Slashdot Mirror
P4 3.2GHz Reviews
Nathan writes "The Intel 3.2GHz Pentium4 has passed its NDA with reviews coming out over the net, including this one at MBReview, This one at HardAvenue, This one at TweakTown and this review at HotHW." Yay. Benchmarks. Wowee-zowee.
"I also reserve the right to mock you for paying $300 for an extra 200MHz." -- Scott Wasson, TechReport.
Yay. Benchmarks. Wowee-zowee.
If it isn't important, if it doesn't matter, then don't post it.
There is no excellent beauty that hath not some strangeness in the proportion. -- Francis Bacon
$760 for it... A bit much for that "little extra," isn't it? You could build a fairly powerful AMD (or even Celeron) machine with that money... twice.
Let me guess... It's a few percent faster than the 3.0ghz, and costs more.
Do I win a prize??
Overclocked to 4GHz? Not bad. I'd love to see this against these supposed new Macs coming out today.
Numerous times I've seen people accuse moderator and editors of being on crack. This one by Hemos might just have me convinced. :-)
zWhat would an EWOULDBLOCK block, if an EWOULDBLOCK could block would? -- me
Is Intel trying to get laid by the best of the PC market by showing how fast it can swing by?
What happened to the days when CPU's would take their time, and get the jobs done the right way.
It's not like it can make your PC scream any faster or louder, or can it?
like 800MHz FSB (albeit 200MHz quad pumped) and Hyper-Threading.
"I just can't sit while people are saying nonsense in a meeting without saying it's nonsense" J Watson, Sci Am 288:(4)51
Whilst I would extend my sincerest thanks to dear Intel for yet more predictable inching up of the top speed for x86, I would like to point out that a far more interesting processor revolution is to take place today at 17:00 UTC, in the form of the PowerPC 970.
64bit for the consumer and the world's most beautiful OS or a meagre increase for a 32bit chip with Microsoft Windows. I know what I'll pick...
iqu
The pentium 4 architecture (heck the x86) is getting long in the tooth. I foresee intel's next market move :)
Intel Employee #1: We can't make our design any better! Intel Employee #2: Surely you jest. Intel Employee #1: No, but I have an idea. Intel Employee #2: What? I'm clueless! Intel Employee #1: Lets up the clock speed! Intel Employee #2: Touche!
(note this is not meant to be a flame, just a little humor)
- tristan
Find some German Reviews at www.hardtecs4u.com, www.computerbase.de, www.hartware.de und www.hardware-mag.de.
;)
Looks, as there is no chance for an AMD 3200+ Systeme to win a round. Hope it will change with the athlon 64
Oh really, I support AMD. But I have to ask, how the hell did you determine that? ...What's that, you pulled it out of your ass? I thought so, Troll.
- Xenius
Picture this....
Salesman: and this toaster makes toast .5 seconds faster
Me: great, how much?
Salesman: its double the price of the standard model
Me: Hmmmm
(Yes, fellow pedants, I am aware that "Pentium" was used for the chip following the 486, as Intel couldn't copyright a number and stop their competitors using the term "586".)
Seriously though, how long have successive generations of Pentium technology lasted? Is it just me, or was the PIII the primary product line for longer than the PII, and when will the P4 break the PIII's record?
When I am king, you will be first against the wall.
Yay. Benchmarks. Wowee-zowee.
If it's that boring, why include it on the main page as a story?
Trolling is a art,
Just for the record, that's from the Tech-report.com review. Forgot to link in my eager to be Mr. funny-fastposter-man.
Compared to the older pentiums the new pentium IV performs all the same instructions in exactly the same way. You may sense a small speed increase; however you are not likely to notice it (unless you are upgrading from a 486DX2-66).
Integer performance has increased by (New Speed-OldSpeed)/OldSpeed * (OldBenchmark Score) - OldBenchMarkScore, as has floating point. However, the electricity bill also rose by the same percentage.
Pros: No one ever got fired for buying Intel!
Cons: It costs more than a used car!
If you run a toast shop, and you're making 5000 slices of toast a day...
http://twitter.com/onion2k
There are many organizations that do not have a budget or process for replacing obsolete/outdated equipment. Like rain in the desert, money for new equipment comes in infrequent deluges. When money is available, you buy the top-of-the-line computer. You may be using it for the next ten years.
Mea navis aericumbens anguillis abundat
Come on beyond a few people this sort of speed really isn't necessary is it!
For most people when processors hit 750 mhz that was enough for them. And then MS released XP but that only raised the stakes a slight bit. 1.2 ghz is enough for 90% of people out there!
Yet some people still crave speed, I have an aunt who does nowt more than send a few emails a month and play minesweeper and (much to my annoyance as I may use it for maybe 5% of my tasks) she has a faster cpu than me!
On a side note, what's happeneing with AMD these days? they seem to really be losing it at the high end, it terms of both value and performance. there 3200 seems only about as good as a p4 2800 of so.
Still they still are the better choice at the same end of the pricing scale below the curve of insanity!
Personally I'd much prefer some nice advances in some other area, cpu's are dull these days and I doubt 64 bit will convince me otherwise.
+----------------- | What is the question!
Mac with PowerPC 2 GHz is coming out today.
That's what I'm waiting for. I weant to see head to head comparisons versus this P4.
It's right here: Pentium 4 starts to die, all hail the Pentium 5.
"INTEL HAS RELEASED a 3.20GHz Pentium 4 microprocessor which will likely be the last of its kind until it introduces its "Prescott Pentium 5" design later on this year."
The same way that old PPC chips are faster. The performance per clock cycle is a lot higher, of course if you can't clock them high enough this doesn't really help.
Look at the price performance ratio though, you can build a whole AMD based PC for the cost of high end P4 processors.
Extreme Overclocking: they actually overclocked the engineering sample. ha! kind of a pricy risk if you ask me. More reviews here, here and here.
Nothing newsworthy in that really.
Thomas S. Iversen
Oh, I agree completely. AMD's chips are twice as efficient as intel. Isn't the XP's pipeline about half that of the P4? And yes Intel is obviously more expensive. With the big name comes a big price to boot. I dunno where the post I replied to is, but it was just another unintelligent post giving AMD fans a bad name. P4's price/performance ratio sucks compared to AMD's. Hands down.
- Xenius
I find it irrelevant whether the speed of an existing type of processor has increased by less than ten percent, although looking at the price compared to the 200MHz lower clocked variant, maybe this would fit under "It's funny, laugh".
;)
However, this processor does seems very suitable for overclocking (4GHz, yikes!). Did anyone manage to come close to that with the 3GHz model, or has Intel increased the therapeutical window of their processors slightly?
yes of course, but the toast shop making that volume of toast would not be buying toasters built for home use. They would have a toaster farm probably running 'open toast' insted of winToast
It only fits if you do it with a French accent.
Mod me down and I will become more powerful than you can possibly imagine...
My Ti Powerbook G4 running at 800MHz runs just fine and it gets 6 hours of battery life. When are PC users going to realize that you don't need any more performance than that? Power savings is more important these days.
No, I don't give a flying fig about buying a 3.2GHz P4, but once it's out, the price of the 3.0GHz model (i.e. the new next-to-fastest) will get much more reasonable. And I'm planning on building my next system soon, too.
What's with the sig?
8 of 13 people found this answer helpful. Did you?
real world perfomance doesn't seem to make a lot of difference, for what you have to pay extra and who is will to cough up that extra $$$ to see UT 2003 jump from 223 to 242 fps, you can't even see the difference with your naked eye!
Something is missing, oh yea, the Intel vs AMD benchmarks, WTF, how can you compare your own CPU's to each other, I wanna see how they hold up to AMD
If there is nothing left worth living, what are you willing to die for?
If you need that kind of toast-making performance, you're luch more likely to either build a toaster-farm with dozens (or maybe even hundreds) of inexpensive run-of-the mill toasters, or splurge for a big, heavy-duty continuous-feed made-to-order beltway toaster.
Sort of like getting either a cluster of cheap middle-performing x86 boxes, or a big-iron type machine from Sun or IBM, come to think about it.
I mean, how many apps really critically need that 2% parformance increase, but do not benefit from a dual or quad-cpu machine, a cluster, or a big non-x86 Unix machine?
Trust the Computer. The Computer is your friend.
I'm waiting for the big announcement so I can finally "switch".
Bye, Microsoft. I never want to see your name on my computer again.
Hi, Apple! Nice to meet you!
What the fuck is an "atahalon" I mean, I could understand it if you were the common or garden lesser moron, but you actually managed to spell it correctly in your subject. You must be some sort of special moron.
I mean, how many apps really critically need that 2% parformance increase, but do not benefit from a dual or quad-cpu machine, a cluster, or a big non-x86 Unix machine?
A quad-cpu machine with the new CPU will still be faster than a quad-cpu machine with the old CPUs.
When you run thousands of jobs on a hundred dual-cpu machines overnight, 10% increase in speed is significant! It means hundreds more completed jobs per night.
When the fastest available is too slow, then any increase in speed is quite welcome.
Never underestimate the bandwidth of a 747 filled with CD-ROMs.
Of course it's predictable, it's been on their roadmap for some time :)
Heck, you'll be able to "predict" the next few releases as well!
It's really up to Intel as far as when they change the name. Presumably they could do it when the Prescott core rolls out later this year, but they could also just say it's a P4 with a Prescott core. Though that wouldn't mean much to the consumer, so they'd probably name it the P4 1MB cache or something, like with the change to the Northwood core.
What about the AMD Athlon64/Opteron? Soon Windows will have a 64-bit consumer version, and Linux already does have.
:-)
And you get x86 compatability too. I'll leave it upto the reader if that is a good/bad thing!
Ok, I have a small rant concerning benchmarks. I'm in the sciences and often look at graphs of data. I am getting SO TIRED of benchmark results being posted with y-axes that go from 2500 to 2600 showing the relative "improvement" of newer, faster cpu's when they ought to be scaled from 0 to X "mips", "flops" or whatevers so that you can see at a glance that the changes are or are not significant.
Better yet are plots showing how much they have "improved" relative to simple clock speed increases (if at all!) and normalized "mips/dollar" for cost evaluation....
I've spent several thousand dollars improving my car's acceleration, brakeing and handeling. Would the average car driver notice? Maybe. Would they be willing to pay what I did for the improvement? Nope.
Electronics is the same as well...high end audio and video systems have a wide range of options.
Makes perfect sense.
Blar.
For those who care, there is also a comparison of AMD 3200+ to P4 3.2 GHz at tomshardware: here
There are a thousand forms of subversion, but few can equal the convenience and immediacy of a cream pie -Noel Godin
Modern chess engines represent the board as several 64bit bitboards, one for the white queen, one for the black queen, one for the white pawns etc.
This as opposed to the good old days with a 64 byte array containing 1 for the white queen, 2 for the white pawns etc.
Bitboards really benefit from 64bit registers and 64bit (integer) arithmetic.
"There is no substitute for thinking" - Bjarne Stroustrup
While not earth-shattering news this is still good news for people who use computers for more than an excuse not to interact with live humans.
Yet every single time there's a news item like this, some moron like you comes along and wonders aloud why he needs a faster computer just to use Mozilla. Guess what: you don't! Guess what else: other people have real work to do, and we DO need faster CPUs.
So to you and everyone else who keeps asking that same dumbass question: STFU already! Just because you're an idiot doesn't mean you have to advertise it.
Thank you.
...Service Pack with this thing. Real life processor benchmarks:
Time to load bloated Office Apps
Time to load ridiculously large email cache
Time to load a half gigabyte MPEG
Ruger
Posted by Hemos on 8:42 23 June 2003
from the cut-and-paste dept.
Nathan writes "Someone else asked us to redirect traffic to their site. We told them of course."
Processor design needs to change. Just to put things to rest, I'm a Macintosh zealot. Intel keeps pushing the clock rates higher, which places more demand on power requirements (the chip itself and cooling), hence most windows users (secretaries, cublicle workers) in an business office environment never need to have space heaters under their desks to keep their legs warm in the winter time. The PowerPC RISC processors are going in the right direction, but let's take a look at the graphics card processor chips. These chips run at lower clock rates, use less electricity, and move MASSIVE amounts data and calculate a metric ass-load of computations. Processors need wide (128-bit or more) and shallow pipelines to get *the best* performance. Looking at the graphs from the article (yawn), well, they look pretty linear. Ramping up the clock rates with a 800 megahurts FSB (PPC 970 has 1GHz FSB) is eventually going to lead to a starved processor (i.e. Motorola PowerPC G4). Well, enough ranting. Intel marketing (girls dancing, chip technicians in space suits doing the disco) prevails.
14.4 was just fine for me, I didn't bother with a 56k modem for at least a couple of years.
now I have a 2 meg ADSL connection and it could do to be a bit faster sometimes.
thank God the internet isn't a human right.
I recently got a 2.4 GHz Pentium IV box, courtesy of my employer. Much to my surprise, the benchmarks for the stuff I am doing (cryptographic algorithms) tell me that this box requires more than twice as many CPU cycles to complete the job as my 2-year old 1 GHz Athlon box.
Mind you, this is exactly the same code. Even when running the same assembly language code Athlon seems to be much more efficient. I wonder if they have fixed this in the new Pentium IV?
OK, who has one of these and want to share your experiences running Linux (2.4 and 2.5) on them.
Does AGP work? DMA? the onboard NIC? onboard Sound?
But then you can usually get the same 10% speed increase less expensively by adding some more non-bleeding-edge machines instead.
Point is, the difference between today's greatest, and the next-to-greatest in speed is small enough, and the price large enough, that it really is very difficult to motivate going for the bleeding-edge solution anymore.
Back in the days of 100Mhz Pentiums, things really were different; the difference was often the ability to bew able to run an app or not, between the latest and the year-old stuff. The speed difference was typically huge. That is no longer the case.
Trust the Computer. The Computer is your friend.
Toaster farms are unrealistic. I worked for a startup during the bubble that wanted to revolutionize the market with huge toaster farms. They went bust in under 18 months after burning (no pun intended) $20 million in cash.
It turns out that except for the big guys like Melba, nobody needs that much toast. And the big guys already have their own custom big-iron industrial toasters. They're not going to outsource their bread-and-butter business to some pimply startup with an array of flimsy Sunbeam toasters.
In hindsight, that was a stupid business plan.
While those reviews are more than adequate, I am surprised that Tom's Hardware review is not mentioned. While I would not mention it blindly just because THG was one of the first sites to offer in-depth reviews, after reading it I gained more insight than from the other "here are the benchmarks, mam" sites. Here's the synopsis:
"Intel launches the last P4, with 3.2 GHz for FSB800 and Dual DDR400. Its rival AMD fights back with the Athlon XP 3200+ and Dual DDR400. With the Pentium 5 and Athlon 64 waiting in the wings, it's a historic duel." [tomshardware.com]
Must-not-watch TV!
I have to agree. They're comparing a 2.2GHz AMD chip with a 3.2Ghz P4 and the AMD chip is holding it's own against a CPU that is an entire 1GHz faster in terms of clock speed. If they would A.) Compared the AMD chip with a 2.2 GHz P4 or B.) Compared the 3.2GHz P4 to a 3.2GHz Athlon, if one existed, then AMD would be far ahead. Which is why we recognise the Athlon as the supperior CPU.
---- "Excuse me. Where's the children's gun section?"
There are many organizations that do not have a budget or process for replacing obsolete/outdated equipment. Like rain in the desert, money for new equipment comes in infrequent deluges. When money is available, you buy the top-of-the-line computer. You may be using it for the next ten years.
Ahh so you must also be working for a city govt too!!! Oh well, at least while we city govt IT workers may get stuck with using dinosaur equipment and technology, which another poster commented is "short-sighted" business practice, at least we are still working with computers and drawing a steady paycheck, instead of working with the "technology" of operating deep-fryers and drive-thru point-of-sale equipment and intercoms while wearing goofy hats.
"What's the use of having an artifically-intelligent toaster IF NOBODY EVER WANTS ANY TOAST?!"
Not to mention the power requirements and heat output.
I can't stand noisy computers, my Athlon 2100+ is inaudible from a distance of around 10 feet.
Mod parent off topic. We're talkin' about toast now..not cpu's. :)
--
"I'm not bright. Big words confuse me. But Wanda loves me and that should be enough for you." - Cosmo
Uhmmm.... Would that be a beowolf cluster of toasters then?!? ;o)
I agree that this article wasn't that interesting, but I did find the following passing interesting
Next up, weâ(TM)ll be taking a look at FutureMark's 3DMark2001SE. With the recent debacle surrounding NVIDIA and FutureMark, I have chosen to exclude 3DMark2003 from our benchmarking suite for those of you wondering why you arenâ(TM)t seeing any results for it. (from here)
We've all read how NVIDIA fiddled with the results and how FutureMark became complacent with it. Now here's the result.
-no broken link
There will be dozens of people saying that they're sub 1 GHz processor is "fast enough". Why bother saying that. Some people want faster computers. Simple as that. It's their money, let them spend it. Personally, I haven't upgraded in 3 years and I could use more speed to process digital video.
I'd like it if they compared on equivalent price (since that's what many people including me are limited by... there's only so much I'll pay for a processor).
Athlon XP 3000 (Barton) = £175
P4 2.8Ghz Northwood 533Mhz = £175
A benchmark comparing to two would be interesting.... not entirely equal as P4 mobos are a bit more expensive than AMD ones.
Once you get into dual processor of course, it gets silly:
Athlon MP 2600+ Tbread x2 = £292
Xeon 2.0ghz Prestonia x2 = £304
It's obvious in that comparison the AMD would crap all over the Xeon.
I disagree about your theory of the toaster farm. We have real-life data: Krispy Kreme.
OK, it's not toast, but when a large company wants to sell donuts in every town, they custom build a mass-production donut machine, rather than installing lots of little donut machines.
The whole analogy is toast, just because both alternatives (cheap toasters and expensive slightly-faster toasters) would not be used in real life.
Hell, even restaurants and bagel shops use Hobart brand belt-fed toasters, which happen to be far more expensive than ordinary consumer toasters.
144l. ph34r my 133t l3g4l 5k1lz!
I can think of MANY tasks you can't do well with a 450mhz cpu. The question is -- how many of these tasks apply to *YOU*? By the same token, I'm sure there are people out there who are just fine using their 486 66mhz computer with 4 megs of ram. I can't fathom why some people assume because *THEY* have no use for something faster, that others wouldn't.
SearchIRC - Now with live chat directory!
I did mention belt-fed toasters (used the wrong word for 'belt').
:)
And yes, of course the analogy is silly. When has that stopped anyone?
Trust the Computer. The Computer is your friend.
You're never going to put this processor in a laptop anyway, so who cares. Apart from the electricity bill, it's not like I'm going to run out of battery life on my desktop. OTOH maybe your argument was that one computer can act as both desktop and laptop, since you don't need more performance than you get with a low-power chip.
BTW, last week it was +3C here in the middle of the night, and this is summer. 90% of what my computer consumes I can subtract from the space heaters... Personally, my biggest gripe is with noise. More power use = more heat to be transported away = more and louder fans. And you can't get a cheap low-power chip for the desktop, as it would eat out of the high-margin laptop market.
Kjella
Live today, because you never know what tomorrow brings
Much to my surprise, the benchmarks for the stuff I am doing (cryptographic algorithms) tell me that this box requires more than twice as many CPU cycles to complete the job as my 2-year old 1 GHz Athlon box.
Heh. You should try a G4 (or, if they come out today, a 970). Altivec has a couple SIMD instructions good for codes, such as SIMD rotate. See distributed.net for more info. (Their stats are for RC5, which is particularly nice on a G4; the improvements with AES, TWOFISH, etc would be somewhat less; with RC4 there's basically no difference).
I hereby place the above post in the public domain.
x86ers would like it very much if 64bit wasn't relevant for the consumer, because they're not going to get it for a little while yet, but in truth, it really is huge.
Really?
Maybe all readers here think "800 MHz is enough for everyone" but I demand more.
Consider xmame, the UNIX/Linux port of MAME. Sure, 800 MHz will play Pac Man full-speed. It should even play early '90s games full-speed. It will not cut the mustard beyond that. Emulating games from 1995 need a top-of-the-line CPU or better. I doubt a CPU exists on earth that can run crusnusa full-speed. Maybe if the P4 hit 5-6 GHz?
xmame is very much CPU-bound. It is single-threaded and changing this will not help performance. And aside from optional OpenGL front-ends a fancy video card won't accelerate anything in the core.
On x86 there is now a fancy "dynamic recompiler" for MIPS code that finally brings MIPS-based games like kinst into the realm of playable on faster AthlonXPs and P4s, but this does not exist for other CPUs, so Mac owners are out of luck. You'll need that PPC 970!!
MAME reports average framerate. It would be very interesting to see kinst and crusnusa benchmarks--with dynarec turned off on x86 of course. Then we could really see how the PPC 970 stacks up to the P4 and AthlonXP. Since xmame and presumably MacMame are all open-source (derived from the core MAME code) you can optimize the heck out of the compiles for each platform as well.
From what I understood no it isn't. Both the address AND the data registers are 64-bit, and it is the data registers where you would expect to see the speed boost.
The early pentiums were called P54, and then P54Cs. Right now we have the Pentium IV C steppings, which is right on the dot.
The P55 did introduce MMX, so we should be expecting a useless feature-set designed mostly for marketing with the next core revision that will introduce Pentium 5.
In accordance with Intel's heavy use of "Pentium" "5" and "S" (from SSE) I'm expecting the new instruction set to be named PIS5.
... that the P4 has pretty much stalled around 3 GHz fr a while now. They were really ramping up the MHz about a year ago. Sheesh, are the suffering from motorolia? As a Mac user, I feel your pain.
"Politicians find new names for institutions which under old names have become odious to the people."
Intel released their new chip, 3.2ghz P4 which is actually 1.6ghz P3 renamed. The other option is 3.2 ghz P4 HiSpeed, which is actually 3.0 ghz P4 with speed overclocked by 5% and price overclocked by 50%.
the major advances in civilization are processes which all but wreck the societies in which they occur - A.N. White
I mean, how many apps really critically need that 2% parformance increase, but do not benefit from a dual or quad-cpu machine, a cluster, or a big non-x86 Unix machine?
Seven.
Oh, was it a rhetorical question?
-- Repeat with me: "There is no right to profits".
http://www.inminds.co.uk/boycott-news-0194.html
http://www.inminds.co.uk/boycott-intel.html
Like... Multics? (circa 1960)
In many ways, Unix mmap() (or the Windows equivalent) probably counts too though.
DNA just wants to be free...
Boeing did 707, 727, 737, 747, 767, 777. (Maybe I missed few.) Now they are promoting the 7E7.
Just call it "Pentium^2"
"Sic Semper Tyrannosaurus Rex."
I had planned to buy a 3ghz P4 machine next month, but now that I know the 3.2 is the last P4, I don't think I'll do that. It would be silly for someone like me, with limited funds, to buy top of the line when an entirely new chip will come out next year. I'd rather buy a cheaper 2.4ghz and save up in anticipation of the new hotness.
Because as all 31337 hax0rz know, Instructions Per Clock is the ultimate measure of superiority.
wide (128-bit or more) and shallow pipelines
One more thing: Pipelining is actually a *kind* of parallelism. Gpu's and Vector computers are not only very wide, but generally very *deeply* pipelined. People think of pipelining as only a way to increase clock speed, but it's more accurately a way to have multiple instructions executing simultaneously without having to duplicate hardware. Being able to increase clock speed is just a consequence of this.
The only time shallow pipelines help you are when you have control and data dependencies that cause the pipeline to stall -- iow, when you have non-parallelizable elements. And usually a deeply piplined version with these potential stalls will still run faster than a shorter pipeline -- especially when the code has very few dependencies. Also, how bad your hazards (potential stalls) are to your performance depends more on the structure of your pipeline than simply its length.
The following sentence is true. The preceding sentence was false.
it's not my business, if the athlon xp would be faster than the intel processor, IF they would get the same clock speed... the fact is: ATM there is no athlon with higher clocks than 2200 mhz and that's not enough to win against big-intel... but go on dreaming... AMD needs money to start selling athlon 64 cpu's ;)
...but he got +5, Insightful and you didn't. Who's laughing now?
The x86 series has had 64-bit instructions since 1997. I suppose if you want to exhault Mac for finally catching up 6 years later you're certainly welcome to.
Now they come up with an inferior implementation of SIMD (eg, MMX) instructions more than six years after Intel, and it's all the Apple fanboys can do to keep from creaming in their pants! What universe are you living in?
Why would you buy Xeons? Regular P4s work just fine in dual configs, no need for a specially priced model, like the Athlon MP.
"Hot lesbian witches! It's fucking genius!"
"how many apps really critically need that 2% parformance [sic] increase, but do not benefit from a dual or quad-cpu machine..."
Almost everything in the field of chip design fits your description. Our software can cost $100k per year, per copy. If you can get 2% (in reality it's more like 5% for 3.06 -> 3.2 GHz) speedup, well hey, you've saved the equivalent of $2000. If the delta-price of the chip is a few hundred dollars, you saved money! That discounts any time-to-market advantage you might get (which could be enormous) -- if your schedule is a year and you can shave a few days off, that's worth money too.
And yes, our jobs do not multiprocess/multithread well. Many, many jobs are like this. That is why people buy machines like these...
Hmm, what do you know, apparently P4s can't be doubled up... wonder why I thought they could be... I assumed that since the P3 worked fine, so would the P4.
"Hot lesbian witches! It's fucking genius!"
I have a friend in that business, and their gate-level simulations and VHDL seem to run reasonably well on multiple cpu:s. I know they run big runs over the weekend so they can farm out the work over multiple machines at the company without disturbing normal workflow.
And if single-cpu speed is so critical, at almost any cost, why run on x86 at all? Why not get a big Sun machine; it's not like not most chip design tools don't run on them anyway.
I did not say _nobody_ needed that speed increase (and, unless you fit your critical loop and data in cache, the increase _is_ more like 2-3%); very, very few do actually fit the profile of 1) needing maximum possible speed on 2) one cpu only, and 3) it must be x86 and nothing else.
Trust the Computer. The Computer is your friend.
Somehow this comment at the conclusions of the TweakTown review feels a bit funny, as does the difference between 3D Mark 2003 performances (4998 vs. 4765) commented as "3DMark03 shows that the gaming performance of the Pentium 4 3.2GHz is much more advanced than the AMD Athlon XP 3200+.". These folks seem to have a slightly different definition of "much more advanced" than I. Personally, I'd say that difference of about 5% is somewhat minor.
IMO performance of both CPUs is more than enough for a large majority of current games and applications.
Everyone who makes generalizations should be shot.
I'd like to know what tool they are using if they actually a multiprocessing a single job. Otherwise, they are doing what we do, use multiprocessors one-at-a-time to handle many different jobs in parallel (there is a big difference). When you do this, your total run time is dictated by the run time of the longest single job. Getting this time down requires a faster single-CPU result.
"Why not get a big Sun machine" -- obviously you have not looked at the SPECint performance of Sparc CPUs lately. Fairly old Athlons can trash 900MHz Sparcs. After long delays there are now available 1.2GHz Sparc CPUs but they get demolished by new-ish P4-Xeon boxes (and don't even get my started on Opteron, good lord!). If you don't believe me, look at www.spec.org and check out the CPU2k results. If integer performance is your thing, nothing beats x86.
No, that would imply that in the rest of the world you cool your coolant system. Which is phenomenally stupid.
Rather, you'd want something along the lines of "In Soviet Russia, CPU cools coolant system!"
"Never attribute to malice that which can be adequately explained by stupidity." -- Hanlon's Razor