Even *single-cell* flagellates have what can be considered a rusimentary sense of smell, and the capability of changing their locomotion in order to lead them to food. That sort of ability is present all the way up through the multicellular ladder, and "smell" (or response to airborne chemical signals) have been well-known for quite some time in plants.
Frankly, I'm susprised that they didn't start out with an assumption that smell was involved.
I'm not entirely positive that they *did* upgrade. I haven't followed the hardware in a while, but it's entirely possible that some models are still running on those 33 MHz CPUs.
If they did upgrade, one reason would simply be that whatever chips are "commodity" at the moment are cheaper. Sort of like DIMMs, it's cheaper to buy 400 MHz DDR memory than plain old PC133.
That is true in some situations, but not in all of them. When you look at RDBMS work, and are talking about handling thousands of transactions per second on many-gigabyte data sets - which can't be pre-cached, and involve tying together a great deal of data sources in rather complex ways, it's very demanding on hardware. Developers of most RDBMS systems go to pretty good lengths to tweak out all of the performance they can. (Whether developers who write the SQL do is another matter.) In that situation, though, it's usually not the CPU that matters - it's the bandwidth. The beauty of the Opteron is that every time you add another CPU, you add another 128-bit memory interface, like the "big iron" provided well before AMD brought it to a commodity level.
There are plenty of mathematical simulations where memory latency is the key - no matter how much you tweak the software, the interdependencies of the various elements mean that you trample all over your memory sets. That's a hardware limitation, not application.
Even when you get into CPU-bound operation, there are plenty of instances where the hardware makes the difference. Video encoding is still the classic example where you just can't have enough horsepower.
Show me where video cards have been replaced with software. General-purpose isn't ever going to take over that. Physics is another area where general-purpose CPUs (even multi-core) don't have what it takes. (Whether you *need* that level of physics is another matter).
If a general-purpose CPU can handle your task, it will come out much cheaper. But there are still tasks where general-purpose CPUs just can't cut it.
Look at video encoding and decoding - a $20 ASIC can compete with a $300 CPU in terms of performance, while being produced on a massively inferior technology. Put a few of those ASICs on a 65nm CPU running at multi-GHz, and for a very small amount of real estate, you'd be able to encode dozens (hundreds?) of video streams in real-time. Again, whether enough people need that is another matter.
The only place that specialized hardware loses is in cost.
General-purpose often wins in value - because of the economies of scale. Specialized hardware always wins in performance. In markets where the "specialized" hardware can be manufactured and sold in quantity, then it wins in performance and is still price-competetive.
Look at a $50 video card. You simply could *not* produce 3D graphics like that from a general-purpose CPU (of any cost) at the same speed.
Yesterday's discussion was about commodity hardware as routers. They win in value until you hit the eventual limitations of general-puprpose hardware, but at that point, then it's like video cards: WIthout specialized hardware, you're just not going to cut it.
"Back in the day", even JPG compression/decompression was considered sufficiently heavy that there was a niche market in ASICs to handle it.
Your comments are a lot more true than many people realize. Specialized hardware always wins.
As an example, people talk about using using multi-GHz machines for TIVO-type appliances, and "getting away" with 600 Mhz or so if your card has hardware MPG encoding. Some of the original TIVOs, because of their reliance on specialized chips and ASICs, used measly 33 MHz CPUs - and worked just fine.
It's not the number of processors or cores that they're afraid of, it's the fact that with the exception of a very few cases, your performance does not scale linearly with the number of CPUS, it is less than 1:1. To make it worse, as the number of CPUs rises, the cost to intelligently, quickly deliver sufficient bits to and from all of the CPUs gets exponentially higher.
Recently, some of our managers wanted to see what it would cost to purchase a system that would significantly outperform our 8-way Opteron for RDBMS work. I got numbers on machines from various manufacturers, and when the managers saw them, the conversation was instantly over.
Didn't Sun try that sort of idea with the UltraSparc T1? If I recall correctly, while the concept of lots of light cores was cool, the real-world performance didn't do any better than Intel- or AMD-based systems.
... of an almost perverse little daydream that I had some time ago. I thought that it would be fun if someone were to made T1, T3, and other interfaces that connected via... USB. A USB connection has enough bandwidth (at least on paper) to run a T3 with ease, and you could pop 32 (or more) USB 2 ports in a machine very easily. And for the lesser-bandwidth interfaces, you could run them off of a USB hub.
So, imagine a single machine with 30, 50, or 60 network interfaces coming out of it, all sprouting USB cables. What a mess.
The more that I think about it, that's not so much of a daydream as a nightmare.
It's not frame-relay, but I run a T3 with Sangoma cards, with per-IP traffic shaping at both ends (a hundred and some-odd different IPs). The machines at either end (one Xeon, one Opteron) can't even really be considered "warming up" when the T3 is saturated.
On a 2621 (which, admittedly, is ancient) with two T1s, the poor thing would drown itself just doing per-IP bandwidth limitation before the T1s were full. Not to mention that you're quite limitted in the total number of IPs for which you can perform individual bandwidth limits on the 2621. Perhaps the more modern models are different in that regard.
$9 of electricity in 5 minutes. Where I live, at 10 cents per KWH, that makes for 90 KWH. That much energy delivered in 5 minutes means a rate of about one megawatt. That means that a half-dozen of these cars could tap out an entire power plant.
Even if you've stored the energy up in another capacitor over time to provide the recharge, it still sounds a little unpractical. Let's say that it's storing at 1000 volts - you'd still need to deliver about 1,000 amps, which would require a conductor about as big as your arm.
... it'll be more available than a land-line. Since my cable modem seems to have a 10% chance of dropping for an hour or two exactly at midnight on any given weeknight, and a 20% chance on any given weekend night, that sounds like it would be *very* reliable. ComCrap seems to have picked midnight as the time to perform any maintenance they want, without informing their customers.
I've been buying AMD since before it was cool, for at least 10 years. But I recently bought a laptop with a t2300e chip for the lower power draw (15 watts), and have been completely blown away at the performance. This puts my X2s and Pentium Ds to shame. A friend of mine who does video rendering on a quad-core mac was pretty impressed at how fast this bottom-of-the-line CPU can crank through video encoding.
Regardless of the name-calling on both sides, the Core 2 Duo is a *very* fast chip.
You run into a dark, cloudy storm, and lose power. And the worse the storm is, the less chance you have of developing any power. For some reason, that doesn't sound like much more than a one-off gimmick to me.
Yes, you could store energy in batteries, but storing enough power to get that boat very far in storm winds means a LOT of weight, which means a lower draft, more resisitance, and the need for more panels....
A few years ago, I had a phone that I really, *really* liked, but had used it so much that I wore the face off of the buttons. So I bought another on eBay, and took the buttons out and installed them in my old phone. But first, I powered up the phone just out of curiosity. It was still activated in the previous owner's name, the address book was still populated, etc.. They hadn't even bothered *trying* to erase any data.
What I would like out of an OS...
on
A New Kind of OS
·
· Score: 2, Interesting
Some time ago, I worked on a friend's computer that was running Windows 95 on a Pentium 166. I was astounded at how fast and responsive it was. Windows XP on an A64/P4 barely keeps up, yet offers very little more to me in terms of usefulness. Neither Windows, MacOS, nor XWindows particularly fits #4, at least not for me.
I will say, in terms of scalability, XWindows is a *real* screamer on a quad-Opteron with 8 gigs of RAM and a nice, fast SCSI array.
Back when I worked tech support (more than a decade ago), you did *NOT* leave your lunch in the fridge. If you did, it wasn't even worth looking for at lunchtime.
... it's how the cows are raised and treated. I've dined on cows grazed freely and fed grass and grain in the Austrian Alps, and it simply can't compare to the beef from a cow that was simply fattened up with no thought to its well-being, simply to increase profits.
Also, a friend of mine raises calves each year, and they're also free-grazed, grain-fed, and cared for properly. You simply *can not* go to a supermarket and find beef like this, no matter how much you pay. And guess how much the cost works out to, including the steaks, sirloins, filets, etc.: $2.00 per pound.
Wow. That's 26 CPU-years for one frame. 1/5,000,000th of real-time. At that point, you would think that custom hardware for rendering would be much faster and more cost-efficient. If only there were some sort of processer dedicated just to rendering 3D scenes... oh, yeah. They're called video cards.:-)
Don't get me wrong, I'm sure that Pixar knows what they're doing. But at the same time, the PC and video game industry learned a decade ago that it was a waste of time to fiddle around rendering 3D images on a general-purpose CPU. I'm surprised that there hasn't been more leverage of GPU capabilities (or even customs chips) for that sort of thing.
I have a dual-core A64 and a dual dual-core Opteron system both running XP 64, and they are astoundingly stable. As well as poor driver coding, don't forget to look at your hardware for the cause as well. It's been a long time since I've bought anything from Creative, but as I recall, they've always had a knack for turning out cards with weird little bugs.
The question, however, is this: Is it WORTH it to put the entire computer on a chip?
The memory controller is already on the chip. Looking at the rest of the motherboard stuff, like the PCI/PCI-X/PCI-E busses, audio, serial, ethernet, etc., it's not cost-effective. They can be manufactured to fully sufficient speeds on much less expensive processes.
The only other chip that would make any sense would be a GPU - but GPUs aren't usually so much cycle-limitted as bandwidth-limitted. Besides that, quad-core chips are clearly server-oriented chips, and most servers don't need that sort of video power.
Photoshop can max out the four cores in my dual dual-core Opteron setup. Admittedly, I don't do that often, but that's still one app which *can*, and that's just a desktop app. Most server-oriented applications, however, are designed to take advantage of multiple CPUs.
I just bought an 8-socket Opteron. The fans on that are practically large enough to levitate a hummer. When you power the beast up, the fans spin up sequentially to avoid too much draw on the 12V line. When they're all spun up, before the "Cool-N-Quiet" kicks in, it sounds like a jet aircraft about to take off. In actual use, however, they're comparatively quiet. There's plenty of room even in that chassis for removing double the heat if need be.
Slashdot Mirror
Even *single-cell* flagellates have what can be considered a rusimentary sense of smell, and the capability of changing their locomotion in order to lead them to food. That sort of ability is present all the way up through the multicellular ladder, and "smell" (or response to airborne chemical signals) have been well-known for quite some time in plants.
Frankly, I'm susprised that they didn't start out with an assumption that smell was involved.
steve
I'm not entirely positive that they *did* upgrade. I haven't followed the hardware in a while, but it's entirely possible that some models are still running on those 33 MHz CPUs.
If they did upgrade, one reason would simply be that whatever chips are "commodity" at the moment are cheaper. Sort of like DIMMs, it's cheaper to buy 400 MHz DDR memory than plain old PC133.
steve
That is true in some situations, but not in all of them. When you look at RDBMS work, and are talking about handling thousands of transactions per second on many-gigabyte data sets - which can't be pre-cached, and involve tying together a great deal of data sources in rather complex ways, it's very demanding on hardware. Developers of most RDBMS systems go to pretty good lengths to tweak out all of the performance they can. (Whether developers who write the SQL do is another matter.) In that situation, though, it's usually not the CPU that matters - it's the bandwidth. The beauty of the Opteron is that every time you add another CPU, you add another 128-bit memory interface, like the "big iron" provided well before AMD brought it to a commodity level.
There are plenty of mathematical simulations where memory latency is the key - no matter how much you tweak the software, the interdependencies of the various elements mean that you trample all over your memory sets. That's a hardware limitation, not application.
Even when you get into CPU-bound operation, there are plenty of instances where the hardware makes the difference. Video encoding is still the classic example where you just can't have enough horsepower.
steve
Show me where video cards have been replaced with software. General-purpose isn't ever going to take over that. Physics is another area where general-purpose CPUs (even multi-core) don't have what it takes. (Whether you *need* that level of physics is another matter).
If a general-purpose CPU can handle your task, it will come out much cheaper. But there are still tasks where general-purpose CPUs just can't cut it.
Look at video encoding and decoding - a $20 ASIC can compete with a $300 CPU in terms of performance, while being produced on a massively inferior technology. Put a few of those ASICs on a 65nm CPU running at multi-GHz, and for a very small amount of real estate, you'd be able to encode dozens (hundreds?) of video streams in real-time. Again, whether enough people need that is another matter.
The only place that specialized hardware loses is in cost.
General-purpose often wins in value - because of the economies of scale. Specialized hardware always wins in performance. In markets where the "specialized" hardware can be manufactured and sold in quantity, then it wins in performance and is still price-competetive.
Look at a $50 video card. You simply could *not* produce 3D graphics like that from a general-purpose CPU (of any cost) at the same speed.
Yesterday's discussion was about commodity hardware as routers. They win in value until you hit the eventual limitations of general-puprpose hardware, but at that point, then it's like video cards: WIthout specialized hardware, you're just not going to cut it.
"Back in the day", even JPG compression/decompression was considered sufficiently heavy that there was a niche market in ASICs to handle it.
Your comments are a lot more true than many people realize. Specialized hardware always wins.
As an example, people talk about using using multi-GHz machines for TIVO-type appliances, and "getting away" with 600 Mhz or so if your card has hardware MPG encoding. Some of the original TIVOs, because of their reliance on specialized chips and ASICs, used measly 33 MHz CPUs - and worked just fine.
steve
It's not the number of processors or cores that they're afraid of, it's the fact that with the exception of a very few cases, your performance does not scale linearly with the number of CPUS, it is less than 1:1. To make it worse, as the number of CPUs rises, the cost to intelligently, quickly deliver sufficient bits to and from all of the CPUs gets exponentially higher.
Recently, some of our managers wanted to see what it would cost to purchase a system that would significantly outperform our 8-way Opteron for RDBMS work. I got numbers on machines from various manufacturers, and when the managers saw them, the conversation was instantly over.
steve
Didn't Sun try that sort of idea with the UltraSparc T1? If I recall correctly, while the concept of lots of light cores was cool, the real-world performance didn't do any better than Intel- or AMD-based systems.
steve
So, imagine a single machine with 30, 50, or 60 network interfaces coming out of it, all sprouting USB cables. What a mess.
The more that I think about it, that's not so much of a daydream as a nightmare.
steve
It's not frame-relay, but I run a T3 with Sangoma cards, with per-IP traffic shaping at both ends (a hundred and some-odd different IPs). The machines at either end (one Xeon, one Opteron) can't even really be considered "warming up" when the T3 is saturated.
On a 2621 (which, admittedly, is ancient) with two T1s, the poor thing would drown itself just doing per-IP bandwidth limitation before the T1s were full. Not to mention that you're quite limitted in the total number of IPs for which you can perform individual bandwidth limits on the 2621. Perhaps the more modern models are different in that regard.
steve
$9 of electricity in 5 minutes. Where I live, at 10 cents per KWH, that makes for 90 KWH. That much energy delivered in 5 minutes means a rate of about one megawatt. That means that a half-dozen of these cars could tap out an entire power plant.
Even if you've stored the energy up in another capacitor over time to provide the recharge, it still sounds a little unpractical. Let's say that it's storing at 1000 volts - you'd still need to deliver about 1,000 amps, which would require a conductor about as big as your arm.
"For all controls, select the safest (to prevent loss of data or system access), most secure value by default."
In other words, treat the user like they don't know what they're doing. Slow *everyone* down, in order to save the idiots.
I really like knowing that when a dialog box pops up, the enter key will usually complete the task that I requested in the first place.
steve
I've been buying AMD since before it was cool, for at least 10 years. But I recently bought a laptop with a t2300e chip for the lower power draw (15 watts), and have been completely blown away at the performance. This puts my X2s and Pentium Ds to shame. A friend of mine who does video rendering on a quad-core mac was pretty impressed at how fast this bottom-of-the-line CPU can crank through video encoding.
Regardless of the name-calling on both sides, the Core 2 Duo is a *very* fast chip.
steve
You run into a dark, cloudy storm, and lose power. And the worse the storm is, the less chance you have of developing any power. For some reason, that doesn't sound like much more than a one-off gimmick to me.
Yes, you could store energy in batteries, but storing enough power to get that boat very far in storm winds means a LOT of weight, which means a lower draft, more resisitance, and the need for more panels....
steve
Waitaminute... Mars' atmosphere is 95% CO2. Wouldn't it be easier to just get rid of a hundred PPM here?
A few years ago, I had a phone that I really, *really* liked, but had used it so much that I wore the face off of the buttons. So I bought another on eBay, and took the buttons out and installed them in my old phone. But first, I powered up the phone just out of curiosity. It was still activated in the previous owner's name, the address book was still populated, etc.. They hadn't even bothered *trying* to erase any data.
1. Non-intrusive.
2. Stable.
3. Efficient.
4. Intuitive.
Some time ago, I worked on a friend's computer that was running Windows 95 on a Pentium 166. I was astounded at how fast and responsive it was. Windows XP on an A64/P4 barely keeps up, yet offers very little more to me in terms of usefulness. Neither Windows, MacOS, nor XWindows particularly fits #4, at least not for me.
I will say, in terms of scalability, XWindows is a *real* screamer on a quad-Opteron with 8 gigs of RAM and a nice, fast SCSI array.
steve
Back when I worked tech support (more than a decade ago), you did *NOT* leave your lunch in the fridge. If you did, it wasn't even worth looking for at lunchtime.
steve
... it's how the cows are raised and treated. I've dined on cows grazed freely and fed grass and grain in the Austrian Alps, and it simply can't compare to the beef from a cow that was simply fattened up with no thought to its well-being, simply to increase profits.
Also, a friend of mine raises calves each year, and they're also free-grazed, grain-fed, and cared for properly. You simply *can not* go to a supermarket and find beef like this, no matter how much you pay. And guess how much the cost works out to, including the steaks, sirloins, filets, etc.: $2.00 per pound.
steve
Wow. That's 26 CPU-years for one frame. 1/5,000,000th of real-time. At that point, you would think that custom hardware for rendering would be much faster and more cost-efficient. If only there were some sort of processer dedicated just to rendering 3D scenes... oh, yeah. They're called video cards. :-)
Don't get me wrong, I'm sure that Pixar knows what they're doing. But at the same time, the PC and video game industry learned a decade ago that it was a waste of time to fiddle around rendering 3D images on a general-purpose CPU. I'm surprised that there hasn't been more leverage of GPU capabilities (or even customs chips) for that sort of thing.
steve
I have a dual-core A64 and a dual dual-core Opteron system both running XP 64, and they are astoundingly stable. As well as poor driver coding, don't forget to look at your hardware for the cause as well. It's been a long time since I've bought anything from Creative, but as I recall, they've always had a knack for turning out cards with weird little bugs.
The question, however, is this: Is it WORTH it to put the entire computer on a chip?
The memory controller is already on the chip. Looking at the rest of the motherboard stuff, like the PCI/PCI-X/PCI-E busses, audio, serial, ethernet, etc., it's not cost-effective. They can be manufactured to fully sufficient speeds on much less expensive processes.
The only other chip that would make any sense would be a GPU - but GPUs aren't usually so much cycle-limitted as bandwidth-limitted. Besides that, quad-core chips are clearly server-oriented chips, and most servers don't need that sort of video power.
steve
Photoshop can max out the four cores in my dual dual-core Opteron setup. Admittedly, I don't do that often, but that's still one app which *can*, and that's just a desktop app. Most server-oriented applications, however, are designed to take advantage of multiple CPUs.
steve
I just bought an 8-socket Opteron. The fans on that are practically large enough to levitate a hummer. When you power the beast up, the fans spin up sequentially to avoid too much draw on the 12V line. When they're all spun up, before the "Cool-N-Quiet" kicks in, it sounds like a jet aircraft about to take off. In actual use, however, they're comparatively quiet. There's plenty of room even in that chassis for removing double the heat if need be.
steve