Slashdot Mirror
Pentium 4 2.8GHz
DigitaBiscuit writes "The new 2.8GHz Pentium 4 has been officially launched by Intel today.
Sporting a 533MHz System Bus, this new P4 looks to put the hurt on AMD's new Athlon XP 2600+. Benchmarks and a full review with
performance versus AMD's new chip,
can be found here." The NDAs must be expiring today, since we already have another review submitted as well.
My next prediction: AMD will release a processor that's even faster! Nobody will expect that one.
;-)
Distant future (more than one week): Intel will release a processor that's faster than that one!
What if the video chipset industry was the same way? Whoa.
I'm typing this on a homemade computer that uses an AMD K6-3/350 overclocked to 392 Mhz. For what I do (web browsing, word processing, email, simple games like Tetris and listening to streams and MP3's) it works great! Why would I want to buy one of these? I'd much rather take a trip to Hawaii with my honey then blow a couple of grand on a Pentium 4 computer. Besides, let's get real...in about three months this will be obsolete...replaced by the newest whiz bang 3.0 GHZ processor that works about 7% faster then this one does. I think I'll wait until a real reason comes along to justify my spending a lot of $$ on a processor this powerful. For what I do, I simply don't need this much power. Do you?
not to mention price.
an intel cpu may only represent $100-$300 difference premium over AMD...but in todays commodity priced computers....that can represent beer money, motherboard money, video card money or ram money....or all of the above!!!!
most of my friends and family would be absolutey impressed with an XP1500 system i could build them for less then US $350 (sans monitor) anyway.
Learn your gambling terms, kids, or they'll laugh you out of Vegas.
-Kevin
You know, system speed is not all about what frequency your CPU is clocked at. The memory system (FSB speed, cache size) matters too!
This is, IMHO, what all these benchmarks show. It is no surprise that a Pentium FSB running at 533 MHz can beat an Athlon with a FSB at 266 MHz. I'm actually more impressed that the Athlon managed to beat the Pentium on some benchmarks.
)9TSS
OK, I officially need sleep now.
PIV 2.8GHz should be about 72.6W (based on the numbers for 2.53GHz @ sandpile.org), AMD 2600+ is 68W, and my 1.2GHz Athlon @ 60W died already.
:(
1.4GHz+ certified fan, was still running after it died, fan still in place, no airflow blockage, but 30C outside, 40C in my room, then some in the case and running at 100% load. Sigh... back to Duron 700
Kjella
Live today, because you never know what tomorrow brings
Depends on who you're aiming at. Somebody running on a P4 processor is already going to have a good idea what 2.8 gigs is gonna be like. An AMD user isn't going to make as much sense of it, I agree.
Frankly, I think they should benchmark these systems on time critical processes. I can't outrun Office on my P2 350, I dunno why they think I care about benchmarks involving that. Quake FPS can be improved (to an extent) with a new video card. Us dudes using LW/Maya/Softimage/Max are the ones who REALLY need the speed/performance rating tested. Time = Money. Benchmark on those progs.
"Derp de derp."
That is the oldest line in the book. Historically it has been proven resoundingly wrong. Apps use new power and resources and new applications are enabled by faster CPUs and more memory. People have been saying "we don't need this much power" since the 486DX 20MHz, and how many people use those today? Damned few. Get over it, stuff gets faster and people keep buying the stuff and developers write their applications for the newer stuff.
barring the fact that Intel will ALWAYS be one CPU RELEASE ahead of AMD (which is fine, I'd never buy the absolutely latest of either, your an idiot to pay exponentially higher for THE most current processor from either)
...WITH OUT A DOUBT, IN NAME AND IN PRICE.
that aside lets look at the facts (from most benchmarks)
1. Your statement, Jacer, looks pretty much dead wrong. How can you explain that a CPU that has a lower CPU frequency AND a freaken lower bus frequency, perform as good or better then it's competition.
because jacer, i look at each cpu release from both intel and amd, and i examine the COST, and which model is supposed to compete with each other.
for example: just a gross comparison
US $146 athlon xp2200
US $246 p4 2200/533
a quick look at ANY site with benchmarks shows that the xp2200 actually holds up VERY WELL against the p4 2200/533.
whether INTEL likes it or not, AMD targeted the p4 2200 with the XP 2200
if the P4 cpu and 533 bus is so superior.....
i'd like to see where..
show me you prick. show me where the 533 is making a dramatic difference.
if anything, the only thing i see, is that INTEL can release one step quicker on their speediest cpu.
$546 - Pentium 4 2.8GHz from pricewatch
no thanks, i'll pass on the latest $600 cpu from Intel.
i can build an ENTIRE system for that much money, and it will perform within 20% of a p4 2.8
I'm sorry (not), but I should NEVER have to wait for anything. Not with a P4 2.x Mirkwood. Or AMD 2xxx+ GTZ. I want instant reaction like Beos had. On my pentium 225. Click, Click. Off I go.
Hard drives are fast(?) and cheap, but still saddled with the bloat code that gets written for this new stuff.
OT, but I would like to see an office suite written by John Carmack. That would rule. Misspelled words would have 3d blood dripping out of them, and fast, fast, fast.
Ok. Time for sleem.p
If voting were effective, it would be illegal by now.
Anyone with enough data to answer?
"People have been saying "we don't need this much power" since the 486DX 20MHz..."
I always found that type of comment amusing. They don't realize that the extra horsepower goes into making the interface better.
Technically, word processing could easily be done on a >486. However, what the extra mhz gets you today is real time spell checking, graphics, and a buncha extra interface options that just weren't possible on slower machines. You can make today's software work on earlier machines, but the optimization to do that will throw away a lot of the UI stuff we take for granted.
"Derp de derp."
"Troll" means someone who is saying something he doesn't believe, merely to cause trouble.
There is no "myopic" or "shortsighted" option for the moderator to choose, so he probably chose the next best way to mod the post down.
I knew that my P200 was getting old when they released this newfangled "AGP" slot. And I realized that it was obsolete when GHz processors started coming out. But approaching 3 GHz?!? That's just rubbing salt in the wound.
Good article with benchmarks over at Aces Hardware
What I like is how the AMD 2600+ is very close on most games either 1-2FPS behind or ahead, and the 2800+ isnt out yet. Go AMD! P4 2.8 $570 or AMD 2600+ $265
This is slashdot! Do people on car enthusiast sites talk about the fact that most people don't need 500 HP cars? I mean, seriously, what planet are you living on. Luddite ;)
Well if I was going to put together a fast desktop system, I can tell you it wouldn't be built around an Intel Pentium 4 Processor (insert jingle here) - as far as I am concered, Intel price their CPU's so far off the scale it isn't true. Add to that the fact that AMD's processors no longer have issues with stability or floating point speed (like the old K6/K6-2) - I cant see any reason to buy such a top of the line Intel chip unless you were absolutely *desperate* to eek every last drip of performance out of a system. But at 2.6GHz and beyond, people aren't really counting - right?
The thing that bugs me is still the stigma attached to AMD.. its similar to the old 'No one got fired for buying IBM' - it is the same with Microsoft, and the same with Intel. People still avoid AMD because they consider them to be inferior..
"Hey! Unless this is a nude love-in, get the hell off my property!!"
With this trend, I can resume writing applications in Qbasic.
But seriously: with processor speeds like these, efficient programming will be even less appreciated. Sadly.
Sigged!
I sincerely hope for everyone's sake you aren't an employer!
I have been in hiring positions throughout my career and will be hiring for a new position that's opening up in the near future. In those years of hiring, I've learned that people that don't keep their own computers reasonably up-to-date, seldom have the skills and enthusiasm to make them stand-out employees.
Note the word "reasonably." I don't care if someone is running a 1ghz Duron or a 2.8ghz P4, but something with an old Socket 7 motherboard is absurd as a computer professional's primary home system.
Read the review. They mentioned in the review which benchmark was the one mentioned as favoring Intel. They posted that along with the other well known benchmarks. They did not hide the fact one of the benchmarks is being disputed as favoring Intel. If the results were bad on all benchmarks except the one favoring Intel, it would stick out like a sore thumb. The other benchmarks also favored Intel, even though they didn't have the optimum MB and memory for the test, even when the AMD setup was optimum. I would love to see the test again with the preferred hardware for the P4.
The truth shall set you free!
I would not hire a doctor who is enthusiastic about his tools.
Fine. Get laser eye surgery by someone using old equipment and outdated techniques. And you don't need a doctor who's enthusiastic about new medicines (the primary tools of the medical profession). He can treat whatever ails you with aspirin and penicillan.
I would not hire an engineer who is enthustiastic about his tools. I would not hire anyone enthustiastic about his tools.
Then I hope you get a mechanic with no modern computerized diagnostic tools and he f***s your car up. Every craftsman, whether a woodworker or a computer professional, takes pride in having good tools. He takes an interest in finding the best tools for his line of work.
I want a person looking for a solution to whatever problem i have, not a nerd who seeks a way to use his newest toys.
Great. Then maybe I can develop a BASIC program for you on a Commodore 64 to meet your needs. You wouldn't want me to use anything too modern.
Tell me about it. X86 clock speeds are now officially *twice* as fast as the CPU in my home machine, which was brand new last year. I installed my CPU when it was the best of the bunch. In a years time, they are twice as fast.
This fight to have the highest clock speed is driving me nuts. It doesn't even really matter any more, does it? Do we need to to shave a few extra seconds off of compile times?
Anyways. It looks like the Athlon holds its own pretty well, considering the near 700 MHz speed difference. But it doesn't matter to me. Opteron will be my next upgrade. When that sucker comes out, it will be a battle for blood between AMD and Intel.
Hrm, you know, the computing world is bigger than office+quake+rendering. There are plenty of cpu hogging stuff besides rendering. But of course, your advice is entirely correct. Benchmark the stuff that's important for you. Not what all those hardware sites think that is important for their on average 14-year old readers.
Thank you for that sane, and well-considered reply. I'm tired of reading the rationalizations for the notion that computer professionals should use outdated, underperforming computers. You shot holes in this particular argument rather nicely. Such rationalizations are just self-serving tripe meant to justify cheapness or laziness.
Can you imagine these people as doctors? "I don't need any of those new-fangled medications or diagnostic equipment. By not keeping up on new drugs and medical equipment, I have more time to spend with my patients and that's what's really important."
Basically we've seen an order of magnitude raw CPU clockspeed increase in 4-5 years. What have we accomplished with that?
It seems that the faster we make the chips the more we squander their power.
Who cares that their processor is inefficient, poorly designed, and expensive? Not the ones who buy it certainly; there is a market fpor it, and they should not be penalised for serving their market - they are a business after all.
For all those arguing that these tests ar 'not fair' (memory, RAMBUS, blah, blah, blah) - you are missing the point. Boo hoo, they are using different equipment; I could equallly argue that AMD is shooting itself in the foot for not utilising the fastest memory architecture available. For most people, 700 or 800 MHz is more than necessary to do almost anything - above that only specialised areas will see any real benefit. Is it really any benefit to be able to play games at 32 bit compared to 24? Can you actually tell the difference at speed? Isn't it more to do with the graphics card anyway? Scientific applications, yes - these can be markedly improved with faster processors. But most readers here do not work in a render farm in Hollywood.
But back to the original point, we shouldn't be so aggressive towards them just because of who they are. They are serving a market, doing if very successfully, and for those people who do have $$$/£££ to spend, they represent the maximum performance. I will continue to buy AMD because I think they give more value, and my XP 1500+, although now slow compared to newer processors, is far faster than I need, even for compiling Mozilla or running KDE3, WinXP or Serious Sam 2. But that doesn't mean I should refuse to talk to people with an Intel chip in their machine.
And don't mod this down as flame/troll just because you disagree -use your points properly and mod up someone you agree with. And stop being small minded too...
This idea was invented by Shampoo.
From the article's title:
Sheesh, you would think that such an obvious malaprop would have been seen by at least one person.Or, perhaps they were referring to the P4's lack of performance over the PIII...
--
void life();
That what was all this school was for... to teach us how to solve our own problems. -- janeowit
According to them, "older legacy code based applications" are applications without Pentium 4 optimizations.
Will we ever get reviewers that aren't incredibly biased... and stupid? Of course P4s do better on software with P4 optimization! And software w/o it isn't "older legacy software"... it's software that isn't written to favor a particular chip in the marketplace...
Gotta love it.
-jbn
The cost is where AMD shines, granted. My point was if AMD could get a faster FSB, they could easily pull ahead, not nessicarily in the CPU frequency. You're an idiot good sir, if you don't the the advancement of the proc is a significant step. With a faster FSB you get more responsive ram, and PCI slot, combined with the fact that the 2200 isn't 2200mhz, it'll make it that much better, closing the gap on 20% margin you're bragging about.
--fetch daddy's blue fright wig, i must be handsome when i release my rage
What? How on earth could you have even seen a "Pentium V"? First of all, the first silicon release of this P4 won't happen until the end of the year. The "P5" is still in its design phase. As far as AMD doing something on the front side bus, that idea is moot. AMD is moving very fast to their star-configured Hypertransport FSB. Once they've moved into that direction you won't be able to compare the P4 FSB to the K8 FSB.
Here's one reason why I like a faster frame rate. I don't know about most people, but I can clearly see a difference between 25 FPS and 60 FPS. I'm one of those people who gets headaches from florescant lights and monitors that are running at 60 Hz. With a first person shooter (or similar game) running at anywhere below ~60 FPS (even if the monitor is at 85 Hz) I tend to get motion sickness after maybe 20 minutes of play.
Maybe it's the fact that I'm using such a large monitor, but it really gets irritating.
I know this is Offtopic, but just thought this would be interesting to think about. ENIAC, the first general purpose digital computer was used in part to calculate artillary arc tables. ENIAC could do in 30 seconds what it took a human 12 hours to do, so it was approx. 1400 times faster than a human. The amazing thing is that ENIAC ran at 0.1 Mhz.
Just think now, a 3 Ghz machine is 30,000 times faster than the first computer. Amazing how far we've come in 40 years.
It seems to me that processor speeds have relly shot up in the last 6 months. Before, there was always a nice tie between the advances in processor speed and the extra needs of Windows. This arrangment worked well for both parties in that everyone felt they had to have both to be right with the world. Has Intel (and AMD) now said to heck with MS we're going to put all the speed we can into our machines? Or, are we seeing a MHz race between Intel and AMD? With a viable competitor, Intel can no longer trickle out improvments at their own pace, they have to deal with someone else matching and exceeding their pace. What does this bode for the future of processors?
THIS SPACE FOR RENT
Damn straight. If you look at those charts, the AMD Athlon 2200+ (which I am going to get next week for my new system), performs just as good as most of the higher end pentium's, and cost like 200-300 less! I'm a college student, and that couple hundred can go to so much more tuition, beer, car payment, etc... Guess what the intel name isn't worth that much to me....
I want my rights back. I was actually using them when our government stole them after 9/11.
Non-Disclosure Agreement
This article Pandering to the Masses: Does Engineering Still Matter? explains how the Pentium III beat the Pentium 4. If the P3 and the P4 are run at the same clock speed the P3 performs much better than its anemic younger sibling.
One quick question though, why do I need a noisy heat generating GHz machine on my desk when all I do with it is "ssh" into the server - dual 2.2 Xeon - in my storage room? I am probably missing something ...????
Yes, you are. You have a real, modern, decent computer. And, like me, you have some older stuff that you use occasionally, too. No problem. The guys I worry about are the ones that have only ancient stuff and no modern machines.
My God! Would you hire me?
Absolutely! Anyone who can create a sophisticated web page like yours is a shoe-in at any tech company. I mean, it not only has a link labelled "Test", but clicking on that link brings you a message of "Hello World." Wow!
Last time I checked, the P5 was introduced about eight years ago. It started at 60 MHz and got pushed to 233 MHz before Intel decided to move on to other things.
YHBT. YHL. HAND.
20 January 2017: the End of an Error.
Lets take hard drives as an example. By your logic, by now you should be able to buy, say, a new 1 gig hard drive for $5. In reality, it doesn't matter much at all what the drive capacity is. It costs $50 to manfac a hard drive. You can't manfac a drive for $50 then sell it for $5 just because it's only one gig in size. There is an absolute minimum charge that the manufacturer has to charge to stay in buisness. This is true for CPUs, MB's, Cases, CDROMs, everything.
So you are right, we have reached a price plateu (actually a price valley). We've reached the minimum possible price for manufacturing all the parts necessary to build a computer. (THat seems to be about $300, without a monitor for a general purpose PC machine) The parts just continuously get better, but the lowest end prices are not going to fall from where they are now.
Contrary to popular belief, coding is not all free blow-jobs and beer. Those things cost MONEY!
My understanding was that the P4's x87 performance sucks eggs...that it's not only slower than the Athlon, but that it's even slower than the P!!! it was supposed to replace. Maybe they've improved things in more recent cores, but you'd think there would have been more mention of such an improvement if that had happened. (Yes, I know there's SSE2, but faster performance there does bugger-all for performance in most apps, since most apps use x87.)
20 January 2017: the End of an Error.
Correct me if I am wrong. Are you suggesting that the decision to design the P4 so it accomplished less per clock cycle, but was able to do them more quickly was purely a technical decision? Are you suggesting that it wasn't influenced at all being able to exploit a foolish confidence in megahertz?
You seem to know what you are talking about, so I would welcome learning how you came to have such confidence in Intel's ethical standards.
Isn't intel the company that tried to slip CPU serial numbers past us, on the Pentium III?
Isn't intel the company that told their customers they would have to prove to them they would have to prove they needed flawless floating point before they would replace defective CPUs? Let me quote from the Doctor Dobbs Journal article:
And how about the intel 487? Didn't they introduce an expensive 487 floating point co-processor, to augment the 486sx cpu, which was actually just a 486dx in disguise, that totally disabled and replaced the user's existing 486sx?
I do belive that turbo-C and pascal, and many other IDEs for that matter, do something before you compile the program while you edit. Somethings are quite easy, like tokenization, pre compile header files. The integration could gain a second or two also by not compiling the same header 3 or 4 times for diferent files in the same project. Gcc is a stand-alone compiler and cannot have this type of integration (could?).
Maybe gcc team could create a gcclib to make it possible to access some lower level structures and doing things ahead in ides.
[]'s Victor Bogado da Silva Lins
^[:wq
Look at the SNK NeoGeo(tm) system - it's got a fraction of the processing power of my main machine, and yet it is the most successful video arcade game system EVER BUILT, based on the length of time it's had new games released for it.
You are confusing the development system and the target system. I've developed code to run $5 microcontrollers -- but I didn't develop the code on $5 microcontrollers. A developer's workstation should be fast, responsive, and capacious.
The argument that nobody who is dis-interested in the technology can be any good is invalid. On the contrary, I think that anybody who doesn't take a step back and think that using a 1000 MHz CPU as a word processor is stupid.
And the person that does take that step back is probably someone who is intensely interested in technology. I've been advocating a stripped-down, low-power notebook for simple word processing, e-mail, and maybe web browsing for several years. Preferably something with a flash-based storage system rather than a hard drive and, perhaps, an ARM family CPU. The average notebook used for word processing work is insanely overpowered, overpriced, overweight, oversized, and battery-limited. Basically, something more than a PDA but much less than a standard laptop is what I'd like to see. But, if that machine is built, it's going to be designed and programmed by engineers using modern development systems.
The integration could gain a second or two also by not compiling the same header 3 or 4 times for diferent files in the same project. Gcc is a stand-alone compiler and cannot have this type of integration (could?).
I don't see any reason why this couldn't be done standalone. Back when I used an Amiga, my compiler of choice (DICE)--a standalone compiler, by the way--had an option to precompile header files and store the results (macros, structs, typedefs, etc.) in a separate file, and then read the results of preprocessing/compiling those headers from that file directly into memory. This sped up compilation of large projects by 2-3x IIRC (and this was in the days of multi-hour compiles).
fmaxwell, all I can say is that you have demonstrated the lack of ability by the people you hire to develop a solution successfully on anything but the latest development kit.
No, you have shown that you know practically nothing about professional software development. My own computer has an Athlon CPU that now sells for under $75 (an XP1700+). The hard drives are a pair of modest 80GB Maxtor IDE drives. My CD burners are half the speed of the current state of the art. Hardly is that "the latest development kit."
Don't lecture me on this. You have neither the programming experience nor the management experience that I do. I've developed code on front panels ("switches and blinky lights" to you), punch cards, Z80 systems running at 2.5mhz, and on 20 year old Intel MDS development systems.
If you ever run a successful project (which seems highly unlikely), you will learn that there is an appropriate balance between personnel and equipment expenditures. Just because you CAN develop an application on an 80286 running MS-DOS does not mean that you SHOULD develop it on that. The fact that you CAN debug hardware and software on an embedded system without an ICE (In-Circuit Emulator) does not mean that you SHOULD.
If you want to hire some guy that has so little interest in computers that he has not upgraded his own system in four years, feel free to. I'll stick to hiring people who have a genuine enthusiasm for computers and who keep their hardware, software, and skills up-to-date.
Hehehehe, I think it's really funny how you're getting so wound up about this.
This coming from someone who just posted the longest message of the thread...
You seem to be under the mis-impression that there is some magical 'brick-wall' which separates 'still usable' kit from 'non-usable' kit. That makes no sense.
I've been very careful throughout this to differentiate between "state of the art" and "reasonably modern." I'm fully aware that there is no "brick-wall" for system usability when talking about megahertz (vs. architectures). That said, if someone has an ancient system that is rarely updated, there are many applications and development tools that they simply won't be able to experience.
All I can say is that you have probably contributed more to the mountain in China than I have, which was my original point.
Since you continuously harp on this, I'll explain. I use my retired components and peripherals to construct PCs for people who haven't the money for new machines. The last one I built went to a cash-strapped family. The mother is divorced (for very good reasons) and supporting two children and her elderly mother on a single, modest income.
Do you think you could lose the attitude of moral superiority now?
Try re-reading my post - I was mocking people who use fast processors for the end application, not the development system.
Nowhere in my many postings on this subject have I suggested using fast processors for target systems. In fact, I have made it a point to stress that development systems are what need to be fast (within limits) while target machines may be very humble -- depending on the application.
Unlike you, who seems to have no respect for me
I apologize. I have become rude and unpleasant as a result of the flak I've gotten in this thread. I was wrong to allow that to happen.
What are you going to do? Fire somebody, because they have too many wait states in their BIOS?
This is just another example of this being blown all out of proportion and taken to absurd extremes. My point all along has been that a computer professional that does not keep his hardware, software, and, hence, skills, reasonably up-to-date is generally not a good hiring risk. I contend that such a person is less likely to be a computer enthusiast. It's not that way 100% of the time. There are plenty of incompetent people with fast machines and talented people with slow ones.
I would worry about hiring an interior decorator whose home was furnished with bean bag chairs and lava lamps. It's analogous.
OK, well apart from not keeping my own hardware up to date, we seem to be in agreement on most other points
Glad to hear it and, again, I apologize for my unpleasant and uncalled for remarks. Going back through the thread, I realized that I had been, in my mind, attributing things to you that had actually been 'said' by other anonymous posters.
As to keeping your PC's hardware up-to-date, look into it. With very high performance CPUs going for under $60 U.S. (e.g. Athlon XP1600+) and motherboards and hard drives at all-time-lows, it's fairly inexpensive to have a very fast system. Then you can not only do your work under EMACS, you can also freely sample everything from 3D rendering to modern video games to CPU-intensive simulations. Never forget that computers can be fun, too.
Peace.