a bad sense of direction before the age of GPS navigation. This is not a new phenomenon.
Not to mention Whidbey Island Naval Air Station. There's a LOT of civilian and military air traffic around here. I'm having a hard time picturing most native Seattlites being all that concerned about this. And I'd imagine the current administration has better things to do than to order military flights to mess with the Seattle populous.
I'm not sure why people are surprised by this. SIGINT is part of the the modern military's war-fighting capabilities. Yes, they're probably monitoring Seattle's communications to practice interception of and analyzing said traffic. That they'd be training their people seems like a perfectly plausible explanation to me.
What next, people freaking out that out in the desert somewhere, the military is practicing with guns, bombs, rockets, and other weapons that can kill large numbers of people? Or that the military flies simulated bombing runs over civilian targets? This is what they're paid to do - preparing for all aspects of warfare.
What are you going to ditch Linux for? Windows, MacOS?
No, the rule is you have to switch to something with less marketshare. That leaves FreeBSD. Once that project offends in some manner or gains too much mainstream support, it's off to GNU HURD.
Apple earns its billions from hardware sales. You're thinking of a different megacorp that injects advertisements into everything as it's primary revenue source.
I've got a six year old i7 (four cores), as my primary Windows dev machine. I also have other machines for porting to Mac and Linux, plus a music/audio workstation. All use Intel processors. I went independent a few years ago, so I don't have a lot of cash to burn on new hardware. These machines have to last me until I (hopefully) get my game out the door soon and start earning an income again. They work fine for me, since my game is targeting lower-spec machines anyhow.
I've always pretty much stuck with Intel for my CPUs. Even so, I'll certainly check out AMD CPUs when its time to upgrade, but so far, they really haven't produced anything that can compete effectively with Intel. I'd love to see that change, as competition is good for everyone.
Until Ryzen takes over and 6 to 8 cores and hyper-threaded goodness become standard then why bother optimizing?
It's important not to conflate average numbers on Steam versus what your game's particular demographic tends to use. When you say "Steam gamers", for instance, that's pretty much meaningless, as it's just an average across a *very* wide swath of gamers and gaming hardware.
The latest AAA games will not simply not run well on a dual core laptop with 4GB of RAM and an integrated Intel GPU. The target audience for those types of games tends to be someone who has a moderately powerful desktop machine or a *very* beefy laptop. Those machines are likely to have 4 cores and eight hardware threads at *minimum*. So, threading optimization is obviously very much worthwhile for these types of games.
Now that I've gone indie (I left my professional job about four years ago), my own game's target hardware is much more modest. So, in that case, I don't have to put in the same degree of optimization effort in my own engine as large studios do, since I'm a one man studio and can't put in the resources they do. There are plenty of older or lightweight indie games that those WalMart laptops will run just fine.
In short, it really all depends on your game's minimum requirements and intended target audience.
As a professional game developer, I can assure you that videogames are the exact opposite of an "embarrassingly parallel" problem. There's a LOT more to videogames than graphical processing: AI, pathfinding, physics, audio, resource and memory management, animation, world updates / occlusion systems, etc. These are all CPU-intensive subsystems, and many of them are inherently bound to global data (a virtual world simulation), which makes it extremely tricky to split off into independent threads (probably the only exception being audio).
That means that each subsystem in the game must be carefully and painstakingly optimized for threaded performance. It's not possible to trivially split up all these subsystems by thread either. Many of these systems tend to interact with each other and the global world database, and that means the gains tend to be smallish and non-scalable in nature.
Generally speaking, it's an extremely difficult problem, and one which I don't think the industry has really cracked yet. Believe me, if it were trivial to do, we've have done it a long time ago.
Slashdot Mirror
As game developers, we call our automated agents "AI" in a long tradition of overloading and bastardizing words from other fields, but we all understand it's not actually real "AI" of any sort. I mean, even pathfinding goes under the term "AI" for our purposes. So, yeah, this is deep learning, but no more "AI" than what we do inside the games. Very often, we actually have to work to make our opponents *less* effective, because computers have so many advantages over players, especially in any game at all where reflexes count, or broad analysis of lots of details is important.
But more to the point, Elon keeps talking about regulating AI to prevent it from destroying the world. Every time he talks about this, he sounds like an unhinged lunatic that has some irrational fears about something he doesn't deeply understand. I still haven't heard a realistic scenario about how AI is going to go about doing this. And let's be honest... the perception of his capacity for rational thought on matters outside his domain of expertise was NOT helped by his declaration that he's 99.9% certain we're living in a computer simulation.
Besides which, how exactly would one "regulate" this, short of simply banning AI development by private enterprises? Massive governmental oversight requiring a programmer to pinky swear or sign in blood that they'll use those neural networks for good instead of evil? I honestly don't get it.
There are plenty of home-grown cynics, now vastly expanded in numbers since the US started wholesale spying on its own citizens, and even now continues to fight to push backdoors into our encryption "for our own protection." I'd posit that our three-letter agencies and colluding leadership have done far more to alienate its own citizens' trust in their government than the collective effort of all the paid 50 Cent Army shills over the past decade.
You say "watered down", I say "liquid cooled".
Continuing skepticism in the face of overwhelming evidence doesn't make you look clever or thoughtful. It makes you look like a partisan zealot refusing to admit when you're wrong. Here, you can read it right on their own Twitter feed:
https://twitter.com/KHayhoe/st...
https://twitter.com/KHayhoe/st...
https://twitter.com/bobkopp/st...
https://twitter.com/bobkopp/st...
The scientists who authored the report are confused as to why this is a story, and you're going to quibble over "publicly available" semantics? No, it was put up on public sites for "public comments". That's hardly equivalent to a "leak", and you damn well know it. I can't believe you're seriously making that argument. Or else, you're doing an awesome job of trolling me right now.
And seriously, "fake news" is an alt-right conspiracy? Are you kidding me? That was a narrative started by the left/mainstream media to explain how Hillary could have possibly lost the election. But the left gets hoist by their own petard whenever something like this happens, as reporters occasionally demonstrate a complete disregard for the most basic fact-checking before they breathlessly run a story that just happens to show the current administration in a bad light. We saw it happen with the Washington Post and the "Russians hacking the electrical grid" nonsense (they didn't even bother contacting Burlington Electric for a statement), and we see it now again with the non-existent fact-checking of the New York Times before publishing a completely made-up story.
Would I call that "fake news"? No, not really. Just plain bad journalism. But in the end, it really amounts to the same thing - misinformation presented as news.
“It's not clear what the news is in this story,” Robert Kopp, a climate scientist at Rutgers University who is listed on the report as among the lead authors, said on Twitter.
Another scientist who authored the report, Katharine Hayhoe, a professor at Texas Tech who leads the school’s Climate Science Center, also emphasized that the report is already publicly available.
“Important to point out that this report was already accessible to anyone who cared to read it during public review & comment time,” she tweeted. “Few did.”
This is not some "leaked report". New York Times reporting at its finest, trying to drum up controversy where none exists. The report isn't fake. Just the reporting.
You forgot Cinemax. Yes, I covered that with "plus a number of smaller specialty channels", in case you missed it. And you make all those options sound like a bad thing. I disagree. I'm not interested in most of that, and I'm thankful I don't have to pay for the channels I don't want. On the other hand, the specialty channels are awesome at giving you lots of what you DO want for a reasonable price.
I subscribe, like you, to Netflix and Prime, which gives me plenty of "normal" TV. I also subscribe to a couple of anime channels, which again, gives me a smorgasbord of anime to watch at any time. All this for under $35 a month, which is far less than what I'd pay for basic cable, and all for a much better TV experience.
What's more, I can choose to pay any range I want, from $10 a month up to $150 (or more?) if I was a television glutton, or just had infinite money to spend on stuff I didn't need. This sure seems like "a la carte" to me. A perfect system? Of course not. But far better, in my opinion, and demonstrably cheaper than cable has been for the last decade or so.
Yes, a la carte was what we wanted. But we didn't want it to come with dozens of monthly bills, we wanted to still have just a single bill.
We don't have "dozens". At the moment, we have FOUR. There's Netflix, Hulu, Amazon Prime, and YouTube as the current big four, plus a number of smaller specialty channels. Now add Disney as well, if it succeeds. But most people are not planning to subscribe to ALL of them, just as few people subscribed to ALL the channels available on cable.
If you were getting a single bill, it would mean there wasn't any competition between these services. A single monolithic company that serves everything you want is a pipe dream. That's what the cable companies were, and that's exactly WHY a la carte programming was never offered. I'll take the lessor of the two evils, I think.
Really, all this fragmentation will kill the business. Without one stop shopping and reasonable prices, it's better to just go back to bittorrent.
I'm not sure I agree. Isn't "a la carte" what we wanted and were unable to get with the cable companies for many years? Well, now we have it. These individual services have to compete with the current price models and with each other, meaning that you can still subscribe to four services easily and still pay far less than the vast majority of cable packages.
I don't think we'll ever have more than half a dozen major services, and the rest will be highly specialized or also-rans. I'm totally fine with that, so long as the services don't collude to bring their prices up collectively.
A better title would be, "Researchers fool Google's TensorFlow library in laboratory tests".
As it turns out, they did NOT test this against actual self-driving vehicle image recognition, but a generic deep neural network library. This seemed obvious, as there are still no commercially available fully autonomous vehicles, but I skimmed the paper to confirm it.
There was another issue I noticed as well. They resized all their training images down to 32x32 pixels. I admit I'm no expert in neural networks, but this seems like it would greatly favor the ability to fool classification algorithms. Maybe someone more knowledgeable can correct me if I'm off base here. Still, my suspicion seems to be confirmed by this little gem:
"Our final classifier accuracy was 91% on the test dataset."
So, their baseline algorithm only worked properly slightly better than 9/10 times. Should we believe that this represents the state of the art that will be applied in actual self-driving vehicles? It seems like the researchers didn't even have a highly robust classifier from the start.
I believe the merits of the paper lie in demonstrating this as a theoretical concern, but this should in no way be construed to represent a definitive threat against actual vehicle systems. You can't necessarily blame the researchers for the crappy headline, of course, as the title is "Robust Physical-World Attacks on Machine Learning Models". But I wouldn't necessarily rate this as the most robust research I've ever seen either.
No, I'm talking about how there's a position called "VP of Diversity", and the dogma that gender imbalances MUST be caused somehow by latent sexism - nevermind that most of the workers are good, open-minded, progressive liberals. The notion that perhaps women, as a group, aren't as interested in careers as programmers can't possibly be entertained, because such a notion would undermine many careers built around promoting diversity.
I agree with some of the conclusions in the memo, but not the way he argued his point. This is a common failing with many men. They often argue points dispassionately and logically, but human beings are not completely dispassionate or logical (not just limited to women). The feelings of your co-workers matter. Companies that allow bad apples to get away with anything they want simply because they are "top performers" are beyond contemptible (see: Uber).
So, yes, he "dispassionately" argued his points, all backed up by carefully sourced scientific data, and ended up inflaming many passions who didn't care about those scientific sources, and it cost him his job. Personally, I felt that the firing was an over-reaction, not to mention the intolerance displayed by people who were so repulsed by a different viewpoint that they would have refused to work with him. But I wouldn't argue that Google didn't have the right to fire him over the brouhaha he caused.
Seriously, stop blaming conservatives for this. I've personally upmodded your comments even when I disagreed with them because I saw that happening to you. Ironically, I believe you also complained about conservatives modding you down in that same thread, which was a bit galling. I've had plenty of occasions when I've been modded down as well for my own views. It's not just you. It happens both ways, and it's equally annoying. /end rant
Do you really think Google's VP of Diversity gives two figs about scientific studies? Pointing out an inconvenient fact will get you fired. That's the reality.
This is political correctness taken to it's (il)logical extreme. He called women names and made them feel bad about themselves. Therefore, he deserves execution, but unfortunately, there seem to be laws against that. The worst they can do is fire him, as it turns out.
Well, here's what Wikipedia had to say about this trait.
Neuroticism: (sensitive/nervous vs. secure/confident). The tendency to experience unpleasant emotions easily, such as anger, anxiety, depression, and vulnerability. Neuroticism also refers to the degree of emotional stability and impulse control and is sometimes referred to by its low pole, "emotional stability". A high need for stability manifests itself as a stable and calm personality, but can be seen as uninspiring and unconcerned. A low need for stability causes a reactive and excitable personality, often very dynamic individuals, but they can be perceived as unstable or insecure.
Okay, so he called women "emotionally unstable" or "insecure". I'm not certain how reading the Wikipedia definition makes that sound any better.
Oh damn, I can see why he was fired. I just read his memo in full, and while I don't necessarily agree with him on many of his points (even as a fairly conservative male), he made a lot of sense in some areas.
But "biological differences" between the sexes is a massive hot-button topic. I can't speak for gender diversity advocates, but I suspect they believe that's used as an excuse for perpetuating injustice and sexism. Many males, on the other hand, likely view it as a statistical or scientific issue, and don't see how we could possibly *ignore* biological differences when looking at root causes of gender inequality in many fields.
If I had to say what it was the put it over the top in terms of firing, I'd have to say when he claimed women suffered from a greater degree of neuroticism than men. Honestly, given most of the fucked-up personalities I've worked with have been men, it's hard for me to figure out how he came to that conclusion. Pointing out that women tend to be less assertive, leading to fewer promotions and leadership positions probably didn't do him much of a favor either.
Anyhow, after reading that paragraph, I concluded he'd have been less likely to be fired if he spit a loogie at the VP of Diversity.
It's not all that surprising that gaming benchmarks don't scale as well to large numbers of cores. Videogame programming isn't a field in which performance can simply scale nearly linearly based on the number of hardware threads available. That's because the CPU is performing a huge number of very diverse tasks among all it's engine components, and there's a great deal of global coordination that occurs on a central database. It's essentially a heterogeneous workload, and those just don't scale as well.
Thinks like a 3D renderer or video encoding benchmark, in which you can divide up portions of the screen or encode successive frames on different threads - those sorts of things will scale nearly linearly with the number of threads, because it's a largely homogeneous workload.
So, it's not necessarily about synthetic vs read world benchmarks. It's really about how well the application in general scales to multiple threads. For some things, it's relatively easy. Others... not so much.
I should have figured. /facepalm
Someone should ask the leader of those nutjobs (a retired firefighter) if he thinks it's nefarious that firefighters go around intentionally burning houses down. What? Yep, houses that are scheduled for demolition are occasionally burned down by the fire department. Because, you know... they need to actually practice their craft under as realistic circumstances as they can manage. It's sort of a win-win, as it saves on demolition costs for the owner in addition to helping train local community firefighters.
It's the same thing here. The military and firefighters are perhaps a bit similar in that they don't typically have enough real life "on the job" training, and when it happens, it's very dangerous. So, they need to do a lot of training and simulations to prepare professionally as best they can.
A moron navigating with a GPS is going to still be a moron navigating with a map and compass. It's not the tools that turn people into idiots. Plenty of people got lost or had a bad sense of direction before the age of GPS navigation. This is not a new phenomenon.
Not to mention Whidbey Island Naval Air Station. There's a LOT of civilian and military air traffic around here. I'm having a hard time picturing most native Seattlites being all that concerned about this. And I'd imagine the current administration has better things to do than to order military flights to mess with the Seattle populous.
I'm not sure why people are surprised by this. SIGINT is part of the the modern military's war-fighting capabilities. Yes, they're probably monitoring Seattle's communications to practice interception of and analyzing said traffic. That they'd be training their people seems like a perfectly plausible explanation to me.
What next, people freaking out that out in the desert somewhere, the military is practicing with guns, bombs, rockets, and other weapons that can kill large numbers of people? Or that the military flies simulated bombing runs over civilian targets? This is what they're paid to do - preparing for all aspects of warfare.
What are you going to ditch Linux for? Windows, MacOS?
No, the rule is you have to switch to something with less marketshare. That leaves FreeBSD. Once that project offends in some manner or gains too much mainstream support, it's off to GNU HURD.
Oh God. 200MB disk footprint, 1GB memory footprint, and requires an i7 processor or better to run without lagging when you type too fast.
It's destined to happen, I'm sure. Because everything's better when written in Javascript, right?
Apple earns its billions from hardware sales. You're thinking of a different megacorp that injects advertisements into everything as it's primary revenue source.
I expect that if all cars were banned, he would "discover" a major source of air pollution due to shoe-leather wear, or farts or exhaling
Too late.
I've got a six year old i7 (four cores), as my primary Windows dev machine. I also have other machines for porting to Mac and Linux, plus a music/audio workstation. All use Intel processors. I went independent a few years ago, so I don't have a lot of cash to burn on new hardware. These machines have to last me until I (hopefully) get my game out the door soon and start earning an income again. They work fine for me, since my game is targeting lower-spec machines anyhow.
I've always pretty much stuck with Intel for my CPUs. Even so, I'll certainly check out AMD CPUs when its time to upgrade, but so far, they really haven't produced anything that can compete effectively with Intel. I'd love to see that change, as competition is good for everyone.
Until Ryzen takes over and 6 to 8 cores and hyper-threaded goodness become standard then why bother optimizing?
It's important not to conflate average numbers on Steam versus what your game's particular demographic tends to use. When you say "Steam gamers", for instance, that's pretty much meaningless, as it's just an average across a *very* wide swath of gamers and gaming hardware.
The latest AAA games will not simply not run well on a dual core laptop with 4GB of RAM and an integrated Intel GPU. The target audience for those types of games tends to be someone who has a moderately powerful desktop machine or a *very* beefy laptop. Those machines are likely to have 4 cores and eight hardware threads at *minimum*. So, threading optimization is obviously very much worthwhile for these types of games.
Now that I've gone indie (I left my professional job about four years ago), my own game's target hardware is much more modest. So, in that case, I don't have to put in the same degree of optimization effort in my own engine as large studios do, since I'm a one man studio and can't put in the resources they do. There are plenty of older or lightweight indie games that those WalMart laptops will run just fine.
In short, it really all depends on your game's minimum requirements and intended target audience.
As a professional game developer, I can assure you that videogames are the exact opposite of an "embarrassingly parallel" problem. There's a LOT more to videogames than graphical processing: AI, pathfinding, physics, audio, resource and memory management, animation, world updates / occlusion systems, etc. These are all CPU-intensive subsystems, and many of them are inherently bound to global data (a virtual world simulation), which makes it extremely tricky to split off into independent threads (probably the only exception being audio).
That means that each subsystem in the game must be carefully and painstakingly optimized for threaded performance. It's not possible to trivially split up all these subsystems by thread either. Many of these systems tend to interact with each other and the global world database, and that means the gains tend to be smallish and non-scalable in nature.
Generally speaking, it's an extremely difficult problem, and one which I don't think the industry has really cracked yet. Believe me, if it were trivial to do, we've have done it a long time ago.