Intel, OEMs Face Lawsuit For Megahertz Marketing 509
prostoalex writes "A group of PC owners filed a lawsuit against Intel, Gateway and HP, stating that companies spread misleading information about Pentium 4 processor performing faster than Pentium 3 or Athlon. The complaint alleges that 'the Pentium 4 is less powerful and slower than the Pentium III and/or the AMD Athlon.' PC World has more details in its story." I wonder if the same litigants have a suit against the USPS for ads leading one to expect prompt service from courteous, competent employees.
What damages are they claiming? (Score:2, Interesting)
Re:What damages are they claiming? (Score:4, Insightful)
I mean, the Pentium 4 is shit, I think, but at least I've done my homework and I know better than to just look at the number listed before "GHz" as the basis for buying my computer...
Re:What damages are they claiming? (Score:3, Insightful)
Madison County, IL, where the suit was brought is a class-action mecca now for its jurors willingness to award anyone money for anything.
Caveat Emptor (Score:4, Insightful)
In latin, Let the buyer beware. It's also a central principle in common law. Courts have recognized since the Romans that the buyer has a responsiblity to ask the right questions. The courts can only intervene where there is a blatent attempt to decieve.
This is just like automakers marketing SUV's as safer than sedans [when hitting a wall straight on]. Sure they are safer when you hit a wall straight on. Now, rolling over, tire blowouts, and repair costs, they are not included in the benchmark. Nor is fuel economy.
But as a bonus, you can get one of those funny propellers for the tow hitch, and 0% financing...
I'm sure some one beat me to this but remember... (Score:5, Funny)
Re:I'm sure some one beat me to this but remember. (Score:2)
Re:I'm sure some one beat me to this but remember. (Score:3, Insightful)
I have a 486-50 laptop and have occasionally used it to browse the web away from home. It barely works with Opera, and is impossible with IE.
You can connect as fat a pipe as you want to the machine, for fat Flash-infested web pages, a Pentium 4 does give you quicker access.
Fast does not always translate directly to 'bandwidth.'
It's just another sign of geek politics that everybody chooses to make these marketing claims into jokes rather than acknowledge they know what they mean.
Re:I'm sure some one beat me to this but remember. (Score:3, Funny)
I remember a 6-year-old kid in my neighborhood who used to think that "cool stickers" made bikes and cars run faster.
I bet he is purchasing a P4 right now based on such an ad.
(However, he is probably also a successful PHB because he thinks like the CEO.)
Re:I'm sure some one beat me to this but remember. (Score:3, Funny)
Re:I'm sure some one beat me to this but remember. (Score:2)
In related news... (Score:5, Insightful)
Seriously though...
WTF? So AMD doesn't even use Mhz rating anymore so they get away with saying 'mines's is better?'
But guess what? the P4 DOES bench faster on some benchmarks than the p3 and Athlon, likewise, the p3 does better in a few, and Athlon does the best in still other things.
Anyway, its not like the processor's slowing the machine down. "It's the DRIVES, stupid!"
Re:In related news... (Score:5, Funny)
That means... (Score:2)
Those athlons really rock!
Re:In related news... (Score:2, Insightful)
Re:In related news... (Score:2)
(I'm being a bit sarcastic here of course, but the clock speed is an honest measurement at least (usually; some components are actually lower clock speed that have been o/ced, like the FSB on Asus(?) motherboards) while this misrepresentation is something that bugs me. But I wouldn't sue over it...)
I hated it too... (Score:2)
But I got over it, what is wrong with these people? I smell money grubbers.
Quite frankly, AMD should step up to the plate with Intel on this and so should every other CPU maker incase this ever comes back on them. Esp. AMD with their current PR 1900+ lingo. Check out Ars's coverage of this story where you see what I have seen, the lacky sales clerk saying No no, 1900+ means 1.9 GHz, even though the sign says different.
Re:I hated it too... (Score:2)
Your comments on AMD are the only place where I could even see this type of case going. You can't sue Intel for selling a 1.4 GHz chip when in fact it runs at 1.4 GHz. It would be like suing Dodge because the V12 in a Viper can't pull the same amount of weight that the V10 in a Ram can, even though 12 is a higher number than 10 so you thought you should be able to pull more, it doesn't work that way. The other limits in the computers (RAM, HD, NB, SB, etc) and vehicles (engine literage, gear ratios, tire composition, etc) alter the performance of the selling numbers (GHz, Vxx) and have to be analyzed along with the selling numbers. This case should absolutely go nowhere in court unless you get some clueless judge up there.
However, AMD with its PR rating system is held to a different caliber because no longer are they selling their products based upon a particular specification of their product, but they are now selling their products based on an interprational 'performance' specification that is not a hardcoded part of the product. I could see people suing AMD if they bought an AMD 1600 and found it to be slower than their former Intel 1400 MHz computer. I would still think that a case along these lines would be completely absurd (why should we ever reward people's ignorance, suits should be reserved for cases where damage was actually done, or people were mislead to the point of damage in their life) - but, I could see it happening.
BOFH was right... (Score:3, Funny)
Re:BOFH was right... (Score:2)
if Intel's just measuring in speed (Score:2, Insightful)
For example, let's get a 4cyl engine next to a 8cyl engine. You COULD redline the 4cyl at say 6000RPMS and only run the 8cyl at about 5000RPM. Most likely, the 8cyl will still perform better than the 4cyl running at 1000RPM faster. (Just an illustration, I dont know how accurate an actuall test would result)
While I don't agree that clock speed soley determines the overall performance of a computer, Intel may be telling the truth when they say they have the FASTEST CLOCKED cpu, but other claims after that may get them in trouble. Sure, their P4 runs at 2.2 or whatever the max speed is now and if you were to gauge it, it'd be correct. I think this is just a case of consumers needing to be more educated in shopping for computers.
This does bring up an question. If we disreguard cpu speed as a selling point and use overal performance rating, judging computers becomes more subjective. Just changing out RAM, or chipsets, or some other small item can make a significant difference in a PC's value. More reason to build your own system.
Marketing holds back progress? (Score:3, Interesting)
We've already seen that this silly chase for faster clocks has caused certain processor makers to abandon computational efficiency in favor of getting to 3ghz as soon as possible. What other engineering breakthroughs have we missed out on because we're too obsessed with fast clocks?
--
Preview should do a spell check. It can't possibly be more then 30 or so lines of code. Highlight the potential misspellings, and provide a list of suggestions below the comment. They wouldn't even have to do the hard part, since there are great scriptable spell checkers already available for free. I'm tired of cutting and pasting my posts through ispell
Hmmm, interesting. (Score:5, Insightful)
Second -- it's a generally established principle ("puffery") that commercials are allowed to exaggerate to some degree. Chevy can claim that their vehicles are tough, "like a rock", which is a far less specific claim than, say, "this product is so tough that it can be driven two hundred thousand miles without maintenance" or "its windows will withstand sustained 9x19mm fire: perfect for the urban gangland outing". "Making the internet run faster"
If they
Re:Hmmm, interesting. (Score:2)
-Paul Komarek
Re:Hmmm, interesting. (Score:2)
Then again, as Bill Gates pointed out in the MS antitrust trial, the computer industry is so incomprehensively different from any other industry there's no point in wasting a judge's time on computer industry lawsuits.
Re:Hmmm, interesting. (Score:2, Insightful)
People confuse the law with "you shouldn't do that."
Re:Hmmm, interesting. (Score:3, Interesting)
I agree. I think the claim is more pointed at the claims about being better than their previous processors, not on Mhz. I mean, a P-4 2.2 Ghz runs at 2.2 Ghz, but it doesn't perform X percent better than a P-III.
I read some reports when the P-4 first came out that for office applications, a P-II 400Mhz was faster than a 1.5 Ghz p-4. They cut off the Level 1 cache, and they use Level2 closely coupled cache cause it cuts costs. Despite the fact that the 8K of L-1 cache they left behind could be overflowed by one horizontal line on a screen at 1024X768.
But basically, Intel has won both the Mhz war and the marketing war. Think of it this way - what's their target audience? Certainly not people who research before they buy a computer with a certain processor. If it tells you anything about the target demographic, the Intel Pentium Four is the P-4 and not the P-IV because they felt that not enough people would know what IV ment, and would call it the Pentium EyeVee.
So, for their target audience, they nailed it.
~Will
Re:Hmmm, interesting. (Score:3, Funny)
If they want to be less misleading, they should then say, "The P4 makes spam more distracting".
Re:Hmmm, interesting. (Score:2)
Doesn't it feel weird that a 8 years old chip has a name "higher" than one not yet released?
OTOH, I don't recall for sure if the original Pentium (or derivatives, like P54C) were known as "P5". If they had, then it's already time for a name change for the successor of the P4.
Oh, and about those commercials.... in the one where the blue aliens modify some digital pictures (remove nose, etc.), doesn't it seem bizarre that when they add color, it's done in a slow, progressive pass? I would have thought that Intel's P4 was faster than that and could do so in the blink of an eye. Thoughts? (I know it's probably some marketing thingy, it's just odd to me)
Re:Hmmm, interesting. (Score:2)
P3 was the code name for the 386, and there was no confusion when people started calling the Pentium III, the "P3". Similarly, P4 was the codename for the 486, and when the Pentium 4 was released, there was no confusion. I would guess that if Intel releases a processor called "Pentium 5", the P5 as the codename for the original Pentium will have been long since forgotten.
Re:Hmmm, interesting. (Score:2)
The "P5" refers to the original, 5 volt, socket 4 pentiums, that ran at 60 and 66 [sandpile.org] megahertz. The P54 were the 3.3 volt, socket 7 pentiums, from 75 -> 200 [sandpile.org] megahertz. The P55 was the split voltage Pentium MMX, from 133 -> 300 [sandpile.org] megahertz, which also used socket 7.
It's true even on the P4 Xeon level. (Score:5, Interesting)
Just recently I had a neighbor hire me to do a concept animation of a machine he was going to build. I used truespace 5.2 [caligari.com]. It was insanely detailed down to individual links on the bicycle
chain drive.
The poly count got so high that my P4 was going to take 3 days to render it. My computer could hardly handle moving around in the scene anymore. I told the neighbor I had brought the scene as far as it could go on my P4
and I couldn't go any further without a new machine. He gave me $2500 to work with so this was what I built.
Dual Xeon P4 2.0ghz
1 Gig RDRAM
Maxtor 80gig IDE drive
DVD-R(by his request)
The system definetly cut the rendering time down, to 24 hours,but something just didn't feel right about the new render time. I could
have bought 2 more p4 1.4ghz and accomplished the same for less. What really got me was when my friend rendered the scene on his single athalonMP 2200.
14 hours
A single athalonMP 2200 was smokin my dual xeon setup! Well, this is all it took for me to write off intel forever. Intel fuck you and your shitty CPU's, you've lost my trust forever!
Anyone that is even considering using a Intel solution as a renderstation, please don't waste the money. You can do a lot more with a lot less using AMD.
Re:It's true even on the P4 Xeon level. (Score:5, Insightful)
This is software dependant. I think it is Lightwave that is clearly faster on the P4s due to it having been specifically optimized for the P4.
As always, benchmark on your application. Weigh price vs performance, then buy.
Re:It's true even on the P4 Xeon level. (Score:2)
Re:It's true even on the P4 Xeon level. (Score:4, Interesting)
Anyone can tack on vector instructions to a CPU. The problem is the underlying architecture of the P4, which isn't as easy to fix.
The AMD Hammer series will have those same SSE2 instructions AND a superior architecture (to even the AthlonXP).
Where will the Pentium IV be then?
Re:It's true even on the P4 Xeon level. (Score:2)
Another thing I have found - after working in the biz (working with everything from amiga's to what you guys are talking about) is that the more complex the scene the more the application relies on high speed i/o (meaning memory to cpu performance).
We used to benchmark the amiga based on the lightwave texture scene. Back then it was a complex scene - these days computers could render that in real time - on the video card.
I don't know - but I'd be willing to bet that when you scale the scene in polygons that the AMD cpu might close in on the P4 (in the caligari case twice as quick as the dual xeon). But if all your doing is flying logo's (again the kind of thing almost any 3d video card could do in real time these days) from the benchmarks I've seen the P4 is faster.
Re:It's true even on the P4 Xeon level. (Score:3, Interesting)
Re:It's true even on the P4 Xeon level. (Score:2, Informative)
Nobody, that's who.
Have you asked? Sure your local PC store might not, but I've gotten to borrow machines from HP, Fujitsu Siemens, sgi and a couple of local box shops in the past. And I was working for a tiny company buying maybe 3-5 high end workstation at a time. If you can convince them that you are seriously considering throwing money their way, you'd be surprised how far they might go.
Re:It's true even on the P4 Xeon level. (Score:3, Informative)
The trick is there was no "might buy." This guy was buying a system from the shop that was going to let him run some tests. It helped that the shop had a rackmount Xeon system that they were getting ready to deliver to another customer. But if they didn't we were willing to wait until they had some stuff worth testing. It was worth the shops time to install winnt, 3D MAX, and load the big models. The only caveat is that most shops won't have top of the line sitting around. Instead you have to test on the best they have and extrapolate. In your case a single test on an Athlon probably would have shown that it was top dog. Remeber that you already had a dual pent to test on.
This dosen't help you now, but next time you need a workstation consider making the sale conditional on a few tests. If you are buying a high end system you are a customer the vendor should want to work with. It is cliche: make the vendor work for the sale.
Re:It's true even on the P4 Xeon level. (Score:2, Interesting)
Now that I've said that, I've got a Dell 530 MT workstation at work. It's a 1.5GHz P4 Xeon with 512 MB RAM and SCSI HD/DVD drives. Pretty nice machine, no doubt. At home I have a 667MHz/512 MB RAM PowerBook G4 laptop. It has a 5400 rpm ATA HD and an ATA CD-RW drive. Clearly not in the same class as my CAD station at work.
At work, I can rip a CD with CDex in about 16-18 minutes per disk using the SSE enabled Lame encoder. On my laptop at home, it takes less than 5 minutes to rip a CD with iTunes.
What gives? I know my Mac's got Altivec and all but shouldn't a Xeon with all that on-chip cache, SCSI interface and a clock speed and bus over twice as fast as my Mac at least be able to keep up? After that, I too began to question just how bad Intel's chips have become in the pursuit of clock speed.
Read speed; ripping CDDA straight to MP3 (Score:2)
At work, I can rip a CD with CDex in about 16-18 minutes per disk using the SSE enabled Lame encoder. On my laptop at home, it takes less than 5 minutes to rip a CD with iTunes. What gives?
How much of that is the physical speed of the CD-ROM drive? My PlexWriter 12/10/32A burner reads data at 10x to 32x (CAV) but reads audio at 10x across the whole CD, limiting me to an 8-minute rip.
But still, I've never understood how people can just rip and encode to MP3 simultaneously. Without an intermediate step where the recording exists as a wav file, there's no chance to fix up pops in the audio, silence explicit language for a play-in-front-of-your-parents edit (I'd rather not pay twice for the clean and dirty versions), or remove leading or trailing silence.
Re:It's true even on the P4 Xeon level. (Score:2)
Re:It's true even on the P4 Xeon level. (Score:2)
It *was* unethical (Score:5, Insightful)
How many advertisements from the companies in question had lines like, "Tired of that old 1GHz PC? Get the latest 1.5GHz screamer!"
I believe that the primary complaint was that people were being misled into thinking that, say, a 1.6GHz P4 system is 60% faster than a 1GHz Athlon or P3, which is definitely not the case unless the only application the system runs is Q3, or a few of the rather limited number that the P4 runs very well. While I don't believe any vendor really explicitly stated anything similar to "a 2.0GHz system is necessarily twice as fast as a 1.0GHz system!", the companies did imply such a conclusion by comparing clockspeeds (without coming to any conclusion except the higher clockspeed is fast, though not saying "faster") or by using ads with lines that implied the same.
One can be misleading without blatantly lying.
Whether the companies in question were just unethical or did something illegal is the question. I would hazard a guess that the lawsuit has no strong legal grounds.
Re:It *was* unethical (Score:2)
At least they're not like hard drive manufacturers who LIE about their sizes.
>>I would hazard a guess that the lawsuit has no strong legal grounds.
IANAL, but it doesn't.
Re:It *was* unethical (Score:2)
I'm not a big fan of the way HDD manufacturers label their products, but there are legitimate reasons for doing it the way they do.
For one, hard drives are not organized or built around binary trees, so it is more convenient to use the definition of "megabyte" which refers to 1,000,000 bytes rather than 2^10 bytes. Additionally, one could argue that using "megabyte" to refer to 2^10 bytes is actually the measurement that is lying because mega, giga, tera, exa, etc. are all standard prefixes that refer to powers of 1,000; not powers of 2^10. Computers simply adopted these prefixes because 2^10 happens to conveniently be fairly close to what a real "kilobyte" would be. (1000 vs 1024)
I don't think it's a big problem (Score:4, Insightful)
Re:It *was* unethical (Score:2, Offtopic)
There are many parallels in other industries. For example, makers of hobby telescopes often use "power" (magnification) to compare scopes. However, magnification is a misleading benchmark. The most important metrics are the main apature and the quality optics, but most people don't know this. (The term "precision ground" is supposed to mean something in the business, but enforcement is weak.)
One can manufacture a $20 scope with 1000x magnification, but it would be useless because the image would be dim and blurry.
Manufactures end up including an eyepeice with useless magnification so that they can put a big number on the box. Hopefully the kit also includes some usable eyepeices in the mix.
Re:Bullshite. If someone "doesn't have the time to (Score:2)
Even then, ask yourself: Are there any *unreliable* but popular sources for hardware information? No?
Read a review from Tom's Hardware or C|net recently?
Re:Bullshite. If someone "doesn't have the time to (Score:2)
-Paul Komarek
Re:Bullshite. If someone "doesn't have the time to (Score:2)
Re:Bullshite. If someone "doesn't have the time to (Score:2)
-Paul Komarek
How about Apple? (Score:2, Flamebait)
Running OS X, it takes about double the clock on the Mac side to equal the speed of Windows 2000 on a PC. (Thus, it takes a Mac at 1ghz to run OS X as fast as a 500mhz PIII runs Windows 2000). This is the exact opposite of Apple's claims.
It really is not fair to the consumer, especially the more novice-type users who tend to buy Macs. I recently visited my family who has two Macs, and they could not believe how fast web browsing was on my wintel laptop. Keep in mind my laptop is about 2 years old!
Re:How about Apple? (Score:3, Interesting)
Remember, how 'fast' you can browse the web has more to do with the efficiency of your web browser and your bandwidth and very much less to do with your processor and your operating system. To say that browsers under Windows 2000 render wab pages faster than browsers under OS X is quite possibly true depending on what browser you are using. But that doesn't say shit about how fast your hardware is. I would bet you that my OS X machine 'browses the web' using lynx faster than your Win2k machine does using Netscape. Does that man that my Mac is faster then your PC? No. It means that my web browser is faster and more efficient than yours. And shall we not get into the relative differences between the way OS X and Win2k draw the screen? X is harder on system resources and takes more processing power to accomplish similar tasks (drawing windows, moving windows, etc). This says nothing about how fast the processor is, only that OS X is hard on resources.
The next time you want to compare processor speed between platforms try and pick a good benchmark. The seti@home client is probably a good benchmark, rendering graphics or video is probably a good benchmark, integer or floating point tests are probably good benchmarks, Q3 is probably a good benchmark. Rendering web pages is probably not a good benchmark because it isn't dependent on processor speed so much as it is on rendering engine efficiency - that's why IE and Opera and Mozilla on identical systems will render identical pages in different times. Some browsers are faster than others, even on identical hardware. This says nothing about the speed of your hardware.
Re:How about Apple? (Score:2)
Re:How about Apple? (Score:4, Funny)
Re:How about Apple? (Score:2)
Do you mean that OS X makes simple integer operations take longer to get through the PowerPC's short pipeline than it takes to get through an Intel P4's long pipeline?
Yes Motorola do have a fair few things to do to make the PowerPC stand up to the x86 architectures that are around these days but IBM have started moving in thte right direction and word is that Motorola will do so soon.
Also remember that Apple clarify all of their statements with details of what they are refering to, Flops, Photoshop processing, MPEG encoding etc and for the most part - in the arteas where Mac software excels or where Macs are the prevelant technology they are a good choice. For people like me - who have grown weary of the Megahertz wars - Mac OS X aon an iBook makes the perfect machine - putting the characters I'm typing up on screen just as fast as I type them
Anything more than that is overkill.
In Related News (Score:2)
Give it a break. I'm not a fan of the P4, but Intel states they run at X Mhz. They run at X Mhz. Different processors produce different performance per clock tick. What else is new?
Shall I sue Intel because my 85 Mhz Sparcstation runs Unix better than a 300 Mhz PC?
The car analogy (Score:2)
Since we often use the "hood welded shut" analogy when talking about Linux, it makes sense to use it in other ways with computers. If you are buying a car, do you walk in the showroom, tell the salesman that you want the car with the highest horsepower rating and then promptly buy that car? Even if what you need is a pickup truck with four wheel drive and an extended cab?
And even if we are just talking about pure performance do you really still only buy the car with the biggest engine and highest horsepower? Or do you also look at transmission, airflow, tires, cockpit ergonomics and skidpad ratings? Of course you do because a car buyer who wants performance educates himself about what factors create high performance.
Sounds to me like the people bringing this lawsuit are either stupid or out for money.
Re:The car analogy (Score:2)
That's not really the best analogy. It's more like comparing two cars based on the maximum revs the engines can perform. I mean sure, everything else being equal a car revving at 10,000 rpm will beat one at 5,000; but that's only if things are equal.
Anyone who implies "hey buy this car - it revs to 8,000 rpm and that makes it faster!"; should get sued if another car has more power but only revs to 5,000 rpm.
Go for it Timothy (Score:2)
I wonder if the same litigants have a suit against the USPS for ads leading one to expect prompt service from courteous, competent employees.
Why don't you start one?
Just Like Monitor Sizes (Score:2)
That could be an interesting precedent for this case.
We Should Be Suing Software (Score:2)
For me it was KDE2. But I didn't pay for KDE. From my word processor to the OCR that came with my scanner, there is so much junk sortware it's depressing.
Mozilla 1.0 came close.
Who can blame them? (Score:3, Interesting)
The facts ... (Score:4, Interesting)
What Intel has been doing to make chips faster ever since the 486 has been adding more execution units. The 386 had 1 execution unit, 486 had two, PII and PIII had 4, and I *think* the P4 had 8 units?
Why did Intel do this? They were scared because AMD beat them at their own game. Intels self esteem was damaged -- So they launched an agressive marketing campaign, and used these tactics to maniupulate the marketing metric, MHZ. Ceartinly sleazy.
You'll notice now that Intels best P4 is faster then AMD's best part right now -- they've backed off the agressive advertising. However, they burned enough geek karma that I'll never buy intel again.
To remedy the situation, processors ratings need to be measured in IPC*MHZ [instructions per cycle] for both integer and floating point operations. Then it would be pretty clear to consumers what was going on.
Re:The facts ... (Score:5, Insightful)
Nicely confabulated! When making stuff up to prove a point, you might as well go for the jugular. The pipeline lengths in question are 20 [realworldtech.com] and 10 [heise.de], not 40 and 7. Incidentally, a long pipeline has nearly nothing to do with "context switch", at least as that term is commonly used (i.e., switching from one process context to another). Any pipeline issues caused by a context switch are dwarfed a thousand times over by cache and TLB issues.
Aside: what is with the short pipeline fetishism on the part of AMD partisans? You guys realize that they had four- and fiv-stage implementations of MIPS CPUs back in the early '80's, right? Imagine how brilliantly fast a MIPS r2000 would be in 3.0GHz! Oh, wait, you can't make an r2000 run at that clock speed. Hmm. Maybe pipeline length is just one parameter in a complicated design space, and we should look on manufacturer variations as differing technical solutions. After all, that's how we treat cache design, functional unit choices, and myriad other microarchitectural parameters.
No, that sounds complicated. It must be an Intel conspiracy to corrupt our precious bodily fluids...
Why did Intel do this? They were scared because AMD beat them at their own game.
Then in a few sentences, you say:
You'll notice now that Intels best P4 is faster then AMD's best part right now...
Umm, so how was Intel "beaten at its own game"? A bit of history, for perspective.
The Pentium III is the same core that was originally sold as the Pentium Pro. That core was introduced in 1995, and Intel is still squeezing performance out of it. At the beginning of the PPro's lifetime, it was an extremely ambitious design for the physical processes then available; people called it a too-hot, too-big, too-transistor-intensive monstrosity that would never be practical. Towards the middle of its life, in the years '97 to 2000 or so, the PIII was nicely matched to the physical parameters of then-current fab technology, and Intel produced modest shrinks and speed bumps seemingly at will. Those were the salad years of the PIII. Now physical technology has moved further down the road, and the PPro core is showing its age. It's leaving performance on the table that could be scooped up with transistor-intensive techniques like trace caches, more functional units, issue width, etc.
Like almost every other design generation of every CPU, ever, the P4 has a more complicated pipeline than its predecssors. Just as in 1995, the first year showed pretty "meh" performance, with much armchair punditry claiming that it's a monstrosity. Now, about 18 months after its introduction, the P4 is scaling well. AMD, on the other hand, is struggling to wring a few more modest speed bumps out of the K7 before it limps along to the end of its design life. The AMD partisans hold out hope for the K8, generally forgetting that the K8 is a K7 with a 64-bit bag on the side.
It saddens me to type this on my Athlon, but there's a strong likelihood that AMD's years in the sun are over. Five years hence, we might be looking back at the years 1999-2001 as a lost golden age of competition in the x86 CPU space.
To remedy the situation, processors ratings need to be measured in IPC*MHZ [instructions per cycle] for both integer and floating point operations. Then it would be pretty clear to consumers what was going on.
Any simple attempt at measuring performance will end up being simplistic. The big problem with your proposal can be summed up as: which instructions? NOPs? SIMD floating point? The instructions that make up Quake III, or gcc, or my LISP stock market prediction application? What about when the instruction sets of the CPUs differ, ala SSE2? Performance characterization really is difficult; anybody who claims otherwise is trying to sell you something.
Benchmarking: Intel vs. AMD (Score:4, Informative)
The article later states that benchmarks would be more reliable. However, I've seen some benchmarks saying that the Athlon is a lot slower than the P4 (at least on Tom's Hardware).. Of course, this is comparing the P4 2400 vs. the Athlon XP 2100. Article here [tomshardware.com].
Tom's hardware mentions that you still get more processer power for your money, but it concludes that Intel is faster (at least in this comparison).
I doth quote:
"In the last "AMD vs. Intel" comparison, the Athlon XP 2100+ took the leading position by a nose, but now, the Pentium 4/2400 easily overtakes its arch rival. Meanwhile, you should keep in mind that that the P4 has a 666 MHz core clock advantage over the Athlon XP. "
So "whats up" with this article? Did the plaintiffs read this before they filed the lawsuit?? Is Tom's Hardware just another victim of the megahertz marketing machine? (Actually, the tests would seem to indicate no). By the way, I'd love to see the plantiffs win, because I get really sick of the megahertz crap that they ramrod down everyones throat. Not to mention, any computer illiterate person knows that "Intel is better" because of this.
At any rate, I don't really think benchmarks are the answer- everyone knows you can make a benchmark say whatever you want (see for instance the Pet Shop application debate w/ Java vs.
Re:Benchmarking: Intel vs. AMD (Score:3, Informative)
Thou, I do like the number of benchmarks Tom uses, lame, quake, scisoft sandra, pcmark, sysmark, specview, (I wish he would use madonion also).. But trying not to repeat a few posts, when an application is compiled towards the cpu, it will be faster. Look at Flaskmpg, AMD compiled version showed an incredible speed up. Same with GCC 3.2 (check the changes), they said an average of 8.7 (with 2.6 or something from 2.95) so thats around a 11+ percent increase, Average! 3DNOW or SSE2 Optimzation really makes a big difference on bechmarks, programs should support both.
Cant wait to see what happens when AMD starts its 3000+ chips, and the 64bit hammer comes out.
P4 is faster (Score:2, Insightful)
My older machine is a P3 650 mhz with 512 mb SDRAM.
The P4 is at least 2-3 times faster when I load windows and applications.
There are a few things to consider in addition to the processor speed.
First, the speed of the memory bus is important. That determines how fast it can move around pages of memory. If Rambus hadn't tried to screw everyone then Intel wouldn't have had to scale back the memory bus speed in the P4s by bringing back SDRAM. As a result, using a motherboard with SDRAM slows down the P4.
Second, the amount of memory hasn't improved much. The P4 boards have the same number of RAM slots as the P3 boards did. If you have a lot of programs open or a huge 600 mb file, then 512 mb on a P4 will feel like a Pentium Pro when it starts having to use the hard drive for swap space.
Third, check your hard drive bus speed. Is it a 66 or 100? Mine is older than that and i'm sure I takes a performance hit.
Re:P4 is faster (Score:2)
when it comes to running everyday type apps, processor speed (once you go over 500 mhz or so) just isnt a factor, its everything else that counts.
Just LIke the Old Days (Score:2, Funny)
Q: What is the difference between a car sales man and a computer sales man?
A: The car sales man knows when he is lying.
tada da boom
apples and oranges? (Score:2)
Um....wow. You mean to say that Intel's 2.53 Ghz chip "edged" past AMD's 1.75 Ghz chip (that's only marketed to beat an Intel 2.1 Ghz chip) --- real impressive there guys...thanks for pointing out how quickly that gap is closing.
???
--noah
Its all about the Chipset. (Score:2)
I now am typing this on a p4 1.8, and am quite happy with the performance, and stabillity with the intel chipset. Had Via solved the PCI bus bottleneck problem they were wrestling with, I would have gone with AMD. No dice, though.
This is much faster than my P3 800, (and light years ahead of my p1 225) so I'm fine and happy with my purchase.
Misleading clock frequencies! (Score:3, Insightful)
Intel didn't design the chip just so that it would have a higher clock frequency and therefore mislead people into thinking their chips were faster. They came up with a whole new processing architecture, that simultaneously created a large efficiency drop in instructions processed per clock cycle but allowed for much higher frequency operation. The end result was faster processors, but the clock frequencies didn't correspond. Not their fault.
Further, end users should have been used to the idea that clock speed and processing speed didn't correspond; AMD's processors had been outperforming pre-P4 processors, clock cycle for clock cycle, for a while. AMD didn't start their "processor equivalent" labeling scheme 'til the P4s came on the market, though.
can you pay a group of people to file lawsuits? (Score:2)
this sounds suspiciously like AMD is astro turfing (or a variation, involving filing lawsuit rather than calling your local congressman) some bad PR for intel to me...
Wrong Headline! (Score:2)
It should be "A pair of lawyers engaged in extortion...: Of course, that is so common that it doesn't need a headline.
No, this isn't meant to be a troll. Most people don't realize that "a class action suit filed on behalf of X million..." usually results in tiny rewards for those million (or no reward) while it results in vast sums for the lawyers who file the suit. Furthermore, because of the absurd state of US tort law - especially in some tort friendly states - Texas and Louisiana.
Note that the complaint claims that the total aware will be no more than $75,000. Of course, this does not include lawyers fees! My guess is that the lawyers put this in so that a court will find it easier to give them a win, or so that the companies will settle.
Once that is done, the real fun will begin. Having already either lost one of these cases, or settled one, the companies will then be attacked in Texas or Louisiana or another state where the tort lawyers routine win obscene settlements. They will cite the previous attack, and pocket zillions of bucks in the resulting easy win.
What will PC owners get? Probably discount certificates allowing them to buy a new processor from the defendant at a lower cost. This is how a typical american class action consumer lawsuit works!
Note that none of this has anything to do with the merits of the case. Personally, I think the case has no merit. The companies didn't lie(although AMD *does* act in a more deceptive matter - did you know that an AMD Athlon 1700+ does NOT have a clock speed near 1700 Mhz?). The consumers weren't deceived, unless they fooled themselves!
Overnight to England takes Three Days (Score:2)
Re:Overnight to England takes Three Days (Score:2)
USPS Response (Score:2)
Re:the article (Score:2)
$75,000 x every intel PC purchased in the last year = ouch.
Re:They have a point (Score:2)
Too wonderfully right. Nothing expressed this more to me than seeing the difference between a 1992 spec machine (33Mhz bus, 33Mhz ram, 33Mhz cpu, 33Mhz cache...etc) and a current spec machine, where every one of those (add in the drive speed and a few others here and there) are all operating at wildly different speeds. In one 2Ghz machine, how many different hertzes(!) do you have?
dana (having a horrific grammar day - but you get the idea
a grrl & her server [danamania.com]
Re:They have a point (Score:2)
Re:They have a point (Score:2, Funny)
Re:Can we sue AMD too? (Score:3, Interesting)
Whatever it was, the rating means, a chip of the older architecture would have to be at or above the rating MHz (2100+ in your case) to give the same performance.
It's actually a decent representation of performance, unlike the Intel higher clock speed but lower bandwidth.
Re:P4 vs P3 (Score:2, Insightful)
(a 486 133 is faster than a Pentium 60)
Re:Way too stupid (Score:2)
Re:Great (Score:5, Insightful)
People have always paid more for Intel CPUs, just as they pay more for many other "name brand" items. If you were to legally pursue every well-known company that produced an item that cost more, yet had inferior quality to competing brands - you'd be in court with just about everyone.
It's *always* up to the buyer to do his/her research. If he/she still decides they prefer Intel just because they like knowing their chip is backed by the largest CPU maker in the world - so be it.
(And anyway, there's more to it than Mhz. Some people, like myself, went with a P4 because we preferred the overall options and quality of the motherboards. AMD had problems getting the "tier 1" motherboard makers to build boards for their CPUs for quite a while.)
Re:Great (Score:2)
To be fair, AMD acted well on this issue and inclues a temperature probe in their chips. It's still up to the MB to read it and turn off/down (I don't know which) the power, which only two have done so far. But any new MB introduced now will have to have it to become AMD certified.
Re:Great (Score:2, Insightful)
Anti-benchmark EULAs (Score:2)
All you have to do is run some benchmarks to prove [that the Athlon is faster clock for clock than the P4].
In that case, Microsoft SQL and Oracle are infinitely slower than MySQL and PostgreSQL because I can't even get past the stupid EULAs that make me promise I won't release benchmark results to the general public. Watch for Intel to start pulling the same sh*t when the AMD Opteron trounces both the Pentium 5 and the Itanic [intel.com].
Re:So what? (Score:2)
You're right, except they have embraced this false association and now present it in their marketing, e.g. "...with a blistering fast 2GHz Pentium 4 processor..." MHz/GHz numbers are the only thing in most ads that even come close to a performance rating, so by focusing on them as the sole performance metric, Intel and the OEMs are implying that this is a valid way to measure overall computer speed, even across platforms and/or different processor architectures.
Just because the public at large believes something, that doesn't mean those beliefs can be presented as fact in marketing materials. There just may be some merit to this case.
~Philly
Re:USPS? What about NYC MTA (Score:2)
"Didn't anyone tell you about the 18 minute rule"?
"What 18 minute rule?" I replied
"You can't use the card twice in any 18 minute period."
"No, nobody explained that, I just used the machine over there to buy one. I went onto the wrong platform. Can you let me onto the platform please?"
"No."
After a lot of arguing I thought 'fuck it' and got a taxi. On the whole I really enjoyed NYC, the only two things that pissed me off were that woman, and the fact the platforms aren't air-conditioned. It was like a furnace at the 34th Penn station.
Re:USPS? What about NYC MTA (Score:2)
Re:Other Companies may sieze upon this lawsuit... (Score:2)
Apple? "Photoshop benchmarks are us?" That Apple? The one that's too embarrassed to post any benchmark results to the Spec organization (http://www.spec.org)? Don't make me laugh.
C//
Re:What's wrong with the USPS? (Score:2)
Worked fine for me. I've gotten plenty of mail sent to my old address, and I did move across the country. Are you sure you filled out the form in the correct language?
But in all seriousness, there are certainly crappy post offices filled with workers whose sole desire is for you to go away and leave them alone. Try a different one.
Re:What's wrong with the USPS? (Score:2)
Re:Slower? (Score:2)
Of course, the 2200+ isn't claimed to be as fast as the 2.53GHz P4. It does match up nicely with the 2.2GHz, though. I think the lawsuit comes from earlier P4's, in the 1.4-2Ghz range, which -were- slower despite Intel trying to make it -seem- like they were faster (though carefully not specifically stating such).
I think it's a stupid lawsuit, but the facts of the case are there.
Re:USPS (Score:2)
I think Timothy was just using the USPS reference as an example of something obvious. Most post offices I have used do indeed have long lines and employees who don't give a shit.
'the Pentium 4 is less powerful and slower than the Pentium III and/or the AMD Athlon.'
This is the relevant example, but it is probably only obvious to a smaller crowd - those of us that are actually interested in the processor speed wars. Timothy's point was probably that people who chose the P4 should have known what they were getting into, hence the comparison with the USPTO. Personally, I would have written the blurb differently, though.
Re:Confusion between ... (Score:2)
Yeah, but clock speed efficiency (IPC) in computers is meaningless -- and nobody cares about it.
According to SPECint the fastest CPU right now is Pentium 4 @ 2.53 GHz, and the second fastest is McKinley (aka Itanium 2) @ only 1 GHz. They are roughly equivalent in speed, but by IPC McKinley is much better. But so what? You can't run McKinley any faster than 1 GHz, and you can run a Pentium 4 at 2.53 GHz, so why even make the comparison?
Modern CISC CPUs are emulators (Score:2)
MGhz for Mghz a RISC chip kicked the shit out of CISC and stole their lunch money. If I'm not mistaken, they still do.
Not especially. Modern CISC CPUs such as the Athlon, the P4, and the Crusoe recompile CISC bytecode into RISC micro-operations internally. The problem with the P4 is that the decoder isn't fast enough (one micro-op per clock for non-cached instructions; three micro-ops per clock for cached instructions) to feed the P4's nine functional units.
Re:If its one thing people don't like to admit... (Score:5, Funny)
Rumor has it that the next release of M$ Office will have minimum requirements that are close to this.