It's less true than people think. The power consumption penalty of x86 is low single-digits. While it exists, Intel compensates for it with better sleep states and superior silicon.
"I think you're the only one that gets it. ARM will displace x86 in the server space, not because their cores are superior to Intel but because they have a superior business model that's amenable to building heterogeneous systems which will lower system power and increase system performance."
I don't think you understand. In an era when moving data is extremely inefficient due to power constraints, you need superior cores. Want heterogeneous many-core? Buy Xeon Phi. Want strong-single thread? Buy high-end Xeon. Want maximum power efficiency with leading single-thread performance? You buy Bay Trail.
Need a lower power interconnect? Buy parts with photonics (in 5-6 years, anyway).
Can ARM creates products that fit well in this market? But they aren't going to waltz in and displace Intel. They may pick up 10-15% of the market; they're not going to sweep.
From the article: "Ultimately I don't know that this data really changes what we already knew about Clover Trail: it is a more power efficient platform than NVIDIA's Tegra 3."
This single data point doesn't mean 32nm Atom is more power efficient than any other ARM device, but it illustrates that the gap you imply is device-specific -- not inherent to the microarchitectures. There are going to be ARM devices that are more power efficient than Atom and Atom devices that are more power efficient than ARM.
Current ARM processors may indeed have a role to play in supercomputing, but the advantages this article implies don't exist.
Go look at performance figures for the Cortex-A15. It's *much* faster than the Cortex-A9. It also draws far more power. There's a reason why ARM's own product literature identifies the Cortex-A15 as a smartphone chip at the high end, but suggests strategies like big.LITTLE for lowering total power consumption. Next year, ARM's Cortex-A57 will start to appear. That'll be a 64-bit chip, it'll be faster than the Cortex-A15, it'll incorporate some further power efficiency improvements, and it'll use more power at peak load.
That doesn't mean ARM chips are bad -- it means that when it comes to semiconductors and the laws of physics, there are no magic bullets and no such thing as a free lunch.
I'm the author of that story, but I'm discussing a presentation given by one of the US's top supercomputing people. Pay particular attention to this graph:
What it shows is the cost, in energy, of moving data. Keeping data local is essential to keeping power consumption down in a supercomputing environment. That means that smaller, less-efficient cores are a bad fit for environments in which data has to be synchronized across tens of thousands of cores and hundreds of nodes. Now, can you build ARM cores that have higher single-threaded efficiency? Absolutely, yes. But they use more power.
ARM is going to go into datacenters and supercomputers, but it has no magic powers that guarantee it better outcomes.
You don't get to decide when a word is pejorative to a group that's historically been targeted with it. I agree strongly with George Carlin when he talks about the ludicrousness of "bad words." There are no "bad words." But you know what there *are?* There are words that have been used offensively against a minority group so often that they've become hurtful *to* that group of people.
You have a right to use those words anyway. You have a right to not care. You have a right to claim that because YOU don't find the word offensive, no one else has a right to do so, either.
You also have a right to decide that decades of discrimination against a particular group were so awful, you'll avoid using a word or two -- not because those words are "bad," but because they serve as reminders of abuse, insults, and ignorance. You have a right to decide to change your speaking habits *ever* so slightly as a way of demonstrating to this person or persons that you don't agree with the way those words were used against them.
You have a right to decide that empathy and acknowledgement is more meaningful than saying a certain collection of phonemes.
What are you going to do with a slice of spectrum? For that matter, what would *I* do with a slice of spectrum? And what would "public" ownership even mean in this space?
"Why is it that folks still have issues when a "Free" service suddenly is removed?"
Because having a service yanked out from under you is annoying? Furthermore, your use of the word "free" is somewhat limited. Google makes money off of users. It monetizes your search traffic, your emails, and tracks your site visitation patterns. It monitors which ads you click on, which you don't, and how to best use that data to better sell more ads. It leverages its share of the search and web services market in a number of ways to support these endeavors.
When you become part of the Google ecosystem, you are agreeing to share data with them that is incontrovertibly *valuable*, even if they never put a value on it, and no money changes hands. So you're right. No such thing as a free lunch. But when I use Google services, I'm paying them with my own personal usage data -- and they're obviously quite happy to use that data in a great many ways to "enhance" product offerings.
I'm not arguing that Google SMS deserved to live, or that Google is morally or legally in the wrong for closing it, but Google is compensated with information when I use its products. It may not cost me any money, but if I give you something you find valuable in exchange for a good or service, there's still an exchange taking place.
I'd like to propose two changes to how we talk about pretty much everything.
1) No more use of the phrase XXX-industrial complex. 2) No more attaching the "gate" suffix to scandals.
We can talk about the growth in surveillance technology and *all* of its associated problems without resorting to a term that was originally coined to discuss the complex relationship between Congress, the military, and the industrial base that supported both. According to Wikipedia, the industry covered by the "military industrial complex" was worth some $600 billion in 2009. Surveillance, while no doubt a booming business, isn't exactly in this league.
That's why, despite record stock gains, real wage growth is flat. Improvements in the unemployment rate overall are much smaller once you count the number of discouraged workers or consider the underemployed. The jobs being generated don't pay as well as the ones people lost, and they don't include the same level of benefits.
Is there any chance of seeing the B5 episodes remastered? Modern rendering, even on consumer hardware would hugely improve the visual effects. This would go beyond a simple re-issuing of the show in HD / 4K, and I imagine the models and renders could be time-consuming -- but there are fans who'd definitely contribute to make this a reality.
Even a modest cluster could deliver a vast improvement over the show's original visuals.
26% of children live with one parent. If you're going to single out that trend as being generally responsible for the decline of American...everything,despite the fact that it's a minority of total family arrangements, you really ought to highlight the fact that of that 26^% group, 26% of *them* are being raised by fathers, while 74% are raised by their mothers. You pour out plenty of vitriol on those "selfish" single women, but don't even blink at the selfish men who are raising kids on their own.
As I see it, you've got two options: Revise your previous post to be equally offensive, stupid, and insulting to both women and men, or adopt an opinion that reflects objective reality and requires a basic grasp of math.
You can't buy a faster system for $300. Intel's cheapest quad core is $179. Toss in $100 for the motherboard, $50 for RAM, $150 for the GPU ( assuming you opt for a suboptimal price point) and you're up to $480. A little shipping, and that's $500.
The point of this article isn't "does $500 buy a better upgrade than $270 (GPU + Q6600?
The point of the article is: Can I upgrade my old desktop and see a decent performance boost?"
Answer: Yes.
Is spending $500 better? Yes. Not everyone has $500.
Because the point was to test a system that was assembled using upper-midrange configuration in 2008. Back then, a majority of customers were still using 32-bit Windows and while 2GB DDR2 DIMMS were available, 1GB were the sweet spot.
My first configuration was a Q6600 with a GTX 260 and 3GB of RAM. I swapped in the E6850 to settle the dual-core question.
Also because that's all the DDR2 I still had on hand after so long.
But 3GB is reasonable. It's enough RAM that someone who upgraded to 64-bit Windows 7 (the OS I tested) might not have felt the need to upgrade more.
It mystifies me that so many intelligent people are so wrong about where content comes from. I'm an online journalist. I've been an online journalist for 11 years. Advertising revenue is what pays the bills.
Blocking ads isn't immoral. It doesn't make you a bad person. But if everyone did it, a whole heck of a lot of websites wouldn't exist. This point stands regardless of what anyone thinks of my content or the content at websites I've written for. If you like a website because it publishes solid, well-researched articles, those articles take time and money to create. Good journalism takes time and money to create.
If you block ads and *don't* subscribe or cut a check every so often, than yes, that's a problem. There are stories that don't get written because investigating them is too costly. It sucks to be in a situation where you've got your hands on something interesting, but you literally can't afford to follow it up. Opting out of advertising has an impact on sites you care about.
1) Being a local, small business is not a synonym for a neutral, objective, or ethical company.
2) Best case, this leads to labels like those already plastered everywhere in the state of California. "The State of California has determined that this product contains chemicals known to raise the risk of cancer" or some shit. Except you know what the problem is? Those labels are *everywhere.* They're the first thing you see when you get off the jet. They're in the jetway. The jetway that you literally *have* to walk through.
When you take a message and plaster it everywhere, on everything, you destroy its ability to communicate effectively. Add to that the complete lack of evidence that any of these changes produce any kind of negative health effect, and you have a scare tactic that accomplishes nothing and only frightens people based on zero scientific evidence.
This isn't the same thing as defending Monsanto (I don't) or calling for crazy-ass levels of genetic engineering in food with no oversight. I'm in favor of neither.
That's a distinction that the average user doesn't make. At the end of the day, I don't care if the front-end secretly passes the video to a collection of manatees who perform FFT calculations using colored balls they pick out of a pit. The criteria was a piece of software with easy-to-use presets that produces decent-quality video after I push "Ok."
If Program X does that, and Program Y doesn't, then Program X wins. The reason *why* is interesting and pertinent, but the question wasn't "Why do two different front-ends give different results using the same encoder?"
No, the article says that GPU encoding software runs the gamut from outright awful to simply broken and limited. Quick Sync video is great in Arcsoft, terrible in Cyberlink, unsupported in Xilisoft, and looks decent in MediaCoder. Check the GTX 580's output in Xilisoft for plenty of proof that no, you don't need insane bitrates to create decent-looking output.
I set out to test presets. Specifically, I set out to test the presets of software packages which are sold on the purported *strength* of those presets. I say so in the first paragraph:
" Our goal was to find a program that would encode at a reasonably high quality level (~1GB per hour was the target) and require a minimal level of expertise from the user."
That's why MediaCoder results weren't included.
The entire article came about because Cyberlink's iPhone 4S preset yielded files that were 1.4GB if I used CPU encoding or a GTX 580, and 188MB if I used Quick Sync. That disparity is what I noticed when I went to check encode quality for the initial IVB review.
Can you build custom profiles in CME and create outputs that avoid these problems? You can -- though some options aren't available. That, however, is not the point. If I'm going to build my own custom profiles, I can download a copy of MediaCoder for free and do it with a more powerful piece of software that offers a huge number of options.
I did a review of "Software that claims to automate the GP encode process." I did not do a review of "Can Cyberlink MediaEspresso EVER create a decent image?" Given what I set out to evaluate, my ability to tweak profiles to achieve a satisfactory result is not a valid criteria for my conclusions.
Simple reason: Because DVD Fab never came up. I Googled several variations on the term and asked Nvidia, Intel, and AMD for their own recommendations as far as products were concerned. Cyberlink and Arcsoft were recommended by multiple sources. Badaboom, I knew about and was familiar with. Xilisoft and MediaCoder were added as a result of additional research.
I never came across DVD Fab. That's not a judgment on its quality or output.
You pretty much nailed my problem with the output.:P That's the reason why Arcsoft, with compatibility problems, ultimately ranked above Cyberlink. Arcsoft doesn't do very good work on the Radeon 7950 and it can't handle CUDA, but it at least gets something right. Quick Sync video is very good.
Cyberlink got nothing right anywhere. And it's the program most-often recommended to reviewers as a benchmark when we want to review GPU encoding.
Because 3000-word articles with PNGs at ~300K per large image and 100K per preview image aren't fun reading in a single go. There's ~1.5MB of imagery just on the third page . Pages 3-8 have about the same, and that's with the images only loaded as thumbnails.
If you've got a fast net connection, you won't care. If you don't have a fast net connection, loading 16MB of images at once isn't a lot of fun.
Visual quality comparisons are one area where you can't use low-quality JPGs. A 9-page article at ET is a real rarity, it's not something we do because we want to spam ads.
Slashdot Mirror
It's less true than people think. The power consumption penalty of x86 is low single-digits. While it exists, Intel compensates for it with better sleep states and superior silicon.
"I think you're the only one that gets it. ARM will displace x86 in the server space, not because their cores are superior to Intel but because they have a superior business model that's amenable to building heterogeneous systems which will lower system power and increase system performance."
I don't think you understand. In an era when moving data is extremely inefficient due to power constraints, you need superior cores. Want heterogeneous many-core? Buy Xeon Phi. Want strong-single thread? Buy high-end Xeon. Want maximum power efficiency with leading single-thread performance? You buy Bay Trail.
Need a lower power interconnect? Buy parts with photonics (in 5-6 years, anyway).
Can ARM creates products that fit well in this market? But they aren't going to waltz in and displace Intel. They may pick up 10-15% of the market; they're not going to sweep.
"The point is that an ARM processor can provide, say, 75% of the performance for 25% of the power compared to x86. "
http://www.anandtech.com/show/6529/busting-the-x86-power-myth-indepth-clover-trail-power-analysis
From the article: "Ultimately I don't know that this data really changes what we already knew about Clover Trail: it is a more power efficient platform than NVIDIA's Tegra 3."
This single data point doesn't mean 32nm Atom is more power efficient than any other ARM device, but it illustrates that the gap you imply is device-specific -- not inherent to the microarchitectures. There are going to be ARM devices that are more power efficient than Atom and Atom devices that are more power efficient than ARM.
Current ARM processors may indeed have a role to play in supercomputing, but the advantages this article implies don't exist.
Go look at performance figures for the Cortex-A15. It's *much* faster than the Cortex-A9. It also draws far more power. There's a reason why ARM's own product literature identifies the Cortex-A15 as a smartphone chip at the high end, but suggests strategies like big.LITTLE for lowering total power consumption. Next year, ARM's Cortex-A57 will start to appear. That'll be a 64-bit chip, it'll be faster than the Cortex-A15, it'll incorporate some further power efficiency improvements, and it'll use more power at peak load.
That doesn't mean ARM chips are bad -- it means that when it comes to semiconductors and the laws of physics, there are no magic bullets and no such thing as a free lunch.
http://www.extremetech.com/computing/155941-supercomputing-director-bets-2000-that-we-wont-have-exascale-computing-by-2020
I'm the author of that story, but I'm discussing a presentation given by one of the US's top supercomputing people. Pay particular attention to this graph:
http://www.extremetech.com/wp-content/uploads/2013/05/CostPerFlop.png
What it shows is the cost, in energy, of moving data. Keeping data local is essential to keeping power consumption down in a supercomputing environment. That means that smaller, less-efficient cores are a bad fit for environments in which data has to be synchronized across tens of thousands of cores and hundreds of nodes. Now, can you build ARM cores that have higher single-threaded efficiency? Absolutely, yes. But they use more power.
ARM is going to go into datacenters and supercomputers, but it has no magic powers that guarantee it better outcomes.
You don't get to decide when a word is pejorative to a group that's historically been targeted with it. I agree strongly with George Carlin when he talks about the ludicrousness of "bad words." There are no "bad words." But you know what there *are?* There are words that have been used offensively against a minority group so often that they've become hurtful *to* that group of people.
You have a right to use those words anyway. You have a right to not care. You have a right to claim that because YOU don't find the word offensive, no one else has a right to do so, either.
You also have a right to decide that decades of discrimination against a particular group were so awful, you'll avoid using a word or two -- not because those words are "bad," but because they serve as reminders of abuse, insults, and ignorance. You have a right to decide to change your speaking habits *ever* so slightly as a way of demonstrating to this person or persons that you don't agree with the way those words were used against them.
You have a right to decide that empathy and acknowledgement is more meaningful than saying a certain collection of phonemes.
Or not to.
What are you going to do with a slice of spectrum? For that matter, what would *I* do with a slice of spectrum? And what would "public" ownership even mean in this space?
"Why is it that folks still have issues when a "Free" service suddenly is removed?"
Because having a service yanked out from under you is annoying? Furthermore, your use of the word "free" is somewhat limited. Google makes money off of users. It monetizes your search traffic, your emails, and tracks your site visitation patterns. It monitors which ads you click on, which you don't, and how to best use that data to better sell more ads. It leverages its share of the search and web services market in a number of ways to support these endeavors.
When you become part of the Google ecosystem, you are agreeing to share data with them that is incontrovertibly *valuable*, even if they never put a value on it, and no money changes hands. So you're right. No such thing as a free lunch. But when I use Google services, I'm paying them with my own personal usage data -- and they're obviously quite happy to use that data in a great many ways to "enhance" product offerings.
I'm not arguing that Google SMS deserved to live, or that Google is morally or legally in the wrong for closing it, but Google is compensated with information when I use its products. It may not cost me any money, but if I give you something you find valuable in exchange for a good or service, there's still an exchange taking place.
I'd like to propose two changes to how we talk about pretty much everything.
1) No more use of the phrase XXX-industrial complex.
2) No more attaching the "gate" suffix to scandals.
We can talk about the growth in surveillance technology and *all* of its associated problems without resorting to a term that was originally coined to discuss the complex relationship between Congress, the military, and the industrial base that supported both. According to Wikipedia, the industry covered by the "military industrial complex" was worth some $600 billion in 2009. Surveillance, while no doubt a booming business, isn't exactly in this league.
As for the --gate thing, it's just plain stupid.
The top 1% of the US captured 121% of the wealth generated during the "recovery." The bottom 99% actually got poorer.
http://boingboing.net/2013/02/13/economic-recovery-in-the-us-ac.html
That's why, despite record stock gains, real wage growth is flat. Improvements in the unemployment rate overall are much smaller once you count the number of discouraged workers or consider the underemployed. The jobs being generated don't pay as well as the ones people lost, and they don't include the same level of benefits.
Facts. They kick ass.
Is there any chance of seeing the B5 episodes remastered? Modern rendering, even on consumer hardware would hugely improve the visual effects. This would go beyond a simple re-issuing of the show in HD / 4K, and I imagine the models and renders could be time-consuming -- but there are fans who'd definitely contribute to make this a reality.
Even a modest cluster could deliver a vast improvement over the show's original visuals.
You're aware that teen pregnancies in the United States are down 41% since 1990, right?
Or that 48% of US families contain at least one multi-generational adult (blowing your whole "Single woman only" idea out of the water?)
http://www.pewsocialtrends.org/2010/03/18/the-return-of-the-multi-generational-family-household/
26% of children live with one parent. If you're going to single out that trend as being generally responsible for the decline of American...everything,despite the fact that it's a minority of total family arrangements, you really ought to highlight the fact that of that 26^% group, 26% of *them* are being raised by fathers, while 74% are raised by their mothers. You pour out plenty of vitriol on those "selfish" single women, but don't even blink at the selfish men who are raising kids on their own.
As I see it, you've got two options: Revise your previous post to be equally offensive, stupid, and insulting to both women and men, or adopt an opinion that reflects objective reality and requires a basic grasp of math.
As the author:
You can't buy a faster system for $300. Intel's cheapest quad core is $179. Toss in $100 for the motherboard, $50 for RAM, $150 for the GPU ( assuming you opt for a suboptimal price point) and you're up to $480. A little shipping, and that's $500.
The point of this article isn't "does $500 buy a better upgrade than $270 (GPU + Q6600?
The point of the article is: Can I upgrade my old desktop and see a decent performance boost?"
Answer: Yes.
Is spending $500 better? Yes. Not everyone has $500.
As the author:
Because the point was to test a system that was assembled using upper-midrange configuration in 2008. Back then, a majority of customers were still using 32-bit Windows and while 2GB DDR2 DIMMS were available, 1GB were the sweet spot.
My first configuration was a Q6600 with a GTX 260 and 3GB of RAM. I swapped in the E6850 to settle the dual-core question.
Also because that's all the DDR2 I still had on hand after so long.
But 3GB is reasonable. It's enough RAM that someone who upgraded to 64-bit Windows 7 (the OS I tested) might not have felt the need to upgrade more.
The next question is how many Americans are now listed as part of a 'terrorist group' by the government for their support of OWS?
Get some historical perspective and look at the stings the FBI ran on MLK Jr and the Civil Rights Movement. This is nothing.
It mystifies me that so many intelligent people are so wrong about where content comes from. I'm an online journalist. I've been an online journalist for 11 years. Advertising revenue is what pays the bills.
Blocking ads isn't immoral. It doesn't make you a bad person. But if everyone did it, a whole heck of a lot of websites wouldn't exist. This point stands regardless of what anyone thinks of my content or the content at websites I've written for. If you like a website because it publishes solid, well-researched articles, those articles take time and money to create. Good journalism takes time and money to create.
If you block ads and *don't* subscribe or cut a check every so often, than yes, that's a problem. There are stories that don't get written because investigating them is too costly. It sucks to be in a situation where you've got your hands on something interesting, but you literally can't afford to follow it up. Opting out of advertising has an impact on sites you care about.
And for the record? I hate ads, too.
Compared to here: http://en.wikipedia.org/wiki/Usage_share_of_web_browsers
StatCounter, W3Counter, and Wikimedia all report IE at ~30%, +/- a couple of percentage points. Only W3Schools has a 15% number.
That alone makes me dubious of its accuracy.
1) Being a local, small business is not a synonym for a neutral, objective, or ethical company.
2) Best case, this leads to labels like those already plastered everywhere in the state of California. "The State of California has determined that this product contains chemicals known to raise the risk of cancer" or some shit. Except you know what the problem is? Those labels are *everywhere.* They're the first thing you see when you get off the jet. They're in the jetway. The jetway that you literally *have* to walk through.
When you take a message and plaster it everywhere, on everything, you destroy its ability to communicate effectively. Add to that the complete lack of evidence that any of these changes produce any kind of negative health effect, and you have a scare tactic that accomplishes nothing and only frightens people based on zero scientific evidence.
This isn't the same thing as defending Monsanto (I don't) or calling for crazy-ass levels of genetic engineering in food with no oversight. I'm in favor of neither.
Sure! Send me one and I'll test it. :)
That's a distinction that the average user doesn't make. At the end of the day, I don't care if the front-end secretly passes the video to a collection of manatees who perform FFT calculations using colored balls they pick out of a pit. The criteria was a piece of software with easy-to-use presets that produces decent-quality video after I push "Ok."
If Program X does that, and Program Y doesn't, then Program X wins. The reason *why* is interesting and pertinent, but the question wasn't "Why do two different front-ends give different results using the same encoder?"
No, the article says that GPU encoding software runs the gamut from outright awful to simply broken and limited. Quick Sync video is great in Arcsoft, terrible in Cyberlink, unsupported in Xilisoft, and looks decent in MediaCoder. Check the GTX 580's output in Xilisoft for plenty of proof that no, you don't need insane bitrates to create decent-looking output.
I set out to test presets. Specifically, I set out to test the presets of software packages which are sold on the purported *strength* of those presets. I say so in the first paragraph:
" Our goal was to find a program that would encode at a reasonably high quality level (~1GB per hour was the target) and require a minimal level of expertise from the user."
That's why MediaCoder results weren't included.
The entire article came about because Cyberlink's iPhone 4S preset yielded files that were 1.4GB if I used CPU encoding or a GTX 580, and 188MB if I used Quick Sync. That disparity is what I noticed when I went to check encode quality for the initial IVB review.
Can you build custom profiles in CME and create outputs that avoid these problems? You can -- though some options aren't available. That, however, is not the point. If I'm going to build my own custom profiles, I can download a copy of MediaCoder for free and do it with a more powerful piece of software that offers a huge number of options.
I did a review of "Software that claims to automate the GP encode process." I did not do a review of "Can Cyberlink MediaEspresso EVER create a decent image?" Given what I set out to evaluate, my ability to tweak profiles to achieve a satisfactory result is not a valid criteria for my conclusions.
Simple reason: Because DVD Fab never came up. I Googled several variations on the term and asked Nvidia, Intel, and AMD for their own recommendations as far as products were concerned. Cyberlink and Arcsoft were recommended by multiple sources. Badaboom, I knew about and was familiar with. Xilisoft and MediaCoder were added as a result of additional research.
I never came across DVD Fab. That's not a judgment on its quality or output.
Fuzzy,
You pretty much nailed my problem with the output. :P That's the reason why Arcsoft, with compatibility problems, ultimately ranked above Cyberlink. Arcsoft doesn't do very good work on the Radeon 7950 and it can't handle CUDA, but it at least gets something right. Quick Sync video is very good.
Cyberlink got nothing right anywhere. And it's the program most-often recommended to reviewers as a benchmark when we want to review GPU encoding.
As the author:
Because 3000-word articles with PNGs at ~300K per large image and 100K per preview image aren't fun reading in a single go. There's ~1.5MB of imagery just on the third page . Pages 3-8 have about the same, and that's with the images only loaded as thumbnails.
If you've got a fast net connection, you won't care. If you don't have a fast net connection, loading 16MB of images at once isn't a lot of fun.
Visual quality comparisons are one area where you can't use low-quality JPGs. A 9-page article at ET is a real rarity, it's not something we do because we want to spam ads.
As the author of the story, that's an error that slipped past in formatting. I'm uploading the proper graph right after I hit "Reply" on this.