whats so good about 64bit cpus?

[corvetteguy]

SHUT THE FUCK UP!!!!!!
PEOPLE BEING IDIOTS AND NOT KNOWING WHAT THEY ARE BUYING DOES NOT MAKE 64-BIT BAD!!!!!!
THINK ABOUT IT FIRST BEFORE SAYING WE'RE STUPID!!!!!!
BY MAKING TWO VERSIONS OF WINDOWS ALLOWS PEOPLE WITH OLDER COMPUTERS TO GET THE NEW OS TOO, INSTEAD OF FORCING UPGRADE!!!!!
OBVIOUSLY SOMEONE BUYING A NEW SYSTEM WITH A 64-BIT CPU WON'T BUY THE 32-BIT VERSION UNLESS THEY ARE INCREDIBLY STUPID AND CHEAP LIKE YOU!!!!!
YOU ARE A FUCKING TARD AND YOU SHOULD STFU!!!!!

 

[MarcusL]

Then why is Windows Vista home edition limited to 8gb, thats not unlimited. 32bit can also have a nearly unlimited flat virtual memory model.

Nesck,

You are mixing two unrelated topics... WindowsXP memory limit has nothing to do with 32bit/64bit computing performance and 32bit/64bit memory addressing.

Vista starter is for the guy in a developing country with a very basic PC. He can't afford and doesn't need a commercial-strength datacenter OS. Vista starter will be very cheap or even free. Why would MS force him to pay for a full capability OS when he will never use it? MS is doing him a favor by offering a cut down version at very low price.

At the same time, MS does not want the datacenter manager to buy 1 copy of Vista starter for $10 and then run his compute farm with 10,000 servers and expect enterprise-level technical support.

There are differnent OS versions between starter and datacenter to fit most users. You pay for what you get. You buy the OS that fits your needs and pay the appropriate price for it. MS is not evil for having multiple OS versions even if there is only one line of code different between them. It is smart marketing and responding to the needs of the customers with features and pricing that is appropriate for each.

Just because some code is on the CD that could enable high-end features doesn't mean you are entitled to it if you knowingly purchased a product with a lesser feature set for a lesser price.

64-bit computing will show its benefits when more mainstream programs are compiled for 64-bit. A lot of enterprise software is already fully 64 bit and would be crippled by a lesser architechure. The hardware has to come first for the software to be developed. In 3-5 years, 64bit will be the norm and 32bit will be forgotten just as 16bit is now. Currently there are not many advantages to buying a 64-bit CPU just for the 64bit features since you clearly don't run any 64bit software. In a few years, a switchover will occur where most new software requires 64bit, you can switch then.

In the meantime, you can console yourself in the fact that new 64bit CPUs do run your 32bit software faster than your old 32bit CPU thanks to other CPU improvements. It just doesn't make sense for Intel/AMD to keep developing the 32bit line when its days are obviously numbered.

 

[Heyyou27]

OK, let's put it simply.
You're saying moving from 32-bit to 64-bit does not inherently make your processor better. I see that you like your analogies, so I'll give you one.

Let's say your 32-bit is a 4-cylinder car and your 64-bit is an 8-cylinder car. If they have the same horsepower, then fine, no big difference. However, having the 8-cylinder increases your "capacity" for improvement.

What's the cost of having 8-cylinders? Less gas efficiency: you need more memory to power this 64-bit processor. However, as you may or may not see, some cars have the option of running on 4-cylinders.

Note: this analogy is not perfect (duh?)

That's a horrible analogy; the move from 64-bit to 32-bit has very few negative effects, if any. I challenge you to find one realistic negative of switching to 64-bit from 32-bit other than the initial upgrade.

 

[corvetteguy]

its a good analogy except for the negatives lol

 

[gman01]

Rofl, you call your OS starting 1 second faster is a big difference, does that justify the cost of buying a duel core?

Who cares about startup.... WIth Dual core, my operating system can run 2 programs/threads at the same time instead of 1.... That is why 2 cores are 80% faster than a single core of the same Ghz...

How can you possibly think 2 cores is not faster than 1 core?

 

[garnin]

Negatives? You put more resources in to get a higher output. It doesn't scale exactly all the way, but it's good enough to upgrade. In the car's case, it's gas...

In the 64-bit, it's memory. If you want to say add 1 and 1, it's still going to take up twice as much as in 32-bit. The binary instruction will have a lot more useless space in it, but it's still "waste." Luckily for memory though, we don't have to keep buying more to feed into the comp.

Although I do understand all the negative connotations associated with gas and its rising prices, I did not mean it in that particular way.

 

[nesck]

Thank god you don't have any power in the industy.
How are we going to progress technology if we don't embrace breakthroughs?

http://www.dailytech.com/article.aspx?newsid=3295
IBM Says It's Lucky to Get 10% to 20% Yields on Cell Processor
You call getting low yields a breakthrough?

64bit might not be useful to everyone now, however the day we do need it we can't just click our fingers and expect it to be there. It will take time to make it perfect. Making it available now means that people can adopt it as and when they like. Also it gives software developers a chance to make apps 64bit ready.

Really, have you seen the difference between 32bit and 64bit apps? And hardly anybody encodes / decodes movies, not even the gamer does.

There is so much software out there nowadays and there is no way you could just convert it all over night.

Ever think of the thought that some software won't be converted at all? Why would software be converted for no reason at all?

You know everything anyway, so you should be capable of getting a very high paid job, and the worries of those precious pennies being lost will be gone!

Yup, I even made progs ( usually based off other progs) to do... nefarious things... :twisted:

 

[corvetteguy]

i was joking about heyyou's comment because i thought it was a good analogy. lol

 

[nesck]

Rofl, you call your OS starting 1 second faster is a big difference, does that justify the cost of buying a duel core?

It's spelled "dual core".
No one is forcing you to buy anything; you're against innovation and the advancement of technology and have no place here. Seeing as a 64-bit processor can be had for $60, your argument makes no sense. http://www.newegg.com/Product/Product.asp?Item=N82E16819104249

What makes u think that a 32bit edition of the processor wouldn't be on sale for $55? PAE is innovation.

 

[corvetteguy]

64-bit is a better innovation tard.... :roll:

 

[threep]

nesck, why do you have to resort to name calling to get your point accross? Are we 13yr olds in Jr. High?

WRONG, its in hex FFFFFFFF - 1, not an additional instruction.
With PAE, more memory can be seen on the RAM, not pagefile on the harddrive!

Um, no, you are wrong, and you missed my point. 2^32 = 4294967296. In programming, you start with zero, (doesn't matter if you're using decimal, binary, octal, hex, or bobal). I already subtracted 1. 4294967295 equals FFFFFFFF in hex.

Same with PAE, its called double buffering u tard.

What does double buffering (http://en.wikipedia.org/wiki/Double_buffering) have to do with PAE??? (http://en.wikipedia.org/wiki/Physical_Address_Extension) According to this, PAE is still limited to 64GB. With a 64-bit processor, you are limited to 18-thousand-terrabytes, without any software tricks. That was my point.

Will you ever see 100 cpus on one processor? Think the yields cell is getting. 100 cpus would have to be to underclocked or they would make to much heat, what use is a 100mhz cpu good for a process?

I would love to see a processor that has 100 or more cores. Can you imagine what that would mean? It means you could theoretically have your video processor, your memory controller, your memory storage, your sound processor, your I/O and everything else on one chip. Imagine a computer more powerful than anything we have now that fits in less space than the "Q" key on your keyboard. There is no evidence today that a dual-core AMD CPU generate any MORE heat than a single core, in fact, that's one of the selling points!

I already installd windows longhorn on an athlon 3200+ that I have, I uninstalled it becaue it used 400mb of ram running useless junk, its basically windows xp with style xp.

Well, your experience is extremely limited. If you want a FREE 64-bit operating system, with low overhead try Linux, like RedHat or SUSE. They've been out for years.

And stop calling people names. It makes you sound like a child.

8) [/quote]

 

[lschiedel]

OK, let's put it simply.
You're saying moving from 32-bit to 64-bit does not inherently make your processor better. I see that you like your analogies, so I'll give you one.

Let's say your 32-bit is a 4-cylinder car and your 64-bit is an 8-cylinder car. If they have the same horsepower, then fine, no big difference. However, having the 8-cylinder increases your "capacity" for improvement.

What's the cost of having 8-cylinders? Less gas efficiency: you need more memory to power this 64-bit processor. However, as you may or may not see, some cars have the option of running on 4-cylinders.

Note: this analogy is not perfect (duh?)

That's a horrible analogy; the move from 64-bit to 32-bit has very few negative effects, if any. I challenge you to find one realistic negative of switching to 64-bit from 32-bit other than the initial upgrade.

Here is a a realistic negative of switching to 64-bit from 32-bit.
64-bit addresses take up twice the space of 32-bit addresses.
64-bit code is almost always larger that 32-bit code because of the number of addresses used in the assembly code.
Thus more cache is used up by the same "amount" of code (# of instructions) and more memory bandwidth is used up in transfering that code.

SO. Here is my take on 64-bit vs 32-bit CPUs.

Migrating from 16-bit to 32-bit was a big step.
So is migrating from 32-bit to 64-bit.

Going from 16-bit to 32-bit didn't improve things for existing applications (except a few that used slow hacks to pretend to be 32-bit), it instead ALLOWED NEW APPLICATIONS!

Thus in 32 bit..
- you could now multi-task much better.
- you could now support better graphics (larger resolutions, bit depths, etc).
- you could now run entirely new kinds of software that were unfeasable in 32-bit.

BUT. this only occurred when most/many applications were bumping up against the limits of 16-bit.

SO. going from 32-bit to 64-bit will allow an entire new class of software to be run.
- video editting of high quality (ie loading the 4+ GB movie in ram, or editting HD movies, 35+ GB)
- server operations involving massivly higher users/database sized (databases of 100 GB cached in ram)
- data manipulations unheard of on desktop 32-bit machines (weather prediction, true AI)

The point is, very few people needed 32-bit while most software was 16-bit. But after moving to 32-bit and getting a whole new class of applications, very few people would even think of going back.

The same will be true a few years after moving to 64-bit applications.
Imagine what virtual worlds our 64-bit games will be 5-10 years, and how impossible they would have been in a measly 2 GB of ram.
Imagine in 5-10 years when you are making 3-d movies in HD resolution, changing the actor in a scene and having it re-render the entire movie IN RAM! Then imagine going back to only 32-bit.

The point is, you only need to move from one capability to a larger capability when you have exhausted the potential of the previous generation, and the cost is now not prohibative. ie ram is cheap enough you can afford > 4 GB.

So for the person that said that why don't we just go to 512-bit NOW, the answer is because we can afford to max out the capabilities of 64-bit now, but we can afford to max out 32-bit, or will in a couple years. When everyone had 1 or 2 meg of ram, much of it barely accessable do to 16-bit cpus, then the time was ripe to move to 32-bit. When most people have 2 GB of ram, and many have 8 GB using hacks, then the time is ripe to move to 64-bit.
Also to use a 512-bit CPU, you would need a 512-bit DATA bus which would require more pins then our technology can support.

PS, very little of the Conroe is still 32-bit anyway.
The memory bus is now 64-bit even if the address bus isn't.
The SSE memory bus is now 128-bit.
The integer registers may be 32-bit, but the floating point registers have been much larger for years.

The only thing that isn't 64-bit in our 32-bit mode is integer registers, address bus and address sizes. Remember that the 8-bit CPUs used 16-bit addresses, and the 16-bit CPUs (8088, 80286) had 24 effective addresses. (though it used 16-bit segments).

We could have gone to use 16 bit segment numbers and 32-bit segments. But our transister budget now allows us to go to full 32-bit addresses without paying TOO much of a penalty.

It is possible though that we need to migrate to a 128-bit data bus to take true advantage of 64-bit CPUs. But that could have been done even with current 32-bit CPUs that use 128-bit data such as floating point and SSE.

 

[JonathanDeane]

OK, let's put it simply.
You're saying moving from 32-bit to 64-bit does not inherently make your processor better. I see that you like your analogies, so I'll give you one.

Let's say your 32-bit is a 4-cylinder car and your 64-bit is an 8-cylinder car. If they have the same horsepower, then fine, no big difference. However, having the 8-cylinder increases your "capacity" for improvement.

What's the cost of having 8-cylinders? Less gas efficiency: you need more memory to power this 64-bit processor. However, as you may or may not see, some cars have the option of running on 4-cylinders.

Note: this analogy is not perfect (duh?)

That's a horrible analogy; the move from 64-bit to 32-bit has very few negative effects, if any. I challenge you to find one realistic negative of switching to 64-bit from 32-bit other than the initial upgrade.

Here is a a realistic negative of switching to 64-bit from 32-bit.
64-bit addresses take up twice the space of 32-bit addresses.
64-bit code is almost always larger that 32-bit code because of the number of addresses used in the assembly code.
Thus more cache is used up by the same "amount" of code (# of instructions) and more memory bandwidth is used up in transfering that code.

SO. Here is my take on 64-bit vs 32-bit CPUs.

Migrating from 16-bit to 32-bit was a big step.
So is migrating from 32-bit to 64-bit.

Going from 16-bit to 32-bit didn't improve things for existing applications (except a few that used slow hacks to pretend to be 32-bit), it instead ALLOWED NEW APPLICATIONS!

Thus in 32 bit..
- you could now multi-task much better.
- you could now support better graphics (larger resolutions, bit depths, etc).
- you could now run entirely new kinds of software that were unfeasable in 32-bit.

BUT. this only occurred when most/many applications were bumping up against the limits of 16-bit.

SO. going from 32-bit to 64-bit will allow an entire new class of software to be run.
- video editting of high quality (ie loading the 4+ GB movie in ram, or editting HD movies, 35+ GB)
- server operations involving massivly higher users/database sized (databases of 100 GB cached in ram)
- data manipulations unheard of on desktop 32-bit machines (weather prediction, true AI)

The point is, very few people needed 32-bit while most software was 16-bit. But after moving to 32-bit and getting a whole new class of applications, very few people would even think of going back.

The same will be true a few years after moving to 64-bit applications.
Imagine what virtual worlds our 64-bit games will be 5-10 years, and how impossible they would have been in a measly 2 GB of ram.
Imagine in 5-10 years when you are making 3-d movies in HD resolution, changing the actor in a scene and having it re-render the entire movie IN RAM! Then imagine going back to only 32-bit.

The point is, you only need to move from one capability to a larger capability when you have exhausted the potential of the previous generation, and the cost is now not prohibative. ie ram is cheap enough you can afford > 4 GB.

So for the person that said that why don't we just go to 512-bit NOW, the answer is because we can afford to max out the capabilities of 64-bit now, but we can afford to max out 32-bit, or will in a couple years. When everyone had 1 or 2 meg of ram, much of it barely accessable do to 16-bit cpus, then the time was ripe to move to 32-bit. When most people have 2 GB of ram, and many have 8 GB using hacks, then the time is ripe to move to 64-bit.
Also to use a 512-bit CPU, you would need a 512-bit DATA bus which would require more pins then our technology can support.

PS, very little of the Conroe is still 32-bit anyway.
The memory bus is now 64-bit even if the address bus isn't.
The SSE memory bus is now 128-bit.
The integer registers may be 32-bit, but the floating point registers have been much larger for years.

The only thing that isn't 64-bit in our 32-bit mode is integer registers, address bus and address sizes. Remember that the 8-bit CPUs used 16-bit addresses, and the 16-bit CPUs (8088, 80286) had 24 effective addresses. (though it used 16-bit segments).

We could have gone to use 16 bit segment numbers and 32-bit segments. But our transister budget now allows us to go to full 32-bit addresses without paying TOO much of a penalty.

It is possible though that we need to migrate to a 128-bit data bus to take true advantage of 64-bit CPUs. But that could have been done even with current 32-bit CPUs that use 128-bit data such as floating point and SSE.

You bring up a good point with BR disks and HD DVD disks on the way working with 35-40GB's of data will pretty much force people to move to 64bit or they will suffer while a disk image is chopped up and processed one peice at a time. The closest thing I can remember is Win9X and its 2GB's limit for a file size... ugghh what a pain that was.

 

[choirbass]

...do you just enjoy arguing??... ...if you look at how many people have responded to you, you would see that they are trying to help you understand ... ...but youre shunning virtually every single persons answer away, and disputing it with them, nomatter what they say practically... ...and i dont know how many other people are reading this thread either... ...but it would seem you are rather alone in your opinion, as there hasnt been anyone thats seemed to really agree with what youre arguing for... ...just arguing for the sake of arguing...

 

[Fox_granit]

Your being focused on a "personal consumer" level of thinking. Where are most of the machines in this world? in business, mostly as servers....as far as the expansion in the MS OS world, the only reason servers are "allowed" that much memory is for ..... wait for it.......wait for it......clustering. yes there is the answer. CLUSTERING. I don't think you've even heard of the term. What does clustering do you ask? It allows multiple proccessors , "gasp", with their own memory, upto 8gigs ecc for each proccessor, "gasp x2", to run as a SINGLE MACHINE! Do you think google has just one machine giving you their website?

BTW most of your MAJOR companies use 64 software on a regular basis. Microsoft, AT&T, Sprint, anyone that has a major website, all use 64bit OS becuase it moves faster. Its like a highway man, the more lanes, the more traffic you can hold.

 

[djgandy]

Thank god you don't have any power in the industy.
How are we going to progress technology if we don't embrace breakthroughs?

http://www.dailytech.com/article.aspx?newsid=3295
IBM Says It's Lucky to Get 10% to 20% Yields on Cell Processor
You call getting low yields a breakthrough?

So you're a quick quitter too.
Since when can you buy one of those for your pc? You've gone from athlon xps to windows to ps3 processors! Might aswell give up on quantum computing too, they haven't made it effective yet so they obviously never will by your thinking.

64bit might not be useful to everyone now, however the day we do need it we can't just click our fingers and expect it to be there. It will take time to make it perfect. Making it available now means that people can adopt it as and when they like. Also it gives software developers a chance to make apps 64bit ready.

Really, have you seen the difference between 32bit and 64bit apps? And hardly anybody encodes / decodes movies, not even the gamer does.



Gamers are the minority of PC owners. I love how all sites always talk about gamers. Gamers are nothing to the market.
You obviously misunderstood my point about integration time too.

There is so much software out there nowadays and there is no way you could just convert it all over night.

Ever think of the thought that some software won't be converted at all? Why would software be converted for no reason at all?

You know everything anyway, so you should be capable of getting a very high paid job, and the worries of those precious pennies being lost will be gone!

Yup, I even made progs ( usually based off other progs) to do... nefarious things... :twisted:

Well if you were a real programmer and not a script kiddy you'll know that code is written to be re-used. Each time "new" software is released, code from older versions will be re-used. Software doesn't write itself yet, and if you are developing for a platform that platform will need to exist so you can test.

Try reading the Intel architechture books. Yes to write an o/s you'll need to read all that. How can you read it if it doesn't exist?
For it to exist the CPU firms need to make the processor. They aren't going to just make 5 so microsoft can write an o/s for 64bit.
If they did that, trust me you'd be paying a lot more for your 32bit processor.

 

[nesck]

Here is a a realistic negative of switching to 64-bit from 32-bit.
64-bit addresses take up twice the space of 32-bit addresses.
64-bit code is almost always larger that 32-bit code because of the number of addresses used in the assembly code.
Thus more cache is used up by the same "amount" of code (# of instructions) and more memory bandwidth is used up in transfering that code.

SO. Here is my take on 64-bit vs 32-bit CPUs.

Migrating from 16-bit to 32-bit was a big step.
So is migrating from 32-bit to 64-bit.

WRONG, The migration of 16bit to 32bit is different that the 32bit and 64bit difference your seeing now. What you are trying to do is compare DoS progs with c++ progs, its not like that between 32bit and 64bit. The difference between 32bit and 64bit is c++ and c++, almost exactly the same. You think that the instructions for the cpu to do are going to change a lot from 32bit to 64bit?

Hint: Compare windows x64 with windows xp.

Going from 16-bit to 32-bit didn't improve things for existing applications (except a few that used slow hacks to pretend to be 32-bit), it instead ALLOWED NEW APPLICATIONS!

Thus in 32 bit..
- you could now multi-task much better.
- you could now support better graphics (larger resolutions, bit depths, etc).
- you could now run entirely new kinds of software that were unfeasable in 32-bit.

BUT. this only occurred when most/many applications were bumping up against the limits of 16-bit.

WRONG, most applications reach this limit cuz they want more graphics and gui, its completely possible using 16bit color.

SO. going from 32-bit to 64-bit will allow an entire new class of software to be run.
- video editting of high quality (ie loading the 4+ GB movie in ram, or editting HD movies, 35+ GB)
- server operations involving massivly higher users/database sized (databases of 100 GB cached in ram)
- data manipulations unheard of on desktop 32-bit machines (weather prediction, true AI)

-I don't do video editting,
-server operations that need that many users / database already have 64bit,
-Your dreaming data manupulations are completely false and rely on programming, for example on Half-life 2 "True AI" etc, all bs.
Weather prediction is already done by ur local meteorologist.

The point is, very few people needed 32-bit while most software was 16-bit. But after moving to 32-bit and getting a whole new class of applications, very few people would even think of going back.

Very few people know what 16bit even is. We are talking about the general public here. Not every person knows what the company Nvidia / ATI even is.

The same will be true a few years after moving to 64-bit applications.
Imagine what virtual worlds our 64-bit games will be 5-10 years, and how impossible they would have been in a measly 2 GB of ram.

All what you said is biased and false.
The first sentence of that statement assumes that in the next few years *people* will be moving to 64-bit applications, your unable to predict the future? You think virtual worlds will come up in 5-10 years? I remember a yahoo tech article saying we would have Shriek (movie style) graphics by 2006 in games, another false prediction. Impossible with 2GB of ram? Your still giving me that windows XP limit if PAE did not exist, the reality is that DEP requires PAE, yet the 4gb limit is still on windows xp with PAE enabled.

Imagine in 5-10 years when you are making 3-d movies in HD resolution, changing the actor in a scene and having it re-render the entire movie IN RAM! Then imagine going back to only 32-bit.

Keep dreaming, Imagine in 5-10 years when your unable to fill gasoline in your car due to peak oil. Many things can happen, and its uncertian whats going to happen in the future. I DO NOT RENDER MOVIES, LET THE MOTION PICTURE INDUSTRY FIGURE IT OUT! Did you know that not EVERY SINGLE PERSON MAKES MOVIES!

The point is, you only need to move from one capability to a larger capability when you have exhausted the potential of the previous generation, and the cost is now not prohibative. ie ram is cheap enough you can afford > 4 GB.

Once again, PAE allows up to 128gb of ram,
http://www.microsoft.com/whdc/system/platform/server/PAE/pae_os.mspx

When DEP is enabled, PAE must also be enabled, why didn't any of you talk about that as a big deal?

http://support.microsoft.com/kb/875352

So for the person that said that why don't we just go to 512-bit NOW, the answer is because we can afford to max out the capabilities of 64-bit now, but we can afford to max out 32-bit, or will in a couple years. When everyone had 1 or 2 meg of ram, much of it barely accessable do to 16-bit cpus, then the time was ripe to move to 32-bit. When most people have 2 GB of ram, and many have 8 GB using hacks, then the time is ripe to move to 64-bit.
Also to use a 512-bit CPU, you would need a 512-bit DATA bus which would require more pins then our technology can support.

"When most people have 2 GB of ram, and many have 8 GB using hack"
PAE is no hack, however they might need hacks to get passed the limitation PURPOSELY put in by microsoft.

PS, very little of the Conroe is still 32-bit anyway.
The memory bus is now 64-bit even if the address bus isn't.
The SSE memory bus is now 128-bit.
The integer registers may be 32-bit, but the floating point registers have been much larger for years.

Conroe has good performance in 32bit, its good performance is not from its support of 64bit.

The only thing that isn't 64-bit in our 32-bit mode is integer registers, address bus and address sizes. Remember that the 8-bit CPUs used 16-bit addresses, and the 16-bit CPUs (8088, 80286) had 24 effective addresses. (though it used 16-bit segments).

Still comparing 8-bit and 16-bit?

We could have gone to use 16 bit segment numbers and 32-bit segments. But our transister budget now allows us to go to full 32-bit addresses without paying TOO much of a penalty.

Still comparing 16-bit and 32-bit?
the difference in applications from 16-bit to 32-bit is extreme, but the difference between 32bit and 64bit is hard to see.

It is possible though that we need to migrate to a 128-bit data bus to take true advantage of 64-bit CPUs. But that could have been done even with current 32-bit CPUs that use 128-bit data such as floating point and SSE.

The main reason for the 64bit push is not enough ram, and whats the point of a 128-bit data bus when its not needed in the first place.

 

[nesck]

You bring up a good point with BR disks and HD DVD disks on the way working with 35-40GB's of data will pretty much force people to move to 64bit or they will suffer while a disk image is chopped up and processed one peice at a time. The closest thing I can remember is Win9X and its 2GB's limit for a file size... ugghh what a pain that was.

Wow your funny, Why is my windows xp able to read my dvd which is over a 3gb file? And you really think people are going to be transering 35-40GB HD DVD movies via progs such as bit torrent? 35-40GB of data is 10% of your hard drive! Usually 20% in most cases!

 

[Action_Man]

nesck, STFU and get over it. People have already gone over it and you wont accept the truth. Go argue somewhere else.

 

[corvetteguy]

your such an idiot... why did you creat this thread? to bate people into an endless arguemnet.... this arguement isn't even a funny one like the endless ones we use to have with maypepnecro... who was an idiot like you...but not as defiant.... just STFU and end this thread..... its as gay as you.

 

[nesck]

Um, no, you are wrong, and you missed my point. 2^32 = 4294967296. In programming, you start with zero, (doesn't matter if you're using decimal, binary, octal, hex, or bobal). I already subtracted 1. 4294967295 equals FFFFFFFF in hex.

Who cares what 2^32 equals exactly, the issue is with 32bit and 64bit.

What does double buffering (http://en.wikipedia.org/wiki/Double_buffering) have to do with PAE???

(http://en.wikipedia.org/wiki/Physical_Address_Extension) According to this, PAE is still limited to 64GB. With a 64-bit processor, you are limited to 18-thousand-terrabytes, without any software tricks. That was my point.
[/quote]

The wikipedia article is incorrect, here is proof

http://www.microsoft.com/whdc/system/platform/server/PAE/pae_os.mspx
32BIT Windows Server 2003 SP1, Datacenter Edition = 128 GB RAM

PAE is no software trick, its been suported since the pentium pro, anyone running windows xp having a cpu less than a pentium pro doesn't need more than 4gb of ram anyways.

http://www.osronline.com/ddkx/appendix/pae_3oyn.htm about double buffering for PAE.

I would love to see a processor that has 100 or more cores. Can you imagine what that would mean?

So, if a cpu with 100 cores supporting 64bit was out, what if it only supported 32bit, your point is?

It means you could theoretically have your video processor, your memory controller, your memory storage, your sound processor, your I/O and everything else on one chip.

I don't care much, unlike you that care if your game gets 0.25 more fps.

Imagine a computer more powerful than anything we have now that fits in less space than the "Q" key on your keyboard.

Imagine the yields of less than 1% to make a working cpu and the cost.

There is no evidence today that a dual-core AMD CPU generate any MORE heat than a single core, in fact, that's one of the selling points!

Thats because their single core counterparts are clocked higher.

Well, your experience is extremely limited. If you want a FREE 64-bit operating system, with low overhead try Linux, like RedHat or SUSE. They've been out for years.

And the difference between the 32bit and 64bit linux's is what again?

And stop calling people names. It makes you sound like a child.

8)

[/quote]

 

[nesck]

nesck, STFU and get over it. People have already gone over it and you wont accept the truth. Go argue somewhere else.

No they didn't go over it, you won't accept the truth because ur to dumb.

 

[djgandy]

The wikipedia article is incorrect, here is proof

http://www.microsoft.com/whdc/system/platform/server/PAE/pae_os.mspx
32BIT Windows Server 2003 SP1, Datacenter Edition = 128 GB RAM

PAE is no software trick, its been suported since the pentium pro, anyone running windows xp having a cpu less than a pentium pro doesn't need more than 4gb of ram anyways.

http://www.osronline.com/ddkx/appendix/pae_3oyn.htm about double buffering for PAE.

128GB is only possible on cpu's with 64bit extensions.

 

[choirbass]

this thread is pretty dead already... its just rehashing the same info over and over again from different perspectives... ...he should stop pointlessly debating like he is... ...and go argue with a book or something, plenty of monologue that can go on there... and if hes not happy with the books conclusion, nomatter how informative it might be... he can go find a different book to carry a monologue along with... ...again, this thread isnt going anywhere anymore, and, it seems its been that way since it started.

 

[Heyyou27]

We get it, Microsoft is an evil monopoly and you never want to upgrade your PC; now please stop ranting on how technology of the past is just as good as technology of the present.