My heterogeneous experience with Cell processor
on
Panic in Multicore Land
·
· Score: 5, Interesting
I've been doing some scientific computing on the Cell lately, and heterogeneous cores don't make life very easy. At least with the Cell.
The Cell has one PowerPC core ("PPU"), which is a general purpose PowerPC processor. Nothing exotic at all about programming it. But then you have 6 (for the Playstation 3) or 8 (other computers) "SPE" cores that you can program. Transferring data to/from them is a pain, they have small working memories (256k each), and you can't use all C++ features on them (no C++ exceptions, thus can't use most of the STL). They also have poor speed for double-precision floats.
The SPEs are pretty fast, and they have a very fast interconnect bus, so as a programmer I'm constantly thinking about how to take better advantage of them. Perhaps this is something I'd face with any architecture, but the high potential combined with difficult constraints of SPE programming make this an especially distracting aspect of programming the Cell.
So if this is what heterogeneous-cores programming means, I'd probably prefer the homogeneous version. Even if they have a little less performance potential, it would be nice to have a 90%-shorter learning curve to target the architecture.
How is this news for nerds or stuff that matters? It's worthless drivel like this that drags the whole site down!
/. is News For Nerds, not necessarily Nerdy News./. is a nerd site, and many nerds like Family Guy. So what's the problem? If you don't like this article, tune it out. Sheesh.
Thank for asking. First of all, those guys getting killed are *definitely* not Ninjas. The odds of a Ninja getting killed in a FPS are like the odds of a movie based on a popular video game not sucking.
So do these definitely-non-Ninjas enjoy getting killed? Only if they're not getting killed by a Ninja. If you get killed by a Ninja, you don't have *time* to enjoy it.
I watched Lessig's video, and my concern was that selfish, clever people will find a way to game pretty much any democratic system. So I'm afraid that Lessig's goals are unattainable.
For example, if campaigns are all publicly funded, then someone will find a clever way to make lots and lots of other campaign speech be volunteer, which is protected by the First Amendment. If there are limits on the sources of money, someone will find a way to sneak in money through cracks in the definitions.
While I applaud Lessig's goals, I'm just not sure they're attainable in a world where many actors lack scruples and seek something other than the common good.
Someone needs to let university network administrators know that students are downloading Microsoft's copyrighted software!
Copyright or corruption as his platform?
on
Lessig For Congress?
·
· Score: 4, Interesting
IIRC, Lessig recently made anti-corruption be his struggle, rather than restoring copyright to something reasonable. On which goal(s) would he focus? I think legislators are often forced to sacrifice one ideal for another during the legislative process.
Most vocal Slashdot'ers, including myself, feel that in the balance between (effective counter-terrorism) and (personal freedom, open government), Bush and Congress err far too much in the (effective counter-terrorism) direction.
Are most private citizens like us in this regard, and it's an authoritarian-vs.-population issue? Or are we/.'ers different from most citizens, and if so, why?
3) I rip and convert my DVDs for use on my portable devices.
...
I also do not use a single bit of pirated or unpaid software (I would say "unregistered," but there are a few free packages like RealPlayer I refuse to register due to spam issues.) No, really.
Then you may be splitting hairs on your definition of "pirated". You're certainly using the DVDs outside of the terms of your license. Are you ready to tell kids, without qualifying any of your statements, that they should thumb their noses at the DMCA, regardless of the consequences?
Only if the only value those journals add is distribution. They will still provide a very important role in aggregating articles of interest, so you're not digging through hundreds that you don't care about. They can also provide a forum for criticism and defense of articles.
I agree. Knowing that an article got published in ACM's Transactions on Programming Languages is a great sign that it's a paper worth making time to read. I think it's terribly important that we somehow retain a set of reviewers who decide what paper are worth broad attention.
But aren't these reviewers unpaid? If so, they could just as well organize themselves outside of the context of some journal. I think all that's really missing at that point is getting a small editorial board to decide what topics are worth focusing on, etc. Of course, since academics' and researchers' careers are measured in terms of their publication history, there is some impedance to changing this system.
I think their theory is that journals that don't allow this will have to change their policy, as they wouldn't want to lose out on publishing articles from Harvard profs.
But wouldn't that just accelerate the demise of those journals, since then there would be little reason to subscribe? I know that lately I've been chaffing at the cost of IEEE and ACM journal subscriptions. The main reason I bother is to get access to the articles I need for my research.
This conversation reminds me of the dilemma faced by newspapers regarding putting their content online, and if so whether or not to require an online subscription. There's no good answer for the newspapers now that the web offers an arguably superior distribution method for the content.
If the great innovation offered by SOM is basically a design pattern or interface technique, do we really need IBM's source code? It seems to me that the great thing about SOM is the idea of how something is done, and that we could pretty quickly write our own implementation of that idea. No?
Maybe you should look into the religious views of the people you reference as religious even from the source you site. Albert seems more of a politically correct (don't want to offend people who look up to him) atheist to me...
Good advice, but in this case I did actually check each one of them out. Einstein was reported to be a vague theist, in that he was sure the universe had a creator. That was sufficient because the OP was arguing for full-fledged atheism, which arguably is incompatible with Einstein's beliefs. I'd be a bit surprised if he was afraid to admit atheism - it wasn't rare in academic circles at the time afaik.
I'm not and I can't dare to express my views in fear of how my family and work colleagues will react.
Yeah - uncomfortable situation to be sure. But I hope you're able to soon, because you're kind of living a lie, and until you're honest about your take on life's most important stuff, they don't really know you, only a facade of you. Good luck with that (seriously).
The problem is that religion is inherently *irrational*.
Shit, we're going to get into this argument again. I guess the only surprise is that it wasn't the "Frist post, niggaz!".
Just because you haven't had any experiences or thoughts that make a religion plausible, doesn't mean that other people haven't. Manyreasonablepeoplehave experiences for which religion seems to them be the best explanation. Perhaps given your experiences atheism seems the most sensible world view. Yeah for you. But don't you feel a little, well, irrational in assuming that no one on the planet has a even a single good reason for thinking that theism might be accurate?
Also, show me 10 people with an apparently irrational belief in God, and I'll easily dig up about 4 other people with an irrational belief in athiesm.
Perhaps like many, I've found Wii's Boxing game to be (part of Wii Sports, iirc) to be very frustrating. It's hard for many of us to reliably throw the punches that we want to, despite carefully studying the directions. Either the Wii's current motion sensor is flaky, or the game is badly programmed. I'd gladly buy a controller with a new technology just to make Boxing work properly.
What's shocking about the report isn't what it covers [...], but what it leaves out
I guess we're all tired of ranting about Bush, but... I'm not shocked that his report left out his failures. Bush doesn't admit failures. (He's only admitted one regarding his work as a President, ever: Making some cowboy-style remark like "Bring it on." regarding terrorists.)
Slashdot Mirror
I've been doing some scientific computing on the Cell lately, and heterogeneous cores don't make life very easy. At least with the Cell.
The Cell has one PowerPC core ("PPU"), which is a general purpose PowerPC processor. Nothing exotic at all about programming it. But then you have 6 (for the Playstation 3) or 8 (other computers) "SPE" cores that you can program. Transferring data to/from them is a pain, they have small working memories (256k each), and you can't use all C++ features on them (no C++ exceptions, thus can't use most of the STL). They also have poor speed for double-precision floats.
The SPEs are pretty fast, and they have a very fast interconnect bus, so as a programmer I'm constantly thinking about how to take better advantage of them. Perhaps this is something I'd face with any architecture, but the high potential combined with difficult constraints of SPE programming make this an especially distracting aspect of programming the Cell.
So if this is what heterogeneous-cores programming means, I'd probably prefer the homogeneous version. Even if they have a little less performance potential, it would be nice to have a 90%-shorter learning curve to target the architecture.
If you don't mind, I for one would be interested to see it.
Was it an email response? I'd love to see it.
it's time to pray for that upcoming Supreme Court decision that will cover the scope of what's patentable.
Cured! Thanks!
... is that I actually want to do Lois. Am I alone in this?
Do you think we can get those politicians to advocate their state as "The G-Spot of Online Safety"?
Perhaps Santorum would make the pitch?
Thank for asking. First of all, those guys getting killed are *definitely* not Ninjas. The odds of a Ninja getting killed in a FPS are like the odds of a movie based on a popular video game not sucking.
So do these definitely-non-Ninjas enjoy getting killed? Only if they're not getting killed by a Ninja. If you get killed by a Ninja, you don't have *time* to enjoy it.
Next question!
Wow, didn't see that coming from a /. reader ;)
I watched Lessig's video, and my concern was that selfish, clever people will find a way to game pretty much any democratic system. So I'm afraid that Lessig's goals are unattainable.
For example, if campaigns are all publicly funded, then someone will find a clever way to make lots and lots of other campaign speech be volunteer, which is protected by the First Amendment. If there are limits on the sources of money, someone will find a way to sneak in money through cracks in the definitions.
While I applaud Lessig's goals, I'm just not sure they're attainable in a world where many actors lack scruples and seek something other than the common good.
Someone needs to let university network administrators know that students are downloading Microsoft's copyrighted software!
IIRC, Lessig recently made anti-corruption be his struggle, rather than restoring copyright to something reasonable. On which goal(s) would he focus? I think legislators are often forced to sacrifice one ideal for another during the legislative process.
Most vocal Slashdot'ers, including myself, feel that in the balance between (effective counter-terrorism) and (personal freedom, open government), Bush and Congress err far too much in the (effective counter-terrorism) direction.
/.'ers different from most citizens, and if so, why?
Are most private citizens like us in this regard, and it's an authoritarian-vs.-population issue? Or are we
When I mis-read it, I thought: "Chuck Norris can make SCO viable!".
I'm glad it's just *Stephen* Norris.
Then you may be splitting hairs on your definition of "pirated". You're certainly using the DVDs outside of the terms of your license. Are you ready to tell kids, without qualifying any of your statements, that they should thumb their noses at the DMCA, regardless of the consequences?
It's supposed to read MS Bytes
I agree. Knowing that an article got published in ACM's Transactions on Programming Languages is a great sign that it's a paper worth making time to read. I think it's terribly important that we somehow retain a set of reviewers who decide what paper are worth broad attention.
But aren't these reviewers unpaid? If so, they could just as well organize themselves outside of the context of some journal. I think all that's really missing at that point is getting a small editorial board to decide what topics are worth focusing on, etc. Of course, since academics' and researchers' careers are measured in terms of their publication history, there is some impedance to changing this system.
But wouldn't that just accelerate the demise of those journals, since then there would be little reason to subscribe? I know that lately I've been chaffing at the cost of IEEE and ACM journal subscriptions. The main reason I bother is to get access to the articles I need for my research.
This conversation reminds me of the dilemma faced by newspapers regarding putting their content online, and if so whether or not to require an online subscription. There's no good answer for the newspapers now that the web offers an arguably superior distribution method for the content.
Hand me your cache!
(Sorry - for humor I go for quantity, not quality.)
If the great innovation offered by SOM is basically a design pattern or interface technique, do we really need IBM's source code? It seems to me that the great thing about SOM is the idea of how something is done, and that we could pretty quickly write our own implementation of that idea. No?
Good advice, but in this case I did actually check each one of them out. Einstein was reported to be a vague theist, in that he was sure the universe had a creator. That was sufficient because the OP was arguing for full-fledged atheism, which arguably is incompatible with Einstein's beliefs. I'd be a bit surprised if he was afraid to admit atheism - it wasn't rare in academic circles at the time afaik.
Yeah - uncomfortable situation to be sure. But I hope you're able to soon, because you're kind of living a lie, and until you're honest about your take on life's most important stuff, they don't really know you, only a facade of you. Good luck with that (seriously).
Shit, we're going to get into this argument again. I guess the only surprise is that it wasn't the "Frist post, niggaz!".
Just because you haven't had any experiences or thoughts that make a religion plausible, doesn't mean that other people haven't. Many reasonable people have experiences for which religion seems to them be the best explanation. Perhaps given your experiences atheism seems the most sensible world view. Yeah for you. But don't you feel a little, well, irrational in assuming that no one on the planet has a even a single good reason for thinking that theism might be accurate?
Also, show me 10 people with an apparently irrational belief in God, and I'll easily dig up about 4 other people with an irrational belief in athiesm.
Perhaps like many, I've found Wii's Boxing game to be (part of Wii Sports, iirc) to be very frustrating. It's hard for many of us to reliably throw the punches that we want to, despite carefully studying the directions. Either the Wii's current motion sensor is flaky, or the game is badly programmed. I'd gladly buy a controller with a new technology just to make Boxing work properly.
I guess we're all tired of ranting about Bush, but... I'm not shocked that his report left out his failures. Bush doesn't admit failures. (He's only admitted one regarding his work as a President, ever: Making some cowboy-style remark like "Bring it on." regarding terrorists.)