There are plenty of architectural improvements yet to be made in processor design. Once feature shrinking becomes even more difficult, clock speeds may begin to change more slowly, but chips will still improve in performance. More effor will be thrown into developing advanced architectural features around existing gate technology.
And don't forget the software side of technology! As more and more software is written towards a multi-threaded architecture, the speed advantages presented by multi-processing and multi-threading architectures will become even greater.
One development I am watching with rapt attention is the transition to Simultaneous Multi Threading Processors (SMT). This is still in the works, but processors such as the Alpha 21464 will be built around this design in the near future.
SMT procs move some of the process table down on to the processor itself such that the processor can fill time while waiting for a cache miss to be serviced by task switching to a seperate thread. Further, SMT allows simultaneous dispatch of instructions from *multiple* instruction streams. This sort of architecture makes much more efficient use of parallelism in hardware than current superscalar processors. Further, executing on the same chip, the different threads can synchronize *in cache* which is far more efficient than hitting memory like current SMP systems are forced to do. Very exciting...
There's no way a processor company would sell the same part for 10 years. If they can't shrink their gates any further, they'll just find new ways to exploit parallelism in hardware with more advanced architectures.
I'm a bit familiar with the research at Rice, and basically they are building gates by deforming nanotubes in certain ways to alter their electrical characteristics. They've been playing with it and playing with it until they can build structure which act like logic gates. The implications boggle the mind.
I say its high time for nanotech logic to start ramping up, so this is very exciting. Silicon is only going to carry us so much further. Once you get down to a semiconductor gate that is 20 molecules across, the Physics get much more interesting, and electromigration starts eating your lunch.
A very interesting field to follow...
--Lenny
Debian should come with warning stickers...
on
Debian Retail on CNN
·
· Score: 3
"Warning! For advanced users only."
Debian is great, but I would never recommend it to a newbie (such as someone who might stumble upon a box in Best Buy). Its really not a "first" distribution. When our LUG does install-fests we use Mandrake, which is nice and cuddly. Most of the members, however, run Debian.
I like the Debian banners that I've seen around the web:
"Debian: your *next* Linux distribution."...I think that sums things up rather nicely.
I must gain objective knowledge of the origin of all space, time, matter and energy, including me, the knower himself, and no rinky-dink little backwater planet overpopulated with superstitious primitives is going to stop me! Igor, the switch!
Beauty. This deserves to be in a fortune file.
It seems ~90% of Slashdot readers side with the scientists, but I wonder if there isn't a real concern here. This is something that you would have to be a scientist just to make a judgement on, though, so I suppose we are stuck with their discretion, whether we like it or not. Let's just hope that they aren't Mad Scientists!
They mention "strange matter" a few times, with no explanation. I am but an engineer, and not very knowledgable about such things. But surely there is a theoretical physicist in the audience who could field this question, and enlighten the Slashdot readership. Please?
"Pentium" I kind of like as a name, but Itanium? It doesn't quite roll off the tongue. I guess I understand AMD thinking "K7" isn't a very marketable name, but "Merced" sounds pretty snappy to me. I guess marketing has convinced them that they *have* to make up a name for their products to avoid trademark hassles.
pity. Not that I like Intel much, but...pity. --Lenny
Amiga has been chasing their tale for years now. They seem incapable of sticking with an idea long enough to create an actual product.
All the waffling has taken its toll, too. Even the old hardcore Amigans are leaving now. There remain precious few die-hards...certainly not enough to constitute a market.
But maybe he can do that bluish ghost thing to get his ideas across...
--Lenny
He sidestepped the clone issue.
on
Be on the G4
·
· Score: 2
All he did was state that they weren't "welcome" on Apple computers. That is wholly irrelevent here. What he was supposedly answering was the question about whether Be was going to embrace the emerging indie PowerPC boards. Notice, these will *not* be made by Apple, and won't even be Mac-compatible clones. Thus, Apple is not involved one bit.
His statement really said nothing at all. He simply reiterated their reasons for not supporting the Mac platform. Those reasons have no bearing on the emerging "no name" PPC market, however. He was hoping that we would forget the question somewhere in the middle of his apparent muddling about x86 and SGI.
It sounds like Be just doesn't want to be on PowerPC anymore. I'm still hoping for a Linux PPC explosion, though.
..is a fairly respected site. I use it on occasion. Anand seems pretty dedicated, as well as being humble and level headed. (traits many claim that Tom lacks)
As for the K7... I've never had doubts about it. I have always liked AMD, and they seem to have produced a truly superior design this go round. What with Merced-delays, they may be able to make a dent in Intel.
That is, if they can get motherboard manufacturers to cooperate...
If there are, indeed, entire galaxies composed of anti-matter, than its not exact a "rare" substance, now is it? Its just a little difficult to come by round these parts.:)
Similarly, the vast majority of matter (and/or anti-matter) in the universe is in a plasma state. Being terrestrial, however, we don't bump into plasma very often.
Scientists have been working on this sort of thing for a while. Its still very interesting. I'm not sure if/how the existence of anti-matter galaxies would effect the Big Bang theory. I am but a humble Engineer. Do we have any Physicists in the room?
1) Mandrake is a co-maintainer of Enlightenment (more than just a contributer) 2) Isn't it VA Linux systems now?
...nitpicks, I know, but this is Slashdot, for heaven's sake. I expect that > 60% of us keep up with Enlightenment.
And remember, kids: Geoff has nothing to do with Linux-Mandrake.
I, for one, would like to hear about the file browser that they are going to add to the Enlightenment "Desktop Shell". I still haven't found one for Linux that suits my fancy. The new KDE fm may do the trick, though...
If I recall, Be left the PowerPC platform citing problems getting access to specs for the G3 processor itself. Apple wasn't being friendly with them, it seems. This information that IBM is releasing is a motherboard design. It shows how to physically and electrically make a board for the PowerPC. Being a software company, Be certainly doesn't care about this. They just want to know how to program for the proc and chipset.
However, most people I have spoken with who follow Be have told me that this was just an excuse. Infact, Be wanted to switch to x86 anyway, and they just sort of smeared Apple on their way out. Be could have gotten the info they needed, from someone else if not from Apple. If nothing else, they could have approached the LinuxPPC people, who have done just fine working with the G3.
In summary: this doesn't help Be, but I don't think that Be really wanted to be helped. If anyone out there follows this stuff more closely, please add to/correct my statement.
If I lived in Kansas, I would promptly leave. I certainly do not want to raise my children in such an environment. I had no idea that any secular education board in the US was so backward as to actually pass something like this.
And this information is being banned at the *university* level? This is truly unbelievable.
--Lenny
Wait! Are you saying it has an OLVWM-style pager?
on
Corel Linux Preview
·
· Score: 2
And it looks pretty appealing to me. The view can be broken into an arbitrary number of frames (a feature that I have been yearning for in file browsers for a while). Further, it is completely built around their object system (Open Parts, I believe), so the handler for any filetype can be run from within a browser frame. This is the sort of thing that OLE and COM promised, but never really delivered.
This filemanager (Konqueror), is the main reason that I am looking forward to KDE 2.0 .
I've heard plenty about the effort to port Linux to IA-64, but I don't think I've come across a single article on MS's efforts. Surely they are working on it as well.
Now, I know that their development model is in no way open, but I'm still surprised that I haven't heard reports about the transition to Win64 (or whatever it will be called). Perhaps they are so bogged down trying to get W2K to a beta stage so they can finally ship the thing that they aren't focusing on IA-64 yet.
1) It doesn't come from a company. 2) It has a really cool name (down with TLA's!)
There's something ironically appealing about a heavy-duty piece of (weapons-grade?) crypto with "fish" in the name. Then there is the whole seussian angle...
My life would be unlivable without the network transparency of X. Whatever finally does displace X (possibly Berlin) absolutely *must* have network transparency.
--Lenny
Yes I do. Bandwidth costs.
on
New Cyberlaws
·
· Score: 1
Do you have any idea how much bandwidth Spam consumes on the net? You can filter it out when you retrieve your mail, or delete it when it gets to you, but that spam still got shoved through the backbone, along with millions of its brothers.
All of that traffic which virtually no one wants ties up "useful" (insert your definition of useful) communication. Its like someone littering your expressway with huge crates. Sure, you can drive around them, but it will slow you down. And do you really want to have to drive around those stupid crates?
Cybersquatting isn't a particular nice facet of the net, but I still say Spam is worse...
--Lenny
Cybersquatting isn't nearly as bad as SPAM
on
New Cyberlaws
·
· Score: 1
The title pretty much says it all. I'm not sure what can be done to truly deter spam, though. This bill about cybersquatting is probably a good idea...
Unix is in great need of reduction. The man/info split is a good example of competing solutions to the same problem (a competition that has never ended), but there are many others. I really yearn for a (GNU) Linux distribution that has the guts to dispense with some of the old baggage and redo some of the older systems in a more unified and sane manner. At times, I've considered doing it myself...
Seriously: what I feel we need is a *progressive* distribution that isn't afraid of breaking a few ties with the old Unix world. Why do we have three versions of grep? Why do we still have "edit" lieing around? Who in their right mind still uses xedit (by choice)? Why is the find command so difficult to use? (and *don't* tell me to read the docs: I've read them all) These are holdovers from the past...largely anachronistic.
Most all distros coming out these days just gank the Red Hat or Debian package list and try to differentiate themselves with a "slick" install program, or "pentium optimized" compiles. Is anyone trying to define a better base for a Linux system?
Now, I understand the benefits of backwards compatibility with Unix, and this ancestry should not be broken without good reason. However, in some cases I *do* think we have good reason. But somehow, what constitutes a "Linux install" has been set for years, and almost no one is questioning that selection. I suppose that they are happy to have some kind of "standard base", but I don't like many of the inconsistencies in that base.
*sigh* I'll get off of my soap cube now. But, please: someone inform me if there is such a project in the works.
--Lenny, who has launched off into day-dream mode.
Actually, I like Sun, but I'm glad there are people who are open about their problems with it...
I'm curious. You mention making your bosses "listen to reason". What are you trying to convince them to do? Do you want to change systems (to another OS), or standardize on a few compilers? Or do you want to get rid of the Sun software altogether (move to GNU perhaps)?
Or are you considering something even more drastic, like buying a few MS products?
SGI Linux Environment Features:
*Includes Red Hat Linux 6.0 *Improved NFS stability with the knfsd patch *Improved Web serving performance *Security strengthening through ICMP denial of service patch
...what is this knfsd patch? If it improves Linux's NFS system, then we want it! (and we want it Open Source) Lack of a journalling filesystem and weak NFS are among the few large deficiencies that still hold Linux back technically. SGI is already helping Linux out in the filesystem department. Perhaps they are going to contribute to NFS development as well. That would be a big win for open source.
The last major problem AFAIK is Linux's scalability to 4 proc and up. And we all know that this is a priority within the kernel bunch right now. By the time Linux 3.0 rolls around, we may have a very serious kernel on our hands.
Slashdot Mirror
There are plenty of architectural improvements yet to be made in processor design. Once feature shrinking becomes even more difficult, clock speeds may begin to change more slowly, but chips will still improve in performance. More effor will be thrown into developing advanced architectural features around existing gate technology.
And don't forget the software side of technology! As more and more software is written towards a multi-threaded architecture, the speed advantages presented by multi-processing and multi-threading architectures will become even greater.
One development I am watching with rapt attention is the transition to Simultaneous Multi Threading Processors (SMT). This is still in the works, but processors such as the Alpha 21464 will be built around this design in the near future.
SMT procs move some of the process table down on to the processor itself such that the processor can fill time while waiting for a cache miss to be serviced by task switching to a seperate thread. Further, SMT allows simultaneous dispatch of instructions from *multiple* instruction streams. This sort of architecture makes much more efficient use of parallelism in hardware than current superscalar processors. Further, executing on the same chip, the different threads can synchronize *in cache* which is far more efficient than hitting memory like current SMP systems are forced to do. Very exciting...
There's no way a processor company would sell the same part for 10 years. If they can't shrink their gates any further, they'll just find new ways to exploit parallelism in hardware with more advanced architectures.
--Lenny
I'm a bit familiar with the research at Rice, and basically they are building gates by deforming nanotubes in certain ways to alter their electrical characteristics. They've been playing with it and playing with it until they can build structure which act like logic gates. The implications boggle the mind.
I say its high time for nanotech logic to start ramping up, so this is very exciting. Silicon is only going to carry us so much further. Once you get down to a semiconductor gate that is 20 molecules across, the Physics get much more interesting, and electromigration starts eating your lunch.
A very interesting field to follow...
--Lenny
"Warning! For advanced users only."
...I think that sums things up rather nicely.
Debian is great, but I would never recommend it to a newbie (such as someone who might stumble upon a box in Best Buy). Its really not a "first" distribution. When our LUG does install-fests we use Mandrake, which is nice and cuddly. Most of the members, however, run Debian.
I like the Debian banners that I've seen around the web:
"Debian: your *next* Linux distribution."
--Lenny
Slackware->Red Hat->Mandrake->Debian
It seems ~90% of Slashdot readers side with the scientists, but I wonder if there isn't a real concern here. This is something that you would have to be a scientist just to make a judgement on, though, so I suppose we are stuck with their discretion, whether we like it or not. Let's just hope that they aren't Mad Scientists!
--Lenny
They mention "strange matter" a few times, with no explanation. I am but an engineer, and not very knowledgable about such things. But surely there is a theoretical physicist in the audience who could field this question, and enlighten the Slashdot readership. Please?
thanks,
--Lenny
"Pentium" I kind of like as a name, but Itanium? It doesn't quite roll off the tongue. I guess I understand AMD thinking "K7" isn't a very marketable name, but "Merced" sounds pretty snappy to me. I guess marketing has convinced them that they *have* to make up a name for their products to avoid trademark hassles.
pity. Not that I like Intel much, but...pity.
--Lenny
Amiga has been chasing their tale for years now. They seem incapable of sticking with an idea long enough to create an actual product.
All the waffling has taken its toll, too. Even the old hardcore Amigans are leaving now. There remain precious few die-hards...certainly not enough to constitute a market.
--Lenny
But maybe he can do that bluish ghost thing to get his ideas across...
--Lenny
All he did was state that they weren't "welcome" on Apple computers. That is wholly irrelevent here. What he was supposedly answering was the question about whether Be was going to embrace the emerging indie PowerPC boards. Notice, these will *not* be made by Apple, and won't even be Mac-compatible clones. Thus, Apple is not involved one bit.
His statement really said nothing at all. He simply reiterated their reasons for not supporting the Mac platform. Those reasons have no bearing on the emerging "no name" PPC market, however. He was hoping that we would forget the question somewhere in the middle of his apparent muddling about x86 and SGI.
It sounds like Be just doesn't want to be on PowerPC anymore. I'm still hoping for a Linux PPC explosion, though.
--Lenny
www.anandtech.com
... I've never had doubts about it. I have always liked AMD, and they seem to have produced a truly superior design this go round. What with Merced-delays, they may be able to make a dent in Intel.
..is a fairly respected site. I use it on occasion. Anand seems pretty dedicated, as well as being humble and level headed. (traits many claim that Tom lacks)
As for the K7
That is, if they can get motherboard manufacturers to cooperate...
--Lenny
If there are, indeed, entire galaxies composed of anti-matter, than its not exact a "rare" substance, now is it? Its just a little difficult to come by round these parts. :)
Similarly, the vast majority of matter (and/or anti-matter) in the universe is in a plasma state. Being terrestrial, however, we don't bump into plasma very often.
Scientists have been working on this sort of thing for a while. Its still very interesting. I'm not sure if/how the existence of anti-matter galaxies would effect the Big Bang theory. I am but a humble Engineer. Do we have any Physicists in the room?
--Lenny
1) Mandrake is a co-maintainer of Enlightenment (more than just a contributer)
2) Isn't it VA Linux systems now?
...nitpicks, I know, but this is Slashdot, for heaven's sake. I expect that > 60% of us keep up with Enlightenment.
And remember, kids: Geoff has nothing to do with Linux-Mandrake.
I, for one, would like to hear about the file browser that they are going to add to the Enlightenment "Desktop Shell". I still haven't found one for Linux that suits my fancy. The new KDE fm may do the trick, though...
--Lenny
If I recall, Be left the PowerPC platform citing problems getting access to specs for the G3 processor itself. Apple wasn't being friendly with them, it seems. This information that IBM is releasing is a motherboard design. It shows how to physically and electrically make a board for the PowerPC. Being a software company, Be certainly doesn't care about this. They just want to know how to program for the proc and chipset.
However, most people I have spoken with who follow Be have told me that this was just an excuse. Infact, Be wanted to switch to x86 anyway, and they just sort of smeared Apple on their way out. Be could have gotten the info they needed, from someone else if not from Apple. If nothing else, they could have approached the LinuxPPC people, who have done just fine working with the G3.
In summary: this doesn't help Be, but I don't think that Be really wanted to be helped. If anyone out there follows this stuff more closely, please add to/correct my statement.
--Lenny, who dreams of an inexpensive PPC clone.
If I lived in Kansas, I would promptly leave. I certainly do not want to raise my children in such an environment. I had no idea that any secular education board in the US was so backward as to actually pass something like this.
And this information is being banned at the *university* level? This is truly unbelievable.
--Lenny
If so, I need to install this *tonight*.
--Lenny
And it looks pretty appealing to me. The view can be broken into an arbitrary number of frames (a feature that I have been yearning for in file browsers for a while). Further, it is completely built around their object system (Open Parts, I believe), so the handler for any filetype can be run from within a browser frame. This is the sort of thing that OLE and COM promised, but never really delivered.
This filemanager (Konqueror), is the main reason that I am looking forward to KDE 2.0 .
--Lenny
I've heard plenty about the effort to port Linux to IA-64, but I don't think I've come across a single article on MS's efforts. Surely they are working on it as well.
Now, I know that their development model is in no way open, but I'm still surprised that I haven't heard reports about the transition to Win64 (or whatever it will be called). Perhaps they are so bogged down trying to get W2K to a beta stage so they can finally ship the thing that they aren't focusing on IA-64 yet.
anyone know?
--Lenny
routinely. They call it "The World According to Alta Vista". Brunching is a pretty good comedy site...
Check it out: www.brunching.com
--Lenny
Damn...you stole my subject line.
I personally am rooting for twofish because:
1) It doesn't come from a company.
2) It has a really cool name (down with TLA's!)
There's something ironically appealing about a heavy-duty piece of (weapons-grade?) crypto with "fish" in the name. Then there is the whole seussian angle...
But I wonder what happened to the other fishes?
"One fish, two fish,
red fish, blue fish".
--Lenny
My life would be unlivable without the network transparency of X. Whatever finally does displace X (possibly Berlin) absolutely *must* have network transparency.
--Lenny
Do you have any idea how much bandwidth Spam consumes on the net? You can filter it out when you retrieve your mail, or delete it when it gets to you, but that spam still got shoved through the backbone, along with millions of its brothers.
All of that traffic which virtually no one wants ties up "useful" (insert your definition of useful) communication. Its like someone littering your expressway with huge crates. Sure, you can drive around them, but it will slow you down. And do you really want to have to drive around those stupid crates?
Cybersquatting isn't a particular nice facet of the net, but I still say Spam is worse...
--Lenny
The title pretty much says it all. I'm not sure what can be done to truly deter spam, though. This bill about cybersquatting is probably a good idea...
--Lenny
Unix is in great need of reduction. The man/info split is a good example of competing solutions to the same problem (a competition that has never ended), but there are many others. I really yearn for a (GNU) Linux distribution that has the guts to dispense with some of the old baggage and redo some of the older systems in a more unified and sane manner. At times, I've considered doing it myself...
Seriously: what I feel we need is a *progressive* distribution that isn't afraid of breaking a few ties with the old Unix world. Why do we have three versions of grep? Why do we still have "edit" lieing around? Who in their right mind still uses xedit (by choice)? Why is the find command so difficult to use? (and *don't* tell me to read the docs: I've read them all) These are holdovers from the past...largely anachronistic.
Most all distros coming out these days just gank the Red Hat or Debian package list and try to differentiate themselves with a "slick" install program, or "pentium optimized" compiles. Is anyone trying to define a better base for a Linux system?
Now, I understand the benefits of backwards compatibility with Unix, and this ancestry should not be broken without good reason. However, in some cases I *do* think we have good reason. But somehow, what constitutes a "Linux install" has been set for years, and almost no one is questioning that selection. I suppose that they are happy to have some kind of "standard base", but I don't like many of the inconsistencies in that base.
*sigh* I'll get off of my soap cube now. But, please: someone inform me if there is such a project in the works.
--Lenny, who has launched off into day-dream mode.
Actually, I like Sun, but I'm glad there are people who are open about their problems with it...
I'm curious. You mention making your bosses "listen to reason". What are you trying to convince them to do? Do you want to change systems (to another OS), or standardize on a few compilers? Or do you want to get rid of the Sun software altogether (move to GNU perhaps)?
Or are you considering something even more drastic, like buying a few MS products?
--Lenny, who still uses OpenLook at work.
The last major problem AFAIK is Linux's scalability to 4 proc and up. And we all know that this is a priority within the kernel bunch right now. By the time Linux 3.0 rolls around, we may have a very serious kernel on our hands.
--Lenny