Has there been any work done on comparing how well TBB performs using simple benchmarks? As in, compared to existing solutions like OpenMP or MPI? The website is rather uninformative as far as benchmarks are concerned.
As many posters have so commented, it is clear that the use of the term Moore's Law was not appropriate. What the article seems to be attempting to purport was that the drive for low-end, inexpensive hardware is going to have a negative effect on the high-end market, and therefor lead to a decline in innovation and technological progression.
The former clause above may be true, but that is still up for debate. As stated, there still exists a very thriving market in the enterprise, media production, and gaming areas for high-end PCs.
The latter derivation is silly to the point of rediculousness. The technology and computer industries will always innovate. Low-end hardware will inovate along with it as the industry must flex to fit whatever the consumer demands.
In the end, if consumers finally realize that they do not NEED a $1000 system to accomplish day-to-day work with their PC, reasonable hardware at low prices will become more ubiquitous, power consumption will fall, better computers will become available for lower income families, and the market will continue to thrive as it responds to this new demand.
Sun actually jumped on this idea of yours (Speed Stepping) a while ago, touting processors with low-power-consumption built into the architecture (eg UltraSPARC T2 advertised as a "green cpu"). From a design standpoint it isn't hard to do; enabling and disabling unused datapaths, control lines, and clocks reduces switching and thus reduces power. This becomes atractive to server buyers, as nobody likes paying for wasted power. It's a suprise to me as well that none of the big graphics card makers have tried this yet: it's clearly effective and not all that hard to do if made a serious requirement.
I know personally I'd love a video card that could claim great performance and superior power management during underuse.
Unfortunately, a compromised system is more than its own worst enemy. Could not a botnet built up of Windows machines still cause trouble for an apache server? This, in my opinion, is the real threat here: vulnerabilities in mainstream OS's that mean trouble for EVERYBODY.
...where it belongs? The summary is right on, this is most important for its symbolism; finally some bodies of government are taking notice of the complexities of this issue and weighing in. Even if they don't rule in favorable ways, it'll be good for the dialogs to start opening up so that voters can get to see where their candidates stand on the issue. I wonder if/when the presidential candidates will touch this. The YouTube debates might be something to watch out for after all.
Well, the analogy I've always heard was "1 woman can have 1 baby in 9 months, but 9 women can't have 1 baby in 1 month."
Lesson here: not everything is as "parallelizable" as digging a ditch. Data dependency in single execution threads means there often simply isn't enough independent work that can be done at once.
Moreover, it is often left up to the user (or third party vendors) to create the application library to take advantage of parallel processing. Almost all code being run at this moment was writen in a serial, higher-level language (such as C++) for serial execution (even if it utilizes threading in the OS). The Cell didn't provide a very good API, and even trivially parallelizable algorithms often have to be rewritten in assembly code to take full advantage of the available hardware. And that just plain sucks.
With process sizes getting smaller and smaller, it is interesting to watch new ideas for as to what to do with that newfound area. The elementary choice seemed to always be "throw on more cores" but the prospects of accelerators and bridges moving into Systems-on-Chips looks like it might have much nicer prospects.
The average parallism factor for most programs tends to hover around four. I think Intel might have figured out that this is a decent stopping point for hardware parallelism as well.
As stated before, the pros and cons are rather complicated. Their economic growth is not at all being exaggerated. However, what most people sneer at are the deplorable labor practices (read: sweatshops) and authoritarian rule which results in articles like this. Then again, the US amassed much of its wealth via slavery. In the end, it comes down to perspective and what people are willing to tolerate in this day and age.
Is Turbo Memory technology hardware that is designed and built around an OS (Vista)? That seems to be a very peculiar (read:bad) idea. What does it mean for other users who intend to utilize different operating systems? Is there a loss of performance or just an added feature that cannot be used?
You're not missing much. This is the argument that MS would use if this case were to ever come to fruition. It's the same way they dodged the Netscape suit: claim that the product being complained about is actually an integral part of the functionality of an operating system in today's computing model.
This worked with Netscape thanks to the sharp rise in internet use by the common user when IE started coming bundled with Windows. At that point, a web browser was indeed an intergral part of the OS and thus not criminal for the OS provider to provide one. This is the line of reasoning that can be leavied against Google: search functions are now necessary for day-to-day use.
But then again, it will never come to that, thanks to Microsoft's clever investments in government.
Re:what chinese see googling for "Tiananmen Square
on
China Censoring Flickr
·
· Score: 1
If other internet references to Tiananmen Square were more applicable to Chinese users, we would at least a different ordering of images in the search results. 18 pages turning into 3 pages by a change of locality is a bit too drastic of a difference to pass off, considering it is all coming from the same internet...right...? oh...
Re:About appliance-like locked down computers
on
How to Save the Internet
·
· Score: 2, Insightful
What could be done to reduce this problem :
-Nothing. Things are going to worsen but there is probably nothing we can do.
-Let OS vendors turn to trusted computing but that would destroy the power and usefulness of General Purpose computer for everybody.
-Hope people will turn to easy-to use appliance like device.
Unfortunately, there is a fourth option: gross restrictions on internet traffic and application usage. This brute-force and lowest-order solution is the impetus for our conversation here. It means loss of net neutrality and the big companies vie for pieces of the online pie. What the public needs to know are that other options EXIST to secure the internet.
Simple, eloquent means of securing general purpose PCs from malware, viruses, and other online threats exist. But, as you said, Average Joe doesn't see them, and can be thus be easily convinced that locking down the internet will be just as effective.
The problem is, if they cave in it may set a dangerous precedent for the rest of the general public to mimic. Then every Tom, Dick and Harry would think they could sue their manufacturers as well (which may very well be their right). No company needs THAT kind of idea getting out.
In all likelihood, however, you're probably right and they'll wind up settling once the media has lost interest.
I just thought I'd take a number-crunching approach to this hilarity...
6807 messages in 1 month (let's say 30 days)
That turns out to be over 9 text messages per hour, assuming she does not sleep and can text-message when/if she is in school. Unfortunately I lack any information to make reasonable assumptions about either of those two habits. And to keep that up constantly for an entire month? I feel like somebody should have taken notice.
What I really wonder if this one bill just came out of the blue or if it was a slow buildup to an $11,000 phone bill and parents who didn't care less. Judging by the smile on her face in the linked article, I'm willing to sway towards the latter. This only made the news because of its extreme magnitude, not because the parents blew their lids.
In an ideal world, the blame would be shared. True, PlusNet made a terrible error due to their own incompetence. There is no reason why they should not be punished accordingly. And, in the end, with all the lost business from people like yourself, their punishment may come in the form of annihilation.
However, I still agree that ISPs are too often taking a reactive approach to these disasters rather than taking simple preventive measures. It's great that PlusNet's director goes on the line now to recommend security software, but why wasn't this already mandated to every user BEFORE this all took place? Until ISPs take serious action to break down botnets, pro-actively blacklist compromised computers, and secure the integrity of their own records, things are going to continue to get worse.
Colleges now take a similar approach to your idea before letting students get onto residential networks. Nobody gets on without first submitting to a virus scan. Computers seen on the network behaving as if they have a virus are quarantined and must scan again. Repeated violators have to talk to the IT/support staff themselves. I work at a personal computing support at a US university that uses such a system, and it has all but solved the problem on our networks. The next logical step is to extend this coverage to malware and spyware not typically touched by antivirus scans.
I fully agree with you, sherriw, ISPs need to take a proactive approach in making sure that their clients take care of their systems, since history has proven that they will not do so themselves. Laissez-faire approaches to maintaining Internet security are obviously not sufficient.
The US is starting to face similar problems thanks to the No Child Left Behind Act. Standardized testing was always a main focus in determining the quality of students, teachers, and schools. This was most prevelant in elementary and junior high school levels. Now, many schools (even high schools) are simply buckling down and teaching their students from a young age exactly how to take these tests, so that their school's ratings won't fall. The result: a new generation of multiple-choice and reading-comprehension masters with little creativity or logical reasoning skills, which are crucial to higher education.
Addendum - For those who do not live in Britain the national curriculum was introduced to supposedly enable a fair comparison via school league tables. The idea was to give parents a choice about where they sent their little cherubs. Then the failing schools would empty out and eventually be closed down when the numbers of pupils attending got below a certain level.
Again, we are not so far apart! This idea echoes a major issue in the 2000 elections: school vouchers. The concept was that if a child is placed in a failing school, the government could award the parents "vouchers" to attend a private institution. The result: US taxpayers pay extra to drain out public schools instead of simply remedying the problem and fixing the problem. A logical solution in the short run which, over the long term, would spell doom.
I might have to agree, at least in part. Requiring a static IP makes this solution not viable for people on, say, a college campus, which is the first thing that comes to mind when I think of areas where power-saving modes could actually add up to something meaningful.
Slashdot Mirror
Has there been any work done on comparing how well TBB performs using simple benchmarks? As in, compared to existing solutions like OpenMP or MPI? The website is rather uninformative as far as benchmarks are concerned.
As many posters have so commented, it is clear that the use of the term Moore's Law was not appropriate. What the article seems to be attempting to purport was that the drive for low-end, inexpensive hardware is going to have a negative effect on the high-end market, and therefor lead to a decline in innovation and technological progression.
The former clause above may be true, but that is still up for debate. As stated, there still exists a very thriving market in the enterprise, media production, and gaming areas for high-end PCs.
The latter derivation is silly to the point of rediculousness. The technology and computer industries will always innovate. Low-end hardware will inovate along with it as the industry must flex to fit whatever the consumer demands.
In the end, if consumers finally realize that they do not NEED a $1000 system to accomplish day-to-day work with their PC, reasonable hardware at low prices will become more ubiquitous, power consumption will fall, better computers will become available for lower income families, and the market will continue to thrive as it responds to this new demand.
Sun actually jumped on this idea of yours (Speed Stepping) a while ago, touting processors with low-power-consumption built into the architecture (eg UltraSPARC T2 advertised as a "green cpu"). From a design standpoint it isn't hard to do; enabling and disabling unused datapaths, control lines, and clocks reduces switching and thus reduces power. This becomes atractive to server buyers, as nobody likes paying for wasted power. It's a suprise to me as well that none of the big graphics card makers have tried this yet: it's clearly effective and not all that hard to do if made a serious requirement.
I know personally I'd love a video card that could claim great performance and superior power management during underuse.
Unfortunately, a compromised system is more than its own worst enemy. Could not a botnet built up of Windows machines still cause trouble for an apache server? This, in my opinion, is the real threat here: vulnerabilities in mainstream OS's that mean trouble for EVERYBODY.
...where it belongs? The summary is right on, this is most important for its symbolism; finally some bodies of government are taking notice of the complexities of this issue and weighing in. Even if they don't rule in favorable ways, it'll be good for the dialogs to start opening up so that voters can get to see where their candidates stand on the issue. I wonder if/when the presidential candidates will touch this. The YouTube debates might be something to watch out for after all.
http://en.wikipedia.org/wiki/Scramjet saying supersonic scramjet is like saying IRS Service
There have been some attempts to streamline the process you describe already (libraries that involve small changes to existing code)r face
http://en.wikipedia.org/wiki/OpenMP
http://en.wikipedia.org/wiki/Message_Passing_Inte
http://www.rapidmind.net/
The Rapidmind project showed some great benchmark results using the Cell.
The water isn't going to be distilled, so odds are it is still electrolytic and thus can just as easily bork a keyboard by itself.
The key step (pun intended) is the air drying. As long as the water no longer bridges contacts, you're fine.
Well, the analogy I've always heard was "1 woman can have 1 baby in 9 months, but 9 women can't have 1 baby in 1 month." Lesson here: not everything is as "parallelizable" as digging a ditch. Data dependency in single execution threads means there often simply isn't enough independent work that can be done at once. Moreover, it is often left up to the user (or third party vendors) to create the application library to take advantage of parallel processing. Almost all code being run at this moment was writen in a serial, higher-level language (such as C++) for serial execution (even if it utilizes threading in the OS). The Cell didn't provide a very good API, and even trivially parallelizable algorithms often have to be rewritten in assembly code to take full advantage of the available hardware. And that just plain sucks.
With process sizes getting smaller and smaller, it is interesting to watch new ideas for as to what to do with that newfound area. The elementary choice seemed to always be "throw on more cores" but the prospects of accelerators and bridges moving into Systems-on-Chips looks like it might have much nicer prospects.
The average parallism factor for most programs tends to hover around four. I think Intel might have figured out that this is a decent stopping point for hardware parallelism as well.
As stated before, the pros and cons are rather complicated. Their economic growth is not at all being exaggerated. However, what most people sneer at are the deplorable labor practices (read: sweatshops) and authoritarian rule which results in articles like this. Then again, the US amassed much of its wealth via slavery. In the end, it comes down to perspective and what people are willing to tolerate in this day and age.
True, but previous search features were never indexed searches which would conflict with google's search functionality.
Is Turbo Memory technology hardware that is designed and built around an OS (Vista)? That seems to be a very peculiar (read:bad) idea. What does it mean for other users who intend to utilize different operating systems? Is there a loss of performance or just an added feature that cannot be used?
You're not missing much. This is the argument that MS would use if this case were to ever come to fruition. It's the same way they dodged the Netscape suit: claim that the product being complained about is actually an integral part of the functionality of an operating system in today's computing model.
This worked with Netscape thanks to the sharp rise in internet use by the common user when IE started coming bundled with Windows. At that point, a web browser was indeed an intergral part of the OS and thus not criminal for the OS provider to provide one. This is the line of reasoning that can be leavied against Google: search functions are now necessary for day-to-day use.
But then again, it will never come to that, thanks to Microsoft's clever investments in government.
If other internet references to Tiananmen Square were more applicable to Chinese users, we would at least a different ordering of images in the search results. 18 pages turning into 3 pages by a change of locality is a bit too drastic of a difference to pass off, considering it is all coming from the same internet...right...? oh...
What could be done to reduce this problem : -Nothing. Things are going to worsen but there is probably nothing we can do. -Let OS vendors turn to trusted computing but that would destroy the power and usefulness of General Purpose computer for everybody. -Hope people will turn to easy-to use appliance like device.
Unfortunately, there is a fourth option: gross restrictions on internet traffic and application usage. This brute-force and lowest-order solution is the impetus for our conversation here. It means loss of net neutrality and the big companies vie for pieces of the online pie. What the public needs to know are that other options EXIST to secure the internet.
Simple, eloquent means of securing general purpose PCs from malware, viruses, and other online threats exist. But, as you said, Average Joe doesn't see them, and can be thus be easily convinced that locking down the internet will be just as effective.
The problem is, if they cave in it may set a dangerous precedent for the rest of the general public to mimic. Then every Tom, Dick and Harry would think they could sue their manufacturers as well (which may very well be their right). No company needs THAT kind of idea getting out.
In all likelihood, however, you're probably right and they'll wind up settling once the media has lost interest.
I just thought I'd take a number-crunching approach to this hilarity...
6807 messages in 1 month (let's say 30 days)
That turns out to be over 9 text messages per hour, assuming she does not sleep and can text-message when/if she is in school. Unfortunately I lack any information to make reasonable assumptions about either of those two habits. And to keep that up constantly for an entire month? I feel like somebody should have taken notice.
What I really wonder if this one bill just came out of the blue or if it was a slow buildup to an $11,000 phone bill and parents who didn't care less. Judging by the smile on her face in the linked article, I'm willing to sway towards the latter. This only made the news because of its extreme magnitude, not because the parents blew their lids.
Just another spoiled brat. Move along.
In an ideal world, the blame would be shared. True, PlusNet made a terrible error due to their own incompetence. There is no reason why they should not be punished accordingly. And, in the end, with all the lost business from people like yourself, their punishment may come in the form of annihilation.
However, I still agree that ISPs are too often taking a reactive approach to these disasters rather than taking simple preventive measures. It's great that PlusNet's director goes on the line now to recommend security software, but why wasn't this already mandated to every user BEFORE this all took place? Until ISPs take serious action to break down botnets, pro-actively blacklist compromised computers, and secure the integrity of their own records, things are going to continue to get worse.
Colleges now take a similar approach to your idea before letting students get onto residential networks. Nobody gets on without first submitting to a virus scan. Computers seen on the network behaving as if they have a virus are quarantined and must scan again. Repeated violators have to talk to the IT/support staff themselves. I work at a personal computing support at a US university that uses such a system, and it has all but solved the problem on our networks. The next logical step is to extend this coverage to malware and spyware not typically touched by antivirus scans.
I fully agree with you, sherriw, ISPs need to take a proactive approach in making sure that their clients take care of their systems, since history has proven that they will not do so themselves. Laissez-faire approaches to maintaining Internet security are obviously not sufficient.
The US is starting to face similar problems thanks to the No Child Left Behind Act. Standardized testing was always a main focus in determining the quality of students, teachers, and schools. This was most prevelant in elementary and junior high school levels. Now, many schools (even high schools) are simply buckling down and teaching their students from a young age exactly how to take these tests, so that their school's ratings won't fall. The result: a new generation of multiple-choice and reading-comprehension masters with little creativity or logical reasoning skills, which are crucial to higher education. Addendum - For those who do not live in Britain the national curriculum was introduced to supposedly enable a fair comparison via school league tables. The idea was to give parents a choice about where they sent their little cherubs. Then the failing schools would empty out and eventually be closed down when the numbers of pupils attending got below a certain level. Again, we are not so far apart! This idea echoes a major issue in the 2000 elections: school vouchers. The concept was that if a child is placed in a failing school, the government could award the parents "vouchers" to attend a private institution. The result: US taxpayers pay extra to drain out public schools instead of simply remedying the problem and fixing the problem. A logical solution in the short run which, over the long term, would spell doom.
I might have to agree, at least in part. Requiring a static IP makes this solution not viable for people on, say, a college campus, which is the first thing that comes to mind when I think of areas where power-saving modes could actually add up to something meaningful.