I'd like to be tied to some railroad tracks and shmucked by a train, I think. Or maybe go skydiving, but have my chute fail. Quick, exciting, and only slightly painful.;)
I'm not American, so forgive me if I get the state wrong, but wasn't it Pennsylvania where several judges jailed innocent kids to get $2m in kickbacks?
There have been several such cases in the past few years.
Yep - either those who target netbooks or those which try to be resource friendly. If one can run on a much older system well then a newer system it should hum, plus not be such a big hit on the battery life.
Unfortunately, this isn't the case, since much of the drain is caused by components that aren't the CPU/RAM/etc.
More code and more overhead results in better power management. It's the same thing we have with DirectX/OpenGL - bigger/fatter drivers do a better job keeping the videocard(s) busy.
I know the feeling. My hometown of ~10k must've had 5 black people in it. It was so rare seeing one, that I couldn't help but stare. Looking back, I wonder if that made them uncomfortable.:P
Okay, we're on the same page - I just worded it poorly.
I guess I'd call PAE more of a "clever back" than "ugly hack".;)
Re: multi-megabyte pages
The point of the hack was that it didn't waste memory. You can read more at the sourceforge page: http://www.7-max.com/
You sound like the kind of person that'd gain something from looking into it.:)
Re: Memory pages
If wasting memory is the primary concern, I think 16-64KB pages would be an improvement without introducing much/any wastage. The smallest Windows process is 16KB, after all. But there's probably a reason 4KB is the default across most operating systems.
Why is/. greenlighting yet another article just for the sake of MS-bashing? Other posters have already pointed out the remaining process memory limits and consumer driver issues that make this a non-starter technically. Not to mention that there's not much (any?) consumer-level 32-bit hardware that has BIOS/Northbridge support for more than 4GB of RAM. Since the 64-bit versions don't cost any more, who cares?
The reason(as he explains) is that Microsoft used to allow more than 4GB of RAM before XP SP2, then they added the limitation. The majority of hardware released in the past 10 years will support > 4GB addressing ranges.
I actually have an old nForce 2 board that claims to support 2GB sticks. (4x2GB = 8GB) It has no 64bit drivers, which means it had to be designed for PAE usage, before the 4GB cap was implemented.
But this guy is saying that Microsoft is doing something "illegal" because he was able to hack his system and enable PAE, even though MS charges extra for that in their server OSes. Is it "illegal" for Ford to sell me a car that's computer governed to 105 MPH even though the engine can get it up over 130? Can I complain to the FTC if I chip the car to remove the limit and then destroy the transmission?
It's not illegal, and I agree that they can do what they want, including putting an artificial limitation in their software to spur sales. Many other companies(Apple, IBM, etc.) have done it too. With old IBM mainframes, couldn't you just flip a switch to get a RAM upgrade?;)
It's their right, and their choice, but there's no harm letting them bask in the fallout.
When XP came out, the average computer had ~256MB of RAM in it. (maybe less) Most games consumed less than that.
Now that Win7 is coming out, the average computer has 2GB of RAM in it. (maybe 4GB) Most games consume close to 2GB.
How many people are going to install 32bit Win7, or receive a 32bit Win7 computer? It seems rather ludicrous setting the limit right at the edge of what is currently being used, don't you think? 256MB -> 4GB was plenty of upgrade room, but with a 4GB cap 32bit Win7 will be obsolete in 2010.
That's an artificial limitation, too. Sun proved you can run 64bit operating systems and virtual machines on 32bit CPUs. Obviously it's slower, though, since 64bit operating systems assume the existence of some speed-enhancing CPU features.
I disagree. Paging is fundamental to memory management. You like having memory mapped to 4KB pages, but call it a hack when those 4KB pages are mapped to different physical memory locations? Why?
Apparently PAE is even more efficient than 64bit addressing. It's the faster option, although you're 100% correct about the per-process limit being roughly 3GB, which means it'll be obsolete shortly. Games, video editing, and virtual machines will be first to benefit from 64bit > 4GB memory usage.
What I've always wondered is why Windows uses 4KB pages. Larger pages are far more efficient. The guy who made 7-zip tried hacking in multi-megabyte pages, and had a 15% speedup. O_o
Errr no, Vista x64 runs with more then 4GB RAM without a problem. Windows 7 32 bit still has same limit. Windows 7 64 bit does not. There is a RAM extender called PAE that is used by all the OSes, Microsoft simply chooses not to install it in their desktop OSes because PAE has been known to break drivers and they want to push 64 bits to desktop. 4GB limit doesn't bother most desktop users and those who have more then 4GB run 64 anyways if they are power users. Yes, I consider 4GB+ users to be power users.
Interesting perception, but incorrect.
Most modern computers - ones shipping with 32bit windows - have PAE enabled because it is required for DEP. The driver argument is hollow, because of that. Clearly there are few enough incompatibilities that PAE/DEP is a good thing when it enhances security, and a bad thing when it deters someone from upgrading their OS. ($$$)
PAE is available in all versions of windows since Windows 2000. It doesn't increase the max available memory in every version - usually only datacenter/enterprise/server versions.
Also, I RTFA. The entire thing. I've also done some research on this in the past. Although I didn't figure out how to beat it for XP, I did bump my available memory up to 3.7GB without PAE.:)
I was just about to comment that the article seems more like single-task juggling.
Maybe I'm wrong, but my impression of multitasking was it was dealing with multiple sources of input at the same time.
So doing all three of those tests at once, and comparing who does the best. I know multitaskers often aren't good at focusing, and as this article demonstrates, aren't that great at quickly switching between single tasks - but how do they deal with multiple simultaneous tasks?
The big issue is the stability of the OpenGL 3.x codepaths on Linux. You'll need some relatively up to date drivers(binary blobs) to get all those new calls working. And no guarantees it won't break, later.
Since OpenGL 3.x and DX10/11 share a lot, it should be more straightforward than it was in the past porting from one to the other. The major differences between OGL and DX have partially been eliminated. Thanks to The Khronos Group, they're both moving in the same direction.
Are you implying that you've never had to upgrade hardware before to run the latest and greatest software? I know the Ubuntu of today uses more resources than the Ubuntu of five years ago. Since Canonical only supports their desktop versions of Ubuntu for a maximum of three years, it's very likely that some computers that ran Ubuntu fine in 2004-2005 will be unable to run the latest version while providing adequate usability for the user. As time moves on, users of that hardware may be forced to seek a distro that uses less resources and will work better for them.
Yes, hardware doesn't magically upgrade itself, but if you want to run the latest and greatest, state-of-the-art distro, chances are you're going to have to upgrade your hardware to keep up with the resource requirements of that distro. In that light, requiring 64-bits for, say, Ubuntu 10.04 does not seem like an unreasonable requirement.
Right now my HTPC/NAS is a crappy old 1.33ghz Via Eden with onboard video. Not 64bit, but it does support SSE3. I'll be bummed if they drop 32bit support.
64bit also vastly speeds up long and double math. It doesn't really apply to a browser, but if you were using 64bit integers to store currency amounts, you'd notice a huge speedup. Adding/subtracting from longs is one thing that SSE probably won't help.;)
Slashdot Mirror
I'd like to be tied to some railroad tracks and shmucked by a train, I think. Or maybe go skydiving, but have my chute fail. Quick, exciting, and only slightly painful. ;)
Only in the US. If you go to a country like Canada, they're both significantly lower. ;)
I'm not American, so forgive me if I get the state wrong, but wasn't it Pennsylvania where several judges jailed innocent kids to get $2m in kickbacks?
There have been several such cases in the past few years.
Death-row organs? $50k a pop? I'm in!
If it's profitable to kill, more will be killed. Simple like that.
Maybe. We don't know what kind of mindset they have over there, towards this sort of thing.
But if it was US judges, then oh yes, you can count on it.
Yep - either those who target netbooks or those which try to be resource friendly. If one can run on a much older system well then a newer system it should hum, plus not be such a big hit on the battery life.
Unfortunately, this isn't the case, since much of the drain is caused by components that aren't the CPU/RAM/etc.
More code and more overhead results in better power management. It's the same thing we have with DirectX/OpenGL - bigger/fatter drivers do a better job keeping the videocard(s) busy.
Meh. I'm Canadian. :P I think most people understood.
I know the feeling. My hometown of ~10k must've had 5 black people in it. It was so rare seeing one, that I couldn't help but stare. Looking back, I wonder if that made them uncomfortable. :P
But I also stared when I met a guy 4" 3'
Okay, we're on the same page - I just worded it poorly.
I guess I'd call PAE more of a "clever back" than "ugly hack". ;)
Re: multi-megabyte pages
The point of the hack was that it didn't waste memory. You can read more at the sourceforge page: http://www.7-max.com/
You sound like the kind of person that'd gain something from looking into it. :)
Re: Memory pages
If wasting memory is the primary concern, I think 16-64KB pages would be an improvement without introducing much/any wastage. The smallest Windows process is 16KB, after all. But there's probably a reason 4KB is the default across most operating systems.
Huh? Since when did Sun prove this?
Sometime last year.
http://www.downloadsquad.com/2008/12/22/virtualbox-2-1-supports-64-bit-guest-operating-systems-on-32-bit/
(first link on google)
I'm just mad that an OEM can get a copy of Windows for $36-60, and I can't for anything less than $100-200
UNR (Ubuntu Netbook Remix) is optimized for small screens, and has an interface similar to Android.
and that is with the standard 3.25Gb limit.
What standard limit? My XP Pro gaming computer has 3.7GB available...
Why is /. greenlighting yet another article just for the sake of MS-bashing? Other posters have already pointed out the remaining process memory limits and consumer driver issues that make this a non-starter technically. Not to mention that there's not much (any?) consumer-level 32-bit hardware that has BIOS/Northbridge support for more than 4GB of RAM. Since the 64-bit versions don't cost any more, who cares?
The reason(as he explains) is that Microsoft used to allow more than 4GB of RAM before XP SP2, then they added the limitation. The majority of hardware released in the past 10 years will support > 4GB addressing ranges.
I actually have an old nForce 2 board that claims to support 2GB sticks. (4x2GB = 8GB) It has no 64bit drivers, which means it had to be designed for PAE usage, before the 4GB cap was implemented.
But this guy is saying that Microsoft is doing something "illegal" because he was able to hack his system and enable PAE, even though MS charges extra for that in their server OSes. Is it "illegal" for Ford to sell me a car that's computer governed to 105 MPH even though the engine can get it up over 130? Can I complain to the FTC if I chip the car to remove the limit and then destroy the transmission?
It's not illegal, and I agree that they can do what they want, including putting an artificial limitation in their software to spur sales. Many other companies(Apple, IBM, etc.) have done it too. With old IBM mainframes, couldn't you just flip a switch to get a RAM upgrade? ;)
It's their right, and their choice, but there's no harm letting them bask in the fallout.
You're forgetting something.
When XP came out, the average computer had ~256MB of RAM in it. (maybe less) Most games consumed less than that.
Now that Win7 is coming out, the average computer has 2GB of RAM in it. (maybe 4GB) Most games consume close to 2GB.
How many people are going to install 32bit Win7, or receive a 32bit Win7 computer? It seems rather ludicrous setting the limit right at the edge of what is currently being used, don't you think? 256MB -> 4GB was plenty of upgrade room, but with a 4GB cap 32bit Win7 will be obsolete in 2010.
I'm still running 32bit XP on my gaming computer. 3.7/4.0GB available, though. :)
That's an artificial limitation, too. Sun proved you can run 64bit operating systems and virtual machines on 32bit CPUs. Obviously it's slower, though, since 64bit operating systems assume the existence of some speed-enhancing CPU features.
Then get a 64-bit OS and be happy.
I agree.
PAE is a hack.
I disagree. Paging is fundamental to memory management. You like having memory mapped to 4KB pages, but call it a hack when those 4KB pages are mapped to different physical memory locations? Why?
Apparently PAE is even more efficient than 64bit addressing. It's the faster option, although you're 100% correct about the per-process limit being roughly 3GB, which means it'll be obsolete shortly. Games, video editing, and virtual machines will be first to benefit from 64bit > 4GB memory usage.
What I've always wondered is why Windows uses 4KB pages. Larger pages are far more efficient. The guy who made 7-zip tried hacking in multi-megabyte pages, and had a 15% speedup. O_o
Errr no, Vista x64 runs with more then 4GB RAM without a problem. Windows 7 32 bit still has same limit. Windows 7 64 bit does not. There is a RAM extender called PAE that is used by all the OSes, Microsoft simply chooses not to install it in their desktop OSes because PAE has been known to break drivers and they want to push 64 bits to desktop. 4GB limit doesn't bother most desktop users and those who have more then 4GB run 64 anyways if they are power users. Yes, I consider 4GB+ users to be power users.
Interesting perception, but incorrect.
Most modern computers - ones shipping with 32bit windows - have PAE enabled because it is required for DEP. The driver argument is hollow, because of that. Clearly there are few enough incompatibilities that PAE/DEP is a good thing when it enhances security, and a bad thing when it deters someone from upgrading their OS. ($$$)
PAE is available in all versions of windows since Windows 2000. It doesn't increase the max available memory in every version - usually only datacenter/enterprise/server versions.
Also, I RTFA. The entire thing. I've also done some research on this in the past. Although I didn't figure out how to beat it for XP, I did bump my available memory up to 3.7GB without PAE. :)
I was just about to comment that the article seems more like single-task juggling.
Maybe I'm wrong, but my impression of multitasking was it was dealing with multiple sources of input at the same time.
So doing all three of those tests at once, and comparing who does the best. I know multitaskers often aren't good at focusing, and as this article demonstrates, aren't that great at quickly switching between single tasks - but how do they deal with multiple simultaneous tasks?
If you're in Britain, you can get a new n810 for about that much.
http://www.expansys.com/d.aspx?i=158518&partner=gadgetoid
I agree. You could build a half decent gaming computer for that much money!
The big issue is the stability of the OpenGL 3.x codepaths on Linux. You'll need some relatively up to date drivers(binary blobs) to get all those new calls working. And no guarantees it won't break, later.
Since OpenGL 3.x and DX10/11 share a lot, it should be more straightforward than it was in the past porting from one to the other. The major differences between OGL and DX have partially been eliminated. Thanks to The Khronos Group, they're both moving in the same direction.
Are you implying that you've never had to upgrade hardware before to run the latest and greatest software? I know the Ubuntu of today uses more resources than the Ubuntu of five years ago. Since Canonical only supports their desktop versions of Ubuntu for a maximum of three years, it's very likely that some computers that ran Ubuntu fine in 2004-2005 will be unable to run the latest version while providing adequate usability for the user. As time moves on, users of that hardware may be forced to seek a distro that uses less resources and will work better for them.
Yes, hardware doesn't magically upgrade itself, but if you want to run the latest and greatest, state-of-the-art distro, chances are you're going to have to upgrade your hardware to keep up with the resource requirements of that distro. In that light, requiring 64-bits for, say, Ubuntu 10.04 does not seem like an unreasonable requirement.
Right now my HTPC/NAS is a crappy old 1.33ghz Via Eden with onboard video. Not 64bit, but it does support SSE3. I'll be bummed if they drop 32bit support.
64bit also vastly speeds up long and double math. It doesn't really apply to a browser, but if you were using 64bit integers to store currency amounts, you'd notice a huge speedup. Adding/subtracting from longs is one thing that SSE probably won't help. ;)
Interesting blog post - but I don't really understand his complaint about the java 7 spec. Doesn't the ISO C spec cost money too?