Is it PowerPoint's fault, or the fault of the Powerpoint creator?
I usually find it very hard to take notes from a PowerPoint presentation. When taking math classes, the lecturer would write on the blackboard prepared sentences, and with a speed that made it easy to follow.
In my CS classes, the lecturer says they've been "forced" by the administration to use PowerPoint presentations. If I simply read them, they're easy enough to follow, but the information doesn't "stick" the way it does when I take notes from a blackboard. If I try to take notes, it's hard because the lecturer very often doesn't say exactly what's on the slide. It also means he can talk a lot quicker and thus not give me enough time to take notes from that slide.
So I have to try to listen to him, read the slide, try to convert what's on the slide into something I can write, and then write the notes, all at the same time. That leaves very little room to actually think about what's being said, and so I don't get the understanding I get when taking notes from a blackboard.
For me, the notes are important not just because they're a good reference for later, but also because the information sticks a lot better when writing it down. Not having good notes due to PowerPoint is one of the primary reasons I find some CS classes hard (especially the math heavy ones, like digital signal processing).
Proper use of PowerPoint would be to use it during a lecture to show complex graphs and images, which may be hard to convey on a blackboard (for instance what histogram equalization does to an image).
The fact that you have just explained it in a way which is subtly wrong supports the idea that it is counterintuitive. An animal does not adapt. It is born with a certain set of DNA, which it cannot change or control, and it lives or dies as a result what DNA it has (along with other factors like chance).
Most organisms don't express all of their genes all the time. They got regulating mechanisms which responds to it's current state and the environment. So in that case I would say it's perfectly possible that one organism can adapt, since it's regulatory mechanisms could mutate and thus respond to some external cue, allowing the organism to survive while it's "brother" dies.
Indeed, very odd. During a visit (with my class) at Ibas, they said they would have little to no problem recovering almost all of the data even after several reformat/rewrite cycles. And while Ibas ain't cheap, I'm pretty sure Ibas would be an order of magnitude or two cheaper than rescanning all the documents, not to mention faster.
What happened to that format, anyway? I was expecting it to be ubiquitous by now.
My guess is patents. And the fact that it only took a few years for bandwidth and storage to negate many of the advantages of jpeg2000 over the plain jpeg standard.
I did some research, turns out the voice is fully synthesized, not a prerecorded message. It was very good, only a few glitches in the flow between words.
There are all sorts of reasons someone might think a drive has failed. They're not all correct.
I can attest to this. One of my hd's started to "crash" (spun down, parked head, spun up again) with shorter and shorter intervals. I replaced it, thinking it was dying. After a while, another hd started doing the same, and a third started to drop out and come back online (as if I had disconnected it and reconnected it). Turns out, when I switched from the chipset sata controller to the onboard raid controller, everything was nice and dandy, even the first drive works like a charm now.
Uhm.. voice recognition and speech-to-text do NOT work. We've got QUITE A WAYS to go. Actually, it's getting pretty good in some cases. My ADSL went down some days ago, and I phoned tech support. Since it was late, I got a nice prerecorded voice saying they had closed for the day, but the "woman" then asked "would you like me to try to solve the problem automatically?". A bit stumped by this, I answered "yes". "She" then asked me to describe with few and short words what my problem was. So I said "adsl internet". "She" then asked for confirmation that I had said there was a problem with my internet connection. After a few more such questions, "she" could tell me that there was a known issue with ADSL in my area, and that it would be fixed by tomorrow afternoon.
So, for limited applications, voice recognition is getting along fairly well I must say.
Doesn't that also mean that if you sue them and lose, then you have to pay their legal fees?
AFAIK, we have a similar system in Norway, but I'm fairly sure I've read about several cases where the judge ruled that the plaintiff should not pay the legal fees (all of it, or the defendants part). So it would be a bit like a slap on the wrist to the defendant, in case they'd been good at covering their asses or something.
If I was AM2 rather than 939, I myself would be down on this in a heartbeat.
I also got a 939 rig, and I haven't quite understood the whole AM2 move from AMD. From what I've seen so far, AM2 doesn't bring a whole lot of improvements to the table, but what it does is equalize the upgrade costs between an AMD system and an Intel system. And in these days, that's hurting AMD bad I suspect.
If AMD needs some easy cash, why not release something for the 939 system? A reasonably priced, speedy dual core for instance? All I can get from my local shops is the X2 3800, which while dirt cheap, is the slowest X2. Why not sell the 4800 you had for not quite as little? I'd buy that as a intermediate upgrade.
There is no such thing as "massively parallel!" It makes no sense! Parallel in qualitative, NOT quantitative! Things are either parallel or they're not, there are no degrees of "parallelness!"
Sure there are. Say you want to find the maximum of 4 integers. You can do that in parallel, but you won't gain much if you have more than two processors (or execution units). Contrast this with say rendering an image using a path tracer, where each ray is independent of each other. First problem is hard to scale up, second one isn't. I'd say that means that ray tracing is a "more parallel" task.
Also, writing algorithms that has to run on 10000 processors efficiently is not exactly the same as one that has to run on 4 processors, in the same way that writing a multiplayer game that handles four players isn't the same as writing one that can handle thousands of concurrent players. So they toss on the "massive" part to separate the cases. At least that's my take on it.
its called a monopoly. It puts companies in a position where they can't say "Fuck off" because doing would result in a massive loss of customers
Well it's more like the prisoner's dilemma imho. If both ATI and Nvidia said "fuck off", MS would have a big problem. If only one of them said it however...
The interference pattern will occur even if there's only one photon in the apparatus at a time (that is, a photon hits the detector before a new one is generated).
I spent two hours every five minutes telling it not to reboot until the render finished. I immediately yanked the internet connection and haven't updated it since.
Alternative is to just move the window off-screen, something like below the taskbar (I usually leave 2 pixels worth so I can get it back if I want to). Then you just have to contend with the annoying keyboard focus change every 15 minutes.
Just because they use "tech" devices does not mean they understand how they actually work. A calculator is an example. Many are taught how to use one, but have no concept of the math involved.
Indeed. Here in Norway, there's an ongoing debate about the rather appalling math level of our school kids. I'm certain it has a LOT to do with calculators. All they learn now is how to punch numbers into a calculator and get some result.
I saw this first hand when I tried to help my girlfriend take some slightly more advanced math. If she encountered an assignment where she was unsure of how to proceed, she would grab the calculator and examine each and every button on it, trying to find that "magic button". In most cases the assignment could be solved perfectly without a calculator.
I like my previous math professor's attitude. When solving some problem on the blackboard, he could say "and then you can punch this into a calculator and get some number, but that's not the important part".
You have gig-e speed connections to the internet? Must be nice. Otherwise I suspect your comment is irrelevant.
Well the article talks about how you can throttle speeds to various departments, which may very well, if not usually, be on the same LAN and is thus likely using gigabit connections (from the server at least).
If so, I thought there was a clear separation between blocks at the pyramids.
From what I've heard, the blocks were so close together scientists were hard pressed to explain how they placed them, as it's exceedingly difficult even today. I guess this "wet stone" theory would explain it rather nicely.
I know the article says what it will be used for, but why do we need to bend the particles path
As found on this page, the magnet is used to measure the momentum of the particles. I'm not into physics, but I imagine it's a bit like rolling (equally sized) balls along a line, and then have a fan on the side trying to deflect the movement. A large momentum (and thus mass, assuming all the balls travel at the same speed) would lead to less deflection.
Except GPU's are at least as high transistor count as CPU's these days and based on transistor count growth GPU's will exceed CPU's in the next one to two generations.
Current GPU's are already ahead of current CPU's: Amd X2 1mbit cache: 233.2 million transistors Intel Core 2 Quad: 582million transistors Nvidia 8800 GTX: 681 million transistors
3/ After all this preloading the user gets their interactive shell and starts an application that is not normally used (time for disk access required) that requires more memory that is available. The OS is not stupid so it moves some stuff that it might not need yet but isn't sure to virtual memory - so the memory your preloaded applications are using gets written to a different part of the disk (which takes time). Bringing it back from virtual memory later will be easier and faster than the initial preload which is why I don't think what would normally be the obvious step of just releasing the memory will be taken as mentioned above - please correct me if I'm wrong.
This is just not how it works. Virtual memory is a bit like how telephones work (well at least used to work). Each local exchange has X users connected to it. However, it's unusual for all people to be talking at the same time. Since it would be very expensive to have a dedicated cable for each user further into the backbone, the central has Y available lines, where Y X, usually a lot less.
When a user picks up the receiver, the exchange picks one of the available lines and routes the line to the user, so that (s)he may dial. In the event that there's no available lines, the user doesn't get a dial tone.
Physical memory is the lines from the backbone into the local exchange. Each application however always "sees" all the memory as available, similar to how you don't worry about if you can dial when you pick up the receiver, you just expect it to be.
The term "virtual memory" comes from the fact that the way the application accesses memory is not directly linked with the physical memory.
The cpu and the OS then cooperates to "route" (map) the physical pages to the pages used by the application, so to speak.
Due to the way this cooperation is done (at least on windows), it is possible for the OS to store some of the physical pages on a different medium, say a hard disk, which is usually called swapping. This is not a requirement for using virtual memory.
So when a program tries to allocate more memory, and there are no free physical pages, the OS can copy a physical page to disk (swap it out), and then clear it and hand it to the program as if it was free all along. When the application who's page was swapped out wants to access that page, the OS get's notified and copies the page back into a free, possibly swapped out, page. The program is not aware that this process is going on, it's handled by the CPU and OS transparently.
Now, here's the deal with the SuperFetch. Since the pages used by SuperFetch represents data that's already on disk, it's already swapped out so to speak. So when the OS needs a free page, all it has to do is to remove the page from its SuperFetch page table, clear it and serve it to the application.
PS: That virtual memory analogy is probably flawed, read a real article if you want to get the full picture.
Personally I think it is stupider than doublespace since memory limited programs like image editors are commonplace now. The annoyance of not being able to print for a couple of minute while the memory swaps out can come to everyone.
But since SuperFetch preloads program data, the data is already backed on disk, it just needs to update the SuperFetch cache table and voila, fresh memory page ready to serve the hungry hippo. Should be almost as quick as a regular alloc. At least that's my take on it.
What kind of hardware they tested it on? If the CPU is 2 GHz, doesn't 20-40 times faster mean that the GPU runs at something like 40-80 GHz?
The 19xx XT(X) core runs at ~650Mhz, and has 48 pixel pipelines. It's a bit like having 48 650Mhz cpu's at your disposal (0.65Ghz * 48 = ~30Ghz, but clock speed is not everything). However, the GPU is highly specialized while the CPU is more flexible, which means that for some jobs the GPU is far better, and for other the CPU is best.
Slashdot Mirror
Is it PowerPoint's fault, or the fault of the Powerpoint creator?
I usually find it very hard to take notes from a PowerPoint presentation. When taking math classes, the lecturer would write on the blackboard prepared sentences, and with a speed that made it easy to follow.
In my CS classes, the lecturer says they've been "forced" by the administration to use PowerPoint presentations. If I simply read them, they're easy enough to follow, but the information doesn't "stick" the way it does when I take notes from a blackboard. If I try to take notes, it's hard because the lecturer very often doesn't say exactly what's on the slide. It also means he can talk a lot quicker and thus not give me enough time to take notes from that slide.
So I have to try to listen to him, read the slide, try to convert what's on the slide into something I can write, and then write the notes, all at the same time. That leaves very little room to actually think about what's being said, and so I don't get the understanding I get when taking notes from a blackboard.
For me, the notes are important not just because they're a good reference for later, but also because the information sticks a lot better when writing it down. Not having good notes due to PowerPoint is one of the primary reasons I find some CS classes hard (especially the math heavy ones, like digital signal processing).
Proper use of PowerPoint would be to use it during a lecture to show complex graphs and images, which may be hard to convey on a blackboard (for instance what histogram equalization does to an image).
The fact that you have just explained it in a way which is subtly wrong supports the idea that it is counterintuitive. An animal does not adapt. It is born with a certain set of DNA, which it cannot change or control, and it lives or dies as a result what DNA it has (along with other factors like chance).
Most organisms don't express all of their genes all the time. They got regulating mechanisms which responds to it's current state and the environment. So in that case I would say it's perfectly possible that one organism can adapt, since it's regulatory mechanisms could mutate and thus respond to some external cue, allowing the organism to survive while it's "brother" dies.
My first call would been to http://www.ibas.com/
Indeed, very odd. During a visit (with my class) at Ibas, they said they would have little to no problem recovering almost all of the data even after several reformat/rewrite cycles. And while Ibas ain't cheap, I'm pretty sure Ibas would be an order of magnitude or two cheaper than rescanning all the documents, not to mention faster.
Exactly. JP2000 was lossless, was it not?
JPEG 2000 supports lossless compression, yes.
What happened to that format, anyway? I was expecting it to be ubiquitous by now.
My guess is patents. And the fact that it only took a few years for bandwidth and storage to negate many of the advantages of jpeg2000 over the plain jpeg standard.
I got a nice prerecorded voice
I did some research, turns out the voice is fully synthesized, not a prerecorded message. It was very good, only a few glitches in the flow between words.
There are all sorts of reasons someone might think a drive has failed. They're not all correct.
I can attest to this. One of my hd's started to "crash" (spun down, parked head, spun up again) with shorter and shorter intervals. I replaced it, thinking it was dying. After a while, another hd started doing the same, and a third started to drop out and come back online (as if I had disconnected it and reconnected it). Turns out, when I switched from the chipset sata controller to the onboard raid controller, everything was nice and dandy, even the first drive works like a charm now.
So, for limited applications, voice recognition is getting along fairly well I must say.
Doesn't that also mean that if you sue them and lose, then you have to pay their legal fees?
AFAIK, we have a similar system in Norway, but I'm fairly sure I've read about several cases where the judge ruled that the plaintiff should not pay the legal fees (all of it, or the defendants part). So it would be a bit like a slap on the wrist to the defendant, in case they'd been good at covering their asses or something.
939=DDR Ram
AM2=DDR2 Ram
That's part of my point, but I fail to see yours. I was talking about taking one of the beefier X2's they had for the 939 and ship it cheap.
If I was AM2 rather than 939, I myself would be down on this in a heartbeat.
I also got a 939 rig, and I haven't quite understood the whole AM2 move from AMD. From what I've seen so far, AM2 doesn't bring a whole lot of improvements to the table, but what it does is equalize the upgrade costs between an AMD system and an Intel system. And in these days, that's hurting AMD bad I suspect.
If AMD needs some easy cash, why not release something for the 939 system? A reasonably priced, speedy dual core for instance? All I can get from my local shops is the X2 3800, which while dirt cheap, is the slowest X2. Why not sell the 4800 you had for not quite as little? I'd buy that as a intermediate upgrade.
There is no such thing as "massively parallel!" It makes no sense! Parallel in qualitative, NOT quantitative! Things are either parallel or they're not, there are no degrees of "parallelness!"
Sure there are. Say you want to find the maximum of 4 integers. You can do that in parallel, but you won't gain much if you have more than two processors (or execution units). Contrast this with say rendering an image using a path tracer, where each ray is independent of each other. First problem is hard to scale up, second one isn't. I'd say that means that ray tracing is a "more parallel" task.
Also, writing algorithms that has to run on 10000 processors efficiently is not exactly the same as one that has to run on 4 processors, in the same way that writing a multiplayer game that handles four players isn't the same as writing one that can handle thousands of concurrent players. So they toss on the "massive" part to separate the cases. At least that's my take on it.
its called a monopoly. It puts companies in a position where they can't say "Fuck off" because doing would result in a massive loss of customers
Well it's more like the prisoner's dilemma imho. If both ATI and Nvidia said "fuck off", MS would have a big problem. If only one of them said it however...
It's not a single photon interfering with itself.
The interference pattern will occur even if there's only one photon in the apparatus at a time (that is, a photon hits the detector before a new one is generated).
See this page for instance.
I spent two hours every five minutes telling it not to reboot until the render finished. I immediately yanked the internet connection and haven't updated it since.
Alternative is to just move the window off-screen, something like below the taskbar (I usually leave 2 pixels worth so I can get it back if I want to). Then you just have to contend with the annoying keyboard focus change every 15 minutes.
Just because they use "tech" devices does not mean they understand how they actually work. A calculator is an example. Many are taught how to use one, but have no concept of the math involved.
Indeed. Here in Norway, there's an ongoing debate about the rather appalling math level of our school kids. I'm certain it has a LOT to do with calculators. All they learn now is how to punch numbers into a calculator and get some result.
I saw this first hand when I tried to help my girlfriend take some slightly more advanced math. If she encountered an assignment where she was unsure of how to proceed, she would grab the calculator and examine each and every button on it, trying to find that "magic button". In most cases the assignment could be solved perfectly without a calculator.
I like my previous math professor's attitude. When solving some problem on the blackboard, he could say "and then you can punch this into a calculator and get some number, but that's not the important part".
You have gig-e speed connections to the internet? Must be nice.
Otherwise I suspect your comment is irrelevant.
Well the article talks about how you can throttle speeds to various departments, which may very well, if not usually, be on the same LAN and is thus likely using gigabit connections (from the server at least).
If so, I thought there was a clear separation between blocks at the pyramids.
From what I've heard, the blocks were so close together scientists were hard pressed to explain how they placed them, as it's exceedingly difficult even today. I guess this "wet stone" theory would explain it rather nicely.
I know the article says what it will be used for, but why do we need to bend the particles path
As found on this page, the magnet is used to measure the momentum of the particles. I'm not into physics, but I imagine it's a bit like rolling (equally sized) balls along a line, and then have a fan on the side trying to deflect the movement. A large momentum (and thus mass, assuming all the balls travel at the same speed) would lead to less deflection.
Except GPU's are at least as high transistor count as CPU's these days and based on transistor count growth GPU's will exceed CPU's in the next one to two generations.
Current GPU's are already ahead of current CPU's:
Amd X2 1mbit cache: 233.2 million transistors
Intel Core 2 Quad: 582million transistors
Nvidia 8800 GTX: 681 million transistors
rename *.jpeg *.jpg
Even if rename wasn't there, it would be easier than unix it seems:
for %A in (*.jpeg) do ren %A %~nA.jpg
(used uppercase A for illustration).
Damn. 300 watts just for a single video card.
From TFA: "Power consumption was measured using a Kill-A-Watt power meter that measures a power supply's power draw directly from the wall outlet".
3/ After all this preloading the user gets their interactive shell and starts an application that is not normally used (time for disk access required) that requires more memory that is available. The OS is not stupid so it moves some stuff that it might not need yet but isn't sure to virtual memory - so the memory your preloaded applications are using gets written to a different part of the disk (which takes time). Bringing it back from virtual memory later will be easier and faster than the initial preload which is why I don't think what would normally be the obvious step of just releasing the memory will be taken as mentioned above - please correct me if I'm wrong.
This is just not how it works. Virtual memory is a bit like how telephones work (well at least used to work). Each local exchange has X users connected to it. However, it's unusual for all people to be talking at the same time. Since it would be very expensive to have a dedicated cable for each user further into the backbone, the central has Y available lines, where Y X, usually a lot less.
When a user picks up the receiver, the exchange picks one of the available lines and routes the line to the user, so that (s)he may dial. In the event that there's no available lines, the user doesn't get a dial tone.
Physical memory is the lines from the backbone into the local exchange. Each application however always "sees" all the memory as available, similar to how you don't worry about if you can dial when you pick up the receiver, you just expect it to be.
The term "virtual memory" comes from the fact that the way the application accesses memory is not directly linked with the physical memory.
The cpu and the OS then cooperates to "route" (map) the physical pages to the pages used by the application, so to speak.
Due to the way this cooperation is done (at least on windows), it is possible for the OS to store some of the physical pages on a different medium, say a hard disk, which is usually called swapping. This is not a requirement for using virtual memory.
So when a program tries to allocate more memory, and there are no free physical pages, the OS can copy a physical page to disk (swap it out), and then clear it and hand it to the program as if it was free all along. When the application who's page was swapped out wants to access that page, the OS get's notified and copies the page back into a free, possibly swapped out, page. The program is not aware that this process is going on, it's handled by the CPU and OS transparently.
Now, here's the deal with the SuperFetch. Since the pages used by SuperFetch represents data that's already on disk, it's already swapped out so to speak. So when the OS needs a free page, all it has to do is to remove the page from its SuperFetch page table, clear it and serve it to the application.
PS: That virtual memory analogy is probably flawed, read a real article if you want to get the full picture.
Personally I think it is stupider than doublespace since memory limited programs like image editors are commonplace now. The annoyance of not being able to print for a couple of minute while the memory swaps out can come to everyone.
But since SuperFetch preloads program data, the data is already backed on disk, it just needs to update the SuperFetch cache table and voila, fresh memory page ready to serve the hungry hippo. Should be almost as quick as a regular alloc. At least that's my take on it.
its not like he's spending more than anyone else, especially considering the intense special effects.
I guess that explains it, after ~100 million has gone to SF, there's not a whole lot left for decent actors and script writers etc...
What kind of hardware they tested it on? If the CPU is 2 GHz, doesn't 20-40 times faster mean that the GPU runs at something like 40-80 GHz?
The 19xx XT(X) core runs at ~650Mhz, and has 48 pixel pipelines. It's a bit like having 48 650Mhz cpu's at your disposal (0.65Ghz * 48 = ~30Ghz, but clock speed is not everything). However, the GPU is highly specialized while the CPU is more flexible, which means that for some jobs the GPU is far better, and for other the CPU is best.