Slashdot Mirror
New Intel Chipset and Extreme Edition CPU Tested
Steve writes "Today sees the launch of both a new CPU and chipset from Intel. The CPU takes the form of a 3.46Ghz Pentium 4 Extreme Edition, running at 1066FSB, and the chipset is the i925XE, the first Intel chipset to support this new FSB. HEXUS.net have a review of both. It looks like AMD still have the lead when it comes to performance, despite Intel's attempts to counter the Athlon 64 FX-55." Hack Jandy links to more reviews at AnandTech, HardOCP, and ExtremeTech.
I think AnandTech summed it up nicely "So there you have it folks - the 1066MHz FSB does absolutely nothing for performance, [...], But with the move to the 1066MHz FSB we have a platform launch that, in the spirit of the 925X and 915 launches, does virtually nothing for performance."
However the real question is, how many decision-makers are reading these review/benchmarks, or do they just buy Intel because it's Intel, or that's what xx-business weekly says?
Rock that crushes, Paper & Scissors that don't matter.
where I clicked it and it said "move along, nothing to see here"
Seemed more accurate.
Intel is finally seeming to realise that the bottleneck nowadays is the bus.
My sig would have been a lot cooler if
Great clock speed and all but is it 64bit?
"There's no set architecture in Linux. All roads lead to madness" -Microsoft
those cards sound pretty sic wit it, but can amd reply with something even better?! i'm sure we will soon see!!
Intel Extreme Graphics! W00t!
anandtech shows a 1% increase in speed over 800mhz fsb in most cases, is this really something to get excited about? will this difference open up in the future?
Marge, get me your address book, 4 beers, and my conversation hat.
As every day goes by
How can we close our eyes
Until we open up our hearts
We can learn to share
And show how much we care
Right from the moment that we start
Seems like overnight
We see the world in a different light
Somehow our innocence is lost
How can we look away
'Cause every single day
We've got to help at any cost
We can bridge the distance
Only we can make the difference
Don't ya know that tears are not enough
If we can pull together
We could change the world forever
Heaven knows that tears are not enough
It's up to me and you
To make the dream come true
It's time to take our message everywhere
C'est l'amour qui nous rassemble
D'ici ?l'autre bout du monde
Let's show them Canada still cares
You know that we'll be there
If we should try together you and I
Maybe we could understand the reasons why
If we take a stand every woman, child and man
We can make it work for God's sake lend a hand
...I'd be twice the sucker I actually am.
:)
Thank god for the ponces and their fast stuff obsession making things cheap for me
Beep beep.
40MHz is fast enough, but 1GHz?
You have my respect
Kind Regards
Simon Harvey
Honestly, who is buying these things? At a price of 999$ US (1000 units lot) and a marginal performance increase compared to other, far less costly solutions (3500+ AMD anyone?), I just don't see a market. Is it just for the performance crown, which they didn't even get to win this time around (or should I say, in the past 2-3 years)?No word on heat, nor power consumption.
AMD all the way. Intel is alive just because of Dell (among others) and a large reserve of cash. They cost more, do less, and heat your bedroom to boot. But it has 'Intel Inside', so I guess it must count for something...
Eureka Science News - automatically updated
Right here. Though, I must admit that I found some of the results to be a little wonky, along with the test bed. How'd they get a FX-51 running on a socket 939 board? Underlocked a FX-53?
http://www.anandtech.com/cpuchipsets/showdoc.aspx
Whenever the offence inspires less horror than the punishment, the rigour of penal law is obliged to give way...
no matter how fast the clock speed the pipline is sooooo long for P4 chips causing a sever preformance degradement. This is why G5 and AMD chips are faster at lower clock speeds because of there shorter pipline. Intels high clock speeds just look good
Intel is slowing catching up. But the fact is that my DP 2.5GHz G5 is at 1.25GHz Frontside bus - per processor.
Yes! I listen to NYC Speedcore and do math at 3AM. I suggest you try it too.
Do you guys think we could, like, halt the march of progress for a year or two? 1066FSB already?
You know, some of us over here have Athlon XPs with 333FSBs, and we're crying our eyes out. Please, think of us. Sometime? Maybe?
Fuck.
occultae nullus est respectus musicae - originally a Greek proverb
AMD still have the lead when it comes to performance
And even more so when it comes to VALUE. Intel just seems to have a problem making the P4 fast but not expensive. I suspect they just need to toss it and come up with a completely new design. Like Pentium M, only better.
Just my sqrt(4) cents.
The difference between spam and poop is that you don't have to dig through septic tanks looking for real food. -- Me
Athlons have always been fairly LOW latency chips, and the memory used (fast DDR memory) is low latency too.
The P4's on the other hand have used Rambus memory for awhile, although that's not really the case anymore. But when they did, they always excelled at memory THROUGHPUT because Rambus runs at high frequencies. Rambus memory however is fairly latent - it's the trade-off.
DDR2 RAM won't be "fast" until we see it in much higher speeds - DDR2-800 most likely. Of course, it will always have more latency then DDR because it uses four banks of DRAM instead of two.. I'm sure you can research all this via google if you're interested in learning more.
- It's not the Macs I hate. It's Digg users. -
The bottleneck is not the bus. If you RTFA (Oh you didn't? How unsurprising!), there is almost ZERO improvement in performance.
queue amd zealots in 3,2,1!
Imagine a Beowulf cluster of these things...
Wait. No, better not.
The difference between spam and poop is that you don't have to dig through septic tanks looking for real food. -- Me
Hastings.
I am NaN
Get off your asses and do something Intel. Why release a slower and more expensive chip compared to the competition, which is far cheaper and faster? Not to metion the AMD overclocks better, I see no point at all for anymore buying this chip. Fools.
I use an AMD box now - I have for some time. It's not even a powerhouse, it's an old XP1800. I've used this CPU on three different motherboards now, and so far I'm still looking for a reason to consider AMD when I finally replace it. I've had an S3 motherboard, a via motherboard, and TWO Nvidia motherboards. The closest I came to having decent chipset support was with the S3 and that's only because the guy who wrote the 3D drivers was able to basically con S3 out of the information he needed in order to do it (ie if you get a mainstream distro his drivers won't be in it due to potential legal issues).
The first Nvidia I bought to try out, then decided I wanted that great whajamacallit sound support so I spent weeks looking for a miniATX motherboard that had this feature. When I finally got it I discovered it has TERRIBLE sound - I mean atrocious, like the crap you would expect from a five year old emachine. Overtones, quantization noise - just horrid. And this is using THEIR drivers, which I cannot use along with THEIR 3D supporting video drivers because of random lockups the two together cause on my mandrake system.
If I get an intel system I at least get decent drivers. So here we have an intel motherboard that offers basically the same performance as the top of the line AMD, meaning "it can be done" and a lesser system (as I would buy) will also be proportionately less expensive. So for a premium of just a few bucks I can get similar performance AND I get open drivers that will work with my linux system?
Where do I sign up?
The FSB on that thing is clocked faster than my CPU....
TODO: Something witty here...
C'mon. This has gotta be useful for all the bachelors out there. Nothing like surfing the net at home while warming the house at the same time. You know that Intel is just waiting for the right time to unleash the hot-plate add on so I don't even have to leave my computer to cook the vast stores of Top Ramen.
While I think all these benchmarks that are being used on the big sites like [H]ard|OCP, Anandtech, and Extremetech are a big part of an overall score when it comes to deciding what to buy and what not to buy; but when it comes to the people that use Photoshop, Premiere, and the other numerous digital content creation applications out there, they're pretty much left in the cold, and then buy a Mac because they know Macs perform these functions excellently. Is it possible for Adobe to make a benchmark based on Premiere or Photoshop? Anywho, this is getting off-topic.. I'd like to see if this chip outperformed the AMD competitors in this arena. Don't get me wrong, I love AMD, I'd just like to see a bit more of the story.
real men use soundblaster audigy 2 ZS Platinum pro packs.... now if someone could just tell me howto configure the remote..
I noticed that the summary points out how the AMD64 is faster. Well, what were you expecting, its 64bit vs. 32bit.
If you want a fair comparison then compare the AMD 32bit processors against Intel 32bit processors.
And compare AMD64s against Intel's 64 bit chips.
Don't complain Intel is slow when you're tilting the playing field. Its just like Microsoft claiming Windows is cheaper than Linux, only on a tilted playing field.
Latency is in clocks. As the click speed goes up, the latency goes down. When DDR2-800 came out, the word was DDR2 would suck until the clock rates got up (and thus the latency down). And now you complain that this is the direction Intel is heading?
I'm no fan of the 915 or this chipset either. But if there's anything that will finally drive everyone away from the 875P, it will be higher FSBs that aren't available on those older chipsets.
I was suprised to hear of it since Intel doesn't talk about it. But Prescott is 64-bit. I found this out when running a test on a friends Prescott that reported it had the 64-bit extensions turned on.
So yes, this chip is 64-bit.
Intel is doing just fine. These chips are for folks that like to spend outrageous amounts of money on a computer for games.
Intels big money will come as Itanium 2 servers gain momentum. Whether you like it or not, Itanium is an improved, faster DEC chip.
AMD will do fine, but their problem is FAB manufacturing. You folks need some business 101.
No. They went off their rocker more than 3 years ago when they designed the P4 and went for GHz for marketing reasons and not engineering reasons. That said the P4 actually served them well for a number of years, but they didn't get off at the right station...
I remember Intel talking about 5-10GHz CPUs. They were probably taking a bet that the process and material engineers would save them.
I'm sure they realized they lost the bet when the Opterons/Athlon64s started spreading their wings and actually flying, but when you have multi-billion dollar fabs, commitments to partners, it takes a while to turn the ship.
If you observe, they've canned a lot of stuff and changed their product milestones/announcements.
They just can't tell Dell, forget the next bunch of P4s, we're going to go Pentium-M NOW (even though we haven't got it fully buzzword compliant)!
Any idiot can realize the Prescott was bad news ONCE it was testing. But by then it's just too late.
AMD has a window of opportunity till at least early 2006. As long as they don't screw up! They better use overwhelming force if they want to win. It's not an easy battle. Intel is no pushover.
I mean - what's there to prevent Intel doing the same stuff as AMD? e.g. Pentium-M with memory controller on CPU?
Meanwhile I'm really curious about the new Intel SMP server chips. What are the power consumption and cooling requirements like?
And here _I_ am with a lowly 266Mhz FSB system!
I never thought I'd be a computer charity case...
Works marvelous under Linux, and I bought mine a couple years ago for less than $60. Onboard sound and 10/100 ethernet, all works perfect under Linux.
I'm a fan of SiS for PC chipsets, and have been ever since I got sick of VIA's instability (and no, I don't overclock). ALi never came up with much, and nForce has been a driver nightmare for some people (never bought one, all my friends couldn't get their boards to work reliably, why bother?).
ECS makes good, stable stuff. No bells and whistles, no overclocking features to speak of in most cases, just good stuff.
But does it make the internet go faster?!?!
When used properly, the above can give quite a hell of a performance boost over a 32-bit x86 solution.
Just good SUPPORT. Provide the fucking documentation we need to support these devices within the community. With rare exception intel has been good about this, but those that provide AMD-centric chipsets have not.
I'm really amazed AMD has such a good rap around here considering the only chipsets (if you want even halfway decent graphics) supported by their CPU are all wrapped up in proprietary (and generally terrible quality) linux drivers.
I don't frag, I don't care about "squeezing out" two or even ten percent more performance - I want something that works and can be supported within the community that has proven itself, time and again, capable of providing the superior product. Unless SiS or Nvidia get their heads out of their butts, my next system purchase will have "intel inside" - and I won't be looking back again.
Anyone that uses onboard audio and is expecting good performance is insane. Get yourself a $20 (maybe even cheaper?) PCI Soundblaster card and cease your worries. I have not had an issue in 3 years of using linux with my soundblaster live card, and it performs extremely well.
Oh yeah, and on my Dell Latitude C610, which uses intel audio, I get a VERY loud hissing on the headphone jack. From what I've heard, intel audio is a PITA in linux, so I'm really not sure what you're talking about. Some links would be good.
other sites have a much more experienced staff than tom's hardware, and a much wider dataset to compare against. eg anandtech.
when reading tom's hardware i often feel like i'm reading press releases rather than a hardware review.
they sunk billions into itanic, thus wasting valuable company resources (engineers, fabs, marketing, etc) that could have been better put to use on their mainstream chips.
then they blew it by designing p4 to purely target a mhz goal, expecting that advances in materials and fabs would easily let them scale to 10ghz.
basically, intel was overconfident and then refused to abandon a ship (itanic) when it was obvious it was in trouble, instead desperately trying to save face and keep it afloat.
while they were busy mending their doomed ship, the popularity of amd64 completely caught them off guard. now amd is eating their core markets for lunch.
in the meantime, intel's itanic partners are beginning to abandon ship. this has to be really alarming to intel.
intel is trying to go too many directions at once, all of them wrong.
amd took the safer bet (amd64) while intel was pooh-poohing and ridiculing them over it. that decision has now come back to haunt intel.
Extreme Edition is Gallatin, not Prescott. So the most expensive Pentium 4 isn't even 64-bit.
There's a short blurb about the hypethreading at the end of the hardOCP article. I think that hypetreading covers a missdesign in the MS Winblows kernel. I haven an Athlon 2000Ghz and i have no problems with multitasking whatsoever when running Linux. So have you experienced improvements through hypethreading while using Linux compared to a similar AMD machine?
Thnks for your answers
A) the MCP-T also provides 1394 connectivity. I wanted a motherboard I could use in a very low profile system and didn't want to have to use a PCI slot as I already had one in use by the tv tuner card and, because it's low profile and all, it's on a riser card laying flat across the other slots.
B) That "on board sound" is controlled by a pretty powerful DSP. Problem is it's completely closed, the drivers suck, and because Nvidia think the sky will fall in if they allow us "lusers' to add features WE want to it, that's not likely to change. I had, however, heard good things (just like the comments here from people who use the thing and "have never had that problem) so I grit my teeth and tried it even against my better judgement.
And no, most onboard sound system are not "crappier than the soundblaster live..." Many of them have very good sound systems, as the people who laid out the board understood to keep the sound traces away from crap like bit clocks and usb buses. That, however, is neither here nor there if one plans to use the system with an outboard DAC - as I had. The problem is the POS drivers resample everything and introduce tones and other distortion (just like that sb live) which render an outboard DAC meaningless - the "noise" is in the digital domain and therefore a "feature."
The AMD 64 is covered by a heat spreader, just like the P4. In a nutshell, short of using a sledgehammer to install the heatsink, it's pretty damn impossible to crack the chip even if you wanted to.
A polar bear is a cartesian bear after a coordinate transform.
This just goes to show that people are more likely to go with the option they know, as opposed to the best ones. This is true with hardware, software (how many computers run Windows?), clothes, music, etc. Mod me Offtopic, Flamebait, Troll if you feel it is necessary.
On the other hand, if microchips are fragile, then Intel and AMD should never have shipped them with the silicon exposed. (And yes, it was Intel that came up with the idiotic exposed flip-chip packaging in the P3, but AMD should have known better than to mindlessly copy it.)
It's not something that happened only to some clumsy (l)user. Even people from big benchmarking sites (e.g., Hard-OCP), which get to play with hundreds of chips, still managed to grind a corner off a chip, or put the wrong heatsink on and burn it.
Basically shipping a chip like that is just as stupid as shipping a remote control that cracks if you push the button too hard. Sure, you could expect the users to be careful, and you could take the lame way of blaming the user, but it still will happen.
And it looks like bad design to me. Good design should strive to prevent problems, not just shrug and call the user names.
A solution isn't even hard or expensive, as the P4's heat spreader (then copied by the A64) proved it. But you need someone to actually think in terms of "how can we solve or prevent a potential problem?" instead of "duh, yet another clueless idiot cracked his chip. Someone should shoot these people."
And that doesn't go only for chips. Software is the same. A lot of it seems based on the requirement that the user is an expert and never makes a mistake. And then when the user does do a mistake it's time to act offended and do a "gee, we let this kinda idiots use computers?" flame.
A polar bear is a cartesian bear after a coordinate transform.
I've handled hundreds of AMD chips, and I've never had one single DOA broken one, ever. The only broken AMD CPU I have owned was an Athlon 900Mhz chip, because I fucked up and busted the chip by putting too much pressure while installing the heat-sink.
I should mention that AMD not only replaced the thing for free, they sent me a 950Mhz chip, and sent it overnight delivery, no charge.
Since then, I have been a lot more careful installing heat sinks. The Pentium 3 "flip chips" were *EXACTLY* the same as the Athlons, and you could break them just as easily.
I call bullshit on you for that FUD. And not to mention, the Athlon 64's all have a heat spreader on them now, so your point is moot.
AMD Processors have always treated me very well. I never have problems with them, and they always run how I expect them to run, plus some. Intel makes good CPU's too, and I use them as well.
I'm not sure why you weren't fired after snapping the 10th CPU? Or the 100th?
- It's not the Macs I hate. It's Digg users. -
The p3 was a hell of a processor. A 1ghz p3 will make a p4 nearly twice as fast feel like a raw deal, but intel decided they had to crank the mhz even while talking about all this multicore stuff.
Too bad they didn't just spend that money figuring out how to cram more PERFORMANCE into the chip instead of more mhz. A four core 1GHz P3 with 2MB of fast (shared) l3 cache deployed on 90nm tech would probably still be smaller and cheaper than any of the P4s they've introduced this year... and it would also, quite likely, walk all over them.
I think one of the reasons intel were very quiet about the whole thing is that this move towards higher memory speeds isn't for the current generation of cpu's ( even if it was part of their roadmap over a year ago ). The major idea behind dual core is that multiple applications can share time on the two cores, this is going to require a fast memory interface. As far as intel are concerned, the sooner more people use fast DDR2 memory now, the more people will be in a position to use intels dual core cpu's later.
AMD chips used to cost very little, yes, but nowadays they're pretty much on par with Intel.
... 899 Euro ... 849 Euro
... 174 Euro ... 184 Euro
... 189 Euro ... 184 Euro
E.g, since we're talking about the P4EE, a fair comparison would be the Athlon FX. A quick look at an online shop here (www.alternate.de) says:
Athlon 64 FX-55
Athlon 64 FX-53
Not exactly a budget chip either, eh?
But let's look at something more mainstream:
Athlon 64 3000+ (socket 754, 2 GHz)
Athlon 64 3000+ (socket 939, 1.8 GHz)
Pentium 4 3000 GHz (Northwood)
Pentium 4 3000 GHz (Prescott)
So it looks to me like they're very much on par, as price goes.
Now this isn't a scientific study or anything, and I didn't even try to find the lowest price or anything. I just stopped at the first online shop that came to mind.
A polar bear is a cartesian bear after a coordinate transform.
Not you, schmuck. Nobody asked you.
The grandparent very clearly asked why the *great-grandparent* needed the 64-bit goodness. We all know the goodness already, thank you.
Even as an AMD man I must ask... is a 32bit processor really suppost to outperform a 64bit one? Wouldnt a more fair comparison have been a 32bit AMD processor??? Isn't it actually a BAD thing for AMD that a 32bit processor can come so close to their 64?
Matt
You have 1 Moderator Point! Use it or lose it! Is that a threat? -vapid
"This Gallatin core CPU that houses the 3.46EE is simply hot. It runs hot and the motherboards that support it will run hot. Components on our Intel motherboard were showing surface temperatures of 150F."
From http://hardocp.com/article.html?art=NjgyLDY=
Have you metaroderated recently?
Unlike SPARC64, the big advantage an AMD64 chip has is that there are double the number of registers. This has quite a significant impact on performance, as the i386 is register starved.
Despite the larger space needed for pointers, this can give significant performance improvements in real world applications. For those of us running 64-bit Linux, this is NOW!
I don't even think he has a given bias as such. He'll just say or do anything that brings clicks.
Whether that means trolling a certain user group (e.g., the insulting editorials about AMD fans), trolling about other popular (and far more competent) sites, or saying what the majority wants to hear, or some cheap lame publicity stunt. (E.g., the video about an Athlon burning horribly without a heatsink at all. Except what he won't tell you is that it only happens without a heatsink at all, so outside of such el-cheapo publicity stunts it never happened.)
Add to that a good dose of self-adoration and downright verbal masturbation. I've yet to see a THG article in the last years that stays on topic, and isn't just an excuse to brag about how Tom's Hardware is the greatest ever. And recap how cool his past publicity stunts were.
Often not even in the right topic. "Today we talk about hard drives. But first, remember kids, we were the first ones who showed you an Athlon burn without a heatsink! And we were the first to include a 33 MHz 486 in a P4-vs-Athlon benchmark! We're soo cool!" Well, not an actual quote, but that's the impression THG articles leave me with.
Basically Tom's is just sucking up to the majority for clicks. He's probably only pro-Intel as long as most folks buy Intel. Since most folks started buying AMD retail lately, expect him to do an about-face turn in no time. He's sure done such about-face turns before, about other issues.
Either way, some results you can't downright spin into victories, but you can downplay. Trumpet as complete victories those cases where your favourite is 1% faster, but when it loses just spin it into sounding like only a marginal loss or an unlikely-to-happen scenario.
E.g., let's say you have two CPUs (or GPUs), let's call them CPU-A and CPU-B. Let's call them Bench-A and BenchB.
And let's say that the scores are as follows:
Bench-A: CPU-A=100, CPU-B=105
Bench-B: CPU-A=1050, CPU-B=1000
In both cases the difference is 5%. How do you spin that into sounding like CPU-A still is the fastest across the board? Easy. Use lame games like "CPU-B is barely 5% faster and only in this particular case" when talking about Bench-A, but trumpet Bench-B as "Here we see CPU-A wipe the floor with CPU-B, scoring a whole 50 points higher!"
Think it doesn't happen? Well, THG's video card reviews do just that every single time. (And those on some other low standard sites too, of course.)
A polar bear is a cartesian bear after a coordinate transform.
the fucking drivers worked.
I have to keep the amp off when I'm not listening to something because of all the shit coming through the speakers. I'm using nvidia's drivers from their website and the only thing that makes their drivers "better" than generic intel 810 sound drivers is the nvidia drivers allow the sound card to be detected on boot thus preserving my volume settings.
This is a well reviewed shuttle motherboard. The one I had before was a well reviewed Asus. Neither have been any better than my old S3.
I'll never buy one of these piece of shit nvidia motherboards again. And if AMD doesn't put some pressure on SOMEONE (anyone) to provide a chipset that's not locked up in NDAs and other bullshit, I'll never buy another AMD.
Want to buy a dualhead miniatx nvidia motherboard?
" the fact that the number of general-purpose registers has doubled from 8 to 16, greatly reducing the amount of register variable swapping needed. Again, most apps simply do not care if they can fit huge numbers in a register, because they do not need them."
And here's your free clue for the day: you only get those in 64 bit software. Because that's part of those 64 bit extensions.
And, indeed, that's largely the reason why software compiled in 64 bit mode rocks on an Athlon 64. (As opposed to actually running slower, as happens on an Ultra-SPARC.)
That said, it only makes the comparison fair after all. Both CPUs were using 32 bit registers and the same number of registers too.
A polar bear is a cartesian bear after a coordinate transform.
Comment removed based on user account deletion
Comment removed based on user account deletion
Comment removed based on user account deletion
they tested 32 bit apps on a 32 bit OS, the 64 bit capabilities of the athlons were not used
Slashbots think exactly how they are told to think.
Slashbots like AMD.
Athlons have always been fairly LOW latency chips, and the memory used (fast DDR memory) is low latency too.
The scary thing about these Opteron results is that the Opteron is bitch-slapping the P4EE at 32-bit performance, AND THAT'S NOT EVEN WHAT IT WAS DESIGNED FOR!!!
The 32-bit circuitry in the Opteron is almost an afterthought - the raison d'etre of the Opteron is 64-bit operating systems.
You gotta figure there's some sweating of the palms and some grinding of the teeth amongst the suits in Santa Clara. Or, if there isn't, there ought to be.
Slightly faster...to the MAX!
28:06:42:12 - That is when the world will end...
To be fair, two of those three are advantages of AMD's 64-bit x86 spec, not of the CPU having 64-bit addressing.
tasks(723) drafts(105) languages(484) examples(29106)
The 32 bit circuitry is not an afterthought, it's still the main circuitry used on the chip. On the contrary, the 64 bit circuitry would be considered the afterthought, since the 64 bit instructions are just extensions that expand on the current 32 bit instructions.
As AMD has said, the Hammer architecture is not a revolutionary new architecture from AMD, it's an evolutionary architecture that builds on the K7 design. The K9 will be the chip with the entirely new design.
Sorry, but that's kinda not correct.
AMD designed the Athlon 64/Opterons to be cutting-edge X86 CPU's, and that's 32-bit x86, while at the same time providing 64-bit instructions and other functionality such as large memory addressing (which kind of comes hand in hand with the 64-bit extensions.)
The idea was, build a CPU that will run all of your software today, at the highest performance possible - but at the same time provide 64-bit compatibility with an easy upgrade path.
They've done it with AMD64. Some will argue that it's just extending the life of an already aged platform, but some very smart people at Intel and AMD prove over and over that they can do amazing things with x86. The Opteron has been compared to Apple's new processor and the Itanium, and performs quite acceptably put next to these "true 64-bit" chips. So well, in fact, that even SGI is using Opterons for their latest "SuperCluster Computers."
So no, 32-bit was not an afterthought - it was the primary concern. If the Athlon 64's and Opterons dogged it on IA-32 stuff, nobody would buy it. AMD positioned Opteron to compete against Xeon, and Athlon 64 to compete against the P4 - both 32-bit chips.
The 64-bit stuff is a bonus, an upgrade path, and it's there on all of AMD's chips so when software makers say "okay, this 64-bit stuff is great, but who can run it?" there will already be a base of users with the capability.
- It's not the Macs I hate. It's Digg users. -