Hmm, well, if you ran a fullchip RTL simulation of an Athlon XP on a P4 your speed you be somewhere on the order of 10Hz-100Hz (depending on your simulation software). That's pretty slow.
In my experience, XP is much faster than 2000. When I upgraded my computers form 2000->XP, I defnitely noticed a speed increase. Not just bootup time, but overall performance. I still have one computer with 2000 (necessary because my employer's VPN software only works with 2000), and I just upgraded to 512MB, but it's still slow as a dog. My Celeron 433 Laptop with 256MB RAM is faster with XP, than my Celeron 800 Desktop with 512MB of RAM is with 2000. (But my P4 with XP blows them both away:-)
Yes, but by selling more units, they will produce more units. Assuming that they are selling below marginal cost, they could lose more by producing 10,000,000 units with 0 units of unsold inventory, than by producing 1,000,000 units with 500,000 leftover units. If you don't buy more units, they won't produce more, which puts a limit on the loss.
That's not strictly true. Sales of Windows is good for Microsoft and its suppliers (which, true, is a huge part of the economy). If a suitable replacement for Windows was available at a lower price, it would help Microsoft's customers (which are also a big part of the economy). It would lower the cost of the operating system software, which would increase the amount of money consumers will spend on other things (increasing production throughout the economy), and which would allow businesses to increase profits (or lower prices), since their costs would have lowered (either case would increase production).
The last sentence of that is misleading, if not plain wrong. That is the maximum number of virtual addresses you can form within one task, yes. But those addresses map all to 2^32 bytes of memory.
Chapter 3 of volume 3 of the current IA-32 manual is much more clear on this.
The easiest way to look at this is through paging, which clearly has a 32 bit size.
I'd be curious of a more complete reference for that citation (URL?)
This is wrong. The maximum linear address on IA-32 is 2^32. The maximum virtual address is also 2^32. Segments can't give you above that, they just adjust the base address and the limit. ALthough you can construct an address which goes up to 8GB (which a very large base, and a large virtual offset), the address will wrap around to below 4GB.
And, the physical address is LARGER than 32 bits. It is 36 bits on Wmt. The physical bus on the processors has 36 bits for the address (well, actually 33 bits, since all addresses are chunk aligned, but that's an implementation detail).
FYI, on x86-64, the maximum linear and virtual addresses are 2^48, and the maximum physical address is 2^40.
Why not just use FVWM instead of that fancy IceWM or the other new window managers? It's very fast, small, and configurable. I'm running on a 1.2 GHz Athlon / 128 MB, and FVWM works great for me.
Does anybody actually use those silly little file managers? IMHO, they just get in the way - why not just use the command line?
If a BMW cost only $20.00, then everyone would purchase one, and the rate of theft of BMW's would drop to almost zero. That doesn't mean it makes business sense for BMW to do that.
This is very basic economics. Obviously you are going to sell more units @ $200 than @ $1000. But Wolfram thinks it can make more profit selling the fewer units @ $1000 than @ $200. They can set the price to whatever they want since they are the only vendor of Mathematica (although other math software on the market obviously has some effect on the price).
The student version is essentially textbook price discrimination. They are charging everybody the maximum price they are willing to pay.
There is no truth to 'Rambus exclusivity' whatsoever. Intel has been shipping the SDRAM version of the i845 chipset for several months, and it is the fastest ramping chipset ever in the industry. The only reason got the gap between the SDRAM and DDR versions is the time to validate it.
It's just assumed when dealing with computer components that giga=2^30
Not really. Gigaherz, Megapixel, Gigabit/Megabit are all powers of 10. Hard drive sizes are powers of 10, although file sizes are powers of 2. RAM is always power of 2.
Some things are just different. Like a floppy disk is 1.44MB, where MB=1024000 bytes (huh?)
Think about one of these babies in your living room, streaming music to the stereo, checking email on the couch, maybe even controlling the tv with the right ir port. It'd be very cool.
Why not just use a regular laptop?
Re:I like to accelerate my mac...
on
Dual G4 Mac Cube
·
· Score: 0, Offtopic
AMD is not doing fine. They had a net loss of two hundred million dollars last quarter. Furthermore, they admitted that they lost about 1% marketshare to Intel in that quarter. My predicition is that they will exit the microprocessor business if the PC industry doesn't pick up within 4-5 quarters. Their cost structure just doesn't support it.
Speed doesn't matter. The generic brand of macaroni and cheese could claim "60% the cost, 90% the cheese" (actually in the case of macaroni, I think it's closer to 25% cost, 100% the cheese), but most people will still choose to pay the premium for Kraft. Intel's the same way. It's brand is so much stronger than AMD's that almost everybody chooses to pay a premium for it.
You can't blame to current homogeneity of computers at all on Microsoft - they are the 'rebel' in the industry and the only company which has resisted Unix (though, barely: Windows is barely non-Unix. Hasn't at least one version of Windows been certified POSIX compliant?)
When you look at something like NetBSD, which runs on something like two dozen different architectures, that's the definition of homogeneity. Unix is definitely the originator of 'open systems' which is precisely what killed the diverse lineup of mainframes in the 80's and 90's.
You can also look at this from the hardware side, which has also gotten homogeneous.
Of course, homogeneity isn't all bad. Standardization significantly improves competition, thus improving quality and lowering price.
First of all, Hammer does not have a flat 64 bit address space. Its address space is 48 bits (linear) and 40 bits (physical). Second, Intel's 36 bit addressing does not require segments, it requires paging. It's a flat space from a process's point of view (limited to 4GB per process, of course).
Well, audio switching is nice so you can have one set of speakers shared between all of the computers. Otherwise, what's the alternative?
Ideally, KVM switches would have audio inputs and be selectable between 'mixed' (where all the audio inputs would be mixed, so you can achieve what you proposed) and 'switched' (so it only selects the audio for the current computer).
This is actually a really good point. I've wondered myself why companies that no longer sell/support a product don't just open it up wide. THis goes for software as well as hardware specs. Is there really that much competitive IP tied up in a discontinued product that they can't let it out?
One reason may be that the product may use some component that the company doesn't want to open up even own. For example, a piece of software may depend on some proprietary library which the company doesn't own (and, thus, doesn't have the authorization to open up), and opens itself up to legal trouble if it opens up. A hardware component may be partially designed by another firm and again the company is opening itself to lawsuits if it releases the information. In these cases it is very expensive to research whether or not a product can be opened, and this is a very difficult process if it wasn't originally designed to be open.
Slashdot Mirror
Hmm, well, if you ran a fullchip RTL simulation of an Athlon XP on a P4 your speed you be somewhere on the order of 10Hz-100Hz (depending on your simulation software). That's pretty slow.
In my experience, XP is much faster than 2000. When I upgraded my computers form 2000->XP, I defnitely noticed a speed increase. Not just bootup time, but overall performance. I still have one computer with 2000 (necessary because my employer's VPN software only works with 2000), and I just upgraded to 512MB, but it's still slow as a dog. My Celeron 433 Laptop with 256MB RAM is faster with XP, than my Celeron 800 Desktop with 512MB of RAM is with 2000. (But my P4 with XP blows them both away :-)
Yes, but by selling more units, they will produce more units. Assuming that they are selling below marginal cost, they could lose more by producing 10,000,000 units with 0 units of unsold inventory, than by producing 1,000,000 units with 500,000 leftover units. If you don't buy more units, they won't produce more, which puts a limit on the loss.
That's not strictly true. Sales of Windows is good for Microsoft and its suppliers (which, true, is a huge part of the economy). If a suitable replacement for Windows was available at a lower price, it would help Microsoft's customers (which are also a big part of the economy). It would lower the cost of the operating system software, which would increase the amount of money consumers will spend on other things (increasing production throughout the economy), and which would allow businesses to increase profits (or lower prices), since their costs would have lowered (either case would increase production).
Dude, if the MOV instruction is a load which doesn't hit cache, or a uncacheable store, it will take a heck of a lot more than 50 cycles to complete.
The last sentence of that is misleading, if not plain wrong. That is the maximum number of virtual addresses you can form within one task, yes. But those addresses map all to 2^32 bytes of memory.
Chapter 3 of volume 3 of the current IA-32 manual is much more clear on this.
The easiest way to look at this is through paging, which clearly has a 32 bit size.
I'd be curious of a more complete reference for that citation (URL?)
This is wrong. The maximum linear address on IA-32 is 2^32. The maximum virtual address is also 2^32. Segments can't give you above that, they just adjust the base address and the limit. ALthough you can construct an address which goes up to 8GB (which a very large base, and a large virtual offset), the address will wrap around to below 4GB.
And, the physical address is LARGER than 32 bits. It is 36 bits on Wmt. The physical bus on the processors has 36 bits for the address (well, actually 33 bits, since all addresses are chunk aligned, but that's an implementation detail).
FYI, on x86-64, the maximum linear and virtual addresses are 2^48, and the maximum physical address is 2^40.
You can NOT opt-out from them selling your personal information, including when and where you are calling.
This is incorrect. You can opt-out of subscribing to their service. So what's the problem?
Why not just use FVWM instead of that fancy IceWM or the other new window managers? It's very fast, small, and configurable. I'm running on a 1.2 GHz Athlon / 128 MB, and FVWM works great for me.
Does anybody actually use those silly little file managers? IMHO, they just get in the way - why not just use the command line?
Which benchmark - the one where the HSF falls off?
If a BMW cost only $20.00, then everyone would purchase one, and the rate of theft of BMW's would drop to almost zero. That doesn't mean it makes business sense for BMW to do that.
This is very basic economics. Obviously you are going to sell more units @ $200 than @ $1000. But Wolfram thinks it can make more profit selling the fewer units @ $1000 than @ $200. They can set the price to whatever they want since they are the only vendor of Mathematica (although other math software on the market obviously has some effect on the price).
The student version is essentially textbook price discrimination. They are charging everybody the maximum price they are willing to pay.
Intel is one of Slashdot's buggest advertisers, and is one of VA Linux's biggest (if not, the biggest) investors.
So why would Slashdot sponsor a boycott of Intel?
Are there black helicopters flying in your back yard?
There is no truth to 'Rambus exclusivity' whatsoever. Intel has been shipping the SDRAM version of the i845 chipset for several months, and it is the fastest ramping chipset ever in the industry. The only reason got the gap between the SDRAM and DDR versions is the time to validate it.
Sure -- and who are you going to sue for September 11th? Boeing, or United & American Arilines?
It's just assumed when dealing with computer components that giga=2^30
Not really. Gigaherz, Megapixel, Gigabit/Megabit are all powers of 10. Hard drive sizes are powers of 10, although file sizes are powers of 2. RAM is always power of 2.
Some things are just different. Like a floppy disk is 1.44MB, where MB=1024000 bytes (huh?)
Think about one of these babies in your living room, streaming music to the stereo, checking email on the couch, maybe even controlling the tv with the right ir port. It'd be very cool.
Why not just use a regular laptop?
You're a moron.
AMD is not doing fine. They had a net loss of two hundred million dollars last quarter. Furthermore, they admitted that they lost about 1% marketshare to Intel in that quarter. My predicition is that they will exit the microprocessor business if the PC industry doesn't pick up within 4-5 quarters. Their cost structure just doesn't support it.
Speed doesn't matter. The generic brand of macaroni and cheese could claim "60% the cost, 90% the cheese" (actually in the case of macaroni, I think it's closer to 25% cost, 100% the cheese), but most people will still choose to pay the premium for Kraft. Intel's the same way. It's brand is so much stronger than AMD's that almost everybody chooses to pay a premium for it.
You can't blame to current homogeneity of computers at all on Microsoft - they are the 'rebel' in the industry and the only company which has resisted Unix (though, barely: Windows is barely non-Unix. Hasn't at least one version of Windows been certified POSIX compliant?)
When you look at something like NetBSD, which runs on something like two dozen different architectures, that's the definition of homogeneity. Unix is definitely the originator of 'open systems' which is precisely what killed the diverse lineup of mainframes in the 80's and 90's.
You can also look at this from the hardware side, which has also gotten homogeneous.
Of course, homogeneity isn't all bad. Standardization significantly improves competition, thus improving quality and lowering price.
First of all, Hammer does not have a flat 64 bit address space. Its address space is 48 bits (linear) and 40 bits (physical). Second, Intel's 36 bit addressing does not require segments, it requires paging. It's a flat space from a process's point of view (limited to 4GB per process, of course).
Well, audio switching is nice so you can have one set of speakers shared between all of the computers. Otherwise, what's the alternative?
Ideally, KVM switches would have audio inputs and be selectable between 'mixed' (where all the audio inputs would be mixed, so you can achieve what you proposed) and 'switched' (so it only selects the audio for the current computer).
This is actually a really good point. I've wondered myself why companies that no longer sell/support a product don't just open it up wide. THis goes for software as well as hardware specs. Is there really that much competitive IP tied up in a discontinued product that they can't let it out?
One reason may be that the product may use some component that the company doesn't want to open up even own. For example, a piece of software may depend on some proprietary library which the company doesn't own (and, thus, doesn't have the authorization to open up), and opens itself up to legal trouble if it opens up. A hardware component may be partially designed by another firm and again the company is opening itself to lawsuits if it releases the information. In these cases it is very expensive to research whether or not a product can be opened, and this is a very difficult process if it wasn't originally designed to be open.