I would also suspect a lot of the badly redacted stuff is made on purpose to make people believe the "redacted" info. That or to distract people away from the non-redacted info.
If you had bothered to read my post, it is essentially what I'm saying. There *are* penalties to CISC, but they're not that big on *non* embedded CPUs. And I consider x86 CPUs to be mostly RISC chips with a CISC front-end. They've been like that since the Pentium Pro (hence why I said CISC is dead).
In the x86, there are so many instructions, the chip designers don't optimize all of them equally. If you want maximum efficiency, you will need to use the correct instruction, and it may vary from chip to chip. Whereas with a RISC architecture, it's a lot easier to guess which instruction to use (there may be only one).
The way things work now, it's hard to optimise regardless of whether it's RISC or CISC because the problems have changed. It's no longer about knowing how many cycles each instruction takes. Now, it's about latencies, pipelines, dependencies, cache, branch prediction, and so on. You end up with the same problems on any super-scalar architecture, not just the ones with CISC front-ends.
There really is no advantage to CISC, other than the backwards compatibility of the x86 architecture.
Compilers are still stuck trying to optimize at the CISC level.
No. Compilers are fully aware of what's happening and take into account what happens after CISC instructions are broken down. Of course there is some gymnastics and overhead involved in translating CISC instructions into RISC instructions, but it's not as bad as you make it sound. What's really complex in modern general-purpose CPUs is all the stuff related to superscalar execution: dependencies, out-of-order execution, branch prediction. Those are still there on a Power chip (except the Cell, which is special) or other high-performance RISC CPUs. Personally, I would love to see x86 go away (we have the source code so who cares!), but I don't think we'd see huge gains either.
There are no CISC CPUs anymore. There are RISC CPUs with RISC instruction sets (e.g. ARM) and there are RISC CPUs with CISC instruction sets (e.g. x86). The cores are mostly the same, except that the chips with CISC instructions need to do a little more work in the decoder. It requires a bit extra transistors and a bit more power, but it's not a huge deal for PCs and servers. Of course, for embedded applications, it makes a difference and for those it makes sense to have more "specialised" architectures (from microcontrollers to DSPs, ARM and all kinds of hybrids).
Do you have any idea of the paper work and investigations that get done everytime a cop pulls his taser? they don't ever do it unless the alternative is dangerous to them or the suspect.
I say make that even simpler. Any cop who uses the Taser receives it afterwards. That would create the perfect balance. It would get used when really necessary and not otherwise. (before anyone complains, I know very well it's not implementable)
Although some tools can be used as "weapons" as you're saying, there just aren't many non-lethal applications to H-bombs cruise missiles and chemical weapons.
Although one might argue that our applications we were using to test were not so memory intensive, the fact is it was a computationally intensive task that regularly accessed about 200-300mb of data from ram. I would think that even if everything would be pre-fetched into the 8MB CPU cache before it was used, we should at least see some small difference.
It all depends on how the memory is accessed. If you keep accessing elements at (seemingly) random locations in the data set, then it's likely that you'll hit the memory latency. If you keep iterating over all the data sequentially with few operations per load, you'll hit the memory throughput. On the other hand, if the data is a big matrix and you keep doing matrix multiplications (with a proper "blocking" algorithm), then you can effectively go nearly as fast as your L1 (which is only 32 kB) because you've minimised the loads. So, it *could* still be that memory is irrelevant to the performance of your software.
that's why all the plans involve putting it down somewhere.
If it was stored in gas form at atmospheric pressure, it wouldn't be a problem (it would just be silly). The problem is that if it's stored in highly compressed or solid form, then if something goes wrong and it goes back to gas, it *will* go up and escape, potentially killing anyone in the area.
Actually, my main concern is "what if it escapes?". Considering that CO2 is heavier than Oxygen, I wouldn't like to be anywhere near (i.e. within tens of km if not more) a site that stores thousands of tons of CO2.
There's two issues here. One is reducing the total energy consumed (i.e. not using it at all) and the other is reducing the peak power (choosing when to use energy). The former is always useful. The latter mainly works around infrastructure problems. In terms of reducing emissions, the only reasons I can see for changing when to use energy is to balance the load for "green" energy like wind/solar that aren't available all the time.
Gnome2: "we took Gnome1 and decided what options are best for you". Couldn't resist. I'm mostly using gnome (with kde apps), but that aspect of gnome2 annoyed the hell out of me.
Many, many people cannot grasp the idea that you will be worm food when you die. They instead seek things like Heaven, enlightenment, or Valhalla as a means to cope with what they do not understand.
The two aren't incompatible. A religious person (which I'm not) can believe that their body becomes worm food, while their soul/spirit/coredump goes somewhere nicer. That's probably how eco-burial should be sold (and probably how it already is).
Thanks, very enlightening. There's still one thing I'm trying to understand, though. I can easily understand how the "soft bribed" official would drop his opposition to some MS project. What I'm having a harder time understanding how how that person can end up actively bypassing votes (and other irregularities) for money. i.e. it's one thing to take money to just "change your mind" (which seems to be what your friend did), but it's another to actually commit some kind of fraud (or at least very obviously unethical).
I have a fruit company called "Apple Macintosh", but ever since I've been trying to diversify to computer hardware, I keep being threatened by another company called Apple. This is stifling my capacity to innovate.
Pretty soon, there'll be about 100 million broadband subscribers in the US alone. Multiplying by $5/months and 12 months would make them $60 billions per year for doing nothing. Sounds like a profitable business model.
me: Obviously now the best idea I can think of. Even when only thinking in terms of Palestinians' good. Bad place for a type. I did mean "obviously not the best idea" (you'll see the sentence makes more sense that way).
However, I take exception with your statement of "the military gear Israel uses to kill Palestinians." Just look at the death toll. There's a lot more Palestinians getting killed. Also, the Israeli pattern is almost always the same. Some Palestinian kills X Israelis, so the Israeli govt replies by sending F-16s. Nobody cares whether it achieves anything (they only need to pretend to), but the Israeli population is content that at least the F16s killed a bunch of Palestinian civilians (generally 2X) in return. I mean even many in the military itself went out and said they refused to continue killing Palestinians.
While you're right that Palestinians would (and do) use any method at their disposal to cause destruction in Israel, I doubt you'd see Israelis blowing themselves up or expressing the same kind of genocidal intent as the Palestinians currently are. If Israel is on the verge of being "pushed to sea" (which is how desperate the Palestinians currently are), I have no doubt we will see orthodox Jews and/or Zionist blowing themselves up or just going out to kill as many Palestinians as possible. Desperate times call for desperate measures and that's the sad thing.
It's easy to just sit back and declare that it's all our fault, that our aiding of Israel is the cause for all the problems and violence in the Middle East. It is certainly not the cause as the problem was there long before the US starting messing up there. So most of the original blame has to go with the British empire and other colonial powers. That being said, what the US is doing is certainly not helping. I'd say the US is effectively "buying time" for Israel, both in terms of opposing and UN resolution involving Israel and preventing the economy from collapsing from this untenable situation. I bet if US support was stopped, there would suddenly be a lot more incentive to make peace on the Israeli side.
I doubt that a "majority" of Palestinians want peace. Of course they do, who wouldn't. Of course, wanting peace generally doesn't involve "wanting peace at all cost". Currently, the Palestinians have "nothing" so of course it's not acceptable for them to just say "OK, war stops here with the status quo". Same if they (magically) were on the verge of pushing Israel to sea, I don't think the Israel govt would be happy to "just make peace and stop there". Of course, neither the current situation, nor the "Israel out to see" are in any way acceptable. And none of the sides realise that.
And Why TF would they fake a celebration over the death of school kids?
The celebration isn't fake, it's just taken out of context. So I go to your wedding and shoot some video you dancing and celebrating. The next month, people hijack planes and crash them into the WTC. Then I tell everyone that you supported the WTC attacks and show you celebrating (or course failing to mention that the celebration wasn't related to the WTC stuff at all).
Did you google Pallywood?
Yes. And I find the idea ludicrous. Especially when you consider: 1) How much power the pro-Israel lobby has in the US (i.e. how much money it spends on US politicians). 2) How little money the Palestinian equivalent has 3) In any case, when I see footage from the occupied territories on Canadian TV, it doesn't come from freelance Palestinians, but from Canadians.
Of course, then you see Palestinians dance in the streets, shoot guns in the air and hand out candies to passing cars.
Has it ever occurred to you that: 1) Most Palestinians don't actually celebrate these kinds of attacks 2) A lot of these images are actually either fabricated(*) or at least amplified
(*) It's not unheard of that journalists will shoot some kind of celebration (even in weddings you can see people shooting in the air) and will play it totally out of context to make it look like the Palestinians (or anyone else for that matter) is actually celebrating some evil act.
Slashdot Mirror
I would also suspect a lot of the badly redacted stuff is made on purpose to make people believe the "redacted" info. That or to distract people away from the non-redacted info.
If you had bothered to read my post, it is essentially what I'm saying. There *are* penalties to CISC, but they're not that big on *non* embedded CPUs. And I consider x86 CPUs to be mostly RISC chips with a CISC front-end. They've been like that since the Pentium Pro (hence why I said CISC is dead).
In the x86, there are so many instructions, the chip designers don't optimize all of them equally. If you want maximum efficiency, you will need to use the correct instruction, and it may vary from chip to chip. Whereas with a RISC architecture, it's a lot easier to guess which instruction to use (there may be only one).
The way things work now, it's hard to optimise regardless of whether it's RISC or CISC because the problems have changed. It's no longer about knowing how many cycles each instruction takes. Now, it's about latencies, pipelines, dependencies, cache, branch prediction, and so on. You end up with the same problems on any super-scalar architecture, not just the ones with CISC front-ends.
There really is no advantage to CISC, other than the backwards compatibility of the x86 architecture.
Hence the subject of my original post.
Compilers are still stuck trying to optimize at the CISC level.
No. Compilers are fully aware of what's happening and take into account what happens after CISC instructions are broken down. Of course there is some gymnastics and overhead involved in translating CISC instructions into RISC instructions, but it's not as bad as you make it sound. What's really complex in modern general-purpose CPUs is all the stuff related to superscalar execution: dependencies, out-of-order execution, branch prediction. Those are still there on a Power chip (except the Cell, which is special) or other high-performance RISC CPUs. Personally, I would love to see x86 go away (we have the source code so who cares!), but I don't think we'd see huge gains either.
There are no CISC CPUs anymore. There are RISC CPUs with RISC instruction sets (e.g. ARM) and there are RISC CPUs with CISC instruction sets (e.g. x86). The cores are mostly the same, except that the chips with CISC instructions need to do a little more work in the decoder. It requires a bit extra transistors and a bit more power, but it's not a huge deal for PCs and servers. Of course, for embedded applications, it makes a difference and for those it makes sense to have more "specialised" architectures (from microcontrollers to DSPs, ARM and all kinds of hybrids).
Do you have any idea of the paper work and investigations that get done everytime a cop pulls his taser? they don't ever do it unless the alternative is dangerous to them or the suspect.
I say make that even simpler. Any cop who uses the Taser receives it afterwards. That would create the perfect balance. It would get used when really necessary and not otherwise. (before anyone complains, I know very well it's not implementable)
Although some tools can be used as "weapons" as you're saying, there just aren't many non-lethal applications to H-bombs cruise missiles and chemical weapons.
I believe the OLPC's Geode is actually slower than your 333 MHz Pentium II. Also, IIRC when XP came out, the slowest recent machine were around 1 GHz.
Although one might argue that our applications we were using to test were not so memory intensive, the fact is it was a computationally intensive task that regularly accessed about 200-300mb of data from ram. I would think that even if everything would be pre-fetched into the 8MB CPU cache before it was used, we should at least see some small difference.
It all depends on how the memory is accessed. If you keep accessing elements at (seemingly) random locations in the data set, then it's likely that you'll hit the memory latency. If you keep iterating over all the data sequentially with few operations per load, you'll hit the memory throughput. On the other hand, if the data is a big matrix and you keep doing matrix multiplications (with a proper "blocking" algorithm), then you can effectively go nearly as fast as your L1 (which is only 32 kB) because you've minimised the loads. So, it *could* still be that memory is irrelevant to the performance of your software.
Just to make it clear, this is what I'm talking about.
that's why all the plans involve putting it down somewhere.
If it was stored in gas form at atmospheric pressure, it wouldn't be a problem (it would just be silly). The problem is that if it's stored in highly compressed or solid form, then if something goes wrong and it goes back to gas, it *will* go up and escape, potentially killing anyone in the area.
Actually, my main concern is "what if it escapes?". Considering that CO2 is heavier than Oxygen, I wouldn't like to be anywhere near (i.e. within tens of km if not more) a site that stores thousands of tons of CO2.
A: "Ah, no worries, it has a shitload of space on it."
But even better if you can say: "I've got 3.2 shitloads".
There's two issues here. One is reducing the total energy consumed (i.e. not using it at all) and the other is reducing the peak power (choosing when to use energy). The former is always useful. The latter mainly works around infrastructure problems. In terms of reducing emissions, the only reasons I can see for changing when to use energy is to balance the load for "green" energy like wind/solar that aren't available all the time.
Gnome2: "we took Gnome1 and decided what options are best for you". Couldn't resist. I'm mostly using gnome (with kde apps), but that aspect of gnome2 annoyed the hell out of me.
...and backup everything as soon as you got access to the HD content so you don't accidentally trash anything.
Exactly. Insurance companies make money by "protecting" you against uncertainties. The less uncertainty there is, the less they can insure.
Many, many people cannot grasp the idea that you will be worm food when you die. They instead seek things like Heaven, enlightenment, or Valhalla as a means to cope with what they do not understand.
The two aren't incompatible. A religious person (which I'm not) can believe that their body becomes worm food, while their soul/spirit/coredump goes somewhere nicer. That's probably how eco-burial should be sold (and probably how it already is).
Thanks, very enlightening. There's still one thing I'm trying to understand, though. I can easily understand how the "soft bribed" official would drop his opposition to some MS project. What I'm having a harder time understanding how how that person can end up actively bypassing votes (and other irregularities) for money. i.e. it's one thing to take money to just "change your mind" (which seems to be what your friend did), but it's another to actually commit some kind of fraud (or at least very obviously unethical).
Install noscript, set the FF option to disable animated gifs, disable flash. It's save again.
I have a fruit company called "Apple Macintosh", but ever since I've been trying to diversify to computer hardware, I keep being threatened by another company called Apple. This is stifling my capacity to innovate.
It's too late
Pretty soon, there'll be about 100 million broadband subscribers in the US alone. Multiplying by $5/months and 12 months would make them $60 billions per year for doing nothing. Sounds like a profitable business model.
And Why TF would they fake a celebration over the death of school kids?
The celebration isn't fake, it's just taken out of context. So I go to your wedding and shoot some video you dancing and celebrating. The next month, people hijack planes and crash them into the WTC. Then I tell everyone that you supported the WTC attacks and show you celebrating (or course failing to mention that the celebration wasn't related to the WTC stuff at all).
Did you google Pallywood?
Yes. And I find the idea ludicrous. Especially when you consider:
1) How much power the pro-Israel lobby has in the US (i.e. how much money it spends on US politicians).
2) How little money the Palestinian equivalent has
3) In any case, when I see footage from the occupied territories on Canadian TV, it doesn't come from freelance Palestinians, but from Canadians.
Of course, then you see Palestinians dance in the streets, shoot guns in the air and hand out candies to passing cars.
Has it ever occurred to you that:
1) Most Palestinians don't actually celebrate these kinds of attacks
2) A lot of these images are actually either fabricated(*) or at least amplified
(*) It's not unheard of that journalists will shoot some kind of celebration (even in weddings you can see people shooting in the air) and will play it totally out of context to make it look like the Palestinians (or anyone else for that matter) is actually celebrating some evil act.