Intel's P4 3GHz w/ 800MHz Bus & Canterwood Chips

OldGrayDave writes "Intel steps out today with their new Pentium 4 3GHz chip that runs on an 800MHz System Bus. They've also released "Canterwood", the chipset chipset for the P4 that supports Dual Channel DDR400 memory, native Serial ATA 150, RAID 0, AGP8X, USB2.0 and a host of other bells and whistles. Check out this showcase and performance analysis at HotHardware, to see what all the buzz is about. Intel distances themselves again from the Athlon." Or, you can read more at Hardavenue, mbreview, Tom's Hardware, hardware unlimited, or The Tech Report. I dunno...hardware gets faster, bus gets faster. Tide goes in, tide goes out.

Comments

What value are these new processors?

[The Briguy]

I have to wonder what the point is with some of these new faster processors. At this point, almost no applications can really take advantage of the fastest chips available. My sister uses a 500 MHZ machine at home, and as far as I can tell she has no real issues with its speed. I have to wonder if Intel is just shooting itself in the leg, spending needlessly large amounts on R&D to produce chips that no one actually needs. PS - FP?

Re:What value are these new processors?

[Anonymous Coward]

I have two 2.8 GHz/333MHz P4s overclocked to 3.06 GHz/358 MHz.

First one is for compressing DivX, doing distributed protein folding and for occasional game of Ghost Recon, Rainbow Six or Operation Flashpoint.

On the second, headless machine I prototype my first principles calculations so that I can use all my supercomputer CPU time on actual production runs and not on debugging.

Re:What value are these new processors?

[bmongar]

I think as long as they can convince people that a new computer is needed, software doesn't really have to need the power.

Plus new computers always seem much faster, because when people get them they don't have all that spyware and trojans running on it yet to slow them down.

Re:What value are these new processors?

[Delta-9]

"I have to wonder if Intel is just shooting itself in the leg, spending needlessly large amounts on R&D to produce chips that no one actually needs."

That isn't the case for all the gamers out there. Each CPU cycle you have that makes your machine faster than the other 12yr olds means more frags fer j00.

Re:What value are these new processors?

[orpheus2000]

*Sigh*, this argument comes up every time there's a new processor out.

There may be no pressing mainstream need for these processor's insane speeds now, but there are two things:

1) Niche markets which will utilize the higher speeds (video editing, photo editing, music production, scientific computing) and
2) the Future. Software will always find a way to use that extra power. We call it "bloat" normally, but then we usually forget about that and accept it as the norm and shun everyone who's running less than 2Ghz.

Better now? Move along

Re:What value are these new processors?

[mrtroy]

My 900mhz machine is too slow for my average computer useage, not even including gaming.

Multitasking is a mother...let's say im running my C++ compiler, photoshop, about 4 internet explorers, msn/icq, email client, irc client, etc etc etc.

I leave my computer on 24/7 and do not like to reboot either which only hurts the situation.

If I wasn't a damn student I would run out and grab one of these nice new mobo's (apparently intel is owning amd now...), but I must wait!

Re:What value are these new processors?

[Morgahastu]

Why would there be an app that would require a processor that doesn't exist? Someone has to create that processor before apps expect people to have it.

The average consumer doesn't need anything about 1 ghz but people (and professionals) who want to play cutting edge games, do some 3d modeling, video editing will love this.

Re:What value are these new processors?

[Lao-Tzu]

Someday, I'll be able to write a highly graphical game like Doom 3 in a beautiful language like Python. :)

Really, I think that higher powered computers allow programmers to write software more easily. When you need a piece of software, and an in-house programmer can write it in a few hours rather than a few weeks, but only if you have a 3 GHz machine to run it on... that's muchly worth while. It's possible.

Re:What value are these new processors?

[splerdu]

That really depends on what kind of work you do. At home where all I do is code and surf the internet, my current P3-700 is chugging along nice and fine.

However I have another rig I use for video encoding, usually mastering old VHS and V8 tapes to DivX or DVD and in that setup I need all the speed I can get.

Re:What value are these new processors?

[fain0v]

I work as a microbiologist, and I occasionally do bioinformatics work. I can tell you that the real driving force behind most new biology is the microchip. I am very thankful that intel is putting that much money into R&D otherwise my data mining software might take 10 hours to run rather than 10 minutes.

Re:What value are these new processors?

[Per Wigren]

almost no applications can really take advantage of the fastest chips available ...unless you're into 3D graphics, videoediting or stuff like that..
Or if you're a developer and need to compile a lot of code very often..
Or if you're a warezkiddie and rips a lot of DVDs...
Or if bla bla bla.. There are lots of reasons to have a faster computer.. The target market isn't people like your sister though...

Re:What value are these new processors?

[shepd]

You'd be surprised at the new applications for a high speed processor.

Think software-based, real-time, HDTV decoding, for instance. Combine doing that with doing some other work at the same time and you've a lot of reason to use this speed.

Re:What value are these new processors?

[gbjbaanb]

it isn't just 'frags fer j00', or fps running your favourite first-person shooter.
As the CPU ability increases, the quality of the graphics increases too. Compare the old Wing Commander series (you know, the one that persuaded people to upgrade to a 486) with the latest version Freelancer..

It won't be that long before we have game that have realistic-looking characters, and people will want to play it, even if it is on a console rather than your PC. Games developers will always want to put in more features, it sells more games after all.

There is another attitude you could take WRT ever faster CPUs - the old, obsolete ones just get cheaper and cheaper. I'm thinking of upgrading to a 1.2Ghz duron because they now cost so little.

Re:What value are these new processors?

[Eric Ass Raymond]

Or if you're a warezkiddie and rips a lot of DVDs...

If you rip your music CDs into MP3s or OGGs, does that make you a warezkiddie? I think not.

I've ripped most of the DVDs I've bought into DivX for exactly the same reason why I ripped my music CDs years ago.

When you have all your movies/music on a hard drive...

1) You have instant access to them.
2) You can create playlists to fit your mood.
3) You can stream movies/music over a network.
4) You can store the original discs in a safe place.
5) You can watch movies/listen to music on your laptop without having to worry about the spinning CD/DVD-ROM draining its batteries dry.

Re:What value are these new processors?

[Planesdragon]

I leave my computer on 24/7 and do not like to reboot either which only hurts the situation.

Let's assume that you're running a desktop version of windows, which wasn't designed to be ran 24/7 without reboots--and that you don't leave all your programs open all the time.

I humbly suggest that you install SOME kind of memory management software. Or, alternately, schedule your PC to reboot itself at 5 am ever day, log in automatically, and then lock / turn on its screen saver / whatever else your security mechanism is.

I have a 700 mhz PC, and it's plenty fast for just about anything I do. Of course, I also shut the darn thing down when I'm not using it.

Re:What value are these new processors?

[Flabby Boohoo]

I think that he means that generally people do not do "first principles calculations", compressing DivX and distributed protein folding.

Generally, people surf, play games and office apps (like taxes).

I do all of the above on my P3-700 without a problem.

Re:What value are these new processors?

[MoThugz]

The value of these new processors can only be seen after the next few releases... the prices tend to drop, and thus making it more affordable.

The speed increments nowadays are much less steep than it was in the mid 90s.

So now is actually the time to purchase that 2.5GHz processor that you were drooling over about six months ago.

Re:What value are these new processors?

[Anonymous Coward]

Lol, now we know you are a warez kiddie, getting all defensive :) no one is attacking you kid, you can relax and enjoy you're illegal dvd's.

Re:What value are these new processors?

[Anonymous Coward]

compressing DivX

Well, something like a five years ago not many "ordinary" people (the ones that now surf, play games and use office apps) were compressing MP3s or using their home computer to manage their digital photoalbums with 1200 dpi resolution photos either scanned in or captured with a digital camera. Now it's commonplace.

I can very well see people in near future recording TV programs directly into DivX/Windows Media and managing their 500 GB "digital video library" at home.

Re:What value are these new processors?

[Croatian Sensation]

There's plenty of value in a faster processor. At work, I simulate contaminant transport in ground water. Some of the simulations that we run can easily take hours, days or even weeks.

If I can finish a simulation 20% quicker by buying a new CPU, I can save my clients and myself piles of money.

Re:What value are these new processors?

[Glonoinha]

6) You have a high speed connection to the Internet
7) ???
8) You're a warezkiddie!

Re:What value are these new processors?

[GeorgeWright]

Being someone that has been on the KDE development team for a while, I can say that compiling KDE takes *forever*. Having a fast processor like that would be great as 20+ hours to compile the whole of KDE from CVS is insane... I'd like to be able to compile each night. Every night.

And to the person that wrote the topic: This P4 chip may be distancing the P4 and the Athlon away from each other performance wise, but AMD have basically scrapped the Athlon in favour of developing on the x86-64 stuff, which will totally rock! I'm personally waiting until I can go out and get a dual Opteron :))

Re:What value are these new processors?

[gilesjuk]

Yeh, video encoding takes a while. Trouble is the codecs get more advanced and the filesizes fall resulting in more CPU power needed to encode :)

Anyway, I always tend to stay back from the cutting edge as the heat generated is too vast to be cooled silently. When the cutting edge chips become the average ones the cores have gone through a few changes so run cooler.

Re:What value are these new processors?

[Anonymous Coward]

Dont worry, remember with every new release of Windows, the sys requirements go up exponentially for no reason at all (or maybe to keep Intel and AMD in business?). Just look at XP.

Re:What value are these new processors?

[Anonymous Coward]

Plus new computers always seem much faster, because when people get them they don't have all that spyware and trojans running on it yet to slow them down.

It's called software depreciation. Microsoft wants us to keep buying new computers, so they up the system requirements with each new version of Windows so people need ever faster computers to do the same thing at the same speed. The truth is, my DX4-100mhz system with Windows 3.11 could at the time do everything that most people use their computer for at the same speed as a new computer with Windows XP. The solution for improving performance lays more in software efficiency than increasing CPU MHz - a 90MHz Pentium with properly designed software (think: an OS than runs in less than 6mb of ram like an OS jolly well should) should be well fast enough for today's computing uses. The reason why new computers seem much faster is not because of the higher clock speed per se, but because systems are having to literally chew through gigabytes of bloat borne of incompetence and for-profit-motive intentional bad-design for the purposes of encouraging new computer purposes.

Re:What value are these new processors?

[Jedi Alec]

Dunno, but my Duron 800 running Windows XP managed just fine. Besides, in your example I'd say it's more a case of the memory getting full and the harddisk doing overtime swapping back and forth than the processor getting bogged down. Ofc it might help that I have 1 Mozilla window instead of 4 IE ones and 1 Trillian instead of msn/icq/yahoo all open at the same time. Add Apache 2, a Firewall, an Antivirus, and i can still game comfortable. Got 1 Gig of memory in that box though...

Re:What value are these new processors?

[Lao-Tzu]

The reason why new computers seem much faster is not because of the higher clock speed per se, but because systems are having to literally chew through gigabytes of bloat borne of incompetence and for-profit-motive intentional bad-design for the purposes of encouraging new computer purposes.

Gigabytes of bloat in software is not born of incompetence. Well, maybe a little. Programmers developing that software keep making software easier to write by adding new technologies and developing new programming libraries. For example, Gnome and KDE have grown in size as libraries like gconf have been added to make software easier to develop.

Windows 3.11 was small and from a user's point of view it may have done 95% of the things that average people do with computers today. But no software developer in their right mind would prefer to write new software for Windows 3.11 rather than for Windows XP. Since those days, common libraries like MS's foundation classes have become more bug-free, new common controls have added functionality to the most basic of objects that previously would take days to write yourself, and technologies like COM and ActiveX have permitted modular and reusable components where previously code would have to be re-written and re-developed for every application.

Re:What value are these new processors?

[oingoboingo]

Why would there be an app that would require a processor that doesn't exist?

I don't know, ask Sun...they're the ones who invented Java. It's still waiting for the first batch of 10GHz CPUs to roll off the production line to be useful (drum roll/splash)

Re:What value are these new processors?

[twisty7867]

This is classic wrongheaded thinking. Who cares about consumers?

Where are most computers deployed? Businesses.

I can tell you that my 2.8ghz desktop is barely adequate for my development work - and that's just software development! I can only imagine doing video or scientific work. The desktop market just doesn't drive performance; the server and workstation markets are the real force there.

Re:What value are these new processors?

[be-fan]

Sigh. This happens every time hardware advances, and every time, I have to post to refute this shit. People have been saying "do we need more hardware speed?" since the 386 came out. Back when the Pentium Pro was king of the hill, a guy in PC Magazine (Dvorak I think) refuted this idea. And people are still doing it! No, computers are not fast enough. I do things every day that need lots of power.

1) Compiling. On my 2GHz P4 with 640MB of RAM, it took GCC well over 4 hours to compile kdebindings-3.1 A compile of KDE as a whole takes 6-7 hours. Even a simple 900 line C++ program I'm working on at the moment takes almost 10 seconds to compile.

2) 3D. Even the latest 3D PC games don't like all that realistic. Notably absent are realistic lighting and liberal use of curved surfaces. In an interview with Boot magazine, David Kirk (chief scientist at NVIDIA) once said something along the lines of (don't remember it exactly, it was in 1998): there is tremendous potential for growth in 3D. 'You want to be able to render a tree so well that it's indistinguishable from reality. Then, you want to be able to render a tree so fantastic that it couldn't possibly exist in real life, but is believable because of how well rendered it is.' Beyond that, 3D modeling is still a huge potential market for more power.

3) Scientific computing. It's not unusual for a simulation, even a simple one, to take hours. I remember that a X-ray diffraction project I did in school once took several minutes to process *each frame* on a 1.5GHz Athlon. When computers get fast enough that you can run that simulation in real time, and watch the diffraction patterns change with changes in the crystal, then computers will be fast enough.

4) Mathematics. Computational mathematics is very big, and there are tons of problems that take hours to solve on current computers.

So no, computers are not fast enough. They might be, if all you do is surf the internet and run Word, but a lot of people in the real world have real (demanding) work to do with their computers.

Re:What value are these new processors?

[be-fan]

Maybe on the processor end, but (finally!) bus bandwidth is increasing like anything. We're already up to nearly 5GB/sec sustained bandwidth (from the benchmarks for this sucker)! That's almost as much as the GeForce4Go 440 in my laptop! These recent changes are a whole lot m ore interesting than 100MHz jumps in CPU speed.

Re:What value are these new processors?

[zbuffered]

I'm sorry I don't have any mod points.

Re:What value are these new processors?

[Anonymous Coward]

right... if amd could still come close to intel and they released a 3ghz proceessor everyone would be talking about how cool it is and how much better they are. but because intel is a big company like ms you guys try to bash them. forget the fact that intel has donated millions towards linux for a long time.

Re:What value are these new processors?

[michrech]

Anyway, I always tend to stay back from the cutting edge as the heat generated is too vast to be cooled silently. When the cutting edge chips become the average ones the cores have gone through a few changes so run cooler.

That is what cases like the VapoChill are for. I just tried to go to the web page and it didn't respond for some reason (mayhapps someone already mentioned them in here making them slashdot'ed? Mahapps a huge meteor fell on their web servers.. Or mayhapps even the dreaded Chubacobra (SP?) has chewed all the ethernet cables between myself and afore mentioned web servers.... Dunno, I haven't read through all of the posts yet)... Anyway..

Nice, cool, more silent than 60 80mm 7,000,000,000 RPM fans in your gigantic, waaaay too large for the use of the computer, heavily modded case..

I'm rambling now...

*pop of pill* Ahhh.. Good ol' ridlin.... *drool*

Re:What value are these new processors?

[Anonymous Coward]

And behold, after a heroic effort expended on optimizing the next killer app on the fastest CPU using 1GB RAM as well as RAID0, the disappointment was great that it would not run on Joe Sixpack's PC using Celereon 1700MhZ and 128 MB RAM. Sounds familiar?

Re:What value are these new processors?

[Bombcar]

Even a student should be able to afford a 1800 Athlon XP + mobo for $86 at Fry's (apparently AMD is 0wn1ng Intel on price...)

Re:What value are these new processors?

[zerocool^]

The speed increments nowadays are much less steep than it was in the mid 90s.

Not really. Granted, the speed increases were greater, but they came less often. The path from 100 to 200 Mhz Pentium took about as long as the path from 1.5 Ghz to 3.0 Ghz (which, by the way, has been out, sans the 800Mhz bus for a while).

Re:What value are these new processors?

[zerocool^]

There are always uses for faster processors. The first that comes to mind is video compression. Spin up gordion knot, and let me know how fast your 500 Mhz compresses a DVD. Probably.... 6 hours-ish? That's probably pretty close.

My 1500 Mhz 1800+ will do it in about 2.

But, then, some people want it done in 10 minutes. Even 3.0 Ghz on 800Mhz bus won't do that.

Things that take orders of minutes, not microseconds, will benifit. Video compression, Audio compression, brute force password cracking, video editing, image filters, distributed.net, etc.

Just keep in mind, we're talking about things that are infinately scaleable, or close enough. Distributed.net is infinitely scaleable, basically. It would benifit from faster proc's.

~Will

Re:What value are these new processors?

[siliconwafer]

One word: Games.

Many hardcore gamers want games that look as lifelike as possible. Although we've certainly come a long way since Wolfenstein 3D, there is still a long way to go. I remember the first time I played Unreal, I nearly fell out of my chair when I took a look at the scenes outside. I'm sure video cards will be resonsible for most of the graphics, but high end CPU's will certainly play a necessary role.

Re:What value are these new processors?

[ncc74656]

I have to wonder what the point is with some of these new faster processors. At this point, almost no applications can really take advantage of the fastest chips available.

Good thing you said "almost," as I'm guessing you don't do much video compression or editing. I've cranked out probably hundreds of SVCDs (mostly of TV shows) over the past couple of years, and I just got a DVD burner last week. MPEG-2 encoding eats up all the processor time you can throw at it...I built a dual Athlon MP 2200, and while it was a major improvement over the 1.0-GHz Athlon I was using previously, I could still use more speed.

Re:What value are these new processors?

[Zugot]

Actually the reason KDE takes so long to compile is because of GCC. From what I understand, GCC 3.4's hand written C++ parser will make a big difference in compiling things like KDE.

Re:What value are these new processors?

[sketerpot]

There are some problems that seem to be able to soak up as much processor power as we can throw at them and still want more. Cryptography is a big drain on CPU, and as the CPU power goes up, so will the size of keys, and so will the power demands of cryptography. Simulations for various things seem mired in complexity; the more clock cycles you can spare for fluid dynamics calculations for that new space shuttle replacement that they've been thinking about, the better. Simulating protein folding takes a lot of time for even our fastest computers. Realtime raytracing would be great for games.

If you get the speed, programs will come.

Re:What value are these new processors?

[gilesjuk]

Seen Vapochiil before, but I would rather keep things simple and I would rather AMD and Intel worked hard to produce cooler CPUs.

Currently I only have a CPU fan and a PSU fan running with a 2100+ Athlon XP., CPU is 46oC and case 37oC which isn't bad.

Re:What value are these new processors?

[ergo98]

Add to that the fact that MPEG2 is a rather simplistic encoding technique, and other compression formats like WM9 or Divx are dramatically more CPU intensive. Now that I've got a miniDV video camera I dump these 12GB or so files to the PC, and having it compress overnight for a one hour movie in WM9 format gets a bit annoying: Doesn't really lend itself to experimenting.

Re:What value are these new processors?

[sketerpot]

Now where's the profit in that? You'd need to charge money for your 3117 w4r3ZZ!!!

Re:What value are these new processors?

[Anonymous Coward]

I run winxp and cubase sx.
cubase sx has what are called "soft synths" and dsp plugins.
These soft synths and plugins replace hardware outboard gear with "virtual" gear. This virtual gear, combined with audio tracks can use GOBS of system bus bandwidth, as sound is streamed all over the place and processed.

This 800Mhz bus will allow the CPU to get the stuff moved around faster, eliminating dropouts as the sound card buffers no longer wait for data.

Before I get hammered about setting up cubase and my sound card correctly, I have already followed all recommendations. My dropouts only occur when running excessive numbers of audio plugins and tracks. Neither my CPU or drives are the bottleneck according to the indicators... and the soundcard buffer size is maxed(have tried all possible settings).

I think it is time for 533 or 800 FSB and GHz rambus...

l8,
AC

Re:What value are these new processors?

[clarkc3]

Software will always find a way to use that extra power

thats very true - its called lazy/poor programming technique

Re:What value are these new processors?

[aoteoroa]

I remember having the exact same conversation back in 1993 with my buddy Phil. He said something like:

"486 100!! Who needs that kind of speed? My 386 does windows, and word processors, and spreadsheets. Nobody needs the speed of a 100MHz machine"

I totally agreed with him. Because back then my 8086, and WordPerfect 5.1 for DOS was serving my needs well. There is always going to be new cool things to do with the computer.

Back then tools like pagemaker, and photoshop were driving the need for more horsepower. Today I think it is the desire for better and better 3d rendering.

Re:What value are these new processors?

[cpt kangarooski]

You forgot to mention that people also enjoy watching Divx movies of proteins being folded.

Re:What value are these new processors?

[Dr. Zowie]

Having seemingly-ludicrous amounts of CPU power available allows all kinds of tasks that you might not think of.

I'm a bit of a special case -- I'm a solar physicist, and a fair amount of my work consists of analysing image and spectral data from solar telescopes. My dual Athlon machine with 4GB of RAM is nothing special these days -- but allows me to do incredible analysis tasks interactively. Machine vision, huge wave-propagation studies, numerical magnetohydrodynamic modeling -- there are a huge number of techniques I can use that would be inaccessible without the fast processors.
Ludicrously overpowered CPUs let me do more interesting science, faster.

More everyday applications (that might actually pay for the development of the chips): home studios with real-time noise-gate removal and bizarre sound-morphing effects; better TiVO effects (identifying commercials!); better user agent software; real-time grammar checkers that work; and immersive, 3-D rendered pr0n.

Re:What value are these new processors?

[Skjellifetti]

The current problem of interest is always too big to run on the current generation of hardware.

Re:What value are these new processors?

[Dr. Zowie]

No kidding. I write magnetoconvection simulations in perl for cryin' out loud...

Re:What value are these new processors?

[jawtheshark]

Uhm... How much memory has this system got? Honestly, two years ago (or was it three?) I bought an OEM P-III 800 with 128Meg RAM. (It was a deal I could get over work). Anyways, I ran Win2000 on it and it was slooooow. Very slow: games were unplayable (even older games like Halflife and Unreal). So, I just went out and bought some additional RAM. Problem solved! It now has 768Meg RAM and honestly, it works like a champ.

Also check if you don't have spyware and check for viruses. I have seen good systems crawling to a halt by both spyware and viruses.

Re:What value are these new processors?

[Anonymous Coward]

500GB is very optimistic: ) I can easily see Terabytes in the home. The average new computer now has 80GB drives. 5 years ago it was 4 GB. If you follow that curve out, in 5 years the average home computer could easily have 1.x Terabyte drives OEM'd from the manufacturer.

I usually fill up a stock harddrive within 3 months. This has changed little over the past 10 years through 8 computers. I was one of those people you mentioned who were compressing MP3's back in 1998.

I have found that in order to not need to move stuff to tape before my next upgrade, I need a file server RAID that is about 5x the size of my local hard drive. Funny how it is cheaper to buy a RAID on a UPS than a decent tape backup... by decent I am talking DLT or something similar at 3MB/s or better.... but I digress....

Bottom line: unless intel keeps hammering on new technology, all the people that ask questions such as "Why do they keep making it faster" would be bitching about how slow their games run, or that new 2x larger movie format is playing herky jerky, etc.

Software will always be the catalyst for better hardware, not the other way around. You design software, through time it gets used and taxed more heavily, files get bigger and more numerous, the software slows down because of this, and the hardware gets faster to compensate.

I will *never* be satisfied with my hardware, until I put in a command, and my answer is returned instantly, or I start that file conversion and by the time I press the button and look up at the screen, it is done. Until then, there is always room for improvement. I am an impatient person living in an impatient society. People like me are getting more commonplace and Intel is depending on me and people like me to keep them in business: )

I have 3 intel computers, and when I am working, all three are working at nearly 100% CPU. When I do music they are working this hard every time I press the play button. So the way I see it, Intel motherboards could be 4x faster and I would be happy.

Unfortunately by the time Intel gets 4x faster, the software will be 4x as demanding. So the cycle continues...

l8,
AC

Re:What value are these new processors?

[Anonvmous Coward]

" I have to wonder if Intel is just shooting itself in the leg, spending needlessly large amounts on R&D to produce chips that no one actually needs. PS - FP?"

No one? Have you forgotten about the game market?

Re:What value are these new processors?

[Xrikcus]

That or security, flexibility, stability (theoretically)

Of course it can also mean more layers of bugs...

Re:What value are these new processors?

[Anonvmous Coward]

"thats very true - its called lazy/poor programming technique"

Why's everybody so fucking pessimistic?

Resolutions get higher. More memory's used. The picturer is clearer, more detail, and can do more. Yet, that's all chalked up to lazy/poor programming technique?

Re:What value are these new processors?

[orim]

"I have to wonder what the point is with some of these new faster processors."

I'd say their biggest advantage is dropping the price of processors only 3% slower by about 50%.

Re:What value are these new processors?

[Anonymous Coward]

Toni Kukoc, is that you?

Re:What value are these new processors?

[Anonymous Coward]

We used to make fractals on an a atari 800, took all night to generate just one picture. We would then argue over where to zoom for the next night. We were running an app for a processor that had yet to be created.

Re:What value are these new processors?

[Anonymous Coward]

Statistical model 'science' isn't scientific.

You're just playing 'Sim City' and fooling somone into paying you for it!

Lord help us if everybody just crunched numbers instead of thinking....

Re:What value are these new processors?

[bigbadwlf]

Plus new computers always seem much faster, because when people get them they don't have all that spyware and trojans running on it yet to slow them down.

Speaking as someone who works in tech support, let me assure you that it doesn't take them long to slap kazaa on there and then call me because their computer they bought (I am not making this up) the day before is running slowly and they're getting popups when they're not even connected to the internet.

Re:What value are these new processors?

[SN74S181]

And yet the software to process the problem's data is always at least a generation older and less optimal than the hardware it runs on.

And generally, the hardware is obsolete by the time (if ever) code generating tools are in common use that take advantage of it.

Let's face it, for most purposes, we still live in a 'faster, better 80386' world of software. And how many of us actually write optimal protected mode code?

The bloat continues. It's necessary, so that stupider and stupider people can use computers, i.e. so the markets can expand.

Re:What value are these new processors?

[maraist]

Java / C# / VB => lazy programmer..

But remember, lazy programmers make good programmers. VB's simplicity means business types don't need to "bother" specializing in programming to perform advanced macros (or crappy web sites).

Java/C# is a more robust language than c/c++ (in my opinion), but allows us to lazily ignore resource management (which has it's pros and cons, but mostly cons), and also is designed to be interpreted (even jit's can't fully transcend the interpreted dynamic loadabled runtime API).

Further, "lazy" also applies to the newer and newer graphical abstraction routines, which go back and forth between enabling more graphics cards to work with the advanced graphics of new games and slowing the begeezus out of our systems. These abstractions are mostly because software developers don't want to rewrite their nifty effects to work properly on every graphics card.

Abstraction is all fine and good, but it's only there for lazy purposes. And the cost is almost always performance.

Poor programming is when we choose lazy solutions over specialized solutions when only cost is the knowledge of an API. (e.g. applying VB/Perl/Java to every problem)

That being said, given that human resources are expensive, I'll take faster machines and slower programs any day. :) You can always profile an app.

Re:What value are these new processors?

[4of12]

I have to wonder if Intel is just shooting itself in the leg, spending needlessly large amounts on R&D to produce chips that no one actually needs.

I have to wonder, too.

If you're right, then there's money to be made in using current state of the art chip design to create old style processors with less performance, less cost, and less heat dissipation.

If they start producing such beasts, and they sell, then your skepticism will have been proven correct.

That "marketing matters" is indisputable, though. Cars with much greater performance than can be legally exercised still sell.

Re:What value are these new processors?

[Anonymous Coward]

Yes, an OS should only run in 6MB of RAM. Why don't we all go back to Mac OS 7 or Windows 3.1 so we can achieve that goal?

Please. I'll take my RAM-gobbling, yet stable, powerful, and not to mention pretty OS X over Windows 3.1 or Mac OS 7 any day.

However, I do think that a 3GHz P4 with an 800MHz bus is overkill for ANY home user. There are some applications that need it, the most popular one probably being games, but for word processing, email, and web no one needs even one of the lesser P4s. My sister uses a 450MHz iMac DV+ with 192MB RAM as her daily computer. The OS X eye candy really slows it down, but it's fine for her use (word processing, web browsing, etc.). The only thing that a P4 3GHz would do for her is animate the Windows XP transitions REALLY nicely (assuming a good graphics card for you nitpickers). Hell, my dad ran a 66MHz 486DX2 until it died on him in 2000 after 8 years of use. He currently has a 500MHz Celeron.

But, like another poster said, the lure of having a clean system (which will obviously be faster than a older system loaded with junk) will make Joe Consumer buy computers with these chips inside, even when they don't need them.

Re:What value are these new processors?

[clarkc3]

While that is true - I was thinking more along the lines of people who are 1st learning to program who fail to optimize code because it still runs ok on their system. They start out not caring about optimized code may or may not be since their system runs it fine

Re:What value are these new processors?

[Anonymous Coward]

Rendering the martians on the fake Mars program

Re:What value are these new processors?

[pmz]

...technologies like COM and ActiveX have permitted modular and reusable components where previously code would have to be re-written and re-developed for every application.

Don't forget Java, Qt, and the like. Why not also add portability rather than chaining yourself to Windows? If the MS ship sinks, why not have a life raft handy? Linux? No problem. UNIX? No problem. Mac OS? No problem. Companies should take advantage of portable APIs at every practical opportunity.

Re:What value are these new processors?

[pmz]

We call it "bloat" normally, but then we usually forget about that and accept it as the norm and shun everyone who's running less than 2Ghz.

It seems the niches get smaller and smaller each time. 200MHz Pentiums reached an important threshold of usefulness. For example, I was able to use a AMD-K6 200-based computer up until recently and was tolerably productive (GNOME 1.4, burning CD-R, Mozilla 1.2, etc.)

500MHz seems about right for Windows XP, Office, etc. for good productivity. This is also getting to a point where modest CAD models are comfortable.

1GHz and above...games, simulations, games, video, games, big CAD models, games, etc.

Quite honestly, the simulation, video, and big CAD model markets are niches. Big niches in themselves, but very small relative to office desktops and home desktops. Intel will have to face this eventually, if they aren't already.

Re:What value are these new processors?

[pmz]

Someday, I'll be able to write a highly graphical game like Doom 3 in a beautiful language like Python. :)

When this is possible, Doom III will look primitive when compared to the simultaneously released C-language based games. Games are the one market where high-level languages will probably never catch on.

Re:What value are these new processors?

[bmcent1]

Thank goodness for the steady march of more powerful processors! Faster processors are a boon to many different types of computing. As are bus speed enhancements, RAM density and speed improvements, and storage gains.

As a person who uses Photoshop to enhance and work with 138 MB images, I can attest that there is no shortage of demand for faster processors, or faster anything for that matter, when it comes to hardware to run powerful programs.

I know many scientists whose research demands fast, inexpensive computers. As computing power becomes more affordable, science projects that were out of reach become feasible.

Faster computing resources are very valuable indeed. For many people, they are directly valuable to their work or hobbies. Many more people benefit from the science, ideas, and technology others develop with new computing power at their hands.

Re:What value are these new processors?

[zodar]

Amen. db joins that took all night on our old, dual 200Mhz pentiums (w/1024K L2 cache) now take an hour or so with the new 2.4Ghz P4s.

Re:What value are these new processors?

[Anonymous Coward]

I'm allready writing games in python. So are quite a lot of other people.

I have shadows, and dynamic characters. A lot of games these days is just going though opengl(or direct3d). Which a high level language is awesome for. You get to tweek/debug/prototype lots faster.

Also if anything gets slow, I can just mix in some c/c++ or asm with my python.

Check out http://www.py3d.org/ for more information about all things python+3d.

Re:What value are these new processors?

[Anonymous Coward]

Then you've completely missed the point. Apps don't have to assume a non-existing CPU to take advantage of the extra processing speed and nenory bandwidth. Take encoding (or decoding) video: my current rig will do a DivX of a 2 hour movie in about 15 hours (2-3 passes depending...). Even a 'low-end' 2GHz system would far more than halve that, as some operations are P4 optimized, plus the increased memory bandwidth. I'm looking forward to a 3+GHz CPU as a future upgrade (as soon as my income tax refund is deposited maybe); none of my apps need to 'require a processor that doesn't exist' to realize a huge increase in performance.

Re:What value are these new processors?

[Xarin]

I think Java is a good example of why more powerful processors are neeeded. When Java first came out, a lot of companies had 99Mhz Pentiums and AWT based GUIs were not responsive enough. As the processors speed up, programs could then be written in a more straightforward manner instead of a less intuitive manner that executes faster. Also, CPU intensive features such as garbage collection can further lighten the load of what need to be coded.

Re:What value are these new processors?

[jugaaru]

thats right, my current PC which is P4 2Ghz works almost the same as my older one P III 450Mhz. I wonder all that mega hertz mean anything?

Re:What value are these new processors?

[wet_blanket59]

Computer's are like cars No matter what you have, it needs to be faster Long may it remain so otherwise I'd never have time to wash!!

Re:What value are these new processors?

[Anonymous Coward]

True enough, what do we need with faster CPUs? It is more of a game with marketing at this point than performance. Most people could save a $1000 and buy a 1.8GHz Celeron with lots of RAM and a good hard drive than a 3GHz machine. A dual core, and not a "hypebole" (oops hyper threading) core is needed. And trimming down the cycles required to do something is a great bonus. If it takes 50 cycles to move some bytes, heck I could design a 25GHz CPU that divides it to 1MHz and then does something.

No doubt; need more innovation and less hype.

Re:What value are these new processors?

[Anonymous Coward]

I think changing PC's should be done when processors (as nothing else gets as fast that fast) get a 10 times multiplication in speed. There's a really big change from a 486DX@50Mhz to a PIII@500Mhz, but not that big from the same PIII to a P4@1.5 or 2Ghz. Processor speed isn't everything - memory bandwidth, HDD read/write/seek times also count. There'll always be a +10/+20 increase in speed when a new processor comes out, but that's no reason to spend 800$ immediately on it.

They need to produce a premium product...

[Goonie]

This comment has been made every time a new processor comes out, and the usual reply is that there are plenty of applications that still require more CPU performance (which there certainly are), and sooner or later there will be one which is sufficiently compelling that Joe Sixpack will upgrade.

Alternatively, one could try a reply based on business models. Intel is an R&D-driven company. They don't want to be the next Zilog. If they don't continually introduce new products, that's what they will become, and it's really hard work competing in a low-margin commodity business.

Re:They need to produce a premium product...

[captaineo]

While CPU power is moving ahead, it always feels like internet bandwidth is lagging behind (anyone else's ISP lowered their up/down caps recently?). A good use for those extra cycles might be advanced compression algorithms to squeeze every last bit of efficiency out of our limited-bandwidth internet connections.

Consider that the Linux kernel mirrors just deprecated gzip in favor of bzip2, which provides significantly better compression, at the cost of more CPU power to encode and decode.

Or better video compression algorithms - one of the main reasons newer video codecs (MPEG-4/WMV9) compress better than early codecs (Cinepak/Indeo/MPEG-1) is that they can assume a much more powerful CPU for decoding.

Or even still image compression - JPEG-style DCT algorithms are handily outclassed by many newer wavelet algorithms. Imagine how much bandwidth you could save by replacing all your JPEGs on the web with much smaller wavelet files.

Re:They need to produce a premium product...

[sphealey]

This comment has been made every time a new processor comes out, and the usual reply is that there are plenty of applications that still require more CPU performance (which there certainly are), and sooner or later there will be one which is sufficiently compelling that Joe Sixpack will upgrade

However, what really happens is that the major box manufacturers need to keep their price points between $1500 and $2500 - if the prices for the "basic machine" (as actually sold - not as advertised in the Sunday paper) fell below that they would not be able to survive.

So Dell, Gateway, HP, etc. work with Intel to assure a constant flow of "new, improved, and faster" CPUs, put those in their boxes, and terminate the "older, slower" systems. Then the volumes on the slower chips and their associate motherboards fall and Intel terminates them due to "lack of demand". So the cycle continues.

Other than gamers and a few applications heavily dependent on processing (when was the last time you folded a protein) very few PC users have gained from the speed increases of the last two years. Most businesses I know are putting fresh 20 GB hard drives into their 3 y.o. P3 733s and are quite satisfied.

sPh

Re:They need to produce a premium product...

[sketerpot]

when was the last time you folded a protein

I fold them all the time! Protein folding is a natural part of physiology, remember.

Perhaps there can be a market in cheap computers with older parts so that older chips don't go away too fast for more expensive ones. I think having bulk buyers (schools, businesses, etc.) would be good for a business like that.

MHz Myth

[Anonymous Coward]

3Ghz? Bah, stop drooling!

Everyone knows that my 1GHz G4 is just as fast...right? right?!?!?!

ah, slashdot

[haa...jesus christ]

I dunno...hardware gets faster, bus gets faster. Tide goes in, tide goes out.

ah slashdot, let ye profundity run far and wide.

Re:ah, slashdot

[NewbieSpaz]

"Tide goes in, tide goes out."
<Homer>
"Bed goes up, bed goes down, bed goes up, bed goes down."
</Homer>

:P

Now We Can Test Serial ATA

[MBCook]

Now we can test serial ATA to see how good it REALLY is. Before now it's always been an add-in, either a card or built onto the motherboard. Now that it's built into the chipset it's not at the mercy of the PCI bus, where it has to deal with your soundcard, your tv tuner, that firewire card you bought, and everything else. This will also bring about more SATA drives now that it's going to be on so many motherboards (I know many have them now, but this is an Intel chipset. This will push everyone to do it if they're not already).

That said, I'm disapointed that you only get 2 SATA channels. Remember, with SATA it's only one device per channel, unlike parallel ATA.

Re:Now We Can Test Serial ATA

[GigsVT]

An integrated SATA controller likely still uses the PCI bus. Sorry.

Try a 64 bit PCI 3ware SATA controller. You can get them with up to 12 ports.

Re:Now We Can Test Serial ATA

[shepd]

Not in this case. The block diagram clearly shows the SATA controller gets 150 MB/s bandwidth, 12% more than the PCI bus.

Re:Now We Can Test Serial ATA

[Zathrus]

Now we can test serial ATA to see how good it REALLY is

And we'll continue to see that... shocker... it doesn't make a bit of difference.

The limitation is not on the interface (parallel vs serial ATA), or on the bus (PCI vs insert_chipset_bus_here), but on the drives. There are no drives available that come anywhere close to saturating ATA/100 or ATA/133, so SATA/150 isn't going to help much. Ok, yeah, it'll help for the microsecond that you're reading from cache instead of from the drive itself, but that time period is so absurdly short it's not even statistical noise.

The advantages to SATA aren't in the bus speed arena... the improved cabling, hot swapping, and simplicity of hookup is what it's all about. I would've killed for SATA this weekend after spending an hour fiddling with 3 IDE drives and a CD-RW to get their master/slave jumpers correct (turned out that one was only happy with the master drive as cable select and the slave CD-ROM as slave -- anything else wouldn't be detected. Joy!).

As far as the number of channels go - 2 may be ok for now, but it's going to be deeply inadequate in the future. I'd hope that systems start appearing with 4 channels in 6 months, and 8 within a couple years. By which time standard ATA connectors may be gone entirely. (For more realistic estimates, change 6 mos to 1 year and 2 years to 5 years).

Re:Now We Can Test Serial ATA

[cqnn]

>> Remember, with SATA it's only one device per channel, unlike parallel ATA.

AFAIR, you can design SATA to have more than one device per channel.
(albeit the devices would then share the bandwidth from that channel).

It is only the current design strategies that are forcing the one device limit.

I think that when SATA-2 starts to get more notice (higher bandwidth and
better switching between devices), then there will be more encouragement
to place multiple devices on a single channel.

Re:Now We Can Test Serial ATA

[cheezedawg]

AFAIR, you can design SATA to have more than one device per channel.

No you can't, and you never will be able to. The design of the protocol makes it impossible.

Re:Now We Can Test Serial ATA

[cheezedawg]

Thats very unlikely, actually. In fact, I don't know of any integrated peripherals in Intel chipsets that are really on the PCI bus. From a software perspective, it has PCI cfg space addresses, but in reality it is nowhere near the PCI bus. The only bandwidth limitation that you need to worry about is the bandwidth between the southbridge and the northbridge.

Re:Now We Can Test Serial ATA

[Alan Cox]

Current SATA stuff isn't going to give you a good idea of what SATA can deliver. Most of the really nice stuff is a few months from hitting the shelves.

Re:Now We Can Test Serial ATA

[Strog]

12% more than the PCI bus

True, if it is 32bit, 33Mhz bus. A lot of systems have 64bit and/or 66Mhz+ busses. It could be on the PCI buss if it is > 32bit/33Mhz but your point of having enough bandwidth is still valid.

Re:Now We Can Test Serial ATA

[Anonymous Coward]

The limitation is not on the interface (parallel vs serial ATA), or on the bus (PCI vs insert_chipset_bus_here), but on the drives

Enter RAID 0 Striping...

Re:Now We Can Test Serial ATA

[Anonymous Coward]

quit talking out of your ass moron

Re:Now We Can Test Serial ATA

[Anonymous Coward]

Wow... you've nearly saturated ATA/100 with that!

Yawn.

2 serial ATA devices

[Wehesheit]

Just out of curiosity why is there only 2 SATA controllers on all these new motherboards? No room? Too expensive? No need? Whats the point in having SATA RAID with only 2 devices? I'm looking at building a new UBER-Fileserver for my home and want to use SATA but I want at least 4 maybe 8 HD's in the thing.

Re:2 serial ATA devices

[Anonymous Coward]

Just out of curiosity why is there only 2 SATA controllers on all these new motherboards?

Because at this stage of the game, S-ATA devices are still relatively uncommon. There'd be no point putting 4 or more S-ATA ports on a board today, because:
-very very few S-ATA optical drives are available at present, so
-you'd still have to put the ATA/100 ports there, so space would be constrained, and
-very few people would use them.

That said, boards WILL have more S-ATA ports, but it'll be a phased transition. Later this year, or early next year (precise timing will depend on the market's adoption rate) expect to see boards with, say, 3 S-ATA ports and only 1 ATA/100 port. And maybe, later in 2004, we'll see mainstream boards that feature only S-ATA. Of course, there'll be vendors who offer legacy IDE-free boards before then, but they'll be niche products - just like the "legacy free", no-PS/2-no-COM-no-LPT boards that are available in small numbers today.

Re:2 serial ATA devices

[Wehesheit]

Is there a limit as to how many controllers can be on board? Could someone for example put 12 controllers on one board with not much trouble?

Re:2 serial ATA devices

[cheezedawg]

I think this chip is just the first step in the transition from PATA to SATA- Intel never goes cold turkey on replacement technologies. It still has both PATA channels, plus the 2 SATA channels. That means you can have up to 6 ATA devices with this chipset.

Re:2 serial ATA devices

[NamShubCMX]

Because you are an exception and will probably buy an additional controller if needed?

Another answer: 3$? :P

Re:2 serial ATA devices

[Strog]

You are just being silly now. You are going to need slots to plug these controllers into. It's hard to find boards with more then 5 or 6 PCI slots. You have to start using PCI to PCI bridges on a board once you get over 3 or 4 slots and performance would suffer. There are some nice server boards out there that might come closer to this.

Maybe you need better controllers like the 3ware serial ata raid cards They can go up to 12 port on a single controller.

Re:2 serial ATA devices

[Kevin Stevens]

most people nowadays only use two drives. One HD, and one DVD/CD Combo drive. People like you and me, can just go out and buy an add in card to run our cool RAID systems and whatnot. Give it time, serial ATA is still pretty new and niche.

Re:2 serial ATA devices

[PCBman!]

I think for the most part the limitation is purely physical-financial.

Since you can impliment any number of things to get at controllers, your limit is how many controllers can you fit onto a die (you get to define this, but larger can mean costlier). Also, without thinking about impedence matching lines, you're also limited in board space for ports.

Re:2 serial ATA devices

[Anonymous Coward]

Actually, you can still only have 4 IDE devices. If you use the SATA ports on ICH5 then you can't use the 2nd IDE port.

Re:2 serial ATA devices

[cheezedawg]

Incorrect. Native mode IDE addressing has been around for a while now, and it is supported by most OS's now, and it is also supported in ICH5. You can have all 6 IDE devices (4 PATA, 2 SATA) with ICH5 without any address or interrupt conflicts.

wow

[DrWhizBang]

does anyone else think it's time to upgrade my k6-2 400?

Re:wow

[Atomizer]

Yeah, you wanna to buy my K6-3 450? ;)

Re:wow

[VelvetHelmet]

That's the same processor I've got and I've been wrestling with this question myself lately.

I don't play the latest and greatest games, but I'm starting to have a hard time finding anything remotely new that I can play.

It seems to be fine for everything else though.

Re:wow

[Anonymous Coward]

Check out some of the deals sites for new Dell computers. I use GotApex. You can get a nice P4/2.4 for $400 or so if you watch these sites. Make sure to get as much RAM as you can afford, though (256MB is a bare minimum; I have 1.5GB in my system.)

Re:wow

[sig cop]

I'll give you $1850.00 for it. Final offer.

Re:wow

[bheerssen]

I don't think upgrade is the right term for what you should do with that.

Seriously though, get a new workstation. The k6-2 would work fine as a file server or gateway or something. Trying to upgrade such an old computer is nearly pointless; better to find a new use for the old one, and get a new desktop system. You can put together a nice usable system for not too much money if you do it yourself

And if you simply don't have a use for an old computer, lots of charitable organizations would love to have it.

Re:wow

[Anonymous Coward]

10 FOR (I=DOOM1; I20 IF SPEEDOFSOFTWARE(I)<ACCEPTABLELEVEL
30 BUYNEWCOMPUTER=TRUE
40 ENDIF
50 ENDFOR
60 IF BUYNEWCOMPUTER THEN
70 DO IT AND STOP WHINING ALREADY
80 ENDIF
90 SLEEP 1 YEAR
100 GOTO 10

Re:wow

[DrWhizBang]

I've given up - I use my kids playstation 2 for games now.

Piece of advice

[Anonymous Coward]

While faster processors are needed by a small group of people the general advice regarding upgrade has to be something like:

If you get at least 3 fold increase in performance upgrade, otherwise wait, because you can't "feel" the upgrade only measure it.

I want cheap SMP, not more MHz

[Gothmolly]

No matter how many MHz you have, broken Java code, lame screen redraws in your browser, compiles set to use "make -j4" and countless other programming adventures can pin the CPU at 100%. I want good, cheap, 2 or 4 way SMP on my desktop. I don't want one app to wait for another, and I don't want to have to wait for any of them. I switched to a dual Celeron board some years ago, and there's really no going back once you've gone duallie.

Re:I want cheap SMP, not more MHz

[Anonymous Coward]

What do you think Hyperthreading is? Cheap SMP. Now go back to your cave.

Re:I want cheap SMP, not more MHz

[Denjiro]

Well the newer P4s with the Hyper-Threading should do a little bit of what you want. Not quite as good as a 2 or 4 way SMP system, but should be significantly better than a tradional single CPU system at multitasking.

Re:I want cheap SMP, not more MHz

[Anonymous Coward]

Yeah once I upgrade to a dual Celeron from my 4 way Itanium2, I think I'll kick you then "go back".

Re:I want cheap SMP, not more MHz

[iiioxx]

there's really no going back once you've gone duallie.

I concur. Right now I have four main boxes in operation (not counting laptops), three of them are dual proc (the fourth is a firewall). Since around 1998, my primary workstation has always been dual proc.

The lack of a reasonably priced DP option has been my primary impediment in upgrading to P4 (two of my boxes are PIII-1GHz, one is PII-350MHz, my uni-proc firewall is Cel-500MHz). I'd love to have a DP P4-3GHz for my primary station, I just don't want to spend $3000 on a motherboard and two processors.

Re:I want cheap SMP, not more MHz

[smg_mrBlonde]

Your right once you have gone duallie you will never go back to monogamy...

Re:I want cheap SMP, not more MHz

[Anonymous Coward]

The lack of a reasonably priced DP option has been my primary impediment in upgrading to P4

Two words:
Hyper. Threading.
OK, so it's not quite the same as "real" DP, but it gives much of the benefit, and cost is effectively $0 (there's no price premium for the HT processors; the 3.06GHz/HT part launched at a price in line with what Intel have always charged for their newest-and-greatest). Given time, HT will be available on all Intel processors, top-to-bottom.
My experience has been that HT makes a big difference. My system is much more responsive now, and rogue tasks don't slug the rest of the system like they used to on my old 2.2GHz system..

Re:I want cheap SMP, not more MHz

[whig]

I think the "make -j4" example is a good example of using all your SMP processors for maximum performance, but not really why I prefer a dual system to a single processor. Frankly, one P4 3GHz is still gonna significantly outperform a dual 1.5GHz Xeon system even for this operation..

I have a dual Celeron 533 MHz at home, and while it's by no means a speed demon, I still prefer it to a single processor system of even 3GHz class.

The reason: No matter how processor intensive a background task I may be running, my computer continues to be smooth and usable. And if it's a long-running task, this is especially important. While it might be nice to be able to run the background job in an hour instead of six, if I cannot use my computer for that hour, I'm actually significantly more inconvenienced.

Yes, I could have two systems and a KVM, instead. But really, SMP is so much less cumbersome. And Intel's Hyperthreading does not provide this benefit, so the next system I'm hoping to get will be a dual Opteron.

Re:I want cheap SMP, not more MHz

[Cthefuture]

There's always AMD with their MP line.

Although I like silent systems and trying to cool one AMD stove is too loud let alone two of them.

The Xeons are great processors but I have never seen a MP Xeon board that had an AGP port so it's no good for what I need (CAD/CAM/3D Modeling).

Re:I want cheap SMP, not more MHz

[jonbrewer]

I want good, cheap, 2 or 4 way SMP on my desktop

I have some nice 4-way Pentium Pro systems I can sell you...

Re:I want cheap SMP, not more MHz

[Anonymous Coward]

The Xeons are great processors but I have never seen a MP Xeon board that had an AGP port so it's no good for what I need (CAD/CAM/3D Modeling).
http://www.penguincomputing.com/store/niveus-400 .p hp
They exist. I'm sure this is not the only example.

Re:I want cheap SMP, not more MHz

[pellaeon]

I think the Intel E7505 supports both :-)

Look here

Re:I want cheap SMP, not more MHz

[Cthefuture]

Ah yeah, I hadn't looked in a long time.

I see Supermicro has a board (same one in the Penguin machine?) with 8x AGP port and dual Xeons. Now if they could just get the Xeons up to 3 Ghz, pop in a Quadro... sweet

Re:I want cheap SMP, not more MHz

[mobets]

hmmm... if the two fans spun at the same speed in the oposite direction, whould they cancel each other out and run almost silent?

Re:I want cheap SMP, not more MHz

[Anonymous Coward]

Why is this necessary? If your OS sets priority right, this shouldn't be an issue...

Re:I want cheap SMP, not more MHz

[Cthefuture]

Yes, the 7505 boards.. I didn't know about them. Cool.

I just built a wishlist machine on Newegg:

- Dual 3.06 Ghz Xeons
- PNY nVidia Quadro 750XGL card
- 2 GB of Crucial (Micron) DDR memory
- IWILL Dual Xeon board
- Antec case

$2800... eat that you Apple losers... LOL :)

Crap that would be a fast machine.

Re:I want cheap SMP, not more MHz

[Anonymous Coward]

- Overpriced, overhyped CPUs
- Overpriced, overhyped video card
- Too much memory
- Iwill? HAHAHAHA
- Antec? HAHAHAHAHAHAHAHA

Eat shit yourself, trailertrash.

Re:I want cheap SMP, not more MHz

[evilviper]

I don't want one app to wait for another, and I don't want to have to wait for any of them.

Stop using Windows then. More advanced operating systems have no problem at all, sharing the CPU.

The hard drive is another story, however.

Re:I want cheap SMP, not more MHz

[aes12]

I hate to burst your bubble, but you really ought to take a trip over to Dell... If you look here you'll see that you can get a dual capable workstation for under $1000!! Now, I don't know what you consider reasonable, but add in another Xeon, and you get a dual proc system for $1100; that seems eminantly reasonable to me. I hate to sound like a Dell ad, but you really have no reason not to upgrade, if all you're waiting for is affordable SMP. Sure, the system listed could use some more RAM, but RAM is cheap. I'd be willing to bet you can add enough stuff to make you happy for less than $1500. Sub-$1000 systems are nice and all, but for a niche like dual proc workstations, $1500 seems pretty cheap to me.

Re:I want cheap SMP, not more MHz

[Gothmolly]

Dumbass, how often do you put "make -j4" in a Makefile on Windows. Obviously I'm using Linux.

Re:I want cheap SMP, not more MHz

[Reziac]

Interesting what you say about background tasks etc. I have the KVM and multiple machines myself, but it sounds like SMP would be a way to get more lifespan out of misc. older CPUs and such ... find a used SMP motherboard, fill it with appropriate CPUs, and you've got a cheap and useful upgrade.

(I'm of the "build it from parts in the closet" school of computing, if you can't tell :)

Re:I want cheap SMP, not more MHz

[evilviper]

Ahem, then you should understand. Linux has no problem sharing the CPU among multiple apps. If you've experienced some sort of problem, it is probably due to some other bottleneck, either disk, bus speed, or possibly simple processor speed. It could also be that the software you have on your SMP system is simply a more stable version than you've used elsewhere. Whatever the problem, it's very safe to say that having SMP does not mean better multitasking (ie. better than an equally fast single processor).

Re:I want cheap SMP, not more MHz

[Moloch666]

Well if you've got at least more than 1 computer you may want to consider setting up a cluster. Given, it won't perform like a true SMP system, it's still cool. For compiling you will probably want to use distcc and then install an openMosix kernel. Of course a 100mb switch is recommended. If you've got SMP systems in the cluster then this will work even better.

I have just a p233 and p3 500 running Gentoo Linux with their openMosix kernel. It's much more bearable compiling on the 233 with help of the 500.

I wish I could put my Athlon 1800+ on the cluster, but I have it running Win2k for games. I don't feel like dealing with Wine or dual booting, maybe one day when I have more time.

Re:I want cheap SMP, not more MHz

[pellaeon]

Suhweet...gimme!

(Although I might beat you to part of that once I get a dual Xeon machine for a thin client server at work :)

Re:I want cheap SMP, not more MHz

[Lershac]

Xeons ARE up to 3GHz

Re:I want cheap SMP, not more MHz

[chez69]

Too much memory? for DOS maybe.

Re:I want cheap SMP, not more MHz

[Anonymous Coward]

"nice -19 emerge package" works great here

Re:I want cheap SMP, not more MHz

[iiioxx]

You've got to be real careful with those Dell prices, as they are very misleading. The front page price is after rebates and minus OS license (despite the fact that they won't sell you a computer without a Microsoft (P)OS on it). Also, there are usually components and applications bundled in that you don't need or want, but can't choose to exlude from the order. Just go into the customize wizard, and you'll see the pricetag start to inflate to reality.

I don't know which system in particular you were looking at (because you didn't provide a link), but I just went to Dell, and the cheapest DP system in the cheapest config (2x 2GHz, 256MB mem, etc) is US$1397 (after rebates). Step up to a DP 3GHz, and you're looking at US$3097 (once again, after rebates). Now, today I was curious, so I went and priced out 2 3GHz Xeons, a high-end Tyan m/b, and 256MB of memory. It rang up to around US$2250 (cash, no rebates). So the Dell is not that well-priced when you consider that they are then charging $847 for a mid-tower case, a 20GB hard drive, a CD-ROM, a floppy, a keyboard, a cheap mouse, and a low-grade video card (about US$225 in components). And you aren't getting a motherboard anywhere near as nice as the Tyan. Just to be fair, I also priced out the same config at 2GHz (the min), and it turns out they don't screw you quite as bad, only charging you about $497 for the rest of the bundle, but that still isn't a bargain at over double the off-the-shelf price.

It gets really pricey when you consider that I already own a case, a floppy, a much nicer keyboard and mouse, a nice DVD/CR-RW, and a hell of a lot nicer video card. So I'm paying US$497-847 for inferior crap that I don't need. Also, given the fact that I am currently running a dual-proc 1GHz machine with 1GB of mem for my primary workstation, upgrading to anything less than the 3GHz solution with 1-2GB of mem doesn't make much sense. And to build a system comparable to what I have now, and upgraded with DP 3GHz + 2GB of mem, using all Dell parts (so I don't have warranty issues later when I call for service) will run about US$4950. That's about double what it would cost to upgrade my existing system and recycle 80% of the parts I already own.

So why again would I want to buy this? Oh, right. Because it's easy as Dell. Just like my Inspiron 7500 that cost me US$4500 and that's averaged a major service every 6 months since I bought it.

Sorry. I've been down that road before.

Re:I want cheap SMP, not more MHz

[aes12]

Correct me if I'm wrong, but wasn't the topic about how over-expensive dual proc systems are? I put up one option for a cheap solution. If you've got a better one, great! I priced out an entry level dual proc system that would be aaffordable to most people. You priced out a really high end system; is there any wonder that yourss costs more? You may not have a compeling need to to upgrade to a 2X2 GHz system, but if you were running 2x400MHz as your main system, it might be more attractive to you.

Re:I want cheap SMP, not more MHz

[iiioxx]

Correct me if I'm wrong, but wasn't the topic about how over-expensive dual proc systems are?

Yes, and that point still stands. Go to Dell's website and price the lowest end dual-proc 2GHz and the lowest end uni-proc 2GHz with identical features. The DP box will cost you $1397 and the UP box will cost you $608. More than doubling the total cost of the machine for simply adding a second processor is "over-expensive" in my book, especially when a 2GHz P4 processor retails for less than $200.

You see, the problem here isn't in the computer, it's in the processor: P4's are uni-proc only, crippled by design. If you want to do SMP, you need Xeon procs, which come at over twice the price. The SMP price hike is due purely to artificial inflation. By the way, Intel isn't the only company guilty of this price fixing. AMD does the same thing with their XP vs MP line, just to a lesser degree at 1.5x the price.

You priced out a really high end system; is there any wonder that yourss costs more? You may not have a compeling need to to upgrade to a 2X2 GHz system, but if you were running 2x400MHz as your main system, it might be more attractive to you.

Go back and look at the post to which you replied. I stated clearly that I had a DP 1GHz and that I would like to go to a DP 3GHz. Your suggested system didn't meet my specs, so I priced the one that did. I could get a dual-proc PI-200MHz for $100. That doesn't mean that SMP is cheap. Compare apples to apples and SMP is way overpriced. The other comparisons I made were simply to demonstrate that for smart shoppers, Dell is not the bargain they claim to be.

Recently discussed on /.

[lateralus]

The possible uses for more computing power was explored recently in an article here. The article I am referring to is Paul Graham's "The Hundred Year Language" and should chill the "my 4.77MHZ XT with a 8087 FPU is all I'll ever need!" crowd a couple of degrees.

Re:Recently discussed on /.

[oingoboingo]

"my 4.77MHZ XT with a 8087 FPU is all I'll ever need!"

Pfft. What the hell do you need floating point for? Java runs just fine on my XT.

Re:Recently discussed on /.

[lateralus]

I just had to put that in because I'll always remember putting the FPU chip on said machine and then running AutoCAD for the first time on my new Amber Hercules display.

SNIFF, SNIFF (dabs at eyes with hanky) now look what you did... I'm all sentimental.

Re:Recently discussed on /.

[oingoboingo]

Don't cry...I have a 68881 FPU card for a Mac LC series sitting right here on my desk. You can run Autocad on that if you want.

im a bit sceptical

[m1chael]

about quad pumped ddr or any ddr ram for that matter. i still only have normal sdram. that said any increase in performace is good for the industry itself since competitor (:P) will want to keep at least a step behind the pace.

Another good summary

[TopShelf]

There's another good review of this in the latest issue of Maximum PC.

Macintosh Processor Speeds

[masq]

This isn't meant as a troll, but I'm sure many of the more... sensitive ... Mac users will take it as such anyway.

This speed boost is great for the x86 world. Speed keeps getting better. Intel and AMD keep competing and leapfrogging each other to greater heights. My sorrow is that Apple's offerings really are *years behind* right now. I know, I know, speed doesn't matter when Macs are slower, but when Macs had the speed advantage, the Mac users claimed speed was all-important and there was no problem attacking the PC users based on their sorry speed. Mac users, like everyone else in the world it seems, aren't objective - if the PowerBook is thinner, they claim size (ahem) is important. When the PC world shows us the Superthin Vaio, we say that Size doesn't matter, it's how you use it (ahem again). And that's the problem; that's why Apple doesn't feel the need to force speed increases out of Moto and IBM to keep up with the Joneses - Mac users are so damn faithful, that they don't apply any market pressure to Apple to force them to compete! Instead, the "Mac Faithful" DEFEND Apple's weaknesses, allowing Apple to slack off in the processor department.

Next time a MacZealot defends Apple's 1 Ghz processors on a slow bus, tell him that he's NOT helping Apple. The way to help Apple is to absolutely demand faster processors, and threaten to switch to x86 if they don't deliver. If we give Apple a "Get out of Jail Free card" with regards to processor speed, we'll NEVER be competitive with Intel.

And yes, I've heard the RUMORS about the IBM chips. They'll still be far behind this, RISC or not.

Re:Macintosh Processor Speeds

[Anonymous Coward]

and yet my 600 MHz iBook is the fastest smoothest laptop I've ever used (and I used a lot) and Apple keeps making the software faster ;-). Having enough/plenty of RAM is more important than CPU speed. :-)

Re:Macintosh Processor Speeds

[Anonymous Coward]

Amazing...

First, why do all Mac users begin their comments with "Well my 700MHz TiBook...." ?? I swear there's like two people writing all the Mac comments--they sound that similar.

Second, the above responses are exactly what the original poster was talking about!! No doubt that these people so desperately want to like Macs that they convince themselves that their machines are fast and that's all they need. I chiefly use PC's, but when I do use Macs, I definitely can tell that they are slower (yes, even the dual 1.25GHz machine, which I've had the "pleasure" to use).

Apple and its fanatics remind me of the story "The Emperor's New Clothes". What will it take for Apple, and its users, to finally wake up?

Re:Macintosh Processor Speeds

[Anonymous Coward]

Umm... Could you please use a few more smilies ;-)

When someone uses three smilies, I know that they're gay. You only used two, so I'm not so sure. :-)

However, you are a Mac user, and you did use one winking smilie.

Yeah, you're definitely gay. 8-P

Re:Macintosh Processor Speeds

[Alomex]

Don't waste you breath. I've been telling this to MacZealots for ages... They have too much invested in the platform (monetarilly, emotionally, and knowledge-wise to be objective about it).

Re:Macintosh Processor Speeds

[Jeff DeMaagd]

And yes, I've heard the RUMORS about the IBM chips. They'll still be far behind this, RISC or not.

I agree that Apple needs to get on the ball with speed, but there are a few uniprocessor RISC systems at 1GHz that can slap around a 2.5 to 3.0GHz equivalent x86 systems. I don't think Apple's systems can do this but rather the workstations from Risc vendors which easily make Apple look cheap.

Re:Macintosh Processor Speeds

[Lars T.]

Well, too bad Intel postponed the processor.

Re:Macintosh Processor Speeds

[platypus]

I agree that Apple needs to get on the ball with speed, but there are a few uniprocessor RISC systems at 1GHz that can slap around a 2.5 to 3.0GHz equivalent x86 systems.

I doubt that. Care to give one example? I know that alphas are quite good at fp, i.e. at 1,1Ghz they are roughly 25% faster than an 3 Ghz P4, but at integer the P4 is roughly 40% faster than the alpha. I wouldn't call that "slap around".
And _if_ apple would go with an such an alpha 21364 (this is the fastest alpha as far as I know, which is not very far), I really don't want to know what that system would cost.

Re:Macintosh Processor Speeds

[Anonymous Coward]

Oh thank god!!!

I guess that means the processor will be delayed a month or two, and in the meantime, Apple will unleash its Pentium-crushing G5 4GHz!!!

Get over it. The 3.06GHz chip that's been out for a while now already kicks the crap out of the G4. Apple's been losing the race for a long time. You act like the posters points are invalidated simply because this particular chip is delay. Typical moron-brained Apple reasoning.

Re:Macintosh Processor Speeds

[Lars T.]

Yeah, the 3.06 GHz chip has been out for a while. And the 3.0 GHz chip will not be out for a while. We are seeing real progress here.

Re:Macintosh Processor Speeds

[GlassHeart]

the "Mac Faithful" DEFEND Apple's weaknesses, allowing Apple to slack off in the processor department.

Apple shipped a 500 MHz G4 tower in September 1999. In July 2000, they shipped a dual 500. In January 2001, a 733 or dual 533. In July 2001, a 867 or dual 800. In February 2002, a 933 or dual 1 GHz. In August 2002, dual 1.25 GHz. In February 2003, dual 1.42 GHz.

If you track the dual processor G4 towers over 18 months, you'll see speed ups of 2.0x (dual 500 to dual 1 GHz), 2.3x (dual 533 to dual 1.25 GHz), 1.8x (dual 800 to dual 1.42 GHz). Comparing the 1999 500 MHz to the 2003 dual 1.42 GHz, it's a 5.68x* improvement in 41 months. This is in the ballpark of Moore's Law**, which cannot be achieved by somebody "slacking off".

However, it is true that Apple is falling behind in raw processor speed. This is something that is important to a core group of users, such as desktop audio, video, and graphics professionals. I just don't think that your worry is reality, because I think Apple is well aware of the problem, even if Jobs has to publicly lie about the performance gap.

* Yes, yes, I know you can't compare a dual processor to a uni-processor system. Discount the multiplier accordingly.
** Yes, yes, I know Moore's Law technically only predicts transistor counts.

Re:Macintosh Processor Speeds

[Zed2K]

I already expressed my frustration with apple hardware (prices and performance) 3 years ago by buildings a windows PC for a small fraction of the price it would have cost me to get a comparable Apple machine. I haven't looked back since and I had been using macs since 1990 or so.

Re:Macintosh Processor Speeds

[Anonymous Coward]

your just bitter slowass

Re:Macintosh Processor Speeds

[Anonymous Coward]

What you don't say is that most mac users buy apple hardware because it can run a nice operating system.

If Apple hardware only ran Windows or linux OSes, nobody would buy them, nobody.

I think it is important to consider that most people who buy macs are doing this because they want to buy a machine (crappy or not) that runs mac os.

You could get the fastest pc in the world, it would still not run macos, which is, especially since the latest releases of osx, a stable, user-friendly, compatible system.

I mean when you buy a mac, maybe you get crappy hardware (though Firewire 800, bluetooth or airport extreme are not that crappy IMHO) or at least crappy CPUs, but you get an OS that runs out of the box nice graphical apps and most apps written for x.

With OSX you also get a nice browser, a nice mp3 app, and some free photo/video editing tools which are also nice.

I own a cybercafé in Paris with 35 PCs and one Mac with OSX and when it is available, most clients choose the mac for surfing because they all say that "the web pages always look better on this comp" - this has to do mostly with font anti-aliasing and color accuracy.

If you want free programs, you can get many quality apps for OSX especially since Apple's new X11 was introduced : Star Office or the Gimlp both work like a charm. You also get all the console based utilities from the unix world, which make the macOS quite compatible now (for example look at all the dvd ripping stuff from the unix world now available : mplayer, mencoder, ffmpeg, cdrdao ...dvd ripping is fast enough on a dual machine now)

Developping under macosx is also very nice, gcc makes really nice code for ppc. Most very simple C programs I compiled (for general cpu optimization) are usually faster on ppc-osx than on x86...

Now the only thing that SUCKS and will continue to suck with macs is 3D gaming. I understand that concern, but personally i'am not too much into gaming, and if I want some gaming i can still buy the latest playstation from sony.

The mac is the only user-friendly hassle-free out of the box system. If you want a clean system that doesn't need virii protection and a huge bug and security-hole awareness and such, you have to get rid of windows and install linux, which is boring, especially when you have to update/reinstall important number of machines.

Now the problem is the speed issue, which is an important one for some power-users. I don't think it is an issue for the average user, or for the graphic/video/audio pro since the stability and user-friendlyness of the machine eliminates any loss of time.

Hopefully this huge speed gap (mostly due to the horrible crappy f*cking Motorola) will shorten in the near future when apple ships its first hardware with IBM CPUs (probably july / august).

It is too early to talk about that but I think things may evolve positively for Apple when these new hardware ship. It's only my own opinion...

G5 support?

[digitalgimpus]

I'm still hoping I can get my 3GHz G5 (TBA tomorrow). ;)

Re:Totally off topic but.

[haroldK]

The "sixpack" refers to beer, not muscle. So, yes, I would say that Joe Sixpack really has a six-pack.

3 GHz Chip Delayed

[lorax]

News.com just updated their article on the chip to state that "a possible problem with the 3GHz Pentium 4, discovered at the last minute, forced the company to delay the chip late on Sunday."

Standard Slashdot reply to new cpu tech

[thomanil]

"Well, nobody needs a zillion Ghz anyway! vi and a compiler on my old 286 should be enough for anybody! Bah!"

g++ is SLOW

[Anonymous Coward]

Optimizing g++ would go a long way to increasing the speed of compiling KDE, and many other large C++ apps like Mozilla... Even just precompiled headers would help a lot!

But why optimize when you can just tell everyone to go out and buy a new faster processor. :)

Re:g++ is SLOW

[be-fan]

g++ is decently fast. ICC is about 50% slower, on my machine, than g++ doing an optimized build. Precompiled headers would help a lot, but wouldn't make things more than twice as fast. C++ is just a very slow language to compile, especially when you throw complex templates at it.

Obligatory workstation zealot post

[Anonymous Coward]

This means nothing. The PeeCee architecture is simply not suited for any real computational tasks. Memory bandwith is limited. System bus speed is limited. Graphics performance is poor. Just the other day I performed a task on my Sun SPARCstation 1 that took only 14 seconds...on my friend's 3.06GHz P4 system it took 14 hours!

Yes my friends...the UNIX workstation is still king! None of this PeeCee rubbish for me!

Re:Obligatory workstation zealot post

[Anonymous Coward]

You're right there. I have this 53 year old sparc server that would kick this things arse in the important tasks. It has 87 SCSI typewritting chimpanzees in a RAID array which would still wipe a couple of new SATA drives off the map. Let me just say this is all due to the massive torque which the server has compared to the high revving PC. I don't mind telling you its exactly like comparing a truck to a race car. Don't ask me how, but mathematically, it is an exactly equivalent equation. So what I am trying to say is that a car will go slower than some types of mopeds. Unless that moped is using a pentium sparcalizer. Don't even get me started on my server.

Re:Obligatory workstation zealot post

[Anonymous Coward]

You've nailed that truck vs. PC analogy. I put diesel in my SPARC IPX...that must be what allows it to get over 100fps in the as yet unreleased Doom III. Carmack develops on IPXs and and SPARCclassics because they both have compression braking, whereas PCs are limited to regular old discs. Yep, those IPXs have some stopping power! But they get through the diesel a bit.

Re:Obligatory workstation zealot post

[TeknoHog]

> Just the other day I performed a task on my Sun SPARCstation 1 that took only 14 seconds...on my friend's 3.06GHz P4 system it took 14 hours!

Yeah, but Linux does infinite loops in five seconds!

I thought the new P4 3GHz w/ 800MHz FSB was....

[SSonnentag]

...delayed. Intel has noticed some anomalies and has decided to hold off shipping for the time being. I guess they'll (Intel) tell us more later.

What would you prefer, no more advancement?

[fbg111]

Honestly, what's the alternative? Stopping with the technology we have because it's good enough? I think not. Even if there is no immediate need for more FLOPS and whatnot on the desktop, it's better to advance the state of technology than not. As Edward Gibbon observed, "All that is human must retrograde if it do not advance."

Re:What would you prefer, no more advancement?

[homer_ca]

There's really two directions for development. On one path lies "more power, more power, argh argh!". There will always be a demand for more power whether it's for gaming, video editing, or some other cool new thing that hasn't been done yet. It won't "make the Internet run faster", but if someone bought a top of the line P4 because of that hype, they'd deserve to be parted from their money.

The other path is for people who just want to do web surfing, word processing, play some music and video, and maybe some light gaming. There's a lot that could be done with quiet, low power, small form factor designs with instant-on capability. You do not want an 80W space heater for this application.

Couple of thoughts.

[Glonoinha]

Someone else (Gothmolly) said 'I want cheap SMP, not more MHz.' earlier on. I thought about his reasoning for a while and have to suggest at least considering RAIC as a way to get cheap SMP. Redundant Array of Inexpensive Computers - similar in thought to the original purpose of RAID (RAI Drives) - use a few cheap pieces to get the same or better performance of one wickedly expensive piece of hardware.

Today, if you were to search around a little to scrounge up one of the 15% coupons floating around on the net (www.fatwallet.com for example) you could go to Dell and have a fully loaded system (Celeron 1.8GHz, 128M DDR266, 40G IDE, 8M onboard video, Intel Gigahertz NIC, 48xCD, keyboard and mouse) delivered to your house for $240 after rebate ($100 rebate but they are pretty good about paying them.) I think you can upgrade the hard drive to dual 80G drives (buy one get one free if you upgrade) for another $60, bringing the total price to $300. Add a $50 two port KVM (I use the Linksys, has build in cables) to your existing rig and now your monitor, keyboard and mouse can instantly switch between the two systems. Have a massive process that hogs the CPU, swap over to the other machine to do whatever you want while it runs. I have been doing this for a while and the ONLY drawback I have seen so far is not being able to cut and paste from one to the other. Other than that they are effectively one machine with two discrete workspaces.

As for the new hardware ... nice. Honestly though I am way more excited about the SATA/RAID 0 performance than I am the additional CPU horsepower. Sustained serial reads of 96MB/s and sustained serial reads of 86MB/s - not burst but sustained -:- DAMN. That is easily twice the performance of my current rig, possibly three times the performance. I care less about a CPU running 7% faster; I wouldn't even notice the extra 40fps on Quake3Arena going from 410fps to 450fps but the ability to move data back and forth to the hard drive three times as fast is going to make this one machine worth upgrading to.

IMHO the advances in hard drive performance are the real story here. Running the P4/3G on a 400FSB vs the old 333FSB is nothing compared to getting 3x the performance from the drive subsystem.

Re:Couple of thoughts.

[Anonymous Coward]

synergy might help you with the cut and paste problem

Re:Couple of thoughts.

[Cthefuture]

Multiple machines generate even more heat and noise. I hate noise. All those power supplies... what a waste.

And there's the convenience factor of having all your processes in one place. It's not easy to just switch to another machine. Tou have to restart all the things you were working on. I mean, I often have lots of open editor sessions, VMware sessions, and other applications running. Sorry, but multiple machines is in no way as useful as a real multi processor machine.

Re:Couple of thoughts.

[jeffy210]

"...as a way to get cheap SMP. Redundant Array of Inexpensive Computers."

He just came up with a new way to say "How about a beowulf cluster..."

Re:Couple of thoughts.

[Anonymous Coward]

That is the coolest thing I have ever seen. Mod Parent up +1

Re:Couple of thoughts.

[Anonymous Coward]

Mod this up - I just checked it out and synergy is SUPER COOL.

Re:Couple of thoughts.

[evilviper]

you could go to Dell and have a fully loaded system (Celeron 1.8GHz, 128M DDR266, 40G IDE, 8M onboard video, Intel Gigahertz NIC, 48xCD, keyboard and mouse) delivered to your house for $240 after rebate ($100 rebate but they are pretty good about paying them.) I think you can upgrade the hard drive to dual 80G drives (buy one get one free if you upgrade) for another $60, bringing the total price to $300.

Any chance of a URL? Dell's site is perhaps the least negotible site on the internet. No doubt it's partly because they don't want you to find the cheap systems, but still...

Re:Couple of thoughts.

[ivan256]

Hate to break it to you, but that price on the dell website already has the rebate deducted from it, AND it doesn't include all the features listed. You have to pay more to get all that stuff (Specifically, you don't get the GigE). After all that you'd have a slowass 1.8Ghz celeron. I'd rather buy a 1Ghz P3 off of eBay for less to get the same performance, or build my own athlon 2500+ or a real P4 2.2Ghz for the same price. You're also not going to get 410fps in quake from the dell setup. That intel video chip will give you more like 15fps.

Actually, though, I just picked up a dual Athlon XP 1800+ for $350 (MB, two retail boxed processors, 80Gb HDD, and GeForce 4 Ti4200, Mid-tower case. Re-used my old Keyboard and mouse. Much faster than my old dual Pentium (1) 233Mhz machine.), so I don't think you need to deal with having two machines if you really want SMP and money is what's holding you back.

Re:Couple of thoughts.

[Glonoinha]

http://configure.us.dell.com/dellstore/config.aspx ?CS=04&kc=6W300&oc=pe600sc

The base system is the PowerEdge600sc
1.8GHz Celeron
128M DDR266 RAM
Choice to double the RAM free or get a $100 Mail in Rebate
40G IDE 7200rpm HD
Crap server video (8M ATI Rage)
GigE NIC
48x CD-ROM
kb/mouse
1 year warranty.

Base price $399. Free shipping.
I found a 15% coupon at www.fatwallet.com (browse the forums, hot deals, takes some looking some times) brings that down to $340. It is probably worth paying the $70 to upgrade the first drive to 80G and get the second one free, bringing the system right back to $400. I also recommend the drive upgrade because the Seagate 40G drive I got in one system was way slower than the Western Digital 80G drives I got in another system (your mileage may vary.)

Get the $100 rebate and you are looking at $300 for the system (plus sales tax on $400,) no monitor.

Note - this system doesn't have AGP ports and uses the new 3.3v PCI slots so most of the cards out there are not going to work. It isn't a gamer box and is going to suck for first person shooters (although I have heard of people putting PCI GeForce4mx boards in them and enjoying them.) It is a bit on the loud side because they built it to server (not desktop) spec's. For doing real work on the cheap, however, as a single system or hooked up to a few others in a RAIC, it is a pretty solid machine.

Granted you can throw more cash into it to make it better ... upgrade to the P4/2.4GHz = +$100. Add more RAM. CDRW. Etc... But that wasn't my intent - I have a machine with a GeForce, CDRW, loads of RAM, a caching RAID SCSI card, etc... - I wanted a cheap box to run processes on, a discrete workspace completely alone that is unhampered by whatever is running on my other machine(s). And for $250 - $300 this one fits the spot.

Get creative and you could build a 4 box Beowulf cluster for under a thousand dollars (after rebates) with GigE networking (well you would need to come up with a GigE switch, and a monitor.)

Given that when I was in college I saved for a year to come up with the $1,600 I paid for a 386sx-16 with 2M / 40M (MFM) and a 14" VGA monitor ... I would have LOVED to get my hands on something like this.

Re:Couple of thoughts.

[evilviper]

Hmm, well, the problem for me would be the lack of an AGP slot. If I wasn't looking for AGP, I would get Walmart's $200 1.1GHz Duron.

Tax in CA is quite steep as well, so I try to buy out-of-state when possible, but to get a brand-name PC it would have been worth the extra $$$.

Incidently, I find it very funy that just about all on-line shopping sites provide minimal specs on their system... It's the only source of info available, I'd think they'd want to make it as complete as possible, for those that want to know.

Tide goes in, tide goes out

[HiroProtagonist]

hardware gets faster, bus gets faster. Tide goes in, tide goes out.

"Bed Goes up, Bed goes down.
Bed Goes up, Bed goes down. Weeehee!" - Homer Simpson

Canterwood chips

[nigel.selke]

AMD started naming their cores after horses (Palomino, Thoroughbred, etc). Now Intel has brought out the Canterwood. Is the Gallopwood next? :)

Re:Canterwood chips

[Anonymous Coward]

Next line would be the Peckerwood chips

Re:Canterwood chips

[Anonymous Coward]

Intel CPU codenames are after places in Oregon. Intel chipset codenames are after places in northern California.

Well you...

[Anonymous Coward]

... thought wrong.

400MHz FSB on Athlons is trivial

[Brian Stretch]

I've reclocked my Athlon 1700+ (TBred core) to 8x202MHz (404MHz DDR) on my ASUS nForce2 chipset board, using a single Corsair PC3200C2 DIMM (yes, two DIMMs would be better, but they were too expensive at the time). It's just a matter of selecting the right BIOS settings. I left the voltage levels at their defaults. MemTest86 verifies that the memory is stable at that speed. Red Hat Linux runs until I need to reboot for the usual kernel/glibc upgrades. I went this this approach because I wanted to optimize the performence/power consumption balance, what with the machine running 24x7 and all.

Of course, tweaking speeds like this is not guaranteed to work, yadda yadda, but it generally does if you built your system right.

If you want serious firepower, build a dual Athlon box, which should cost no more than the uniprocessor P4 being reviewed. time make reports a bit over 9 minutes when building Wine with MAKEFLAGS=-j2 on my dual 2400+ (not overclocked). Nice, especially when you forget the --with-nptl switch the first time around (d'oh!).

Of course, next week, the Opterons ship, starting with Opteron DP 240's and 242's. It's unclear whether there will be cheap workstation motherboards available right away or just the seriously nice (and expensive) Newisys-designed 1U rackmount servers. It appears that AMD is going to use the Opterons to slap the high-end P4's around, saving the Athlon 64 until they want a low-to-midrange 64-bit desktop platform. I'm surprised the various hardware site reviewers haven't picked up on this.

Re:400MHz FSB on Athlons is trivial

[gamorck]

If you have a dual box I would recommend using "-j3" instead of "-j2". Typically "-j2" is the default for single CPU systems. Having make create one more thread than the number of processors you have allows you to make more efficient use of processor power. For example on my Dual Xeon with Hyperthreading turned on (looks like 4 cpus to linux and windows) I use "-j5" for maximum processor usage during the compilation process. When I turn off hyperthreading I typically use "-j3". I've found that these settings work very very well especially on Gentoo where you spend a lot of time compiling.

Cheers,

J

Re:400MHz FSB on Athlons is trivial

[siliconwafer]

Why would 2 DIMM's be better than one?

Re:400MHz FSB on Athlons is trivial

[GroovBird]

Because the nForce2 chipset also has a Dual Channel memory bus like the chipset mentioned in the article.

It's one of the reasons why the nForce2 chipset is the fastest chipset for Athlon based motherboards at the present time.

Dave

Re:400MHz FSB on Athlons is trivial

[Anonymous Coward]

LOL... are you related to the iraqi information minister by any chance?

Re:400MHz FSB on Athlons is trivial

[Moloch666]

I've noticed something interesting since I've moved from FreeBSD to Gentoo. FreeBSD recommends make -j4 for single processor systems, but the Linux world recommends make -j( number of processors + 1). FreeBSD's resoning for a higher number is that your processor would spend more time working and less time waiting on the hard drive. Do you have any insight here?

Re:400MHz FSB on Athlons is trivial

[ivan256]

You should try '-j'. It'll be damned close to the fastest your system can do without having to experiment with wether 3,4,5,6... or whatever is the fastest. You'd think that all those threads spawning all at once would be detrimental, but with more recent kernels it hardly slows it at all. You're system will be fairly unresponsive while it's compiling, but it'll be over with quicker.

Re:400MHz FSB on Athlons is trivial

[Anonymous Coward]

The best way to reduce time spent compiling is to use a binary distribution. You're just wasting electricity when everybody has to build instead of just one person.

Re:400MHz FSB on Athlons is trivial

[Anonymous Coward]

He might be right -- but AMD can't ship 400Mhz Athlons because it would make the Hammers (with their integrated 333Mhz memory controller) look bad.

Re:400MHz FSB on Athlons is trivial

[Darren Winsper]

The problem with upping the -j count beyond (number of procs + 1) is you're going to get more and more overhead from caches being flushed and general overheads of context switching.

Terminator II Extreme Edition in HD

[OG]

The newest edition of the T2 DVD will include an HD transfer of the movie that can be run with WMP9. I tried to run a sample HD clip ("Step Into Liquid") on my computer (Athlon XP 1600 w/ 512 MB RAM)--I got a nice slide show of still frames. It ran a little better on my laptop. MS's website recommends 2.4 GHz minimum, and I can see why. T2 will be higher rez than that clip, so I'd expect you'd need something even faster.

HDTV can also benefit, as new tuners like the Fusion HDTV card are inexpensive but have software-only decoding, putting a good strain on the CPU. I want one of these new chips. For the lust factor? Nope. There are applications I'm interested in that will actually benefit from the higher speed.

Change for the sake of change ...

[fygment]

... every article I've seen to help you speed up your PC begins with the directive to "increase your memory" before all else (and at all levels memory is the bottleneck). Is it really cheaper to come out with a whole new CPU than to make sure you have squeezed maximum performance out of the last batch? Why not increase onboard cache size so you could swap out the old 1 Ghz with a new 1 Ghz that would have better performance but wouldn't need a chipset/mobo upgrade?

Re:Change for the sake of change ...

[malIgna]

Intel is a processor company. Well, they make nics and so on, but they really aren't in the business of pushing memory.

Re:Change for the sake of change ...

[Kevin Stevens]

One thing I would really like to see is better use of RAM. With power user's systems frequently containing 512mb-2GB of ram, cant we re-optimize our computing architecture to take advantage of this and store more programs in ram? That way when I start up IE ten times a day every_day, it will come up blazingly fast? Same with word, dev tools, etc... yeah there are Ram disk programs out there, but how about we make this a more fundamental part of the OS?

Re:Change for the sake of change ...

[TheDauthi]

I believe that the real reason Intel doesn't do this is that (most) people think that more Ghz means that the computer is "faster". Techie people know that it's actually a combination of factors, and that speed isn't necessarily performance. You can have a 40 Ghz machine, but it won't do you any good if it can't access main ram faster than 3 hz, or read from a hard drive more than a byte a minute.

Re:Change for the sake of change ...

[fygment]

Processors have onboard memory ... as in on the CPU. It's called cache and there are usually several levels. It's where instructions or data can be kept when needed repetitively instead of repeatedly accessing RAM. Because the memory is onboard the CPU, ie. it is in close physical proximity, it is thus many, many times faster than RAM. More cache is, if managed correctly, better. For a nice simple view see

• http://computer.howstuffworks.com/computer-memor y3.htm

Intel makes CPU's => Intel makes caches. I proposed they make larger caches, etc.

Tom's Hardware article

[panurge]

I read the article (pause for shock)
All that stuff about huge speed increases and sackloads of extra memory bandwidth, reduced clock cycles, RAID...but when you eventually get to the performance testing it seems very little faster than top end current boards. Perhaps if you have a daily compute-intensive job that is slowly growing and currently takes 23hours, you would get excited, but as a developer I guess I might gain a few minutes off my build times (and that's staring into space thinking time anyway.)

I'm not knocking progress: the lower voltage and the ability to use a 4-layer board, plus the serial ATA on-board support look nice, but the number of people with more money than sense needed to get a fast R&D payback isn't that high at the moment.

Or is this a cunning plan to make money through selling compute farms to rogue states that have just decided they need WMDs really fast?

Re:Tom's Hardware article

[Anonymous Coward]

You must not be working on a very large software project. Try compiling gigabytes of source code. You'll want the fastest computer on the planet for that.

When I'm waiting for a compile, I don't stare off into space. I'm usually working on another problem.

What's with the Luddite Coup?

[bill_mcgonigle]

Everytime there's an article about a new faster piece of tech on Slashdot lately there are a dozen comments that get modded up right away about 'Why do we need faster computers anyway?'.

Look, if 640K is good enough for you guys, fine, but let's stop the whining. The rest of us like to do things like compile code and compress video.

Re:What's with the Luddite Coup?

[Anonymous Coward]

You just like dicksize competitions.

'Fess up, dude.

Get a Life

[Detritus]

I have a dual processor 450 MHz G4 that runs on a "slow bus". It is more than fast enough for the average user. So it doesn't run at 3 Ghz. Do I care? No. Am I going to write a letter to Apple, complaining that they are losing the dicksize war? No.

The computer runs my applications without noticable delay, is "user friendly", doesn't crash, and requires minimal maintenance. I'm happy with it.

Re:Get a Life

[Anonymous Coward]

LOL, exactly the response the orignal poster was talking about. When the Mac can't complete in a certain area Mac users say it's unimportant. But a few years ago when Macs were beating PCs on performance that's all we ever heard.

Re:Get a Life

[be-fan]

Well shucks. If you do nothing with your computer, why do you assume nobody else does either? I, for one, would like to get a Mac. I have no ties to stupid Windows software, and could switch on a whim. I could even run Linux on it :) But a Mac is too slow for the work I do, and too expensive for the poor performance you get. I'm sure there are a lot of other people in my position (in relation to Apple's small market share), and Apple is losing a lot of potential users by not being competitive in the speed department.

Re:Get a Life

[siliconwafer]

I have a Mac too. Like yours, mine is fast enough for the average user.

So it doesn't run at 3GHz? Do *I* care? No.

Do people buying Apple's new line of servers care? I'm sure they certainly do, and will in the future.

Although we might not have a personal use for faster Apple's, if Apple wants to stay competitive in the server market, they will probably need to bump up the speed a bit.

Re:Get a Life

[Anonymous Coward]

Yes, and my dad is perfectly happy with his Dell Pentium 90 with 64 megs of RAM. That's not the point. The point is that Apple's new offerings are not competitive.

And it's only a "dicksize war" if you're losing...

Re:Get a Life

[Anonymous Coward]

we're not losing . macs are faster than pcs. we just like our mouse on slower sppeds so it looks slower

Chipset

[devnullkac]

They've also released "Canterwood", the chipset chipset for the P4...

I wonder... How many chips could a chipset set if a chipset could set chips?

Re:Chipset

[theendlessnow]

I wonder... How many chips could a chipset set if a chipset could set chips?

Not very much I bet, if a chipset could set chips.

Whine, whine

[Sycraft-fu]

"But my 8088 scrolls text just fine! Why would anyone want something more powerful?"

Because there is neat new shit that takes more power. Receant example: HD multimedia. The Windows Media 9 HD demos kill a P4 1.6 and really take at least a P4 2.4 to play well. Means that if you have a 2.4, which is pretty good these days, that is about ALL your system can do, not much room for bacground tasks.

Or how about speech recognition? There are some nifty new technologies in speech recognition the integrate it better with text parsing for far more accurate recoginition. One problem: they take loads more power than normal speech recog, which takes a bit itself. Given that ideally this should happen in the background as a normal part of the OS, more power become critical.

Or how about better game AI? I am so sick of 3d bots that get "good" by becomming aimbots or RTS AI that attacks you in teh same predictable way every time. I want smarter AI. Well, to do that it is going to take more processing power. Teh smarter the AI, the more CPU time it needs. all this while still doing all the other calcuilations a game needs (like physics and game logic).

We are not even close to expending the need for more computer power. As power grows, we'll simply find new and creative way to use it that were not before possable.

After all, my 8088 scrolled text like a champ, but I much prefer my P4.

Spreadsheet software: 27,520 bytes

[Futurepower(R)]

It amazes me, but the first spreadsheet software, VisiCalc, was 27,520 bytes.

Re:Spreadsheet software: 27,520 bytes

[sketerpot]

The people who made VisiCalc had to worry about fitting the program on tiny computers; people making spreadsheets now don't, and we can use lots of graphics, features, higher level languages, maybe embed python/tcl/scheme/whatever, and generally have a lot of fun. The VisiCalc programmers actually thought about how much memory the character screen would take.

Go Intel!!!

[Anonymous Coward]

I find it really funny how the people of this forum react to the release of a new proceesor and mobo architecture. When AMD releases something new, some of the people of this forum would like to declare a national holiday, but when Intel releases something obviously superior in quality and processing capability (possibly revolutionizing the industry), all you people do is bitch and moan about the lack of necessity for something that is so powerful and expensive. I've got news for you, by the time AMD comes out with something that can hold its own with this new Intel release, the Intel version will be half the price of the AMD. So as far as I see it, just be happy that technology is advancing, even if it is "THE MAN" (Intel) that is making it happen

Great! Now there's a MHz race on the FSB!

[Anonymous Coward]

Thankfully the AMD Hammer chips will not use a FSB so Intel can race itself instead.

New Chipset, but Intel just pulled the Processor..

[micron]

Intel found a last minute bug on Sunday night, and will not relase a processor that can take advantage this chipset until further notice!

http://news.com.com/2100-1006-996611.html?tag=fd _t op

Noise. Bigger, Faster, Louder.

[Glonoinha]

Ok I didn't mention the noise from all the fans and drives. I didn't say it was a perfect solution, merely a cheap solution. I have to admit that if I need to concentrate for some serious coding I generally pop in ear plugs (the yellow squishy ones you use when shooting guns.)

As for easy to switch between machines, all I do is hit the Scroll Lock key twice and my screen flashes. I have the desktops of both machines pretty much configured the same but the wallpaper is different so I can differentiate between them.

They have become (in my mind) two discrete workspaces of the same machine. If you view 'the computer' as the keyboard/mouse/monitor combo used to interface the machine then adding another machine behind that interface merely adds another dimension to your existing computer.

Granted you can't simply shift the same session you are working on between machines, that isn't what this is about. It is about division of labor. A multi-processor system isn't going to allow you to continue to surf the web or edit code during a reboot and a blue screen pretty much brings all processes to a stop; the SMP box will still be disk bound (sluggish) during very disk intensive activities (ie system backups or DivX ripping / encoding / burning sessions, or whatever ties up your disk subsystem.) A single CPU is dedicated to a single process (even though that process may involve many programs, editor sessions, and apps running) but rather than shared resources (besides the CPU) each CPU gets dedicated RAM, disk I/O, network connectivity, video, CD-ROM, etc ... - all piped through to the same keyboard/mouse/monitor.

It also serves way more useful when creating client / server applications that need to take into account network access, or apps that have multiple simultaneous users accessing the system, particularly multiple users with different levels of access.

Each has its pros and cons, I was just saying that with the exception of single task operations (ie. FPS games, PGP cracking, DVD ripping, etc...) your best bang for the buck is multiple machines. For the cost of a single SMP Dual Xeon 3.06GHz machine (?? $3k) you could set up a dozen of the machines I described earlier. Which is a faster rig? Depends on what you are doing. The Dualie is going to rip and compress a single DVD to .AVI way faster than a 1.8GHz Celeron ... but if I needed to rip and compress a dozen different DVDs ... I would put my money on the Celery farm (no pun intended.) The SMP box is going to show way more fps in a game like Unreal2, but if I had to render 300,000 frames of graphics for the next Pixar flick I would much rather have the farm of mid range machines.

For your particular needs it doesn't sound like the solution, but you did consider it - which was my goal.

One step back

[presroi]

In this heise-article from 15 minutes ago, Intel has withdrawn the release of the new CPU. A google-crappy translation can be found here.

Re:One step back

[Vlad_the_Inhaler]

YUP, I saw that as well. There is nothing wrong with the Motherboard / Chipset, it is the processor itself which sometimes misbehaves.

Intel P4 is delayed...

[dtjohnson]

This ZDNet story
says that Intel has stopped shipping the 3 Gig P4 with the 800 mhz FSB due
to "anomalies."

Not quite a Beowulf

[Glonoinha]

Actually very close, but different. With a Beowulf all the machines actually are working as a single coherent entity, and are throwing all of their processing power at a single process - hopefully a process coded to take advantage of the distributed processing of the cluster. Adding more machines will make the process faster (in theory.)

With RAIC each machine is doing its own thing although they generally all have identical drive mappings to the same places so they can share files back and forth and regardless of which machine is running a program they can all be configured the same way. Adding more machines will not speed up any one process, it just lets you run more (different) processes at the same time.

Compare a single uberSMP box (dual HT 3.06GHz CPUs) with two mid range regular (single 2.4GHz CPU) machines. The uberBox is going to run Q3A, UT2003, and EverQuest way faster, zero video lag, etc... than the 2.4GHz box and adding boxes to the cluster simply can't help the performance in this case. But what if you wanted to bring up a map, or a web page with all the console commands, or Google for instructions or whatever ... no matter how fast your SMP box is you are pretty well screwed (have to close your game to do those things) but you can easily do it on one of the other 2.4GHz boxes by double tapping the scroll lock key, swapping back and forth between the two.

I would eagerly Beowulf the RAIC if I could only figure out how to actually apply that power to the stuff I am running ... but nothing I run benefits from a SMP box, much less even knows how to cooperate with a Beowulf cluster.

Too general purpose for where we need speed.

[Junks Jerzey]

I see the usual replies claiming that 1GHz is more than fast enough and the usual critical replies citing video editing and speed recognition. Now while I can agree that more speed is often good, are these really examples where it makes a lot of sense to buy hot, expensive, and power hungry general purpose CPUs to handle special purpose tasks? 3D graphics didn't wait for processors to hit 20GHz, but a 300MHz graphics processor can outrun any general purpose CPU. Video editing is another good example. Why is it so slow? Because it involves compressing lots of data. A team of graduate students could create an FPGA that runs rings around a P4 for video compression (by, say, a factor of 20). Speech recognition is the same way.

In short, paying $1000+ for a processor that's 9% faster and uses 15% more power is not a good solution for "I need more power for video editing," especially when you should be able to get 20x-50x performance increases for 10% of the cost.

Re:Too general purpose for where we need speed.

[Anonymous Coward]

I guess that explains all those chips out there for video compression and speech recognition...moron.

Video compression and especially speech recognition are constantly evolving technologies. Just look at how many /. articles discuss the latest algorithms. The HD-DVD using Microsoft compression tech was the most recent. But strides are continuing to be made in DIVX and MPEG4. Ditto for speech recognition.

The bottom line is, we are using general purpose processors for these types of applications because once the code is hardwired into silicon, it can't be changed. And FPGA's are still too expensive.

The one is the codec chips found in DVD players. But everytime a new format comes around, we have to buy new players. That's a real great solution.

I'll stick with my fast general purpose processors, thank you very much.

Re:Too general purpose for where we need speed.

[Junks Jerzey]

I guess that explains all those chips out there for video compression and speech recognition...moron.

Look before you post. There are lots of video compression chips out there. The market for them as PC add-ons is deemed high-end at the moment, but I don't expect that to last for long.

The bottom line is, we are using general purpose processors for these types of applications because once the code is hardwired into silicon, it can't be changed. And FPGA's are still too expensive.

FPGAs are cheaper than the alternative. Look, we're talking about people who claim that paying $1000 for a new processor that gives them a 9% speed increase is cost effective.

81.8 watts

[bromoseltzer]

According to Intel, the 3 GHz P4 dissipates 81.8 watts. While that keeps you warm over a long winter night, it's a lot of power to run all the time. I guess it's about $72 a year for energy. You have to put up with noisy fans, dust bunnies in your box, and all that.

I'd like to see speed/power specs advertised, and not just for laptops.

-mse

Re:81.8 watts

[slittle]

That'd be peak output, ie. when your CPU is at 100% that's what it's using.

So stop running shite like SETI/RC5, and it'll stay cool(er).

Reason normal people need SMP

[rsborg]

The reason: No matter how processor intensive a background task I may be running, my computer continues to be smooth and usable. And if it's a long-running task, this is especially important. While it might be nice to be able to run the background job in an hour instead of six, if I cannot use my computer for that hour, I'm actually significantly more inconvenienced.

To add to the list of inconveniences: Shitty programs that for whatever reason, take 100% regardless of their usage (in windows). I mean, lots of games, encoders, oldy-but-goody 16 bit programs, you name it. If people really threaded what needed to be threaded and coded properly, then I would not need a 2nd core.

Another reason is IDE. Sure, it's fast enough, but it's pretty dumb compared to SCSI's capabilities of offloading processing and data transfer. If all desktops had good SCSI connections, massive processor speed would be less of a need.... I still love my 8x SCSI CD-RW :-)

well that's all great but,

[ellem]

will my floppy drive still be really slow?

Processor postponed

[Lars T.]

Says Heise (in German).

A small number of processors had "unexpected results" in FSB800 mode, that Intel has to check out more carefuly now.

In other news...

[Anonymous Coward]

Apple execs comment on the recent chipset from Intel, their fastest to date: "We at Apple feel that computing experience is more important than raw computing speed. Therfore, we have decided that 1.42GHz is fastest that G4 processors ever need to be. In fact, we think 1.42GHz is too fast. Therefore, we will be re-introducing our 600 and 700Mhz lines. Thank you."

MY PORN STILL RUNS ON A 386

[Anonymous Coward]

It's mostly 160x100 avi's.

Oh, and it's a 386DX, not SX. Runs at 33Mhz smooth as shit.

Windows 3.1 boots in a mere 19 seconds.

Re:MY PORN STILL RUNS ON A 386

[AceM2]

I'm sorry to hear that you didn't get modded up for being funny, you seem to need the attention.

MOD PARENT UP

[Anonymous Coward]

While I don't agree with the poster's first point (by and large the zealots' postings are original), I love the last bit about "The Emperor's New Clothes". Apple users should sit and up and think about that.

More info here

[bstadil]

Here is some more information on the problem TheInquirer

Quote from Intel:

"Due to recent analysis, and given our commitment to quality, Intel will be placing the Pentium 4 processor at 3.00 GHz with an 800MHz bus on ship hold temporarily. In the course of our final testing in our validation lab environment, we have observed an anomaly on a very small number of the 800MHz bus processors. We are working to understand and resolve the issue and we hope to ship this new processor as soon as possible."

Re:w00t. p4 3GHZ

[Anonymous Coward]

odd, PII's topped off at 400Mhz. You don't have to lie to kick it. Loser.

Re:Noise. Bigger, Faster, Louder.

[jmacleod9975]

My setup at work is exactly what you describe. It works great for me. I think you really have a good idea, and you get out much cheaper. Also, if you have dual boot on the machines you can be rebooting one while using the other. It really is terrific.

cheap SMP = mosix.

[Anonymous Coward]

collection of linux boxens: check
add compiled kernel with mosix patch to grub: check
remote display pointing to system with monitor, kb & mouse: check
reboot.

Pentiums

[Anonymous Coward]

"Are you a code cracker hacker leet dude with porn?..it's all about the...pentiums?"

"You can just turn the other cheeck, Dare to be stupid. Yeah it's ok it's like saying: You gata buy one if you wana get one free."

They suck don't use them AMD's are compatible and better. So are macs.

Question about multiple drives...

[Anonymous Coward]

A total newbie question, mostly cause I never even tried it...

But could you buy an extra IDE PCI card to supplement the existing on-board IDE controller?
Or would the card "take over" that functionality from the motherboard, like an extra video card "overrides" the on-board video?

Would Linux (Redhat) see all 8 HDs in that case?
Anybody out there tried it?

Re:w00t. p4 3GHZ

[Anonymous Coward]

Nope, he's right, the P2 topped out at 450MHz.

http://www.intel.com/design/PentiumII/datashts/243 657.htm

MOD PARENT UP! COOL SOFTWARE!

[orim]

I join my esteemed colleagues in their evaluation of this software: it f***ing kicks ass :)

Who cares how fast the Mac is?

[Anonymous Coward]

Ask the important question: What color is it?

The Beginning of the End

[endeavour31]

See Ya - AMD!

Instead of ever-"faster" processors, I'd rather ..

[Niet3sche]

... that we took this energy and put it back into writing really small, tight, compact code. It seems to me that we've become too dependent upon slopping software together in this day and age that the artistry, mastery, and elegance of slinging code is long-gone from most code that a consumer will see.

Instead of touting a new 3Ghz CPU with an 800Mhz FSB ... why not actually look at it from a pipeline / instruction setup: "oh yeah? Well *my* code can pull off the equivalent of 12 operations per second with *no* pipeline bubbles, AND shadow the RAM" (just like the C-64 was shadowed to get the "128K" later on in the product cycle.

Crap, it's another case of, "born 15-20 years too late". Last week (or was it 2 weeks ago?) started with an idea of mine (that I was coding up to get a patent for) being slashdotted ... twice ... by someone else ... and this week looks no better.

*sigh*

PS: Anyone know of a FREE C or C-like PIC chip programming software? Yeah, OT-away. :|

[OT] Sig

[Anonymous Coward]

That's because 1/2 == 0.
Modify your expression:

float x = 3.14159; float y = 1.0/2.0 * x;

D'oh

[Glonoinha]

Redid the link.

I forgot about the dreaded space in the middle of the url 'feature'.

Intel stops shipping new P4....

[Anonymous Coward]

Aparently they've already halted shipments...

Here's the link.

Questions about your SMP

[Glonoinha]

I have had the opportunity to use SMP machines in production but that involved little more than moving my programs there and letting them run ... no day to day experience and I am curious :

I know what throttles a single CPU box - swapping to disk when RAMbound, a full page of flash animations, a disk subsystem that is too busy moving massive files around to acknowledge my petty http requests, but what have you found that can throttle your SMP Athlon machine?

As for 410fps you are right, that was a server with crap video - not a particularly good choice for games. It does come with the GigE NIC though, built right onto the mobo. I also agree about the Celeron chip, when I ordered mine they were offering free CPU upgrades so my machine came with a P4/2.4GHz.

Re:Questions about your SMP

[ivan256]

but what have you found that can throttle your SMP Athlon machine?

Disk intensive processes will still slow down on the SMP machine, however the machine will still be responsive during heavy I/O unlike a uni-processor box. I can have a java vm start eating an entire CPU and not notice until I happen to run 'top' for something else. The process that's waiting for the disk will still lag just as much on an SMP machine unless the interface runs in a seperate thread as the I/O, and that's rare. Other than that I can't do much of a comparison for you. (Not trying to brag) Because I do kernel work that all has to be SMP safe, all the machines I use on a daily basis are SMP. I've got a dual P3 800, a Dual G4 450, and a dual 866Mhz alpha at work, and my dual Athlon at home... The only uniprocessor machine I still use is my powerbook, and I basically only use it as an X terminal for my other boxes, so I never really get a chance to notice the lack of a second processor.

Actaully...We have cheap SMP...

[Fallen Kell]

Its called Hyper Threading. This is why all CPU's with HT are actually detected as 2 CPU's by Windows and Linux alike. It is not completely 2 CPU's, but the execution of instructions in the pipeline can be "bypassed" in a sense by allowing other instruction threads to execute while other threads are waiting on things like I/O.

Also, there was no performance increase in the CPU (well sort of, but not really). Just that the bus speed was boosted. This only changes the speed at which data and instructions can be sent to the CPU, but the CPU itself is still limited to the same amout of theoritical instructions per second as it was origionally (in fact, its a little slower 3.0 GHz vs 3.06 GHz). The increase in bandwidth will really only affect operations which are data heavy (wasting CPU cycles while waiting for the data). Anything that was CPU cycle dependent will actually perform slightly worse on the new 3.0 Ghz chip compaired to the 3.06 GHz (since we lose 60 cycles per second, that is 60 more operations that it can not complete).

A real SATA implementation is one of the real benefits to the Canterwood systems. Tests are showing RAID0 systems using SATA are scaling at almost 70-85% performance of each disk in the raid volume. Non-raided SATA disks are also benefitting from the south bridge integrated controler. Initial tests are showing the new Maxtor disks to reaching over 100MBps vs 80-88MBps when used on a PCI bus based SATA controler.

The CSA bus is also a great improvement with the use of gig-ethernet. Only a few of the canterwood manufacturers are actually using the CSA bus (Gigabyte, Abit, MSI, and IWILL), all the others are not using this new bus. The necessity for the CSA bus will become very apparent with the use of math. If a gig-ethernet solution is placed onto a system on the PCI bus, you will actually max the PCI bus itself with this card, and that is not taking into account all the other devices that share this bandwidth (sound cards, RAID cards, etc). The only reason not to use CSA is price. The CSA bus was designed to use Intel chips. The Intel chips are more expensive then some other gig-ethernet chips out there, but anyone who has used the Intel chips will easily see the performance/value of their solution (this is even without the CSA bus, Intel's ethernet chips are by far one of the best available). The motherboard manufacturers who are not using the CSA bus will try and tell people that it is really not needed as the overall bottleneck will be your hard drive for transfering data, but this is really not the case with reguards to the systems this motherboard will be used. Almost all of these systems will have large amounts of RAM that will easily be able to cache the data communications for streaming to the hard drives (the intended use of the Canterwood chipset is for workstation and small server class systems, hence the support for ECC checked RAM) and these systems will usually have at least a few gig's of RAM (mine will have 2).

Basically this chipset fixes SEVERAL bottlenecks in the overall system (integration of SATA to remove it from the PCI bus, use of CSA to remove ethernet from the PCI bus, increase of bandwidth between CPU, RAM and hard disks). The main bottleneck that was not addressed by this chipset was the use of the PCI bus. Even though the PCI bus's usage was offloaded to other buses, the PCI bus is very much outdated and should have been replaced with a new bus. PCI Express is a new viable option, and including it would have been a very good move on Intel's part. It should have done what they did when the ISA bus became outdated, included the PCI bus along with the ISA bus on the board, thus allowing people to still use their old PCI cards for the next few generations of systems while all the hardware manufacturers can start migrating to the new bus architecture (can't remember if PCIX is backwards compatible or not, or if it was another bus which was not compatible). In any case, this should have been included in this chipset to start pushing the hardware manufacturers to start changing their production.

Re:Actaully...We have cheap SMP...

[Glonoinha]

I am eager to get my hands on a machine with the new HyperThreading chips in them, and probably more eager for the SATA RAID 0 setup doing a measured 96MB/s sustained read and 86MB/s sustained write (serial reads and writes) ... but when I said cheap I was talking $250 cheap, not $1,500 cheap :)

All said, I am holding off buying a new 'Primary' machine until Doom III ships and at that point I am planning on spending silly amounts of cash on a new box. Hyperthreading, SATA RAID 0, a ton of RAM, as of yet rumored about but not yet commercially available video, possibly dual video out for twin display, 5.1 audio ... the works. Until then however I will continue to spend time doing work on the several machines I have hooked up via my KVM in a RAIC.

Funny thing is that two seems to be enough boxes - I can occasionally throttle one so bad that response times are really bad and swap over to the other one to entertain myself while the first chugs for half hour, but I don't seem to ever have them both throttled so bad that neither is perky.

Re:Noise. Bigger, Faster, Louder.

[Cthefuture]

if you have dual boot on the machines you can be rebooting one while using the other

That's what I use VMware for. ;)

Lazy programmers?

[FredBaxter]

I know that one of the biggest problem facing the place I worked this summer (a truely enlightened place for programmers) was trying to teach "old" programmers not to spend so much time trying to write concise, small, effeicient programs. Their time and resources can be better allocated in other ways. Not to say that lazy and bad programming is good of course (wow, that's convincing), but it ceases to be AS important since the hardware is far beyond the software. But this is just one isolated example.

Lose the battle. Win the war.

Re:Noise. Bigger, Faster, Louder.

[jmacleod9975]

vmware is great, I really like it, but considering I can buy a new computer for the price of the license, I think I will stick with having a second computer.

Someday...

[Anonymous Coward]

Someday there will be a list containing just a few of the things done by people with really fast computers. That list won't contain web surfing, word processing, and recipe file management. It will contain things like protein folding, simulation of computer chips, market modeling, database searches, virtual reality games, creating 3D annimations at home, multimedia editing, and other similar things. That list will be used to reduce the incidence of posts that say, "My ten year old computer does everything I want and that invalidates the need for any faster computer by anyones else." Someday that list will exist and be posted on popular discussion sites.... but not today on Slashdot.

whats the big deal about raid 0

[QuantumRiff]

Why do so many people drool over raid 0, i could understand raid 0,1 or raid 5, but just 0?? seriously, you slap 2x100GB drives in a system and get a faster 200GB drive, but if you lose a single drive, (Very common in IDE) your out 200GB of data. Seriously folks, think about that..

While backups are the obvious solution, is someone that is slapping together a raid 0 system from cheap IDE or Serial ATA drives going to be able to afford a Tape drive that can hold 200GB of data? thats a hell of alot of CD-R's, or a damn long time to download, even off a T-1.. Hell, thats a handfull of 80/40GB tapes for DLT, and the drives aren't too cheap either..

or not

[egoff]

Considering... of course? Singularity.

Thank You

[Mark Dentari]

If Joe Smoe can't use it then nobody else needs it either. I always hated that idea. Thanks for mentioning my field. I'm into music production and that's one area that will always need more power. In order to run high end pro reverbs a great deal of CPU power is needed. I never seem to have enough and the native reverbs never seem to be as good as the dedicated boxes from TC Works or Lexicon. If the CPU power is there then someone will find a use for it.

You are stoopid

[Anonymous Coward]

These "IBM" chips you are speaking of have been used in everything post 68040 you ass clown. They are all RISC and kick the crap outa x86. x86 just keeps wanting to make the cycles quicker and quicker without doing things more intellgently.

Dumbass

Re:You are stoopid

[Anonymous Coward]

x86 just keeps wanting to make the cycles quicker and quicker without doing things more intellgently.

Dumbass

It's interesting that a guy flinging around the word "dumbass" can't spell "intelligently".

Last Post!

[alpg]

...everything on this earth has a purpose, every disease an herb to cure
it, and every person a mission. This is the Indian theory of existence.
Mourning Dove, (Salish 1888-1936)

- this post brought to you by the Automated Last Post Generator...