corvetteguy
Distinguished
- Jan 15, 2006
- 1,545
- 0
- 19,780
NESK... i think what you need besides meds is to go and get a job with WORLD VISION.... since your so concerned about how much ethiopians pay for windows. MAYBE if you are soooo concerned... you could go out buy a shit load of cd's and burn thousands of copies and give them to them for the price of a cd.... if thats what willl make you feel better and stop the flow of shit outa your mouth... then please be my guest and do it... otherwise STFU!!!!
You sir, are an idiot. You've argued for 10 pages about absolutely nothing because you cannot be satisfied by anyones answer but your own. If you dont understand it, then spend a little time and do some research and self-education. It's also apparent you lack even a fundamental understanding of basic microprocessor architecture as given by your repeated comments on memory. 32 bit CPU's are limited to supporting 4GB natively because they only have 32 bit registers. A software solution to this limit may be possible, but a hardware solution is a much more performance effective means of adressing memory.
deweycd
Distinguished
- Sep 13, 2005
- 846
- 0
- 19,010
If you go back and read articals that have been posted, comments made, you would realize that more then 4 GBs in a 32 bit operating system is a work around for a hardware problem. You can use software, but that requires memory and cpu time that would otherwise be more productive doing something else. You have also begun an argument about 64 vs 32 at the point where the 64 is just starting to come to fruition. If you ask yourself the same question in 2 years, do you think that your going to be able to have the same answers as you do now.
Also, think about the companies making the chips? Why would they produce a 64 bit computer that is the same as a 32 bit? Yes, marketing does sell, but come a couple months after being out on the market, a sham would have been discovered, but it hasn't because it doesn't exist. I say that 64 bit is here for a reason, and is going to stay. If you want to stay in the past, go ahead, but stop arguing about something you have a lacking knowledage of, and that had been proven to be a step forward.
Also, think about the companies making the chips? Why would they produce a 64 bit computer that is the same as a 32 bit? Yes, marketing does sell, but come a couple months after being out on the market, a sham would have been discovered, but it hasn't because it doesn't exist. I say that 64 bit is here for a reason, and is going to stay. If you want to stay in the past, go ahead, but stop arguing about something you have a lacking knowledage of, and that had been proven to be a step forward.
Spenzer4Hire
Distinguished
- Apr 11, 2006
- 7
- 0
- 18,510
I can't believe I read all 10 pages of that drivel.
Nesck has been asking two questions all along. Whenever an adequate response is given to his first question, he flips to his fallback.
Nesck, let me see if I can answer both of your questions in one go.
Question 1: (Topic) whats so good about 64bit cpus?
or
What advantages does 64bit have over 32bit?
Firstly, many people have stated the advantages of 64bit memory addresses. PAE, a hack (no matter how ingenious it might be) to overcome hardware limitations, allows a 32bit OS to address more than 4 GB of RAM. Correct me if I'm wrong, but even with PAE enabled in a 32bit OS, each individual 32bit application is still limited to 4 GB of address space. Although the OS can "see" more than 4 GB, no single program can use more than 4 GB.
Second, I run a home recording studio on a budget, but the choice to go 64 bit was easy. I'm not sure if you understand the principles of digital audio, so I'll lay them out for you. CD quality is 16bit, 44.1kHz. This means the voltage (on a microphone input, for instance) is measured 44,100 times per second with 16 bits of resolution. DVD-Audio is 24bit, and up to 192 kHz. Now, I record my raw tracks at 24bit, 96kHz (the highest quality my interface supports). Say you are mixing together 2 very loud tracks (with individual samples reaching the 24bit limit). The sum of these two tracks could easily exceed 24bits. This is called digital clipping, and it sounds horrible. Any decent 24bit audio editor, therefore, handles the 24bit audio as 32bits, leaving plenty of overhead. For the same reason, many of my audio plugins (and I run alot of 'em) operate in 64bits to avoid truncating (clipping) the 32bit data being passed to it.
Let's pretend I am running an audio plugin that increases each individual sample's volume by 30. In a 64bit environment the CPU makes a single call for the entire 64 bit block. Then it adds 30 to it. It then stores the 64 bit sample, and moves on to the next. To add 30 to a 64bit number in a 32 bit environment, the CPU must call the first 32 bits, then it must call the second 32 bits. The CPU must then add 30 to the lower half of the sample. If there is a carryover number, that must then be added to the upper half of the sample. Then each 32bit number is stored.
Multiply all those extra CPU cycles for one sample by 96,000 (my sample rate) and you can see why a 64 bit CPU (and operating environment, and plugins) has the advantage for my application.
Question 2: "I don't edit video" or in this case audio. Why should I switch?
or
Most gamers don't have more than 1GB of memory (if that). Why should I switch?
First, just because you don't work with multimedia doesn't invalidate a 64bit system's increased efficiency in manipulating it. Second, my examples are limited, but any application dealing with numbers larger than 32bits will benefit from being designed for a 64bit environment.
I'm not a gamer, and you're right, if you polled all gamers in the world, the average amount of RAM is probably below 1GB. I don't game at all, but I've heard of this game Oblivion, apparently it's all the rage, and I hear it can eat up 1.5 gigs of ram without a problem. Although 1.5 gigs is still below the 32bit 4GB (individual application) address space limitation, I can see it being close enough to warrant a 64bit processor. I'm sure with the majority of gamers on 64bit CPUs and the newest Windows OS 64bit Vista, game developers will waste no time putting that extra address space to use.
And if you're none of the above? If you just want to build a system to surf the internet and use office applications, go with the cheapest components that will do the job. You can't really beat the price of a 754 Sempron64 (on clearance EVERYWHERE) and you sure can't beat the price of a free 64bit Linux OS with free Open Office and Firefox.
That's "whats so good about 64bit cpus"
It removes one more level of restriction for hardware manufacturers, software designers, and data sizes. And with current market prices, even when building a budget system, it would be silly to limit yourself to a 32bit processor.
Nesck has been asking two questions all along. Whenever an adequate response is given to his first question, he flips to his fallback.
Nesck, let me see if I can answer both of your questions in one go.
Question 1: (Topic) whats so good about 64bit cpus?
or
What advantages does 64bit have over 32bit?
Firstly, many people have stated the advantages of 64bit memory addresses. PAE, a hack (no matter how ingenious it might be) to overcome hardware limitations, allows a 32bit OS to address more than 4 GB of RAM. Correct me if I'm wrong, but even with PAE enabled in a 32bit OS, each individual 32bit application is still limited to 4 GB of address space. Although the OS can "see" more than 4 GB, no single program can use more than 4 GB.
Second, I run a home recording studio on a budget, but the choice to go 64 bit was easy. I'm not sure if you understand the principles of digital audio, so I'll lay them out for you. CD quality is 16bit, 44.1kHz. This means the voltage (on a microphone input, for instance) is measured 44,100 times per second with 16 bits of resolution. DVD-Audio is 24bit, and up to 192 kHz. Now, I record my raw tracks at 24bit, 96kHz (the highest quality my interface supports). Say you are mixing together 2 very loud tracks (with individual samples reaching the 24bit limit). The sum of these two tracks could easily exceed 24bits. This is called digital clipping, and it sounds horrible. Any decent 24bit audio editor, therefore, handles the 24bit audio as 32bits, leaving plenty of overhead. For the same reason, many of my audio plugins (and I run alot of 'em) operate in 64bits to avoid truncating (clipping) the 32bit data being passed to it.
Let's pretend I am running an audio plugin that increases each individual sample's volume by 30. In a 64bit environment the CPU makes a single call for the entire 64 bit block. Then it adds 30 to it. It then stores the 64 bit sample, and moves on to the next. To add 30 to a 64bit number in a 32 bit environment, the CPU must call the first 32 bits, then it must call the second 32 bits. The CPU must then add 30 to the lower half of the sample. If there is a carryover number, that must then be added to the upper half of the sample. Then each 32bit number is stored.
Multiply all those extra CPU cycles for one sample by 96,000 (my sample rate) and you can see why a 64 bit CPU (and operating environment, and plugins) has the advantage for my application.
Question 2: "I don't edit video" or in this case audio. Why should I switch?
or
Most gamers don't have more than 1GB of memory (if that). Why should I switch?
First, just because you don't work with multimedia doesn't invalidate a 64bit system's increased efficiency in manipulating it. Second, my examples are limited, but any application dealing with numbers larger than 32bits will benefit from being designed for a 64bit environment.
I'm not a gamer, and you're right, if you polled all gamers in the world, the average amount of RAM is probably below 1GB. I don't game at all, but I've heard of this game Oblivion, apparently it's all the rage, and I hear it can eat up 1.5 gigs of ram without a problem. Although 1.5 gigs is still below the 32bit 4GB (individual application) address space limitation, I can see it being close enough to warrant a 64bit processor. I'm sure with the majority of gamers on 64bit CPUs and the newest Windows OS 64bit Vista, game developers will waste no time putting that extra address space to use.
And if you're none of the above? If you just want to build a system to surf the internet and use office applications, go with the cheapest components that will do the job. You can't really beat the price of a 754 Sempron64 (on clearance EVERYWHERE) and you sure can't beat the price of a free 64bit Linux OS with free Open Office and Firefox.
That's "whats so good about 64bit cpus"
It removes one more level of restriction for hardware manufacturers, software designers, and data sizes. And with current market prices, even when building a budget system, it would be silly to limit yourself to a 32bit processor.
Atolsammeek
Distinguished
- Dec 31, 2007
- 1,112
- 0
- 19,280
Well seid Spenzer4Hire and I agree with you. Just remember Nesck is just plain Dumb. I dont think he knows right from wrong. It not a insult it a fact there nesck. So live with it.
drcroubie
Distinguished
- Feb 2, 2006
- 162
- 0
- 18,690
well, i haven't read any of the other posts, but the last sentence of this paragraph just made me laugh.
here's a hint: you don't actually need new hardware, you don't actually need new software. windows 98 coupled with wordperfect 7 and netscapde 7 running on a pentium2-350 served my essay-writing needs fine throughout my whole IB diploma and almost all of my university life.
It's only games, video editing etc, and for the most part, marketing hype, that makes joe consumer think he needs new software/hardware every two years.
as for 64 bit applications, having explicit x86-64 instructions does make certain things faster (ie, one 64-bit add is twice as fast as two 32-bit adds). but the programs have to be written for 64bit, and have to need it (some current 64bit programs aren't any slower on 32bit only hardware).
Personally, i don't think paying extra for a 64bit cpu is worth it. but seeing as it's permeating everywhere, within a year you won't be able to buy 32-bit only cpus, so just buy what you would anyway, ignore the 64bit-ness
(actually, the same goes for NX/XD bit, VT, etc. they're good for people who want them, but not worth the extra pricetag for joe consumer, he probably won't use them anyway. Same with SSE1/2/3/4. no use if you don't play games or watch/encode videos or whatever, but seeing as you can't buy a cheaper version without them, just put up with having extra useless features...)
choirbass
Distinguished
- Dec 14, 2005
- 1,586
- 0
- 19,780
were mainly just boycotting responses to the OP now, as the thread just degraded entirely... ...in all honesty, with all the different answers and variations on the same answers that he was given, to try to answer his question as best as anyone could, throughout all 10 pages... his stance on 64bit, did not change, at all... ...he still sees it as a pointless endeavor for the most part, as he did when he first asked the question... and he argued with every single person that tried to help him, refuting with his own personal biases and fallacies... ...it pretty much became a lost cause after a while... and, i do agree with JumpingJack, this thread does need to be locked, asap.
Wider buses and more processed per clock basically. The reason that you can't see any benefits is that not much software has been written for 64 bit. You may be able to get 64 bit OS's, but most of the software is still on 32 bit.
This is again looking to the future. When transistor based computers where first introduced they couldn't conceivably understand why you would need more than 256Kb of RAM, ever! With RAM chips ever getting bigger in capacity, 2Gb on a stick now, and newer programmes using more and more RAM, the ability to address more RAM is required. The limit of 64 bit addressing is 2^64 addresses, equivalent to 17,179,869,184 gigabytes. As for the page file, it will continue to evolve in future operating systems to facilitate the use of more RAM more efficiently.
But that is not direct processor support, you have to "fool" it into seeing more RAM, which takes up processing clock cycles, reducing performance.
I think your missing the point. 64 bit isn't really about now; it's about the future and the needs then. This has always been the way. How many systems had 4Gb of RAM in 1980 when 32bit was introduced? It's all about updating a 26 year old system to one which will be relevant in the future. Expect your 64 bit versions of you current programs to have more features and run faster.
Just a pause for thought...
Access violations aside 😉
[code:1:15420823ec]
or eax, 0xFFFFFFFF;
mov ecx, [eax];
inc eax;
mov ecx, [eax];
[/code:1:15420823ec]
Welcome to 32bit integer overflow. The mmx and sse registers are larger, but rely on the GP registers so you are limited to 4gig.
PAE Is therefore a kernel hack. It rely's on privileged instructions not available to applications.
I can't be bothered to explain it any further, you cannot move an integer larger than the 4gig limit into a 32bit register. End of.
Access violations aside 😉
[code:1:15420823ec]
or eax, 0xFFFFFFFF;
mov ecx, [eax];
inc eax;
mov ecx, [eax];
[/code:1:15420823ec]
Welcome to 32bit integer overflow. The mmx and sse registers are larger, but rely on the GP registers so you are limited to 4gig.
PAE Is therefore a kernel hack. It rely's on privileged instructions not available to applications.
I can't be bothered to explain it any further, you cannot move an integer larger than the 4gig limit into a 32bit register. End of.
It took me ½ an hour to answer your question Nesck, and I think i did a decent job .... But it seems you are more interested in the discussion rather than the answer ... If you are so convinced you have the answer you shouldn't ask such questions ... I like replying to post if someone is looking for an explanation, but it seems like I just wasted my time ...
Spenzer4Hire:
For your purposes I don't think you will ever need to go the 64 bit route. You are not summing the multiple tracks, you are multiplexing them - each chunk being smaller than 32 bits. Effects, signal processing and encoding your tracks will definitely benefit from it. However, for the recording process you get no benefit what so ever ... I can't see why raw audio would ever need better precision than what can be delivered by 32 bits ...
Spenzer4Hire:
For your purposes I don't think you will ever need to go the 64 bit route. You are not summing the multiple tracks, you are multiplexing them - each chunk being smaller than 32 bits. Effects, signal processing and encoding your tracks will definitely benefit from it. However, for the recording process you get no benefit what so ever ... I can't see why raw audio would ever need better precision than what can be delivered by 32 bits ...
Ok, who cares about logic anymore... Obviously not Nesck
Based on everything said in these 10 pages, I can only conclude on these things:
a) He's from thailand and just got ripped off for $36 on the Windows XP Starter edition.
b) He's a PAE salesman who has just been put out of a job because 64-bit is coming around.
c) He actually has 512 mb of ram and is freakin' pissed that he can only use 256 for his non-gaming, non-multimedia editing, programming in C++ without knowing he has to compile activities.
Clearly, this topic was started to get everyone on the PAE bandwagon... (as if there was one) and failed. It can just die now...
Based on everything said in these 10 pages, I can only conclude on these things:
a) He's from thailand and just got ripped off for $36 on the Windows XP Starter edition.
b) He's a PAE salesman who has just been put out of a job because 64-bit is coming around.
c) He actually has 512 mb of ram and is freakin' pissed that he can only use 256 for his non-gaming, non-multimedia editing, programming in C++ without knowing he has to compile activities.
Clearly, this topic was started to get everyone on the PAE bandwagon... (as if there was one) and failed. It can just die now...
TRENDING THREADS
-
-
-
-
News Introducing the Tom’s Hardware Premium Beta: Exclusive content for members
- Started by Admin
- Replies: 43
-
-
Question Is there anyway to force gsync through integrated graphics?
- Started by nelska
- Replies: 1
-
Question Is there any way to connect this multi-adapter to a UHD disc drive?
- Started by Jay_34
- Replies: 4
Facebook
Twitter
Instagram