• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

interesting article, shows how CPUs HAVEN'T grown that much

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

jonwessel

Member
Joined
Jun 21, 2003
Location
Cincy! Go BENGALS!!
"The Intel 80386 Processor - x86 goes 32-bit
The 80386 heralded the beginning of a new age for the IBM PC. The 386 was the first 32-bit x86 processor. As such, it was capable of breaking the 640 Kb memory barrier and running software written for graphical user interfaces. The 386 introduced a 32-bit architecture while maintaining full backward compatibility with earlier x86 processors. This was accomplished by using two operating modes: "real" mode, which mirrored the segmented memory of the older x86's, and "protected" mode which took full advantage of the 386's 32-bit enhancements.

The 386 began shipping in August 1986, but unfortunately it was several years before PC operating systems could make use of its 32-bit capabilities. IBM's OS/2 and Microsoft's Windows '95 were really the first 32-bit mainstream PC operating systems, and even on them most applications were still 16-bit!"
http://www.pattosoft.com.au/jason/Articles/HistoryOfComputers/1980s.html

the first 32-bit processor was released over 17 years ago, and i believe the first 64 bit processor wasn't released over 1 year ago!! my question is why?? it's now obvious to me that the technology is not growing at such a rapid rate as it used to, this could be because back then everything was so basic, or is it just how it looks? if i would have read this article in August of 86, i would have thought by April 04 that computers would be 256-bit, with 40-50GHz clock speeds, and about 100MB of L6 cache!!! I would have thought that one day u would be able to share files with others from around the world in seconds, i'm talking about Terabyte per second DL speeds!! i think if the technology growth rate would have kept steadly increasing then that would be reachable. this isn't a rant, juss something to get u thinkin and for fun that's all!
 
It IS increasing. At an exponential rate.

http://www.intel.com/research/silicon/mooreslaw.htm

MooresLawgraph2.gif
 
jonwessel said:
the first 32-bit processor was released over 17 years ago, and i believe the first 64 bit processor wasn't released over 1 year ago!! my question is why?? it's now obvious to me that the technology is not growing at such a rapid rate as it used to, this could be because back then everything was so basic, or is it just how it looks? if i would have read this article in August of 86, i would have thought by April 04 that computers would be 256-bit, with 40-50GHz clock speeds, and about 100MB of L6 cache!!! I would have thought that one day u would be able to share files with others from around the world in seconds, i'm talking about Terabyte per second DL speeds!! i think if the technology growth rate would have kept steadly increasing then that would be reachable. this isn't a rant, juss something to get u thinkin and for fun that's all!

If you had read the article carefully then you would have seen why processor-related inventions are not progressing as well in the Intel/x86 et. al. camp.

x86 is shackled with backward-compatiblity with an inferior instruction set that is complex and non uniform. Reduced instruction sets would lead to a greater gain in performance but ALL the software would have to be rewritten ground-up for a RISC architecture and Intel is scared to death of someone 1) coming out with better RISC and 2) losing their market share when they stop manufacturing x86 products. Unlike Apple, who went from the MC68000 (complex instructions) to the PowerPC (RISC) in the early 90s and rewrote all the software.

Technology can move on but because of Intel's reluctance to change an overpatched, overextended and overstressed architecture and instead use bandaid fixes (higher clocks, longer execution pipelines, and 2.2-pound heatsinks) to "fix" the problem it really hasn't.
 
Last edited:
jonwessel said:
and i believe the first 64 bit processor wasn't released over 1 year ago!
x86-64 was not the first 64-bit arch. People who actually needed 64-bit have been using it for years now, with MIPS, Alpha, Sparc, Itanium, Power4 and others. You probably don't need 64-bit even now, and the biggest advantage you'll get from x86-64 itself is the 8 additional GPRs added to register-starved x86 (bringing the total up to 16 GPRs) and the 8 additional SSE2 registers. Unless, that is, you have a great need for 64-bit ints.
 
jonwessel said:
"The Intel 80386 Processor - x86 goes 32-bit
The 80386 heralded the beginning of a new age for the IBM PC. The 386 was the first 32-bit x86 processor. As such, it was capable of breaking the 640 Kb memory barrier and running software written for graphical user interfaces.

This is incorrect. Protected mode was introduced in the 286 (the 386 implementation is almost identical), and the 286 was capable of accesing 16mb of RAM (the 386 pushed this to 4gb). Also, you can run any version of Windows prior to '95 with the exception of WFW3.11 (which dropped support for "standard" mode). I'm not sure whether OS/2 works on anything less than a 386, though, as it was a bit more advanced for its time.

And before anyone asks, yes I DO spend way too much time than what's healthy with old x86 computers :p
 
I thought I had a 286 with 2mb of ram (10 meg hd) but I guess it must have been a 386 in disguise
 
I was j/k about that, I knew it was a 286, you shouldnt write long posts about hings you dont know about (he thinks level 6 cache is better that level 1).
 
286 cpus didnt have the enhanced features the 386 did, I cant exactly remember either. I know windows 95 wont run on a 286 nor alot of newer programs and games that a 386 could run
 
after 86x-64bit I hope we do go RISC. Instead of making the cpu faster we need to make the code faster and more efficient
 
Re: Intel not moving forward

What do you think Itanium is for? A year or two ago, everyone panicked when Intel said they wanted to ditch 32 bit and go to 64 bit. They panicked because it would make everything out there inoperable, so Intel backed off and AMD's 64-bit EXTENTIONS have since taken hold. I am an AMD fan, but even I can see that in the long run, Intel's approach would be better.

But, of course, that's not the end of the story. Not even the giants Intel and MS working together can ditch x86. It requires all the apps-makers to get on board too. Everyone has so much tied up into x86 that the only way to get out of it is for some third party without the inhibitions to make a big splash. Think Apple without the snooty attitude and the ability to ACTUALLY prove they are better. People will slowly jump ship, and if the big names can be shifted to the new system without too much pain, they may take the long view and shift.

But then again, we ARE talking about corps, whose only purpose is to extort the maximum possible from the NOW and pretend there is no LATER.
 
Re: Re: interesting article, shows how CPUs HAVEN'T grown that much

Gnufsh said:
and the biggest advantage you'll get from x86-64 itself is the 8 additional GPRs added to register-starved x86 (bringing the total up to 16 GPRs) and the 8 additional SSE2 registers. Unless, that is, you have a great need for 64-bit ints.
:attn: :attn:
That's pretty much what I'm looking forward to in the A64. How in the world x86 went without an upgrade in the number of GP registers since the FIRST CHIP (yes, the 8086 has just as many GPRs as your shiny P4 :eek: ) is beyond me. Thank goodness for SSE.......


~~~~~~~~~~~~~~~
TheDS said:
What do you think Itanium is for? A year or two ago, everyone panicked when Intel said they wanted to ditch 32 bit and go to 64 bit.

But, of course, that's not the end of the story. Not even the giants Intel and MS working together can ditch x86.
Year or two ago? It was certianly longer than that when Intel announced it's plans for Itanium to replace x86, and have the whole world go 64-bit.

Being the giant is why it's not easy to change. You essentially hold EVERYBODY'S baggage, and if you want to suddenly change without backwards compatability you better somehow keep everybody's baggage afloat for a little bit....

JigPu
 
Tech Advance

Another good reason is a huge portion of the population, though it is hard for people like us to imagine, don't upgrade like they should. People keep computers forever, and wonder why they run like ****zle. There is a huge group of people out there who think "mine is good enough, I don't need anything else" and they don't upgrade. It is going to be hard to get those people to upgrade to 64bit.

These are the same people who only know when there is a new version of AOL out, but have no clue that there are 87 new virus deffinitions, 36 windows updates, and two newer versions of IE than they are running; and could care less. Personally, I think people like that don't even need a computer, but thats just me.

Thats all I got.
 
These are the same people who only know when there is a new version of AOL out, but have no clue that there are 87 new virus deffinitions, 36 windows updates, and two newer versions of IE than they are running; and could care less. Personally, I think people like that don't even need a computer, but thats just me.

I'm a tech support plebian; that's how I make my money at the moment :cool: And everyone should have a computer.
 
Itanium is IA64 architecture, and is what intel wanted to replace x86 with. As I've mentioned in previous threads, x86-64 is just a 64bit hack of a 32bit hack of a 16bit hack of a 25yr old 8bit architecture. x86 has to die sometime, but it's in AMDs best interest to drag it out as long as possible, because of the x86 cross licence agreement they have with intel. Do you think intel would ever give AMD a licence to make IA64 chips? I highly doubt it. What would AMD do then?

Right now I'd like to see LongHorn released for the PowerPC 900 series. Both NT 3 and NT 4 had PowerPC ports, but MS never released NT 5 (2000/XP) for PowerPC. Since the Xbox 2 is supposedly going to be PowerPC based maybe we will see a PowerPC version of LongHorn.
 
Back