Slashdot Mirror
Intel's 64-Bit Pentium 4s Hit The Streets
ThinSkin writes "Nearly 18 months after rival AMD released its 64-bit processors, Intel quietly added its first 64-bit Pentium 4 microprocessors to the market on Sunday. Four versions of the Intel Pentium 4 6XX series were announced at speeds up to 3.6-GHz, a frequency grade lower than the existing 5XX series. Prices will range from $224 to $605. Intel also added the 3.73-GHz Pentium 4 Extreme Edition to its lineup, a $999 chip that is fabricated on a finer 90-nm process than its older 130-nm P4EE components. As Slashdot previously reported, the 64-bit series will likely be the major enhancement to the Pentium 4 line before the introduction of the Pentium D "Smithfield," Intel's first dual-core part, which is slated for next quarter."
He's a good man, I've worked with him for over a decade. He's no Bill Brasky, though/
Now the A46 prices should come down a bit.
Um - A little slow? I've had an EM64T "630" for about 3 weeks now.
Was it just me or did anyone read that as the Pentium D "Seinfeld"?
Not that there's anything wrong with it...
Anyone know how much heat these put out?
Now Intel can cook my toast using more than 4 GB of memory!
I love you suckers out there who are buying these top of the line, bleeding edge chips. It brings the price of "outdated" hardware back to reasonable levels.
Now if you excuse me, I have a 486 DX4 100MHz that I've been keeping an eye on for a while.
Does that come with... uh... Intel Extreme Graphics? Sweet!
It would be interesting to see how much business Intel theoretically lost to AMD as a result of their failed first 64-bit attempt and subsequent delay.
I've only been running one of these chips for two weeks now.
You know, it's just struck me as being somewhat odd that Intel is naming it's chips Pentium 2, 3 and 4. Does anyone have any ideas why they stopped going from 486, 586 to 686 etc. I seem to remember the Pentium being associated with the 586s, and this name has stuck.
The Inquirer is reporting that Intel will counter the AMD Turion 64 Mobile Processor with a Pentium M Extreme Edition. It is an alphabet soup of potential Intel Pentium M releases and you'll probably have to read it through twice or thrice to understand it all, but an interesting and inciteful read nonetheless.
There's another interesting article about the future of 64-bit as it relates to Intel here.
And of course, we can't forget our beloved Celeron.
I just told a family member to "go with Dell" (ie. don't bother me:) ) and they went with a 3.2 GHz P4. Should they phone them up first thing tommorow and see if they can cancel the order? They had actually managed to haggle with the sales person and got free wireless NIC thrown in. Not great but should they hold off and go 64 bit? Will Dell cancel / change
an order the day after?
Are they really AMD-compatible?
Ever notice how Apple was doing really poorly when they were providing dozens of different system configurations on a fairly large handful of Mac platforms, and were suffering because of it?
The problem was that the consumers simply didn't understand which computer most favorably matched their criteria.
I see the same thing here with Intel's lineup. What is what? Why is this M? Why is that Centrino? WTF does "Extreme" mean in relation to a CPU?
It wasn't until Steve Jobs was able to cut through the bullshit and bring the Mac lineup back to 2 basic consumer platforms that Apple was able to enjoy the benefits of the Apple brand. Until Steve came back, it was just another PC outfit. Now, with Jobs at the helm, and through his seemingly infinite ability to grasp consumer wants and needs, Apple is enjoying a resurgence in popularity and relevance.
Without someone with a grand vision like Steve Jobs, Intel is going to continue suffering through doldrums trying to guide the market with its "alphabet soup" (which you so very astutely coined) without actually listening to the consumers.
Why is 64-bit so desirable? Is there much performance gain? Because from what I understand not all programs use 64-bit instructions yet.
I thought newegg had the 600 series out for a while now.....with EM64T. I know I bought a dell a week ago and they were selling 600 series procs with them.....so how is this new news?
I talked to some of the few people who were testing these. Apparently it couldn't keep linux running for more than 4 minutes. Lets hope Intel was able to fix that "issue"
Methink these new pentium64s should pay tribute to Rodney Dangerfield, coz they get no respect from nobody either.
Muchas Gracias, Señor Edward Snowden !
Now it's time for game makers to get with the game [dum dum TISHHHH]. I'm tired of seeing 32 bits of my AMD 64 wasted every time I game, and now that every major player has 64 bit processors succeeding 32 bit, they need to get with the program and stop wasting bits.
The Peanut Gallery, Ubergeek, Biblically Sober
NCAAbbs.com: Thousands of fans, Hundreds of teams, Just one place
Will this be backwards compatible or support 32bit software.
Its nice to be important but its more important to be nice
does it run linux?
(Sorry, no story about new hardware would be complete without it.)
Amen. This x000+ crap has got to stop. I'm t hinking we need some sort of ISO standard for clock speed, something we can compare PowerPCs, AMDs, Intels, and the like together without having to pull out a calculator and Google for benchmark tests. However it wont happen due to some stupid reason or other.
-Eric Smith
Wow its about freaking time! Consumer 64bit computing here we come! On a side note, who bets that these chips will be as defective as the first 386's (double sigma anyone?)
erm.. are you an idiot?
4000+ = 4 gig+ in Intel numbers, the system is simple as hell. AMD processors get more power out of a smaller clock speed because they do more cycles per second.
So no, they just label their products to compare them to the Intel line easily.
I like muppets.
Wow, and only 10 years after Sun's UltraSPARC, 13 years after the DEC Alpha, and 14 years after the MIPS R4000
Yes, I have or something simulations to run!
(is it ok to use * and HTML emphasis?)
What's the plural of emphasis?
Shut up. OK.
This issue is a bit more complicated than you think.
No, you're just an idiot. If you are unable to grasp the simplicity of the AMD naming scheme then you should be forced to use a Cyrix. It just shows how badass AMD is compared to Intel.
Question... do any of these processors come in the socket 478 form factor, or is intel forcing us to upgrade our motherboards yet again? If I have to buy a new mother board, I might as well go AMD this time around!!
Bryan
You Fat balding idiot!
Does Intel have a marketing department for these chips? Doesn't this chip add yet another layer of confusion on the Pentium 4 line that can only cause trouble in the marketplace?
We've got Celeron (shudder), Pentium M, Pentium 4, P4EE, P464bit... P1s and P2s lasted for about 250MHz each line, where the hell is the P5 marketing?
Or is Intel scared of 55?
If you can get all the manufacturers to agree, comparing systems based on FLOPS would be effective, because it would remove the irrelevant clock speed argument, and thus allow you to compare how much work can be done in a time frame by the processor.
Unfortunately, I doubt you'd get the manufacturers to agree to it, since it would make too much sense and allow an easy and unbiased comparison between their products.
Overrated / Underrated : Moderation
Both you and your parent are the reason x000+ exists in the first place, and from the sound of your post, what you are asking for is exactly what the AMD x000+ numbers are. If you rate chips by clock speed, you get numbers inherently non-indicative of performance. Intel recognized this long ago, and this is why you're getting ~4Ghz cpu's that don't run any faster than a ~2.5Ghz from their competitors. They simply broke up their clock cycle and did less each cycle.
Because of pipelining, they can have multiple instructions in the pipeline at a time which, if they could predict future instructions perfectly, would mean you'd have the effect of having a processor which actually did an entire instruction at 4Ghz. However, since they break instructions into such ridiculously small pieces (it takes over 30 clock cycles for one instruction to work its way through), this means they need to be able to accurately predict the next 30 instructions to keep up the speed. Of course they can't, loops and if statements and other types of flow control guarantee this. And so, a 4Ghz processory with 37 stages in its pipeline is not faster than a 2.4Ghz processor with only 20 stages. However, people like you and the parent insist on believing that clock speed = performance, which it clearly doesn't, and so, Intel wins by making design decisions on how to up their clock speed to the max without regards to performance. This has finally caught up to them as people are starting to realize the sham and AMD's marketing nomenclature is countering the misinformation.
That "Stupid Reason" is that clock speed is not everything, Thank you for proving to Intel that it's marketting is worth it though, There are a lot of other factors regarding chip speed.
So what's the speed of an Intel 560 ?
Ya. You should be modded down. Calling Intel the underdog ? You're account should be banished.
You sir, win the Ironic berating of intelligence award.
You call him an idiot, without realizing that a hertz and cycle are synonymous. A processor at 4 GHz has 4,000,000,000 cycles a second.
Furthermore, "power" is unrelated. Power relates to the wattage which the chip draws, and dissipates as heat, aside from IBM's POWER line of chips.
Third, you're thinking of instructions per cycle. Not the cycles per hertz, which is a nice constant 1.
Congratulations for the award!
> I'm thinking we need some sort of ISO standard for clock speed
I know!
We could call it "Hertz" (abbreviated as Hz), as a measurement of frequency where one Hertz means one cycle per second.
When used in relation to CPUs it could be prefixed with the SI multipliers G or M for Giga and Mega, and be used to refer to the speed of the internal clock by which the CPU synchronises its instructions. It however will say nothing about how many instructions happen in a cycle, or what those instructions actually do.
So I'm guessing you don't want a standard for clock speed, you want a standard for performance.
But regardless, if you make your entire purchasing decision off the numbers on the box (no matter what they mean) then you deserve whatever you get.
Advanced users are users too!
The problem is that with current CPU designs the Mhz is only a small factor in the processing speed of the CPU. The design of the CPU has a great impact on the processing speed.
Intel has adopted doing many smaller calculations quickly, while AMD is doing more complex instructions at a slower rate. The net result is the same processing speed but at vastly different clock rates.
I'm sick of idiots like you.
Higher clock speeds do contribute toward performance, but idiots like you think it's all there is to it.
But, the other big factor that goes hand in hand with this is how much instruction it can accomplish per clock cycle. The P4 is much worse on that point (worse than some of their older CPUs), so they have to increase the clock like crazy to get any speed. That's not an advantage at all, and it does not give the chip any more real life speed or anything over the slower clocked ones that can do more per clock cycle. (not counting cache/pipeline/... concepts either, which also makes a real difference)
It just goes to show how you know absolutely nothing about CPUs, not even the very basics.
Perhaps you're buying your car just because of BHP's or RPMs? Or digital cameras solely because of megapixel resolution? At least we know that's the way you're buying your CPUs, by a single number that's unrepresentative of actual speed, but it does make you feel good as you don't understand what it even means.
Donate anything 300mhz or above to your local school district. I am a computer tech for my local district and we need all the help we can get.
The AMD numbers are also useful because they cover more of the factors that affect performance. AMD often sells CPUs with the exact same clock speed but different XXXX+ numbers. The performance difference in that case is due to different cache sizes or CPU architecture generations, but you can get a rough guess for their relative performance just by glancing at the model number. You don't need to be familiar with the details of their entire product line just to get a feel for what each chip will do.
I'd mod ya down if I werent replying, but not because you're pro intel, just because your logic is flawed.
Its called marketing, you do whatever you can thats legal to get people to buy your product.
If a dumbass consumer things 4000+ is 4000MHz, especially when there is a asterik saying that the processor actually runs at 2.4GHz, then they're probably the ones who dont understand the 9/10ths of a cent at the end of all gas prices. Intel made processors all about MHZ and thats what AMD has to compete on. The processor ratings are somewhat accurate (varies depending on benchmark).
And FWIW, a 4000+ does outperform a P4-2.8GHz.
The Doormat
If you're not outraged, then you're not paying attention.
No problem, even with a field of smiths, Neo beats them all.
(before chilling out)
Its VERY hard. Want to standarize on Spec_INT? Intel has optimized their compiler to perform better (that it actually can in real world conditions) in the SPEC benchmarks.
Flops (floating point Ops per second) arent that good because while a processor can process so many Flops per second, stuff like branch mispredicts and caching strategy can affect operation throughput and deliver a lower than theoretical FLOPS figure.
The Doormat
If you're not outraged, then you're not paying attention.
I've always thought they should be rated in something like flops or mips indeed (or that some sites should provide those specs instead of hundreds of benchmarks), but I guess that those tests will only be somewhat significant. The tests used to rate the chips could be tweaked/optimized for certain CPUs, or in some cases, extra cache could give more of a real life boost than the ratings would show...
It used to be pretty simple, but now it's getting a little out of hand, hopefully they'll come up with something soon...
And I thought I was just a low down dirty bastard.
At least according to what I have learned from the Simpsons.
Patty: The easiest way to be popular is to leech off the popularity of others.
Selma: So we propose changing our name from "Springfield" to "Seinfeld".
Intel can start making processors with names that make sense!
Pentium D Costanza!
- This is our newest budget-aimed processor which will deliver today's technology in ways you can't imagine. The Costanza will answer your multimedia needs thru experimentations that one would never dare to even think of.
Pentium D Kramer
- We are very proud of this processor. The main audience are the gamers in general who don't want to spend too much money on something as radical as this. The Pentium Kramer will deliver spontaneous performance boosts thru applications and make random noises whenever the cpu is overheating.
Pentium D Steinbrenner
- Our flagship product, the Pentium D Steinbrenner will deliver performance beyond your expectations. Not only does it come with 2MB of Cache but it also comes with a special software that has been added into the cpu which will warn the user when he's about to click on something that may be a virus. We call this feature the 'Steinbrenner Alarm':
STEINBRENNER ALARM: Of course. Of course you can do it. Things get done all the time, I understand. Don't worry, your job'll be waiting for you when I say so! Get better user 'GEORGE'! Get betteeeeeeeeeer!!!!!!!
Yeah.... now I so want one!
I think your right. I have seen benchmarks on the AMDzone forums which show that while AMD gets a speed boost (on average) from 64-bit mode, Intel takes a performance hit (again, on average) when in 64-bit mode.
Intel is just trying to be compatible with AMD64. They won't have a serious product for another quarter or two (or three).
I am a viral sig. Please help me spread.
Our antenna design group is eyeing the 64-bit products with intrest. Their simulations always use up all the memory in their systems. Well right now, only option for 64-bit with their software is Sun hardware. That's really expensive and it doesn't tend to perform as well (the chips are more server oriented, less vector math). So they are quite interested in a 64-bit Intel or AMD system with 64-bit Windows. If they could get like 8+GB of RAM, they'd be interested. $1000 for a chip is fine, that's downright cheap compared to the Sun solutions.
yes, this is all nice and good, but did you know that INTEL COPIED THE INSTRUCTION SET FROM AMD AND DOESN'T WANT ANYONE TO KNOW ABOUT IT?!?!?!!11? HUH??? DID YOU???????
Reason: Don't use so many caps. It's like YELLING.
--
Stay tuned for some shock and awe coming right up after this messages!
The 64-bit move means about jack and shit for home users and gamers right now. Most game math isn't integer, it's floating point and those units have been larger than 64-bit for some time. 64-bit integer calcualtion won't get you anything, really.
The whole reason for the changeover is access to more memory. 32-bit systems cap out at 4GB of addressable RAM, to get more you have to do some kind of hack like the PAE extensions the Xeon has that isn't all that good an idea.
Well right now, you just don't need that mind of memory. 1GB does great for a gaming system, 2GB is overkill. It's the high end scientific kind of stuff, AV work, etc that can use it. Gaming rigs won't need it for a few years to come. By then, 64-bit should be commonplace and so it shouldn't be a problem.
However right now moving to a 64-bit chip and OS are going to get you pretty much nothing gaming wise.
Just a question...
So many versions of the Pentium4.
So many cores. So many variations. So many significant architectural differences.
Seriously... when it it enough to be the Pentium5? I seriously doubt there is as much difference between the Pentium-3 and the IV (original P4) as there is between ANY other P4 cpu and this one.
Seriously... what's the deal?
Other than the 5-for-$5 jokes (Pentium 5 being a rather redundant name, after all...)
-- Sometimes you have to turn the lights off in order to see.
65C? Under what conditions? What heat sink? Where is it measured?
Most important is the power dissapation figure. That shows how much energy has to be dumped by the cooling solution.
Look, just because it says 64bit, doesn't mean it has a 64bit instruction set!. the EM64T stands for Enhanced Memory Technology. It allows you to access more than 4GB of ram, that's pretty much the end of it. IF you want a 64bit machine, I suggest you go get a Sun, a Power5(not the wimped out G5 machines) or a Mips proc. For those of you on a budget who want true 64bit stuff, go and get a DEC chip...
First DEC and now AMD. http://www.internetnews.com/infra/article.php/3335 981
It was my understanding that the AMD processor numbers were based off a 1Ghz(?) K7.
I guess that is why so many facilities are choosing G5 clusters, because they are so wimpy and not because their terflop to price ratio is so good. I suppose a Power4 with a 128bit SIMD tacked on is whimpy in your book? If you want whimpy, look up the Itanimum 2.
:)
Jelous that the 64bit market is quickly becoming common place and no longer a placid of the super nerds?
Yes we would all love a Power5 or Mips cluster to run e-mail and Maya on the same machine. Yep, that's a great way to blow a small business's budget.
I talked to some of the few people who were testing these. Apparently it couldn't keep linux running for more than 4 minutes. Lets hope Intel was able to fix that "issue"
Assuming of course the issue is Intel's not Linux's. Yeah, heresy, sorry about that.
Actually I should also say assuming the hardware is not flaky. How's the 64-bit WinNT beta run on that system?
Does anyone remember what a VAX-780 cost? That was a 1 MIPS CPU.
Bruce Perens.
Now it's time for game makers to get with the game [dum dum TISHHHH]. I'm tired of seeing 32 bits of my AMD 64 wasted every time I game, and now that every major player has 64 bit processors succeeding 32 bit, they need to get with the program and stop wasting bits.
No, it is not time for 64-bit ports, well at least for game clients. Servers are a different topic. First your CPU is probably not your bottleneck, you graphics card probably is. Secondly, developers don't optimize for the high end, they generally spend what little time they have on optimizing the low end. The low end needs the help, the high end is fine as is.
So why would Intel slap the 64bit name on something that's not truly... 64.
[Fuck Beta]
o0t!
It's sort of like throwing a few billion dollars in the fireplace.
Bruce
Bruce Perens.
Hah! What do you think of that?
Intel EM64T systems have been out since July 2004 in higher end workstations and servers. Itanium with a true 64bit arch well before AMD released a hacked x86. x86 is supposed to be on its way out. Intel knows this. I am sure they will kill x86 soon enough. It has lived enough of a life. Especially with Cell and PowerPC on the horizon.
Don't be surprised if a year from now, Intel buys IBM's microprocessor unit.
All of this is for Joe Sixpack. Not gamers and enthusiasts.
If you want to go 64-bit, pick up an Athlon64 2800 for about $100, or Athlon64 3000 for abt $130. AMD motherboards also work out cheaper, since they have been around for a year and a half.
The 64-bit market is just opening up, expect the pentium prices to come down significantly soon. By 2006, most processors will ship with 64-bit capability. There are not many 64-bit native applications available now. Games are still 32 bit. Windows XP 64 bit is just coming out next month. And Linux still does not support Joe.
If you are price concious, NEVER buy anything quite recent. Save the money, and buy dual-core 64-bit processors a couple of years from now.
Life is just a conviction.
Your understanding is... wrong.
The numbers are based on a "hidden" series of benchmarks used internally at AMD.
Since it's a model number, it's no guarantee of performance, it's just a shopping guide for retail consumers.
AMD's corporate options are labelled more obfuscatingly, for instance the Opterons have numbers like 244 and 424.
I am government man, come from the government. The government has sent me. -- G.I.R.
they're probably the ones who dont understand the 9/10ths of a cent at the end of all gas prices
/Writes petroleum accounting software
Your point is valid, but for the record the 9/10ths has a lot more to do with taxes than marketing.
The problem with comparisons in MIPS (millions instructions per second) is that for different architectures, it takes a different number of instructions to accomplish the same amount of work. This doesn't matter for AMD/Intel (Athlon/64/Pentium/Xeon) as they all use the same (mostly) Instruction set. It would matter if, as the gp said, you wanted to compare Sun, Apple, Intel Itanium, or Intel Pentium, since they use differnt ISAs.
Flops are a little cleaner, but still only test the Floating Point units. What if a chip has good floating point, but really shitty integer (it could happen). It would still suck to use for a lot of things. The only decent way of testing something (AFAIK) is to run it with the programs you intend to use it for and compare.
Does it just seem like it or are all the posters with 700000 UIDs clueless?
immediately they're overtaken by AMD64s and G5s once they get on the freeway. They're really easy to spot with an infrared camera, too. Kinda like overheated, broken down cars.
Get yourself a 64-bit P4, and get the coolest and only 64-bit radiator in existence! J/k, these things send idle commands now, but still, not comparable to cool'n'quiet, though I imagine this'll improve when the current generation of mobile processors replace the current generation of desktop processors sometime within the next 2 years, good power savings.
and another reference here
perhaps someone should edit that then, but I must ask where you got your information
Wow it was 20 years ago the memory protection bit a.k.a. NX was seen in 8086 en 80286, after it is was lost. Intel needed 20 years to find thair new implementation of the bit. the OP-code was always available in the intrstuction set of the intel 80x86 procesor for compatability reasons and never used in the real world.
Does anyone with a soldering iron want to help me replace the P4 in my Dell laptop with one of these new chips (it is a desktop P4 BTW)? I wonder if such extreme hardware hacking is actually possible.
With Athlon64s and Opterons we're fully set up.
Sorry Intel, we're all AMD here now.
"Nine times out of ten, starting a fire is not the best way to solve the problem." - my wife
Will existing x86-64 Linux distributions Just Work on one of these processors?
-- Ed Avis ed@membled.com
All our secrets are out. It's true.
They've been on the market longer than that. We've carried them at Dell in our high end for the past 3-4 weeks.
In short, yes. Any code compiled for x86_64 will run on these chips, which, as previously stated, have actually been out for at least a week. The prescott cores that make up the 64-bit chips also have 2MB of L2 cache now.
In theory of course.
In practice, you can get up to 1GFLOPS on a pretty simple machine, just put all your resources on doing flops asap. Ignore branches in your design, just make it run as many floating point instructions in a row as quickly as possible.
In real situations however, both FLOPS and MIPS say very little. There's a damn good reason it's commonly transscribed as "Meaningless Indication of Processor Speed"
Somewhere around the time of the 3 & 4- 86, competitors started making compatible chips and giving them 3- & 4- ??? names. Intel tried to sue to prevent them from doing this, but the court ruled you can't copyright numbers. Thus their next iteration had a real name, the 'Pentium'.
Someday we'll all be negroes
Well, everyone who does the effort to get as many as 700000 UIDs must obviously be clueless. I mean, I could imagine that there's a value in having two or threee UIDs (although for me personally a single one is clearly enough), but who on earth needs 700000 of them?
:-)
SCNR
The Tao of math: The numbers you can count are not the real numbers.
Oh God no, please avoid ISO at all (cough) "costs".
That qualifies as "active cooling" because liquid helium is not available at room temperature and pressure.
True, but back them you programmed it in assembly, and avoided software bloat, today half the power is spent on eyecandy and gui.
Freedom or George Bush
Don't forget to go socket 939, otherwise you have no upgrade path in the future.
Why would anyone need Central Heating??
For goodness sakes it was 81 degrees in Miami yesterday. We need A.C., but definitely NOT heat!
Thanks, but no thanks!
Woohoo, new P4's. Man, I could really use one about now. Kinda chilly today.
One of the big advantages of 64-bit CPUs for gamers is that 64-bit addressing will open up new horizons in gaming. One of the problems with todays 32-bit computers is the memory limitation. With 64-bits memory sizes, it is possible to put lots of GB RAM, and then have truly huge game worlds.
This is especially important for FPSs. Take HL2, for example. Although it boasts 'big' maps (well, bigger than previous efforts), the gameplay is based on limiting the player into one possible path. Each map is really small if one explores it using 'noclip'. With 64-bit CPUs, game maps could be huge, allowing for multiple routes and much greater variety, and increased re-playability.
it seems to be a blatant ripoff of AMD64, renamed a few things and introduced a few incompatibilities just to give the architecture their own name.
and according to the benchmarks they did a poor job at it, but that's intel for you.
they were working on something different, and paniced when AMD came up with something better.
Rating machines based on MIPS has always been part of the old RISC vs CISC debate. MIPS ratings give RISC machines an advantage because the instructions are simpler and execute faster. If a valid comparison were to be made, it'd have to take into account the decomposition of CISC instructions to internal RISC operations that the P4's microcode/trace cache performs. Good luck getting any information on that, though.
LRC, the best-read libertarian site on the web
I'd mod ya down if I werent replying, but not because you're pro intel, just because your logic is flawed.
But slashdot doesn't have a "-1, Stupid (and Wrong)" moderation, more's the pity...
It's official. Most of you are morons.
that is very false, my old 1333 Tbird exactly equals a 1500+ in performance when not overclocked, when overclocked to 1500MHz, it's equal to an 1800+.
Retail 64-bit windows is coming out very soon.
Those bastards at Micro$oft waited for Intel to catch on with AMD.
It wouldn't be Wintel without that.
I havn't bought intel in 8 years
AMD4LIFE
All of this is for Joe Sixpack. Not gamers and enthusiasts.
Except that Joe Sixpack doesn't understand the point of 64-bits. Okay, you can have a lot more memory, but then you hardly run into any average computer users with more than 512 megabytes of RAM anyway, and *zero* who are maxing out what a 32-bit CPU can handle.
Not to mention Joe Sixpack's reaction when he finds out that upgrading to 8 gigs of RAM is going to cost more than his entire PC did.
I'm thinking we need some sort of ISO standard for clock speed
:)
How about the amount of time it takes a processor to do a full-text search on the entire Library of Congress?
Tired of FB/Google censorship? Visit UNCENSORED!
just bought a dell with at 2.4 ghz celeron / 256mb ram which is about 4 times faster than she will ever need.
Good thing was it was $375 total cost.
Now only if Dell, HP, etc could come out with a small form factor low end machine at $200 or less.
It's a small historical point but the first 64-bit AMD processor was released in April, 2003 which was 23 months ago, not 18 months ago as the article states. I remember this because I started using a 64-bit Opteron 240 workstation with an ASUS SK8N motherboard in July, 2003 which was 20 months ago. The 64-bit Windows XP and Linux OS's as well as a lot of 64-bit software were developed on 64-bit AMD systems since there were no 64-bit Intel x86-64 systems existing.
Something that i have seen here that is a bit of a misconception is that some people seem to think 64 bit processors are twice as fast, or can handle twice as much data as a 32-bit processor. This is not so. 4 types of data run through a processor: Integer Address String Floating Point in a 32 bit processor, the ranges are as follows Integer - 32 Address - 32 String - 64 Floating Point - 128 in a 64 bit processor, however, the ranges are as follows Integer - 64 Address - 64 String - 64 Floating Point - 128 All that is simply saying, by saying a 64-bit processor, is that the processer can handle 64 binary instructions per clock cycle, twice as many as a 32 bit processer, only in the integer and address ranges. so it is not truly twice as fast, and neither can it truly handle twice as much data, just 2 of the 4 types of data have been doubled. theoretically, a 64 bit processor could address 16 exabytes of ram (i believe this is 16 million gigs of ram) but obviously there are limitations to this. and interesting fact i thought i would stick in here :p
ps. - my ranges for floating point and string might have been reversed, now that i think about it i cant remember whether floating point is 128 bit or whether string data is 128 bit.
"Potpourii doesn't taste as good as it smells." - Dark_Link2135
So I guess no one would be interested in a cluster of 64bit P4s?
Seriously though, I'd love to see them start to increase the cores from 2 to 4 then 8 etc.
Could the chip be designed like a parking lot, such that if a processing core is being used the request is sent to the next available core?
Time to make the chips smarter without relying on enhancements from the OS community.
Just trying to punch a hole in this paper bag.
I moved from 1GB to 2GB receantly (I play with samplers that feature multi-gigabyte sample sets, my drumkit alone is 2.4GB) and I haven't noticed any improvements, even in World of Warcraft. It seems that after a point, games just run out of things to eat up memory with and you don't get any benefit from more.
Perhaps this is what was meant?
"As Slashdot previously reported, the 64-bit series will likely be the last major enhancement to the Pentium 4 line before the introduction of the Pentium D "Smithfield," Intel's first dual-core part, which is slated for next quarter."
Must-not-watch TV!
These new chips are not true 64-bit chips. The EMT64 chips just add support for more memory, and will "support" 64-bit operating systems (they're just doing some fancy emulation).
i ons/faq.htm i ons/index.htm
http://developer.intel.com/technology/64bitextens
http://developer.intel.com/technology/64bitextens
We shouldn't see a TRUE 64-bit chip from Intel for a year.
For doing small tasks, such as text editing, 32 bits is better. For task that needs enormous amounts of memory, such as simulations or large systems, such as SAP, 64 bits is better. On servers, choose carefully: sometimes 64 bits mode is slower than 32 bits. Unless you have over fifty simultaneous users, SAP (www.sap.com) is usually faster in 32 bits mode than in 64 bits mode.
32x64 decision is also very processor dependent: AIX 32 bits can address up to 96 GB of RAM memory with up to 4 GB allocated per process. And AIX 32bits does run 64 bits process with no changes. Currently I run 32 bits AIX, with 64 bits Oracle and 64 bits SAP (my current backup system doesn't run on AIX 64 bits).
Like everything performance related, test your particular situation before you switch.
FLOPS are highly inaccurate for indicating true processor power. Let's just leave it to how many fps the thing gets in Quake.
Some relatively new research in economics shows that having more choices can actually lead to less satisfaction with the choice you do make. Have too many choices makes people confused about what to choose and then it makes them unsure that they made the optimal choice. I think Intel has dug themselves this hole by trying to segment the market through product and price differentiation without actually trying to meet the consumers actual needs.
The courts ruled that you can't trademark numbers.
You are wrong.
"have to" means "must", whereas the second "have" mean possess. "have to have" could be paraphrased as "must possess".
In the other phrases you give, the repeated word has an identical meaning. Even so, I don't see what is wrong with them. "I need a license to get a license.", is a well formed and meaningful (if slightly ambiguous) English sentence. I could even see situations in which it could be used - in the UK at least, you need a Provisional Driving License to get a Driving License, and a Driving License to get a Public Service Vehicle License.
It isn't a particularly pretty choice of words, but maybe you'd care to explain exactly what gramatical rule you think this is breaking?
...as long as what you're planning to run software that needs lots of FLOPS. Personally, FLOPS are completely meaningless on my servers; infinite FLOPS aren't going to make buildworld, run my mail filters, or serve Zope pages any faster. You might find that to be a meaningful number but I have no interest in it, and such is the nature of benchmarks.
Dewey, what part of this looks like authorities should be involved?
As a couple of posters have pointed out, I was a little short-sighted on suggesting using FLOPS as a benchmark for processors. Serving web pages, building software, etc. doesn't require floating operations, so FLOPS are irrelevant to that type of work.
So, I propose that manufacturers list
Basically, tell us all about how a chip performs in each area and let us make the decision.
Overrated / Underrated : Moderation
Joe Sixpack doesn't need to need to understand 32 versus 64 bit. Even when I was a kid and didn't know jack about computers, I "knew" that the Sega Genesis was better than the NES, because the Genesis was 16 bit and the NES was only 8 bit. A same level of (almost absent) understanding is enough 64 = 2*32, therefore I want the 64 bit one.
Best Slashdot comment ever