Thunderbird/Spitfire/Mustang etc are almost certainly internal names for the project - Athlon/Duron/Itanium etc are made up 'marketting' names chosen because they're words that (hopefully unless someone screwed up) haven't been used in the English language before - thus avoiding all sorts of potential trademark pitfalls
Is sadly going to be that you will need a special chipset - unlike Intel, where the same memory controller that does single processor work is used for SMP, the AMD memory controller for SMP will have to be a seperate SKU (because it needs an extra set of 90-odd pins to talk to the other CPU) - the upside to this is that it's likely to perform better, the downside is that because this chip will ONLY be used in Athlon SMP boards it's volume will be low and as a result its cost is likely to be much higher (unlike for example an Intel 440BX which is used in many many single CPU boards and a relatively small number of SMP boards).
I currently buy SuperMicro P6DBEs (dual cpu 440BX boards) 10-20 at a time @~$160 a piece - I'd jump at the chance to be able to buy something similar that's Athlon based - but at a similar, or lower, all up (stuffed including CPUs) price - sadly I doubt this is going to happen any time soon
actually if you're not going to have cartridges with built-in cache then the slot1/slotA stuff doesn't make much sense cost wise or signal integrity-wise - I just wish I could buy all those dual-cpu socket370 (or Athlon) motherboards I so desperately need
Basicly most CPUs use a shared CPU bus to memory over which things like cache coherency is done - K7 (and DEC's Alpha) use a bus that's point-to point and rely on the memory controller to manage the cache coherency for them.
I believe the problem is that AMD have yet to release a memory controller with 2 CPU ports so that we can do 2-way SMP - K7 is supposed to be already to work it's just waiting for the chipset.
Sadly, because this memory controller chip will only be used for SMP systems and most systems are single CPU, this will mean that volumes on this chip will be low and it will likely cost more meaning dual motherboards will be more expensive than their Intel cousins:-(
I think your last point is the most important one - we have great (and expensive) CAD tools to build synchronous logic - but none for asynchronous stuff
(having said this I have to spend today building something that's a perfect fit for an async solution - if I had a tool that would do a safe glitch free synthesis - instead I have to waste 100s of gates resolving metastability etc etc to get the signals I need near a clock so I can live in a timing methadology I know works)
strange... the idea of evolution is part of my heritage passed down from previous generations of scientists and is most certainly one of my beliefs.... besides you can see evolution in action every year when a new flu virus comes around - I know from sneezing and wheezing that it very much does happen - every day.
Of course as a (somewhat) adherent to the concept of memes I consider your 'cherished beliefs' and mine to be akin to viruses competing for hosts in order to propagate themselves - may the fitest, most usefull memes win!
I think they're spread thin - they're trying to bring out a new CPU family (itanium) while at the same time they finally have some real competition from AMD at the top end and an erosion of profits in the middle and low ends. Oh yeah and the current jobs environment in the valley means that chip designers are evaporating to networking startups way faster than you can hire them.
Having said that I know there's a often lot of luck in getting on-the-edge chips to yield well - and they are probably really pushing the limits to keep up with AMD (as are AMD).
I thought genes were selfish. Sex was not "created" to increase genetic diversity, although that may be a side effect. Sex is merely a way for your selfish genes to find there way into another survival machine and (hopefully) propagate again.
Well like all things there are levels within levels.... IMHO the 'sex is a means for selfish genes to propagate' only can be applied to the genes that actually code for sex... otherwise you have to say 'the selfish genes that find it usefull to hang around with other genes that code for sex' which starts to sound like an organism rather than just a single selfish gene.
I think you can make valid arguments about this stuff at the gene level, at the organism level and at the species level. For example it makes sense for a species to have lots of different genes in its organism's immune systems so that a disease wipes out just some of the organisms (and genes) but not all. Obviously from the points of those individual genes this however isn't a good idea
Actually I decided a while ago that the main thing that Linux has going for it is 'genetic diversity' - in the sense that we're all using lots of different mailers/browsers/GUIs/etc.
The virus/infection analogy fits well here - consider a genetically engineered corn crop - a monoculture - every plant has identical DNA - and the whole thing will die if a blight mutates to fit just that particular DNA. On the other hand wild corn has tremendous genetic diversity - a survival mechanism evolved to combat just this sort of threat.
Of course that was the whole reason sex was created in the first place - to increase genetic diversity within a species to allow it to adapt better.
So far my experiments in this area have failed... I tried to mate KDE and GNome... but they just stood in the room with their backs to each other arms crossed pouting.... seems their a lot like pandas....
from people at work (using outlook of course)... kmail doesn't care, it harmlessly drops them in my inbox without reinfecting anyone else.
The hard part was writing all those carefully worded notes (all different) to my coworkers letting them down gently and explaining that I don't return their affections...
E-Bay should wake up - it's not a copy - the DCMA is intended to prohibit COPYING which is what a copyright gives you controll over (a photo of a physical item on an ebay page is not a copy - it's a depiction of the original).
Sadly Scientology has more lawyers than god (or the devil) and are used to cowering people inro submission
This is because each RDRAM accesses go to a single chip on a DIMM while an SDRAM access goes to all the chips on a DIMM - worst case for an SDRAM spreads the generated heat load across all the chips, on an RDRAM one chip gets REALLY hot.... you have to spread that heat out to help it disipate.
This is the worst case behaviour of course - normal behaviour of the two is roughly random accesses and is probably about the same.
So probably adding a heat sink is not all that usefull for your SDRAMs unless you know you have a problem - and then more airflow in the case is probably a better bet.
As for an EMI shield? unless your neighbors (or spouse) are complaining about their TV reception what's the point
I bet Linus is doing usefull work - consider that their on-the-fly code recompilation requires there be a hidden meta-OS hanging around managing the emulation world in which an x86 emulation (and linux) is embedded....
if so, can John Q. Linux (ha) get his hands on one?
Probably you have to buy one on a board, or in a box (sounds like a line from GE&H)..... from TMs point of view they are a (very) small company who are trying to do big things - supporting lots of small developers who are buying single chips probably doesn't buy them much - they have to be very carefull where they spend their money (even if they have $88m to play with their burn rate will be high - figure $150k+/engineer/year including benefits - $88m means ~200 engineers for 2 years, a CAD budget of $2m+/year etc - that should last them 'till their IPO..... if they're carefull)
IBM is almost certainly the fab-of-choice because their licensing of the x86 (many years back) gives TM some legal protection against Intel's lawyers.....
the bottom seems to have dried out of the Pentium marketplace in the past few weeks - cheap PIII-550 class cpus are virtually impossible to find - either demand's really high and the Gateway/Dell's of the world are soaking them up - or Intel has production problems too..... or they're just playing with the market to keep their margins high... just like the oil companies.... I suspect that despite AMD's inroads Intel's still big enough that it can do that
Now if only AMD would get it's dual processer bridge chips to market (in MBs as cheap as the cheapo pentium ones) so I can start buying tons of Athlons instead..... then I could dump Intel
I distinctly remember us buying 1.5Mb of core (those little round magnetic thingys with 1uS access times) for a cool million $$ for our B6700 which supported about 30 timeshared users and a large batch stream (from honest to got card readers no less) - any 486 PC would have walked all over it.
Anyway my point is that a mini like a PDP11 that could support 30 timeshared users for tens of thousands of $$ WAS a big deal.
While that SEEMS like a great idea that would work, we already have (some) trouble getting SOFTWARE drivers to work properly for some hardware on Linux, if the company won't release software drivers, what's to say that they'll release firmware drivers?
It IS a great idea - with a 10 year old existance proof - Mac add in cards from the MacII/NuBus onwards came with drivers in their ROMS - you brough a card home, plugged it in and rebooted and (if the driver writer was doing their job right) it worked out of the box. These days PCMCIA, 1394 and USB plug and play stuff SHOULD work this way too - borrowed your friends camera from his Mac? plug it into your Linux box and it should show up ready to serve images without any software installation - everything should work this way - and when enough devices do that customers expect this behaviour vendors wont be able sell buggy devices. Also the combination of flash ROM and the 'net give vendors a great way to fix buggy drivers in the field. From our point of view the real problem is how do you get the hardware vendors to support more than just M$ software - the simple answer is what's happening today Linux/BSD/etc are becoming critical mass - they have to - or lose out to their competitors.
Or what's to say that they won't charge extra for the hardware because it's being used on an 'under supported' 'minority' or 'rebel' OS?
If it's a portable driver format that's supposed to work on lots of OSs how could they know what to charge you?
Now whether they actually support you - or more likely whether M$ will toss some spanner in the works with the intention of nuking the viability such a system - is another question.
Slashdot Mirror
Thunderbird/Spitfire/Mustang etc are almost certainly internal names for the project - Athlon/Duron/Itanium etc are made up 'marketting' names chosen because they're words that (hopefully unless someone screwed up) haven't been used in the English language before - thus avoiding all sorts of potential trademark pitfalls
I currently buy SuperMicro P6DBEs (dual cpu 440BX boards) 10-20 at a time @~$160 a piece - I'd jump at the chance to be able to buy something similar that's Athlon based - but at a similar, or lower, all up (stuffed including CPUs) price - sadly I doubt this is going to happen any time soon
actually if you're not going to have cartridges with built-in cache then the slot1/slotA stuff doesn't make much sense cost wise or signal integrity-wise - I just wish I could buy all those dual-cpu socket370 (or Athlon) motherboards I so desperately need
I believe the problem is that AMD have yet to release a memory controller with 2 CPU ports so that we can do 2-way SMP - K7 is supposed to be already to work it's just waiting for the chipset.
Sadly, because this memory controller chip will only be used for SMP systems and most systems are single CPU, this will mean that volumes on this chip will be low and it will likely cost more meaning dual motherboards will be more expensive than their Intel cousins :-(
(having said this I have to spend today building something that's a perfect fit for an async solution - if I had a tool that would do a safe glitch free synthesis - instead I have to waste 100s of gates resolving metastability etc etc to get the signals I need near a clock so I can live in a timing methadology I know works)
silly person have you not seen the sorceror's apprentice?
Of course as a (somewhat) adherent to the concept of memes I consider your 'cherished beliefs' and mine to be akin to viruses competing for hosts in order to propagate themselves - may the fitest, most usefull memes win!
Having said that I know there's a often lot of luck in getting on-the-edge chips to yield well - and they are probably really pushing the limits to keep up with AMD (as are AMD).
Well like all things there are levels within levels .... IMHO the 'sex is a means for selfish genes to propagate' only can be applied to the genes that actually code for sex ... otherwise you have to say 'the selfish genes that find it usefull to hang around with other genes that code for sex' which starts to sound like an organism rather than just a single selfish gene.
I think you can make valid arguments about this stuff at the gene level, at the organism level and at the species level. For example it makes sense for a species to have lots of different genes in its organism's immune systems so that a disease wipes out just some of the organisms (and genes) but not all. Obviously from the points of those individual genes this however isn't a good idea
The virus/infection analogy fits well here - consider a genetically engineered corn crop - a monoculture - every plant has identical DNA - and the whole thing will die if a blight mutates to fit just that particular DNA. On the other hand wild corn has tremendous genetic diversity - a survival mechanism evolved to combat just this sort of threat.
Of course that was the whole reason sex was created in the first place - to increase genetic diversity within a species to allow it to adapt better.
So far my experiments in this area have failed ... I tried to mate KDE and GNome ... but they just stood in the room with their backs to each other arms crossed pouting .... seems their a lot like pandas ....
The hard part was writing all those carefully worded notes (all different) to my coworkers letting them down gently and explaining that I don't return their affections ...
Sadly Scientology has more lawyers than god (or the devil) and are used to cowering people inro submission
This is the worst case behaviour of course - normal behaviour of the two is roughly random accesses and is probably about the same.
So probably adding a heat sink is not all that usefull for your SDRAMs unless you know you have a problem - and then more airflow in the case is probably a better bet.
As for an EMI shield? unless your neighbors (or spouse) are complaining about their TV reception what's the point
I think not ... it shows how to add a foil EMI shield to SDRAM so it LOOKS like RDRAM ....
WinHec started today - everyone's announcing their latest and greatest homing to get game developer mindshare -
well of course the B6700 I was refering to was in NZ (at Otago) .... and back then a $NZ was worth more than a $US .... sigh
I bet Linus is doing usefull work - consider that their on-the-fly code recompilation requires there be a hidden meta-OS hanging around managing the emulation world in which an x86 emulation (and linux) is embedded ....
Probably you have to buy one on a board, or in a box (sounds like a line from GE&H)..... from TMs point of view they are a (very) small company who are trying to do big things - supporting lots of small developers who are buying single chips probably doesn't buy them much - they have to be very carefull where they spend their money (even if they have $88m to play with their burn rate will be high - figure $150k+/engineer/year including benefits - $88m means ~200 engineers for 2 years, a CAD budget of $2m+/year etc - that should last them 'till their IPO ..... if they're carefull)
IBM is almost certainly the fab-of-choice because their licensing of the x86 (many years back) gives TM some legal protection against Intel's lawyers .....
Now if only AMD would get it's dual processer bridge chips to market (in MBs as cheap as the cheapo pentium ones) so I can start buying tons of Athlons instead ..... then I could dump Intel
Just looking at the front panel switches (after 20 years!) my fingers felt this urge to key in the boot loader ....
Anyway my point is that a mini like a PDP11 that could support 30 timeshared users for tens of thousands of $$ WAS a big deal.
Who will get the Raquel Welch part? :-)
It IS a great idea - with a 10 year old existance proof - Mac add in cards from the MacII/NuBus onwards came with drivers in their ROMS - you brough a card home, plugged it in and rebooted and (if the driver writer was doing their job right) it worked out of the box. These days PCMCIA, 1394 and USB plug and play stuff SHOULD work this way too - borrowed your friends camera from his Mac? plug it into your Linux box and it should show up ready to serve images without any software installation - everything should work this way - and when enough devices do that customers expect this behaviour vendors wont be able sell buggy devices. Also the combination of flash ROM and the 'net give vendors a great way to fix buggy drivers in the field. From our point of view the real problem is how do you get the hardware vendors to support more than just M$ software - the simple answer is what's happening today Linux/BSD/etc are becoming critical mass - they have to - or lose out to their competitors.
Or what's to say that they won't charge extra for the hardware because it's being used on an 'under supported' 'minority' or 'rebel' OS?
If it's a portable driver format that's supposed to work on lots of OSs how could they know what to charge you?
Now whether they actually support you - or more likely whether M$ will toss some spanner in the works with the intention of nuking the viability such a system - is another question.