Slashdot Mirror
1-GHz Pentium III Due This Month
ReviewSeek writes "According to this News.Com article, consumers will be able to buy Pentium III computers running at 1 GHz from Hewlett-Packard and IBM later this month. Volume production and sales aren't expected until third quarter though. " It's strange to me that for some reason that "One Ghz" thing seems important. But ya gotta love fast.
Oh give me a break! This was yesterday's news and I wouldn't be surprised if many ppl submitted it yesterday as well. Also, hitting the 1 GHz marker has no real significance, as a G4 500 MHz will probably still outperform an Intel 1GHz processor.
Were I to make a purchase to speed up my home computing, a good AGP 3D graphics card would probably be a better purchase.
But cool in theory, I suppose.
CPUs may be getting faster, but I can still only type 35 words a minute (on a good day).
but Intel kinda needs to get those 800/850 chips out before worrying about 1Ghz
-Elendale (is fed up)
IANAT (I Am Not A Troll)
Here's an article at The Register that explains it nicely.
And there's a stable 1.1GHz Athlon that was shown at CeBIT. Tom's Hardware has more.
IMHO, the question of when the consumer can buy this chip and what quantities they'll be available in will determine who really was the first to release a 1GHz chip.
Have fun: Join D.N.A. (National Dyslexics Association)
I'm sorry but I can't think WTF I would need 1 GHz for. I'm currently managing with a P133 which, although a little too slow for comfort, still just about manages to cope with my demands. And even if I did want a 1 GHz processor to show off to my mates I'd wait for the 1 GHz Athlon to come out rather than buy a chip based on an outmoded and inefficient design.
And with Intel's recent track record in supplying their processors I think you'll be lucky to get one of the dozen systems that'll be available in the next year :)
If I was intel, I wouldn't rush to put out chips labeled 1Ghz. here's why:
a 333 mhz proccessor sounds a good deal faster than a 200, and clock chip speeds have been increasing at a couple hundred mhz per whatever time span it is (wow, i'm out of the loop)
but a 1.2 to 1.3 gigahertz chip dosen't have that appeal to induhviduals as "massively faster".
Somebody get our flag back!
...is like the year 2000. They don't really mean anything (besides faster chips and slight Y2K computer problems) but it's a nice round number with lots of zeros for people to latch onto. Everyone needs a buzzword to feel cool.
Light a fire for a man and he'll be warm for a day. Light a man on fire and he'll be warm for the rest of his life.
It's strange to me that for some reason that "One Ghz" thing seems important.
What about the year 2000? Does it seem strange to you that it should be so important? Apart from the Y2k bug, which didn't happen, it's just another year...
Actually, come to think of it, if the year 2001 is the Millenium, then would the 1GHz processer be clocked at 1001MHz ???
T.
The 1Ghz figure is important to the computing community for the same reason that the year 2000 was important to society as a whole.
We humans have an irrational interest in what we consider to be 'round numbers' whereby we feel that a year with a 'round number' or, in this case, a processor clocked at one will in some way be extra-better as it were over previous incarnations than if it did not have that round number in it's name.
There is also the psychological effect that canging the name of the unit has. Once we are able to rate processors in GHz rather than MHz, people will subconciously expect them to run significantly faster (the difference between 900 & 1000 MHz is not that big, but the difference between 900MHz and 1GHz sounds like a lot more), so the manufacturer who can hit that 'magic number' first will have a bit of a head-start in shifting increased numbers of units.
And of course, the people who buy those processors expecting increased performance gains, will then brag about them, even if they're not noticable, because otherwise they may look foolish.
What very few people will ask is "do I need this". Personally, I have a P166MMX at home, and it does everything I need. I can run Linux, Star Office and Nyetscrape without difficulty. 1GHz would be nice, but frankly, I don't need it.
--
Listening for the sound of the coming rain...
but you'll only be able to get one if Andy Grove is your granpa. Ever seen a PIII 800MHz box? Intel has to get past the 800 landmark first instead of making those bogus announcements every other day.
The thing is that people gagging to get faster and faster clocks, freezing their CPUs to overclock them etc. are probably running distinctly suboptimal systems anyway, because there are usually a whole load of things you can do to a machine to get the best performance out of it beyond bumping up clock speeds.
A while back when the processor and memory clocks were more closely related, you used to run more slowly with a higher processor clock under certain conditions! A lot of P120 systems ran slower than P100 systems for most applications because they ran the system bus slightly slower. Does anyone know if this is still an issue - is there still any kind of relationship between headline processor speed and the bus speed? If not, when did that get decoupled?
Ian Griffiths
So, what, are they going to have a lottery to decide who gets the few precious chips they can turn out?
Oh! I got it. I bet they'll have competitions as to who gets the chips.
First round elimination: Stupid Bunny Suit dancing competition. Points for garish colors and imaginative dance moves.
Second round elimination: Chili cookoffs, using old P60's as the burner. Is that smoke from the food or the chip? Who cares! It burns, baby!
The PIII Superbowl: The person who actually gets the chip is the person who can come up with the least-stupid-sounding reason why the general public needs to drop 1k on a 1Gz CPU so they can check out porn sites and forward the same retarded joke (headers and all!) to a bazillion people.
They're going to be really expensive and what's the point? So you get an extra 10 FPS in Q3Arena, honestly, I can't tell a difference between my 450 and 550 computers in Q3. One gets about 10 FPS more but it's not visible, they both look about the same. Apps will load a little faster but is that worth the price? I don't see programs coming out that have system requirements for 1 GHz for quite a while so I think I'll stick with my 550 for a while.
kwsNI
You all might want to have a look at The Register and Tom's Hardware to get better information about this story. Both sites have been publishing some interesting stories (specifically The Register) concerning this chip race that we all love watching. They seem to do a good job of getting the dirt on both AMD and Intel, and although they do seem to play favorites with AMD (as I'm sure most of you do as well), they have posted relatively unbiased articles about both companies. When I say unbiased, you have to look at all of their storys, as both sites have taken to Intel bashing as of late, and for good reason.
Intel is still having production problems with their 800Mhz PIII, whether they will confirm that is another matter. The Register contacted a German distributor and he said that he isn't getting the 800Mhz chips, let alone something running at 1Ghz. Also, a story on The Register a couple of days ago said that Intel had told manufacturers to expect 866Mhz's by the end of February. That obviously hasn't happened.
I find it hard to believe that Chipzilla is going to be able to jump 200Mhz in one month. Maybe they'll take a prototype, tape an ice-cube to it and ship it to Dell so they can say they did. You may see these chips in volume by June, but I sure wouldn't believe this information coming out of Chipzilla now...
How about some 1GHz Athlons instead. They'll be available in full volume production too... unlike the P3. Cheaper too.
I've always been one to buy Intel, but I'm thinking my next box will probably have an AMD chip. I expect AMD will have matching or faster chips within a month or two.
Unbreakable toys can be used to break other toys.
The only issue I see here is to make games perform better. My AMD233 is just fine for applications, coding, internet. The misconception I think the average user here is that they think a faster CPU will give them faster internet. Long live Broadband!!!
Really though its quite sad how anti-Intel everyone is, I mean the reason why Intel became Intel was because of its reliability and performance, this is not to say that today Intel doesn't sing a different tune, but still everytime I read about something that Intel is doing to push the enevelope in CPU technology people just bash it away as if its already been done or there is something better coming around the corner, which is always true, but regardless if the industry didn't have Intel who knows what could of happened, some would argue that another manufacturer would of just taken control of the market and, rightly so, but the history of Intel is very impressive and their ability to stay as one of the top manufacturers of chips is rather astounding, and yes I would like to recognize the fact that they have used some Microsoftesque tatics to stay in that lead, but the credit that has been given to them is rather pathetic people will critique Intel for anything that they do, the point I am trying to make is that I am sick of the same post of people just flaming Intel, if your going to flame them I suppose make it original ok ;), not ohh well AMD is blah blah, or ohh big deal I did that yesterday crap that I keep on reading repetively by people. Think Original, Think Spam, as in the actual Spam.
"Real knowledge is to know the extent of one's ignorance" -Confucius
Most people don't know what MHz or GHz means anyways--to them it might as well be a model number.
--
You're still using Windows?
Quando Omni Flunkus Moritati
-Possum Lodge Motto
Simple.
Intel wants to be able to have the bragging right over AMD... a little bit of marketing savvy.
Press releases to the media can proclaim they relased the faster chip.
Mailing to the stockholders can announce breaking the 1 gig mark.
Personal I still won't buy one, but I can understand why they would move to put the chip out prior to being able to mass produce it.
Malk-a-mite
A little AC who finally choose a nick.
Here are a couple of reasons why I could use all the processing power I could get:
If I am working on a project, my code-compile-debug (whatever it's called) cycle is approx 5-10 min plus the time it takes to compile. If in that time I change a .h file, I must recompile most of the project, which takes a few minutes. If my cycle time goes from 10 min to 5 because of a processor that compiles faster, I just doubled my productivity!
As another example, A year ago I was doing some research involving Mandelbrot/Julia sets. Rendering those in large quantities at high quality can take forever even on a fast machine.
I won't even talk about tons of applications in the scientific world--they should be obvious. So if you post that you don't need no faster processor, all that means is you're not a coder.
-- The Sheep --
1Ghz on a 100Mhz memory bus 66Mhz pci bus OVERKILL!!!!!!!
That's right. A G4 at 500 MHz is still cooking that Gigaflop, baby!
Normal FUD. They can only deliver a couple of them but at least it's "official" and Intel has reached the 1 Ghz barrier. Well done.
-Danny
Talk about science fiction! I'm in my mid-30's and remember having conversations with friends where we marvelled at having 64K RAM and a 4MHz processor.
Now Joe or Jane Consumer can easily go out and buy a supercomputer for their family. Yet, in the long run are they doing anything more with their home computer than they were doing 10 years ago (other than surfing the web)? It just seems that all that computing power hasn't really changed what most people do with their PC's -- which is pretty much use it as a glorified electric typewriter, surfing the web and e-mail.
Despite what Intel and Microsoft might say, a 1GHz Pentium III is not necessary, nor does it even enhance the experience of web browsing. It certainly isn't needed for the dreaded paperclip living inside Word.
Maybe I'm just a curmudgeon, but I just don't see where all this computing power is of any practical benefit to the average user.
Simply, because it's a benchmark! I don't understand why people don't understand that. It's the same thing with 2000, or with turning 30. It's a round number, and our human nature is to like round numbers.
Personally, I do think, in all irrationality, that 950 MHz doesn't have the same ring as 1 GHz.
As a side-note, to all the people who say, 'Now, what the hell am I gonna do with 1 GHz??' Gimme a break. I heard people say that when the 100 MHz Pentium came out, and when the 1 GB drives came out. but I bet they didn't count on bloatware and games becoming more and more demanding on a system.
Now, you're gonna tell me most games today only require a killer video card. Sure. And what do you think powers the AI's?
Wow...in a little bit my processer may be running a 4.77 again - at least Ghz not Mhz, however when they get there, the current windows release may make it feel like Mhz.
And I remember replacing my 8088 with a NEC V20 because it was just a tiny bit faster (4.79 or so).
This may outperform the crusoe chips released so far but the point is that the clock speed is pretty immaterial when judging processor performance, much less system performance. We need to see a benchmark figure for processor performance. I expect I could design a processor that does almost nothing but has a clock speed of 2Ghz, we need to establish how much work this thing can accomplish in a given time.
The reason is that Intel is scared shitless of AMD. Intel knows that AMD can best them in the MHz (GHz?) race at any time, and they know that AMD has 1GHz chips. They also know that the Athlon beats the pants off the old 686 core of the Pentium III (which is really only a Pentium Pro that went uptown).
The fact of the matter is that Intel is scrambling to keep its mindshare, so it makes big news about things that will happen six months from now. People that trade stocks and make PCs now have it in their heads that Intel actually has a 1GHz system, and that they were the first ones to break the GHz barrier. Those people forget about AMD and the K7. That's the really Intel's strategy: keep announcing things that aren't here yet so the spotlight never strays too far, even though the PIII is inferior. Make people forget about that "other" chip company.
But don't take my word for it, no. Go try and buy an 800 or even 750 MHz PIII system. Then go shop around for an 850 MHz Athlon system. AMD announces things when they happen, like a company should. Intel is the hardware equivalent of Microsoft and I hope their subterfuge and bully tactics (look through Tom's Hardware for articles about Intel and K7 motherboard manufacturers for a little info about friendly old Intel) come back to bite them in the ass.
-B
Ash and Hickory, straight-grained and true, make excellent bludgeons, dandy for the cudgeling of vegetarians.
What good is a faster chip if it is waiting for: bus, ram, and disk?
AMD will have 200mhz bus with their 1ghz, and Intel will have 133mhz, which one will be faster? the AMD one, Intel needs to get their act together.
(RISC processors get MANY more MIPS than CISC ones, but as they only do a fraction as much, that isn't a measure of anything -real-.)
Unfortunately, there is no genuinely useful measure of performance for a processor, and all the benchtests that exist are catastrophically flawed.
(Typically, a benchtest program will fit entirely in the memory cache, and will probably mostly reside in the processor cache. I don't care if you're using X11R6.3 or Windows 2000, there is no way that a REAL application, or even a REAL window manager would be crammable into that kind of space. If your processor, on a benchtest, has no wait states, but is burning 90% of its time in idle cycles for real applications, then the benchtest is useless.)
Personally, I think the P3 is an over-bloated lump of silicon. I feel that it's time that it got divided into a network of high-speed RISC chips that -pretended- to be a single CISC chip. That way, you'd get the speed of RISC, with the power of CISC.
You also don't need QUITE so many duplicate instructions. Last time I counted, I found over 100 ways to program a jump instruction. That is STUPID! Most RISC chips don't even have 100 instructions in total! It makes no sense to scan through that many instructions, when you could start by determining it's a jump, and figuring things out from there.
(This would involve an instruction tree, whereby related instructions are in related parts of the tree. By following such an approach, by the time you're far enough in to need to do a linear search, you already know what you're doing and what is involved.)
However, this is getting off-track. To get back to the main point, if you are going to use/need a single, simple benchmark, the MIPS rating is far, far superior to the clock speed, because at least it measures how much the chip is doing in that time. A 1 GHz chip could only be doing 1 instruction per second - what use is that to anyone?
It's a small world and it smells funny; I'd buy another if it wasn't for the money; Take back what I paid (SoM)
I personally have never used an AMD chip. I also thought, that as long as we are going to be using operating systems on 808x architecture.. I would rather use the original (INTEL), rather then a emulation chip (AMD).. Thats just how I thought. And thru the years, seeing that there was this or that incompadibility problem with AMD chips for certain applications (ie. There were a handful of games in the past years that needed a patch in order for them to work on AMD chips).. It just reinforced my feelings against having an AMD chip.
:)
Now much time has gone by.. I've honestly not heard a bad thing about an AMD chip in a long time. Perhaps they have achived a perfect duplication of the instruction sets that are in PIII's?
My concern is.. I really want a 1 GHz chip. I actually have some stuff that really could use the speed as my P2-400 is starting to feel sluggish. I'm much more willing to entertain the idea of having an AMD Athlon chip verses a PIII. Especially if AMD releases their 1 GHz in high volume production before this month is over (which looks likely).
SO MY QUESTION IS... What do I gain or loose by choosing one over the other? I've seen chip-enhancements talked about.. Like MMX, 3DMax, etc etc... Does AMD support MMX? Or is that strictly an Intel thing?
Basicly.. what I'm getting at is.. You can't have a sound card without Creative Labs Sound Blaster compadibility.. unless you want alot of games that don't support your sound card (at least thats how it was before DirectX). Are we still in that type of time period with using the most popular chip verses a chip (AMD) that only has 16% of the market?
If someone could list the Pro's & Cons of choosing AMD verses Intel.. or point me to a webpage that has such information.. I would be eternally grateful. Especially if it frees me from my bondage that Intel has me in.
Thanks,
-Matthew
At least Intel and AMD.
Follow my thought: why is 1 GigaHertz important ?
Because we are only human and we are impressed by round numbers, that why there are often prices at 0.9, 9.9 (in any money), because usually people thinks that 1000 is much more than 989 for example.
What does this imply ?
When the price of 1GHz CPU come down, they will sell a lot! That's good for CPU makers..
BUT it has also other implications: people won't be much impressed by 1.3 GHz CPU over 1GHz CPU!!!
So once the gigahertz frontier is passed, people will care less about the speed of their CPU, and will buy more "low grade" CPU... (well if you admit that within 3 years a 1 GHz CPU will be considered as a low grade CPU...).
What do you think ?
PS: once upon a time, radio makers where comparing their radio by the number of transistors included in the radio.
I think that we will reach quickly times where you don't buy a 1.6 GHz CPU over a 1.2 GHz CPU but a computer with Firewire, or a computer which looks good, or which doesn't make too much noise...
The Athlon can outperform any PIII at similar clock speeds, in a few weeks we should start seeing DDR enabled Athlon motherbaords, basically doubleing the memory bandwidth to twice what's available to the PIII. The Athlon is faster than a PIII and it's not even getting all the momory it can handle. With full speed cache (now enabled) and the DDR sytems, the Athlon will make its lead even bigger over the PIII. It will still be cheaper. And with the release of the DDR chipsets will come support for SMP too. (for all those servers) There's no good reason to go with a Cu P3.
Intel's days as the number one x86 CPU maker are numbered.
This is how long things would take (approx) if you or I operated at 1 GHz...
Eating Breakfast (cereal, 40 bites, 10 chews per) : 0.0000004 sec
Dialing the phone: Local: 0.000000007 sec
long distance: 0.00000001 sec
Germany: 0.000000013 sec
Flipping through a magazine: 0.000000136 sec
Flipping through Machinery's Handbook: 0.000002555 sec
Dealing a game of poker (4 players): 0.00000002 sec
Writing 100 pages of text: 0.000090016 sec
Writing this post 0.000000484 sec
-ShelbyCobra
Living life in the right side of the s-plane
my main question is what the bus speed is going to be... i've been fairly disappointed that even the pIII 500MHz models (i've only seen one dual 500, but the same rules applied) only had a 100 MHz bus...
:)
for the people who rip on AMD, even though intel has a larger portion of the market it doesn't make them "better" or "superior", and there is probably not much someone can say to make certain people think otherwise... it's like linux versus windows, some people just like the fact that even though it sucks (windows), it's where the money's at.
-barton
- Weight, in itself, isn't an indicator -
That was true of the old calorie ratings, too.
(olives have MANY more calories than apples, but as that doesn't take into account digestibility, that isn't a measure of anything -real-.)
Unfortunately, there is no genuinely useful measure of value of a fruit, and all the benchtests that exist are catastrophically flawed.
Personally, I think the giant apple is an over-bloated lump of fruit. I feel that it's time that it got divided into a pile of high-calorie olives that -pretended- to be a single apple. That way, you'd get the calories of olives, with the convenience of one apple.
However, this is getting off-track. To get back to the main point, if you are going to use/need a single, simple benchmark, the calories rating is far, far superior to the weight, because at least it measures how much energy the fruit contains. A 1000 pound fruit could only have 1 calorie per pound - what use is that to anyone?
Sometimes "it's just more" is all the argument you need, when you're comparing apples to apples.
So is Linux, so Slashdotters are going to be, for the most part, one-sided on this issue. I myself have always bought Intel chips, but my next box may have an Athlon because one: they're cheaper, and two: they're pretty damned impressive when comparing them to comparable Intel chip performance.
_______
I just wish I could c:\format Internet
OK, I am not saying Intel doesn't actually have a 1 GHz PIII. They probably have, but vary, very few pieces. The yield to get such silicone is low at House Intel, which means they will have one or none ! GHz chips per wafer. Probably less, judging from the "ease" you can buy a 800 MHz PIII.
And maybe Intel can even afford to sell such rare silicon for about US$ 1000, efecctively losing on the deal. They can afford it because, well, they still have money to throw away. Being the first to cross the 1 GHz barrier is probably worth it.
BUT do they really think people are going to drink the story this time? Ask Gateway, that got seriously burnt with Intel. I think NOT! It's to damn obvious Intel doesn't have enough 1 giga, or even enough 800 MHz silicone. Itäs so obvious that the marketeering with the "first to giga" won't work. It just won't, no matter what Intel does, they can't cause a collective brain damage to their potential customers. Oh, wait, Microsoft has been doing this for years, and people are still buying the crappier than crapy releases of Win9x...
Sigged!
How's the heat situation on this 1 GHZ? Do we need a friggin air condition unit for it? :)
Ant(Dude) @ Quality Foraged Links (AQFL.net) & The Ant Farm (antfarm.ma.cx / antfarm.home.dhs.org).
My mommy told me to never buy 1.0 of a product; they're always buggy. I'm waiting for the point releases!
James
oh man, you are a stupid flamer. far inferior to the average troll.
I'm hemos., aka Jeff. Bates.. I help run this site, along with Rob. Malda.. I handle books, and generally posting storie
Athlon is kicking Intel's ass, and there's really not much they can do about it for now. Investors and the technical press all seem pretty well clued in about Intel's current problems, and judging by this forum the tech-savvy (i.e. likely to buy a premium 1GHz system) end users also realize what's going on.... so who are Intel hoping to fool by announcing a 1GHz PIII that everyone knows they can't deliver in anything other than perhaps sample quantity? Apparently AMD *already* can produce the 1GHz Athlon, and are just waiting for the strategically best time to announce (and deliver!) it.
I'm looking forward to Intel's next processor announcement: "The 600MHz toaster over heating element"... they've got problems ahead.
The fact is: The average user is using a chip which at one time, someone said no one would ever need. Saying "no one will ever need" to an improved technology is only revealing your ignorance of this industry and technology in general. At one time, no one needed a car, television, microwave, VHS, phone, DVD, refrigerator, plane flights, or electricity, or I can go on and on. Just because you can't imagine a use for it, doesn't mean someone else has. "Argue for your limitations and sure enough they're yours." - Richard Bach
Really, why the excitement, a much 'slower' CPU could (theoretically) execute programs much faster by beeing better designed. We nead to get away from this oversimplified CPU centric mesures of performance. We all know that the CPU isn't the bottleneck on modern systems, so why is it the every improvment in CPU clock frequence is met with drooling addoration, when improvments in motherboard design, IO thoughput, memory chip speed, and so on are practially ignored.
Thad
Sheesh, I probably will get a 1Ghz in a year. :) What about you guys?
Ant(Dude) @ Quality Foraged Links (AQFL.net) & The Ant Farm (antfarm.ma.cx / antfarm.home.dhs.org).
Unless of course you want to talk scalability. While at home I use Linux, at work we have 64 proc E10K's and 128 proc SGI's. While I love linux as a workstation/departmental server, for the scientific community Solaris/Irix are the best choice right now. Unfortunately, at this time linux does not scale as well as Sol/Irix on a single system. I am hoping that linux scalability will reach the levels where I could replace Solaris7 on my E10K with linux but until then Solaris is still the only choice.
what is this sig I keep hearing about?
If you have been using PCs for more than about 5 or 6 years then you probably have used AMD chips. About half of all 286 and 386 systems produced contained AMD chips. That is how AMD started gaining momentum. Back then, they were really the same chip (as far as blueprint is concerned) and AMD was a manufacturing partner with Intel. Of course, they ended up besting Intel by producing Intel's designs at higher clock speeds then Intel was willing to release for themselves (use of Intel's chip design was part of the bargain). Ever hear of a 386-DX 40mhz? This was the last time AMD came out with a chip that was faster than what Intel offered at the time. Up through the 486, the chips were the same but with a different label. Intel had enough and didn't renew the contract with AMD as of the P5 (which was named such to do away with legalities concerning x86 nomenclature). Since then AMD has been using the knowledge and experience of their former life to try to cath up to Intel and has failed...until recently. The Athlon is a better technology, a cleaner technology, faster at execution of basic x86 code per given clock speed. This will be especially true when the SMP version of Athlon ships. It should put such a gap between it and Intel PIII that it is a joke. But I can't convince you of this unless you know a little about system hardware and it's related bottlenecks. You will have to do the digging yourself and when you understand what goes on with processors and memory busses and the like you will agree that Athlon is a better chip. Until then just be patient and give the benefit of the doubt to us who believe in the Athlon. We are not zealots for any given name brand but rather we are selfish consumers who simply want the most for our money.
Intel will soon (mid 2000) release their "Italium" processor. This will be a rather big reworking of the chipset. As far as I can see they have built the chip from scratch. It still has the ability to run programs for IA32, but I guess they have to in order to sell them.
For more information see the Intel Itanium site.
Phobos - Greek word for fear or flight
Post the same story and switch the vendor to Intel and you get a dozen threads like "Who needs 1GHz?". Lame.
The Chip makers must be resorting to cutting corners somewhere if they are able to move up the release date 2 quarters. Has anyone heard anything about this? How are they capable of accelerating the production time like this, without compromising some feature or security or introducing bugs.
Please...it might be too bad, but in my humble opinion, it is more or less fact: Macs are simply not as stable as other PCs...period. I don't care how "fast" they are. I want a machine that won't crash on me.
Not that we'd need them or probably even buy them, but when the market is saturated with 1Ghz chips, and they sell like beanie babies to Wintel lemmings, the prices of the current 800-700Mhz will drop. Even chips like the 450Mhz (more than adequate in most needs) will drop substantially.
So even if a 1Ghz seems to be a waste to you, let it live, so the rest of us can get cheaper equipment that runs just as fast
Ya Ya Ya, we hear this all the time. Linux can't compare to solaris on high end hardware. Solaris intel sucks, but linux/intel can't beat solaris on sparc hardware in the high end.
Linux/intel works good on 4 processors/max
Solaris/sparc works well on > 64 processors.
Don't make me laugh. Linux is good, but it won't replace solaris,AIX,or SGI at the high end for quite a while.
PHP is the solution of choice for relaying mysql errors to web users.
well...come on, would you rather run a 486 or a 550+ mhz? faster is better, in my opinion.
perhaps fps here refers to "frags per second?"
heheheheeee......
Being a network engineer for a large company I get to deal with high-end servers and the one drawback I see in today's servers is poor IO. I would much rather see a motherboard with a fiber GHz bus than a GHz CPU. Unless you are doing encryption the CPU does not really come into play with say a web server or a firewall. I know that the GHz bus technology exists today just look at most RAS equipment and you will find that most have fiber GHz bus. We keep getting faster and faster memory, hard disks, and network cards and the bus speed stays the same. I can't imagine why anyone would put a GHz Ethernet card in a system with a 100Mhz buss. Am I the only who can see the flaw in this logic?
Encryption: I may not agree with what you say, but I will defend your right to encrypt it...
Games are a current example, but not the only one, of the usefullness of simulation. The fidelity, and thereby effectiveness (convincing emersion in games, accurate results in science, etc) increase with finer grained simulations. The 'bound' on the value of reducing grain size is at least as small as the molecular level. This is effectively unbounded (Games would have to simulate a character by doing a particle sim including every molecule of the character to be at this grain size...might as well be the real thing).
My point is that there will always be use for more computational power (not just traditional CPU cycles, but CPU cycles are included).
However, for the office desktop, with the current metaphors people are addicted to, more CPU is questionable. When this silly 'window/desktop' metaphor finally breaks the CPU will come to bear again in the office.
ps. This isnt the place to go into it, just search the net yourself, but there are many alternatives to the window/desktop metaphor already fleshed out pretty well in discussion/research/prototypes.
- Game
- Program
- Run an MS OS
Or have some other obvious reason to use 1 billion cycles per second. I believe the need that will come along will be called the "3D OS", and it's coming soon ladies and germs. The combination of consumer 2D and 3D cards into one piece of hardware in the last couple of years was the handwriting on the wall, and these processors are the wall itself.
But I don't need 3D, I'm happy with my command line! the curmudgeons cry. Well fine, and you'll probably still be using your P133's and your command lines 10 years from now and you'll probably still be smug. But there are real benefits that could be realized from a 3D OS, or to go a bit further, an immersive OS. The human mind is built to navigate a 3D, full-sensory world, so it's only natural that making the computer more like that in every way would enhance usability, shrink the learning curve and increase productivity.
You probably didn't read it here first, but you read it here last.
It's rare that you're presented with a knob whose only two positions are Make History and Flee Your Glorious Destiny.
let me in on a "cooler" number than zer0!! hehehee...and don't even think of bringing 23 into this - 'cuz that's a whole different thread...
embrace the nothing.
sure, people have this weird fascination with whole numbers and things...why do you think that is? i have an idea about it: things change over time. change is constant. change is (in my humble opinion) very good. one way to keep track of change is to note the numerical reflections of this ongoing cosmic change (eg. years (y2k) or processor speeds (1 ghz)) - what's the big deal? sure, every day is just another day, and every chip set is just another advance - yes, advance - in tech. but, ultimately, these are weird milestones in the ever-unfolding mystery of the Universe, and i don't see what the problem is with acknowledging that. in a few years someone will come out with a 3+ghz processor, or a quantum prcoessor or something like that...and it'll be the same old story. let's just have fun and make the most of what we, as crazy humans, create... let's make the world a better place, eh? why not?
The race to 1GHz is at the expense of things that really matter to people who use computers. I stopped being able to tell the difference in processor speed after about 200MHz or so. After that point, resources would have been better spent on getting a processor that uses very little power and costs $10. In some ways we're not advancing at all. I could boot up a circa 1984 PC into WordStar--from diskette--faster than I can boot a Linux machine and get into Emacs.
Please, can people get a clue and realise that MHz isn't the /only/ thing the worry about?!
/. community also has this idea.
/real/ CPU's instead of these overheating pieces of crap using 10 y/o technology?
.. then maybe we might get some decent technology out of them.
"MHz" is used by both AMD and intel as propaganda and FUD as they try and capture consumers $$$'s. Most consumers nowadays know what "MHz" is, and most are now (unfortunately) under the misconception that "more MHz == better". Sadly, it seems that the
Sure, MHz vs. MHz on a particular processor, the higher the better. But it's not worth singing and dancing about. It's no great giant leap in CPU design - in fact, intel take their L2 cache latency UP to accommodate for higher clock rates. And when you go out and buy your brand spanking new PIII Xeon - have you thought of the fact that its really just a glorified Pentium Pro with a few odds and ends tacked on?
Get real. Intel and AMD aren't giving us anything great by giving us more MHz. (Look at how much intel advertised 66MHz to 100MHz bus speeds - they only made ~= 15% performance increase). It's more cost effective for them to keep upping the clock rates and adding bigger fans - there's not much R&D involved in that - and intel and AMD know that MHz sells.
Why don't these companies invest in some proper R&D and make some
Look at MIPS. An R12000 at 300MHz beats a PIII 700MHz in FP (specbench), and G4's are noticibly faster "MHz" for "MHz". Then again, say the word "altivec" and consumer won't have a clue what your talking about, even though this technology allows a G4 to totally thrash your intel counterpart at half the MHz.
If we all stop worshipping intel and AMD for pushing the MHz barrier
I'm not an "old curmudgeon" and I still prefer the CLI over any GUI for file management. With a few keystrokes and some wildcards, I can do in a couple of seconds what would require 10x as much time and energy with a GUI. 3D would only be worse.
Booting Win2K: 100000 sec
10^1000000000000000000000000000000
Now That's a cool number.
Although I strongly agree that expanding SMP would be a nice thing, Megahertz ratings don't add. They just don't. Not because of SMP overhead diminishing effective usage (mentioned in previous reply), but because two devices acting independently at X MHz do not suddenly act at 2X MHz when placed near each other. This is not unlike claiming that the left half of your processor is running at 500 MHz, while the right half of your processor is running at 500 MHz, together they run at a Gigahertz.
This is a quaint little instance where a scientific field is too strongly influenced by a marketing department. I'm pretty sure if I could overclock, say, my hamster to 1.1 GHz on his wheel, I would get an ASP over $700 for the year. If I put a second hamster on the same wheel, marketing would claim 2.2 GHz. That's why they are marketing.
In a wave you have frequency and amplitude. Multiprocessoring is more analogous to an increase in amplitude than frequency.
-Rothfuss
What kind of cooling will one of these bad-boys require? A refrigerator case?
INTEL CORP PURCHASES KEEBLER TO IMPROVE SUB-MICRON CAPABILITIES
INTEL DEVELOPER FORUM, PALM SPRINGS, Calif., Mar. 2, 2000 - In a move to improve their supply of high end processors to OEM and retail markets, Intel (INTC) has today purchased the infamous Keebler Tree, including all Keebler Elves, for $400 million and an undisclosed number of shares of Intel stock.
"They understand volume," explained Intel Chairman Andy Grove. "They have been a mainstream supplier for many years, with a proven track record. The synergy of our R&D with their sub-micron expertise will allow us to significantly improve yields in our PIII line as well as speed up the ramp for the upcoming Willamette. In addition to these obvious tier one benefits, we believe the Keebler move offers further advantage with regard to technological innovation. Before the takeover, we had a team of Intel geologists do a field study of the soil mineral content near the Big Tree. Copper everywhere. We are excited about how this positions us for the possibility of employing copper interconnects in a future product line we are inexplicably calling AluminumMine. The Elves will be invaluable to us in this effort."
Ernest J. Keebler, the CEO of Keebler responded to the acquisition,
"We were not happy about the takeover, but this is a business and we will proceed according to the best interests of American Elves everywhere. Intel is a savvy corporation and they don't like to be kicked in the teeth by the competition. (Referring to recent market gains by competitor AMD) They saw a need and they saw a solution. The takeover has been in the works for a couple of months now. We are already sampling 0.13 micron,1.1 GHz PIIIs with a deeper pipeline, 512kb L2 cache, copper interconnects and a tasty fudge coating. There are heat issues, but when we refine the process, we will be able to take the voltage down and possibly add a crisp wafer."
Immediately after announcing the acquisition, Intel lawyers fired off a volley of lawsuits against Nabisco for patent infringement, and filed a series of grievances against a leading Taiwanese cookie exporter in an attempt to ban the import of their cookies into the United States. One Intel lawyer explained the move - "We have a cgi script that takes care of it all. We just have to fill in the company name, and maybe an address or two, and all hell breaks loose. We really don't even know how it happens. Computers are shiny."
Stuff it up your huge flaming ass, goatse-boy!
Really Pascal, give it a rest, mkay?
Do you think that the 1GHz would br hotter then Dreamcast?
What I find interesting about all these arguments about faster and faster CPU's is that you CAN right now make your system go quite a bit faster without having to resort to getting a new CPU.
There are three things you can do:
1. Increase the main memory size to the maximum the budget allows. Just going from 64 to 128 MB produces HUGE benefits, because you use your hard drive a lot less as "virtual memory," which speeds up things as much as 50% or more.
2. Get yourself a 7200 rpm hard drive. Higher spindle speeds usually mean faster data reads and writes on the hard drive regardless of whether you're using IDE or SCSI interface.
3. Get yourself a graphics card (if your system has an AGP graphics port) that uses the Matrox G400 or nVidia Riva TNT2/GeForce 256 chipsets. These faster chipsets makes a big difference in many games.
People are usually surprised how much "snappier" their computers get when they following the suggestions listed above.
Raymond in Mountain View, CA
> *sigh...* I suppose it really is just too difficult to actually read the article, isn't it? This announcement is that 1GHz PIII systems will be available, actually available, by the end of the month.
For what it's worth, here's what I see at pricewatch.com right now:
PIII 800 - four vendors listed, one offering at $799, the rest over a grand.
Athlon 850 - fifteen vendors listed, thirteen under $800, the rest between $800 and $850.
Athlon 800 - five pages of vendors listed, prices starting at $522.
Yet somehow Intel is going to jerk the rug out from under AMD's feet in the next four weeks. If you doubt it, you can just ask them.
--
Sheesh, evil *and* a jerk. -- Jade
When Pixar did their first animated short, they were using some old systems they scraped together. Each frame took about 3 minutes to render.
Now that Pixar has money, they use a huge rendering farm of Sun Enterprise servers with some SGI's for other work. Each frame still takes about 3 minutes to render.
For some tasks, two 500Mhz CPUs will run at nearly the same speed as one 1Ghz. Things like compiling or 3D graphics can be easily parallelized, and BeOS is great for SMP systems.
Search for "amd vs. intel" on google.com for some comparisions.
I bought an Athlon system as soon as I saw that it was designed by some of the Alpha 21264 team. So here's the differences as far as I remember - Tom's hardware has far more detail in it's articles on the K7 (Athlon).
Take your pick - I've been extremely pleased with my Athlon 650MHz. In fact the only thing I'm less pleased about is that this MHz race is making my processor look slower much too quickly - I'd usually upgrade once the top of the line processor gets to 2-3 times the clock speed of my current one, but at this rate that will mean an upgrade in Q4 this year :-)
Cheers,
Toby Haynes
Anything I post is strictly my own thoughts and doesn't necessarily have anything to do with the opinions of IBM.
1GHz? That'll help my distributed.net ranking.
The idea of releasing generations of little-improved processors with higher clock rates to give the impression of progress is decadent. What can we do with a 1GHz PIII processor that heats like a furnace and looks like a brick that we couldn't do w/ a 500MHz PIII? Nothin'. But lower the power requirements and heat generated (really the same thing...), and mebbe overall reliability, and you could add uses.
I say, go Transmeta or IBM/Motorola. My PowerBook G3 gets quite a few hours to the charge, without reducing the clock speed as a non-copper "Coppermine" would. I can't wait until Transmeta PDAs come out.
And it's not like anything but the first stage of the pipeline is running at the full speed. Ooh, the new PIII emulates the x86 ISA at 1GHz...
Where is my mind?
mfspr r3, pc / lvxl v0, 0, r3 / li r0, 16 / stvxl v0, r3, r0
Check out Project Upper/Mute, an all-around awesome compiler fra
Stop spreading the FUD
I could care less whether people get an Athlon or an Intel based chip, but at least try not to
look "pathetic" yourself in lying about prices.
A quick check at pricewatch.com will show that you're lying.
Prices on the 733 run from 530-540 (with a low of about 480).
The 800mhz Athlon (notice the spelling since the Athalon does not exist) runs from 520-540 (with a low in the 520's).
Maybe it's because I was a math major in college, but I can tell that the price difference between
(520-540) & (530-540) is NOT $200.
You're right, and you're wrong.
Hertz is "cycles per second".
If I'm making 5 widgets per second (5Hz), and you're making 5 widgets per second, they we're making a combined 10 widgets per second (10Hz).
Two Pentium 500s sitting side-by side, are each doing 500 million cycles a second. Together, they are doing 1000 million cycles per second. (500 each).
He's not arguing that. He's claiming that 500MHz + 500MHz != 1000MHz.
The applications which will use the power of these types of processor speeds are things like speech recognition and real-time video compression/decompression (video conferecing). Games of course always make use of whatever power is available - more speed enables more realistic games in terms of things like 3-D, textures, lighting, object modelling, etc.
Thinking "do I want/need to run my current apps faster" is the wrong question. Sure a few extra FPS never hurt a game, but it's "what will I be able to do that I could never do before" that is the much more interesting question.
BTW, a 1GHz CPU is probably capable of real-time MPEG-2 compression, which is quite a feat!
hhee let me explain why this is really important. cuzz everythings been mega for so long.. giga is somehow exciting damnit :p
Damnit if a 1gig processor dont give you penis envy your sic. Who the hell cares if youve overanalyzed it to death to point out The following things.
:-) Muahaha.. and for most people when they say.. so what kinda computer you got they only recognize maybe one or two terms.. like mega hertz or RAM so when you say.. 1giga hertz. there eyes light up.. Ahh a fast one cool! :-)
A. Processor Speed is Irellevant (whine)
B. My p133 does just fine thanks
C. I cant think of any reason *I* would need one
D. So its only 10fps more
WAAAAAAAAH You guys are destroying all the fun of new and faster processors. Just cuzz you cant think of a use for it dont mean I cant ( I can I promise
Is 1 GHz == 1024 MHz or should we continue to confuse Joe Average :) (and John Doh )
/das Ix
This is my sig, show me yours
Apple used to make a desktop that ran off of the Motorola 68000. Motorola added analog to digital, pulse width modulation (for music or motor control), serial and IR interface, and LCD control on dye. They called it the dragonball and sold it for $12 a pop. It now runs 3Com's PalmPilot.
Anouncements the such-and-such has a prossesor at X Mhz for $1000 at 50 Watts with no mobo ready does nothing for me. When someone releases a Pentium class MPU for $20 with on chip peripherals that runs off of AAA batteries, then I will sit up and take notice. Transmeta may be a step in the right direction but we are not there yet.
Why would anyone go out and buy a 1Ghz processor when you do not even use all the resources on the system that you already have? I believe if someone buys the 1Ghz processor it will be to impress friends and such things as that. Most of the operating systems these days (Linux) work better on more RAM anyway.
If your CPU usage is over 50% at ALL times, an upgrade may be due, but to a 1Ghz system when you are running a 200Mhz... Why bother, wait 6 months down the road and there will be a 2Ghz processor and the 1Ghz will cost $350 or so instead of in the $800 range, and all the other processors will be discontinued if not sold real low...
AMD is already ahead in the speed wars (850 vs 800), and with the 1GHz Athlon will also announce the 900 and 950MHz parts vs Intel's as yet unannounced (850), 866 and 933MHz PIIIs... Never mind the fact that the Athlon is a superior design.
AMD have no need to push lower speed prices down by announcing the faster parts, since they're already ahead and well positioned. Intel on the other hand is playing catch-up and hence is forced to announce faster parts even before it is capable of shipping them in volume (thereby hurting it's sales).
Jeez what's the point in buying a 1GHz CPU and plugging it into a cheezy 5400rpm hard disk.
Where the hell do you all think your CPU cycles go anyway?
I'll spend £1000 on doubling the speed of my disks before doubling the speed of my CPU.
Deleted
I can order a 1.1GHZ machine from Kryotech as a barebones system for around 2200$. Besides its frozen (40 C below I believe), so it will last longer than any other proccesor. BTW I hate AMD. The crappy proccesors they make are unresponsive and slow. Kind of like how a celeron can be clocked at 500mhz, but feels like a 133mhz. I know flame me, but thats my opinion, SO BITE ME
During WWII, they had shortages of all sorts of resources needed for the war effort here in the US. Chrysler came came up with a cool design where they put a number of regular automotive engines (I think it was something like 5-8 of them) on a common crankshaft and turned it all into a single Tank engine.
That's multiprocessing muscle.
People ask what is so special about 1 Ghz? Simple - it sounds cool. Sort of like nano.
Gigahertz is not necessarily better than megahertz, but it has a certain mystique that makes it great to market. Mega was already a common prefix before it came into common usage in reference to processors or other computer things. Thus when using the term mega in marketing, it had certain implications from prior usage.
Giga is different. Giga is not in common usage for things other than computers or electronics. So the companies that make giga products can create whatever meaning they want for it. As a result, giga tends to be marketed as new and better, because, for most people, it is new. Example: a certain cordless phone is toted as having "gigarange". For most people, this has little meaning, other than a new, better cordless phone. In reference to phones, more gigahertz is not necessarily any better - they are just different bands.
For some reason, giga has a certain appeal. My guess is that it is for the same reason we like other things that involve the rolling over of zeroes, like the transition from 1999 to 2000. For reasons I do not understand, people, at this point in time, like this sort of change. I am almost certain people will behave in a similar manner as tera comes into more common usage.
Try not to look "pathetic" by twisting what the original poster said.
700 Mhz Athlon- $260
750 Mhz Athlon- $360
800 Mhz Athlon- $520
733 Mhz PIII - 530-540 (one is at 478!?!)
800 Mhz PIII - 3 listed at $1000, 1 at $800
If you are going to flame, at least take a second to understand what he actually wrote. AMD's are a hell of a lot cheaper than P3s. The 733 P3 is 170-180 dollars more than an Athlon 750, and 280 dollars more than an Athlon 700. Pretty much what the original poster wrote.
show me a single PC that can serve up a one Gbit/sec of data? how fast was your disk again? sheesh.. you dont dedicate the entire network bandwidth to one machiene here...
quake. i turn around in about a half a second (prolly less). thats 15 frames to render a 180 degree rotation. 180/15 = 12 degrees between a frame. you can EASILY see this in quake, i can see it getting over 60FPS.... in addition, while turning, i am bringing new textures into view, and swapping out old ones that i no longer see. framerate drops below average. i hope ppl quit with this "30 FPS is the max" nonsence.. its just not true
Next month, I'll be handing out the bank the final monthly payment for my flashy new P200MMX... you should check those beasts out; they can do great stuff. Mine can play the whole Weezer video on the Win95 CDROM, without missing a beat. Seems like I'll need one of those 1ghz machines in about 10 years.
A 1 GHz Pentium is not a supercomputer -- a supercomputer is a moving target. Today's supercomputers get a few hundred gigaflops peak performance, not less tha one.
Intel's promises of processors to match and surpass those from amd, when they clearly can't deliver already cost them a lot of money. To start off some of their biggest customers, such as gateway did not receive the processors that Intel had promised and to cure their supply problems turned to AMD. Gateway did not sell any coputers with AMD chips until recently and was forced to do so because Intel was not able to deliver the processors that they were advertising. It seems that in this hyped up race to reach the 1Ghz mark, Intel has forgotten good business practices. If they were not able to produce 800Mhz chips in any quantities, what makes them think that they will be able to manufacture 1Ghz chips. The supply problems will continue and they will further alienate their customers.
The second problem in this boneheaded move is that now AMD will be forced to release their 1Ghz chips. AMD however will be able to manufacture them in much larger quantities then Intel. The net result will be that if you want to be the first one with a 1Ghz PC on your block you will most likely end up going for an AMD chip simply because the 1Ghz vapor Pentium chip will not be available for sale anywhere.
As a side note, think long the 800Mhz Pentium has been out and then tell me where I can buy one.
Your argument is semantically logical, but still a bastardization of the intent of the Hertz rating. 2 photons with identical frequencies do not result in a photon with twice the frequency. They add to give a photon of twice the amplitude (assuming coherent phase).
The simple fact that the units are consistent does not mean a property is additive. Take temperature for example. Intel Marketers would be adding temperatures right and left if it helped them sell PIIIs.
"This processor runs at 150F, this one runs at 160F, together, a staggering 310F!"
No.
I do see your point, but electronics rely on rising and falling waves, not a crank turning out widgets. When you deal with EM waves, adding hertz just isn't done.
-Rothfuss
the moderator who marked this as a Troll obviously has no clue what a troll is.
it's offtopic/flamebait, but by NO means a troll.
What really cripples a system be it Mac or PC is the OS running on it. Kill off Windoze, get rid of MACOS, install Linux (if ya gots the balls) and breathe new life into your system.
2000 - Year of the Penguin -- GO TUX!
the world's most intelligent space heater. Intel should be bundling marshmellows and hershey bars with that chip. The pentium III might not actually "make the web better", but I bet it could produce some damned fine smores.
IIIIIIII| HEIL JON KATZ!
IIII|
IIIIIIIIII|The Fourth Reich is Upon Us!
IIII|
IIIIIIII| jonkatz@slashdot.org
I've observed the behavior of quite a few SMP systems lately and I've noticed that theory and practice don't always agree. Simply trying to add MHz numbers together is wrong--it not only overestimates performance, it underestimates it as well. An SMP system is so different from a UP system that benchmarks are more or less meaningless without knowing the structure of the application.
You would think that a 1 GHz processor would run faster than two 500 MHz processors in real life. Certainly, for a narrowly defined set of applications (single-threaded CPU-intensive cache-bound operations) this is true. If you utilize all of your CPU's at full efficiency (in terms of total number of operations completed per second in one machine) then a single 1GHz part will produce more than twice as much throughput as two 500MHz parts in an SMP configuration, all other factors being equal. This is because the 1GHz part doesn't have to suffer bus contention or SMP signalling overhead while the pair of 500MHz parts do, so 1x1GHz CPU is theoretically faster than 2x500MHz CPU even if the speed difference is as small as one clock cycle per second.
However, real life does not always agree with the theory. It all comes down to efficiency: if you can't get 100% efficiency out of your CPU, you will typically get more total efficiency from two CPU's than you will from one. This increase in efficiency will actually make the two-CPU system more than twice as fast as the one-CPU system even at the _same_ clock rate. In other words, two CPU's are usually better than one CPU at twice the clock rate.
Why is this? Well, there's a two major factors, really: inefficient software and inefficient hardware.
The software people typically use (applications written using general-purpose libraries and development tools and operating systems designed to run general-purpose applications) sucks. It's an absolute efficiency nightmare. A lot of the performance features of an OS are workarounds for applications that don't (or can't) know better; e.g. virtual memory relieves an application designer of the obligation to manage RAM efficiently, but it does a very bad job unless the OS model of memory access happens to match the application's actual behavior. You can lose more than half of the processing capabilities of the machine to even the most efficient and powerful operating systems.
Running an OS with pipes, sockets, disk cache, virtual memory, processes, I/O buffers, graphics drivers, network stacks, protected address spaces, and other "conveniences" is really inefficient in terms of extra data copying, context switches, and other bus bandwidth around the CPU. A lot of these constructions are faster when implemented in an SMP system than in a single-processor system simply because one CPU can do the overhead while the other does useful (i.e. non-OS-overhead) work at the same time. It's like adding a hardware accelerator for OS overhead, and it has the same effect on OS performance that most video accelerators have on video performance. Of course, if you write a stand-alone program (i.e. you replace the entire OS) in hand-tuned assembly language designed to accomplish your one task, then it will run faster on the 1GHz 1-CPU system than the 500MHz 2-CPU SMP system--but very few people actually do that, so the advantage is non-existent for most users. Note that "None of the above" has always been and is still the industry leader in the embedded OS market, where every CPU cycle counts.
Enough about bad software, now let's talk about bad hardware. Yes, there is a tiny core of high-speed memory (128K or so, compared to hundreds of megabytes of slow RAM in a good size system) that is actually running at 1GHz. But that memory is attached to a L2 cache running at half that rate, and the L2 cache is attached to a memory bus running at 30% (at most) of the L2 cache clock rate. If your CPU hits the PCI bus, your CPU is stalled waiting for a clock cycle that is a whopping 30 times slower than the 1GHz CPU core. If your CPU hits the ISA bus...well, hopefully if you care about speed at all, you simply don't have ISA cards in your machine running between 125 and 1000 times slower (depending on transaction size and wait states) than your CPU core...
If you want a nice humorous read, read the Intel Pentium Optimization manual. The Pentium III is just an 8008 with lots of new features grafted on to it--features such as an asymmetric parallel executing RISC unit that you can only access through a braindead synchronous CISC instruction decoder. The PIII instruction set is heavily optimized for pointer arithmetic and ASCII character string operations, so it performs really well on benchmarks, but there are still a lot of instructions that cause the execution units of a PIII to stall for a few clock cycles at a time, and a few instructions that cause a stall lasting _dozens_ of clock cycles. This combined with cache locality (i.e. the CPU's aren't trying to use the shared memory bus 100% the time) means that memory access contention in an Intel SMP system is a non-issue for most people since the "other" CPU is very frequently stalled or busy and doesn't need the memory bus anyway.
When your CPU gets stalled doing these slow operations, it can often be running at 3% of its overall efficiency or less. Most of the "slow" operations only stall the CPU for a handful, if any, clock cycles, but even those can reduce efficiency to at most 20%. If you have a single CPU, at any speed, it is going to completely stall and your machine will sit idle for hundreds of picoseconds; however, if you have two CPU's, you can often make use of one while the other is stuck in one of these slow operations. A two-CPU system executing two threads with this kind of bus access pattern can give very close to twice as much throughput as a single-CPU system, all things being equal.
Of course, this assumes that both CPU's are equivalent in efficiency to a single CPU doing the same task, but this is often not the case--in fact, the efficiency of each individual CPU typically goes up because it has fewer context switches and less L2 cache traffic to deal with. These two operations are the big instruction pipeline stall generators. Remember: one cache miss is equivalent to about 100 integer operations.
Two CPU's means twice the L2 cache. Do not ignore this advantage. Main memory is slow--about 10 times slower than the core of a 1GHz CPU. You will never notice the difference in speed between a 500MHz and 1GHz CPU if you're spending all your time waiting for a 133MHz memory bus because you have insufficient L2 cache to do the job you want to do. 512K is good enough for desktop systems (128K or 256K is good enough for toys) but you need 1 meg or more cache in order to really notice how fast a 500MHz Intel CPU can go.
What that means in practice is that we are often comparing a machine running at one execution unit at 1GHz with 20% efficiency with a machine running two execution units at 500MHz with 30% and 90% efficiency. Do the math, and you find that the 500MHz 2-way SMP system is running at 300% of the throughput of the 1GHz UP system.
Yes, that's three times faster with the cheaper parts. Clock rates don't add at all.
Theory gets us again in the end, though. Single-threaded CPU-intensive cache-bound applications do come along from time to time, and on the 2-way 500MHz SMP system these applications will run at half the speed of the 1GHz UP system. Even worse, two truly memory-bound tasks executed in parallel on a 500MHz SMP system will take twice as long to run as the same two tasks executed sequentially on a 1GHz UP system (so instead of 1 result in 1 second you get nothing in 1 second and 2 results in 2 seconds). So it averages out in the end--some applications are much faster while others are much slower.
But the sad fact is, a well-tuned 2-CPU SMP system running a typical Linux desktop will be somewhere on the order of 3-5 times faster than a comparable 1-CPU UP system running at the same clock rate, particularly for an office-suite style of application usage. That means a real person doing office work on a 2-CPU 500MHz system is working not just as fast as a 1-CPU 1GHz system, but typically 1.5 to 2.5 times faster. And of course the CPU parts themselves are somewhat cheaper...
Hmmm...I sense a market here...
-- I avoid spam by accepting only OpenPGP encrypted or signed email at this address. Clear-signed, RFC2015, heck, even
THat all sounds great.. You've got me really motivated to getting one of these chips.
My question though is.. Is there ANY advantage to getting an Intel chip over AMD? Besides sticking to the most popular chip?
Thanks,
-Matthew
The 1GHz PIII is going to use Rambus too, I think. The IGHz Athlon will use SDRAM.
:-)
That's a huge price penalty to put on Intel systems, even with the just announced cheaper RIMM packaging.
Of course, as an AMD shareholder, I'm enjoying every moment of this!
Possibly. In an effort to maintain their lead through other means than performance, Intel keeps inventing new SSE SIMD (Single Instruction Multiple Data) instructions. You will have noticed that the K7 covers all the standard PII instructions now, and has it's own set of SIMD instructions which go under the moniker of 3DNow. So if you get hold of an application which is *very* heavily optimised to use only the latest SSE instructions, you might see a *slight* performance hit. I think Adobe Photoshop is the only major suite I can think of which has gone this route, and even then there is not much to pick between the Athlon and the PIII. Most other application makers have kept out of the SSE/3DNow battle and just support the more basic SSE instructions which are covered in both CPUs and don't hurt older processors much.
Cheers,
Toby Haynes
Anything I post is strictly my own thoughts and doesn't necessarily have anything to do with the opinions of IBM.
03/04/2004 - Intel today announced that the PentiumXII will now be available at 40GHz.
:/) are hopelessly too slow.
Many consumers are looking at the price and asking "whats the point?" Many users feel that their computers run their voice recognition software, video-phone software, neural-net/genetic algorithm AI agent software, applications just fine at around 35 GHz. Apparently these vision-less users see computer software as something that has already reached its peak, that there are no more useful applications that can be developed that will make use of this extra speed, thus making computers more useful.
Personally I think computers can never be fast enough. There are still hundreds of potentially useful applications to be developed for which todays computers (and network bandwidth
I'm with you on the compile-time thing, nothing seems to compile fast enough.
But there are also plenty of consumer-market applications that we are "still waiting for", such as speech recognition and AI software that will make computers a cinch for the public to use one day in the future, when they'll just be able to talk to their computers. (I emphasize the "still waiting for" because a certain company has been making a lot of noise publicly for the past 3 or 4 years about how much work its research department has been doing on speech recognition, but magically haven't come up with a single product .. I am assuming that either it is because there is no real competition in specific field, or it is because current CPU's are still too slow (I remember doing some speech rec stuff back on my pentium120) or it is because software hasn't quite caught up with the concept .. or maybe its a combination of all of these ..)