AMD's Athlon XP 2700+

kraven_73 writes "According to some Taiwanese sources, AMD will officially reveal its Athlon XP 2700+ processor on the 7th of October. Most interesting is that this CPU will have a 333 MHz FSB. The first implementation of this increased FSB on Athlon platform. It is expected that the novelty will be based on the latest Thoroughbred core stepping 1, just like the current Athlon XP 2400+ and 2600+, and will work at 2.17GHz."

Comments

What's the point anymore

[ObviousGuy]

To bring down the price of slower chips to reasonable levels, that's what the point is.

Expensive bleeding edge crap.

Tom's Hardware Comments

[Winnipenguin]

TH has this info and advice:

Once again, Intel wages war on AMD, fighting to attain the fastest desktop CPU. AMD is sure to launch the Athlon XP 2800+ soon (in October at the latest), so that it will be able to keep close on the heels of its arch-rival. Intel has also made preparations of its own, with the P4/3066 up its sleeve.

At any rate, the real winner is the ambitious end user, who will be able to choose between the P4/3066 and the Athlon XP 3000+ by the time Christmas rolls around. Both the successor to the P4 and the AMD Hammer won't be available until next year.

As always, price-conscious buyers who are interested in getting the best price/ performance ratio are a bit better off with an AMD Athlon XP than with a P4..

Link here:
http://www17.tomshardware.com/cpu/02q3/0208 26/p4_2 800-16.html

Glad they chose to up FSB

[McCart42]

Most hardware review sites I've read commented that upping the FSB speed was the best way AMD could reclaim the speed crown...Intel regularly uses much higher FSB clocks with their chips (in the neighborhood of 533 MHz). I may be missing some crucial aspect of AMD's strategy but that seems to be what is holding them back right now, from a high-level standpoint.

Re:Glad they chose to up FSB

[Alan Partridge]

I'm sorry, but Intel does NOT use a 533Mhz FSB speed any more than AMD use a 333Mhz one. The "533" refers to 4x133Mhz (it's a 133Mhz bus with QDR tech) whereas the "333" refers to 2x167Mhz (it's a 167Mhz bus with DDR tech). Incidentally, I think that Apple is the first company (unbelieveably) to have implememted a 167Mhz FSB in their new "DDR" G4 designs. Shame the G4 chip isn't up to using the (otherwise fast) bus in DDR mode. Oh well, roll on the fabled MPC 7470!

Re:Glad they chose to up FSB

[Xeriar]

Intel uses a Quad-pumped architecture, that is, in a given cycle, the flip-flops trigger four times (beginning rise, ending rise, beginning fall, ending fall).

AMD only has a 'double-pumped' architecture, where the flip-flops trigger on both the rising and fallig edge of the clock signal.

Unless AMD licenses Intel's technology, they really can't compete in that arena for awhile. There are other strengths to the AMD platform that help bridge the gap, for example.

Re:Glad they chose to up FSB

[taeric]

Note he said beginning rise, ending rise and not beginning high, ending high. The rise/fall of a signal are not connected.

Now, I am not saying the original poster is correct, as I don't know. However, what you said makes no sense. If the next rise happens as soon as the previous fall ends... then you are saying that the hold states of the clock are as fast as the rise/fall states. I HIGHLY doubt such a statement. This would essentially be saying that as fast as we can make a clock switch is how fast the clock goes. This is CLEARLY not the case.

Did I missunderstand you?

Re:Glad they chose to up FSB

[tshak]

AFAIK - and this is based on earlier P4 systems - although Intel's bus has a higher throughput it also has a higher latency. Only applications that demand throughput gain an advantage, while many applications can actually run slower on these "quad pumped" busses. DDR memory has the same problem when compared to SDR memory, but nowhere to the degree of RAMBUS memory. AMD seems to have found a balance. For the future we can only hope that QDR will not have the same latency issues. However, I share a similar concern and I wonder why AMD didn't go to 400mhz DDR. You don't have the latency of the Intel platform, but you have great throughput. I can only speculate that it's because AMD runs at a much lower clock rate that the extra bus speed would do little for real world performance.

The move to 166mhz bus is nice but

[DeafDumbBlind]

Might as well wait for the Hammer.
The built in memory controller should to wonders for latency. Of course the 64 bit stuff will be a nice future feature to have.

Re:The move to 166mhz bus is nice but

[ackthpt]

Might as well wait for the Hammer.

This, of course, is the risk of having a really sexy new item coming down the pipeline. At some point those Xtreme gamers/programmers/modelers/or just people who like to have the latest and most expensive thing on their desk to play solitaire with, begin to hold back on purchases and wait for that new item.

I'm on the fence, but at the rate I've actually done anything to build my next system (hey, I did buy a cabinet! :-) the wait for the Hammer shouldn't be much longer (why does this name summon the memory of the artwork inside PF:The Wall, hmm, something there, but what...)

Fortunately for AMD, not everyone is holding off and all these really spifftacular improvements of , what will eventually be $60 processors in a couple years, are pretty damn exciting.

Re:The move to 166mhz bus is nice but

[Christopher Thomas]

I'm on the fence, but at the rate I've actually done anything to build my next system (hey, I did buy a cabinet! :-) the wait for the Hammer shouldn't be much longer (why does this name summon the memory of the artwork inside PF:The Wall, hmm, something there, but what...)

I'm waiting too, but that's because I don't think I'll have to upgrade for at least a year.

Remember how long it took for Athlon DDR chipsets to stabilize, and for the prices to drop. I'm not expecting a reliable, affordable Hammer/Opteron system until at least mid-2003.

Now's as good a time to buy as any (just not for the top-of-the-line models, with the speed war going on).

Re:The move to 166mhz bus is nice but

[ncc74656]

Might as well wait for the Hammer. The built in memory controller should to wonders for latency. Of course the 64 bit stuff will be a nice future feature to have.

I thought for a while that I'd do that, but I started getting tired of 12-hour SVCD encoding jobs (which is what you get with a 1.0-GHz Athlon when you use TMPGEnc at some of its highest-quality settings). Besides, a single-processor Hammer setup looked like it was going to be more expensive than the dual-processor Athlon MP that I just put together. With 12-hour jobs cut down to just 3 hours, life is good. :-)

(Whether a single Hammer would be faster than a dual Athlon MP is still an open question, especially with 32-bit apps. I've heard Hammer is supposed to be 10-25% faster at the same clock speed when running 32-bit apps, but one processor would still need to be damn fast (probably 3500+ or better) to keep up with a pair of Athlon MP 2100+s.)

Re:The move to 166mhz bus is nice but

[geoswan]

Might as well wait for the Hammer. The built in memory controller should to wonders for latency. Of course the 64 bit stuff will be a nice future feature to have.

Let me share a budgeting lesson from the game of Monopoly, that I feel is on-topic here.

In Monopoly, practically everyone wants to acquire Park Place and Boardwalk. Sure, when your rivals hit those properties, once they have hotels, they have to dig deep. But Those properties are expensive to buy, and expensive to develop. Whereas Baltic Avenue, and its sibling, are very cheap. Developing houses and hotels on those properties is also very cheap. And yet, when you do the math, the return on investment on those two properties is the best on the board -- better than Boardwalk.

The new machine, the cutting edge machine? You know you have to pay a premium for it. You know its value will depreciate very quickly. Its value will depreciate much more quickly than a computer built around a more mature technology.

Sure, I figure buying the latest, whiz-bang thing at premium prices, so you can have bragging rights, is a fine strategy, if you are rolling in dough. If I won the lottery, I would go right out, and buy a premium machine, with lots of memory, and 512MB of DDR.

But if you are on a budget, I don't think it is a good strategy. A lot of my baby-boomer pals hold off on buying a new computer, until they can pay for a premium, latest whiz-bang thing. When I ask them why, they say, "well, I want it to last me for five years or more. So I have to get a really powerful machine, so I won't be left too far behind."

I figure that, if you are on a budget, you should buy the technology that is a year or three behind technology's cutting edge. It is a lot more affordable. So, you can afford to replace it, or upgrade it, more frequently. I figure, on average, my computer is more up to date if I upgrade it every two years, but only to the level of last year.

My last CPU was a K6-2 500MHz. I paid about $75 CAD (about $50 USD) for it. I used it for about two years. Last week I bought a Duron 1100 for $75 CAD, and an ECS K7A motherboard, for another $75 CAD. Next year maybe I will replace my old PC133 RAM with DDR. Maybe I will get an Athlon 1400, when its price drops to $75 CAD.

No, it doesn't give me bragging rights. But, on average, I figure I am farther ahead than if I blew all my dough on a premium machine I expected would last me five years.

My buddies who buy that latest whiz-bang thing are happy with their bragging rights for the first six months, and then, if they follow their budget, they have to sit through 54 months of feeling their computer was an expensive lemon.

Re:The move to 166mhz bus is nice but

[jonadab]

> In Monopoly, practically everyone wants to acquire Park Place and
> Boardwalk. Sure, when your rivals hit those properties, once they
> have hotels, they have to dig deep. But Those properties are
> expensive to buy, and expensive to develop. Whereas Baltic Avenue,
> and its sibling, are very cheap. Developing houses and hotels on
> those properties is also very cheap. And yet, when you do the math,
> the return on investment on those two properties is the best on the
> board -- better than Boardwalk.

Okay, I'm going to follow your line of reasoning way off
topic here, but rest assured I will bring it back full circle
and return to the topic at hand eventually...

Both are poor investements, if they are the only thing you develop.
Boardwalk takes too long to develop and doesn't get hit with any
frequency, and Baltic and Mediterranean with hotels can get hit
three times and not pay you enough to land once on any serious
developed property. Sure, they pay for _themselves_, but you
can't build a game strategy around that, unless you plan to forego
dice and land on your own property every time.

The light blue, orange, and yellow properties are the ones you want.
The orange ones (New York and so on) are best. Build them to three
houses as quickly as you can for optimal return on investment. When
you can afford it, push them on to hotels for the extra income. The
yellow (Marvin's Gardens and whatnot) are a bit harder to get
developed, and the light blue (Connecticut et cetera) max out too
low, but they still give a good return on your investment. If you
can get both these sets, build the light blue ones up first, and
pray the orange ones don't get built up by someone else before you
can get serious with the yellow ones, because a couple of lands
on St. James will wipe out your chances of building up any
investment capital. In a pinch, you can substitute the magenta
or red ones, but it's an uphill battle, because the magenta (St. Charles &c) cost more than the light blue to develop and don't get
hit enough to pay off like the orange, and the red ones compare unfavourably with the orange and yellow on the same grounds. I
should mention that the light blue set by itself is inadequate
to allow you to compete in the game. However, it can be good
enough to let you get another set developed that you otherwise
could not (say, the red ones).

In the event _two_ powers emerge with sustaining levels of hotel
income, then the properties on the fourth side of the board (green
and dark blue) become important.

If you play with an open market (trades and sales among players
permitted), it is _always_ a good investment to purchase any
bank-owned property you land on except the utilities, because
developable property is worth more than the bank price. (Usually
substantially more.) If you play with a closed market, you have
to be more selective in the early game, so you can afford to get
one complete set. Also: resist the urge to believe that the
rents on undeveloped properties (excepting railroads when there
are no serious (>Baltic&Med) developed properties yet) can have
an impact on the outcome of the game; it ain't so.

> The new machine, the cutting edge machine? You know you
> have to pay a premium for it.
This is true.

> You know its value will depreciate very quickly.
While also true, this statement is meaningless. _All_ hardware
depreciates rapidly, whether it was top-of-the-line or bargain
basement or used. Today's $200 system will be worth approximately
nothing in sixteen months.

> Its value will depreciate much more quickly than a
> computer built around a more mature technology.
Only because it has further down to go. What is more interesting
is not the resale value but the replacement value and the cost
of maintaining it at a usable level.

> well, I want it to last me for five years or more. So I have to
> get a really powerful machine, so I won't be left too far behind

There is merit in this approach. Now, "really powerful" may be
overkill, but you do want to get a system that will be able to
be maintained with affordable upgrades for several years, for two
reasons. First, it means you can get comfortable with the system
and finally get to the point after about two years where you
_don't_ discover every _week_ something you hadn't got around to
installing yet that you need (PAIN), and second because upgrading
is a good deal cheaper than replacing, so the costs ballance out
if you strike a decent happy medium.

Now, it's possible to go to far. A Boardwalk system is not
for the average user. It's price-prohibitive. But it's possible
to get a system that can be developed (upgraded) to a decent and
reasonable level, like New York and St. James, for a pretty
reasonable price, and it will last you a lot longer than a
Baltic system. My computer right now is over four and a half
years old (well, most of it is; some components are newer).
It will be at _least_ another year, maybe two, before I have
to replace the system. (Some components I'll be able to keep,
of course, but I'm talking motherboard and CPU at least, and
probably some other major parts too at that point.) If you
bear with me, the ecconomics of this will bear me out.

Discounting the monitor, which is really a subject for another
thread, I paid $1550 for this sytem new, in 1998. It's a
PentiumII/233 system, but the motherboard was a nicer one with
lots of expandability. I could have got a system for around
$1200 at the time, but it would have been much lower end, not
nearly as upgradeable. For example, when RAM prices dropped,
I eventually beefed up my system to 512MB of RAM. If I'd bought
a $1200 system, it would have maxed out lower than that, and I'd
have replaced it by now; instead of spending $80 on RAM a few
months back, I'd have probably spent $400 on a new system. PLUS
I'd have had the hassle of losing my nice, comfortable system with
everything I use already installed and going back to an out-of-the-
box system with virtually nothing installed, at least two years
sooner than necessary. Compare:
$1200 + $400 + PAIN = $1600 + PAIN
$1550 + $80 + comfort = $1630 + comfort
In addition, I had a somewhat better system ad interim. My
conclusion: Yeah, Boardwalk systems are for people rolling in
dough, but Baltic systems are for people who enjoy pain. Buy St.
James systems (or at least Connecticut systems) and stay sane.

What this means is, you don't have to wait until you can afford
a Hammer system. All you have to do is wait until the news of
Hammer systems hitting the market drives the prices on moderate
Athlon XP systems through the floor, and buy one of those (St.
James) or a good quality non-bargain-basement Duron system
(Connecticut). If you feel guilty about saving money at the
expense of a struggling computer industry, make a donation to
your favourite OSS vendor or something.

Disclaimer:
People who use a lot of CPU power may find that things
break down differently. Most of what I do leaves the CPU
sitting idle most of the time, so I find that things like
RAM and drive space (I'm a multibooter (six OS installations
on the same hardware and counting...), which uses up drive
space several times as fast) are more important. If you do
a lot of raytracing or calculate the factorials of large
primes, you'll have to upgrade the CPU, and that costs more.

In Today's Rumor Mill Report...

[Anonymous Coward]

Sorry, my ears are now tuned to the Hammer frequency.

Subject, of course, to pricing ;-)

Re:In Today's Rumor Mill Report...

[scotch]

Did I hear you right? Is it almost Hammer Time?

Re:In Today's Rumor Mill Report...

[WIAKywbfatw]

In my crystal ball I can see an email inviting Intel marketing people to an upcoming strategy meeting.

The subject line is becoming clear... it says... Stop Hammer Time!

Re:In Today's Rumor Mill Report...

[ArthurDent]

It doesn't matter because you can't touch that!

Slightly off-topic, but...

[Mr. Sketch]

I've been curious, I have an old standard Athlon motherboard (Socket A), and I wasn't sure if it would work with a new Athlon XP processor, or if I would have to upgrade the motherboard too. I thought I remembered reading an earlier slashdot article a while back about motherboard incompatibility, but I wasn't sure. I would just like to know so I can budget a new motherboard, if necessary, in my computer upgrade in a few months.

Any help would be much appreciated, thanks!

WTF you talkin about willis?

[Scrybe]

Right now the AthlonXP has a FSB of 133DDR or an effective 266. Increasing this to 166DDR or 333 gives an additional 25% bandwidth. This means that the IS a point to putting DDR333 ram in an athlon MB and seeing a real performance advantage. I personaly would like to see them skip 333 altogether and go to 400. This would bring them up to one generation behind the P4 in terms of FSB bandwidth and would even out alot of the test scores that ppl are complaining about right now.

In case you have been asleep for the last year the FSB is the AthlonXP's largest bottleneck!

As for overclocking: Remember when the 266 FSB came out and ppl were complaining about the low overclock potential on the new boards? well that will happen again, but the second generation of boards will ROCK for overclocking. I have my money on boards that will handle a 400 FSB within 6 months of volume market penetration for the 333FSB.

Re:WTF you talkin about willis?

[mczak]

nope, the original poster was (somewhat) correct.
It's true the ram doesn't need to run at the same frequency than the FSB, but it doesn't help if it runs faster (DDR 333 ram has a bandwidth of 2.7GB, but the FSB at DDR 266 only 2.1GB). So, the FSB is not the bottleneck if you have DDR 266 ram, but it definitely is with DDR 333 ram. (There is an exception to the rule that higher-bandwidth-than-fsb ram won't do much for performance, this is if agp texturing is used, but this really only matters if you have an integrated graphics chipset.)
That said, tests with the higher (333) FSB show decent, but not really large performance increases - the Athlon XP doesn't seem to be that much memory bandwidth limited today.
I also disagree with the original poster about just using DDR 400. Not only a JEDEC specification for DDR 400 ram doesn't exist (and DDR 400 ram is needed to get really a performance improvement out of a DDR 400 FSB), but first boards which support such rams have some stability problems obviously at that speeds (the new asus kt400 board only allows one (!) dimm at DDR400 speed, and 2 at DDR333 speed). So, this would most likely be a nightmare for board manufacturers.

mczak

Re:WTF you talkin about willis?

[Chris Burke]

That's mostly true, but there is one factor to consider -- the FSB of the Athlon has much better utilization than the memory bus. The theoretical maximum of the DDR RAM is exactly that -- theoretical. You're usually doing really good if you're getting 70% of that. So the 266 MHz FSB can benefit from faster than 266 MHz RAM, to a point -- benchmarks with PC2700 and 266 MHz FSB bear this out. As the RAM frequency increases, you get diminishing returns until you really are getting no benefit because you are saturating the FSB.

In fact, I'd wager that the reason the 333 MHz FSB shows only minor improvements in performance with PC 2700 is because PC2700 was still not completely saturating the 266 bus, so the new FSB only gives a tiny improvement in actual bandwidth. So that plus the small improvement in latency are the only benefits you're getting.

Re:WTF you talkin about willis?

[Chris Burke]

Yes, you can construct an artificial address stream that gets close to the theoretical maximum bandwidth of the RAM. Yes, your theoretical maximum is limited by the FSB as well as RAM. But outside of SiSoft and other synthetic tests, that maximum is rarely achieved. In the case of bandwidth-hungry applications with non-optimal access paterns (nearly all of them), the more efficient FSB will provide more than sufficient bandwidth to handle what the memory bus is capable of sustaining.

I'm not saying that the faster FSB won't be good, or that it never happens that the FSB is limiting bandwidth. I'm saying that most of the time at DDR 333 the bottleneck is still going to be the memory bus, not the 266 FSB.

Re:WTF you talkin about willis?

[tshak]

Remember though that Intel's bus has a much higher latency that limits it's performance in certain applications. When DDR400 becomes available it will be a superior solution then the Rambus solution that Intel used. Like CPU's, the FSB on each platform can't be compared Clock for Clock.

Re:Wtf is with this?

[OrangeSpyderMan]

I completely agree. I am sorry to say that there is simply no news in this article, just a rumour. And a pretty dull one at that. To hell with my karma, this just isn't news, and shouldn't be on here. It's not even the release of a chip (which as others have pointed out happens at a predictable rate every month or so anyway... so probably isn't news either).

I guess it's just a slow news day for /.

How about they release the 2400+/2600+ first

[1000StonedMonkeys]

A 333MHz FSB is all well and good, but until AMD actually delivers the XP 2400+ and XP 2600+ that they supposedly released a week ago, I'm going to take this sort of announcement with a grain of salt.

Not a troll, just a question ...

[KelsoLundeen]

Okay, let me preface this by saying I'm genuinely curious about the answer. So I'm not trying to sow the troll seeds here.

That said, I'm curious about what people are using these super-fast processors for. Apart from upgrading so that you can play the immiment Unreal Tournament 2003 demo ("Only two weeks away!") and hoping to get the jump on a Doom 3 system -- what exactly are people doing with their super-high powered rigs?

I just upgraded to an Athlon XP 2000+ (from a PIII), and while I sorta dug the impressive 3DMark2001SE scores (over 10,000 with a Ti4600), I'm still not exactly sure what I need all this speed for.

For gaming, yes.

But for what else? MS Word still opens in a split-second.

OpenOffice 1.01 still opens pretty quickly.

IE, Netscape, and Opera still open in a split-second.

And, yes, now I run Quake3 with all the settings cranked.

But this sorta of "gee whiz, that's cool" wore off in a couple of days.

Now I'm left with a pretty powerful system, but I'm at loss as to what it has actually improved. Maybe if I were doing a lot of coding, then the compilation speeds would jump significantly, but I guess since my main coding right now is writing a fairly small (only around 6,500 lines) text-adventure in INFORM, I haven't really seen the jump in compilation speeds I'd see if I were compiling hundreds of thousands of lines of code ...

So, I'm curious. I haven't tried NWN yet, so maybe that's the sort of high-powered cybercrank I need to get myself hooked on the slickmercury speeds of AXP 2000+ and Ti4600.

There's always the new Neocron (sp?) beta 4 out ...

Anyone?

Re:Not a troll, just a question ...

[TheWickedKingJeremy]

I'm still not exactly sure what I need all this speed for.

You probably dont ;) If you have an Athlon 2000+ you should be all set for some time... Even games do not require a processor like that, as the graphics card plays a much more important role in game-performance. I have an Athlon 1200 and a Geforce2GTS, and I have yet to play a game that does not run beautifully.

Processor speeds are most important if you do a lot of heavy number crunching, such as video encoding, etc... or if you just want to kick a$$ on Seti-at-home

Re:Not a troll, just a question ...

[VAXman]

Mainly to develop faster microprocessors.

Re:Not a troll, just a question ...

[edremy]

I'm not normal, but here's a few from my background

Video editing. Nothing out there is remotely fast enough for what I want to do, and what I want to do is pretty limited.

Computatational chemistry. Nothing out there (or scheduled for the next ~100 years) is fast enough to do the simulations people are really interested in.

License key cracking for those companies who decide to use encryption. :^)

Re:Not a troll, just a question ...

[killmenow]

I'm still not exactly sure what I need all this speed for.

For finding things

Re:Not a troll, just a question ...

[Telastyn]

Why running windows of course!

heh, but in all seriousness, I use most of my speedy athlon machines for running algorithm confirmation tests, and silly things like mersenne checking.

Re:Not a troll, just a question ...

[gregbaker]

That said, I'm curious about what people are using these super-fast processors for.

I agree; I don't see the point. I just upgraded to a Althon 2100+ from a Celeron 500 and the difference is minimal. Kernels compile in a flash, but other than that, no great improvements. Some lags is a few applications are gone, which is nice.

What I really want is a faster hard drive--the only real wait on my system is for large applications starting up as they come off the hard drive. Opening Openoffice takes about 10 seconds; closing it and opening again (from cache) takes maybe two.

I'm thinking about setting up my /usr partition as a two-disk RAID-0. The throughput should double (small test partition confirms). Sure the probability of failure doubles too, but my /usr is all backed up by my local Debian archive anyway. :-)

Re:Not a troll, just a question ...

[Zathrus]

Most others already said it - video editing, video encoding, high end math. If your compile environment needs that kind of horsepower, then you're probably using bad tools or doing something horribly wrong in your code (duh, there are exceptions, but for 95% of the coding out there?)

In the business world there are often needs for more CPU - although more often than not the issue is I/O throughput rather than CPU.

As for NWN - shrug... I'm running it on an Athlon 750 w/ a 32 MB GF2. Am I missing out on some of the eye candy? Probably, but it still runs just fine.

I do plan to buy a new system in the near future though - but I'm playing the waiting game right now since I don't need a new one yet and some of the parts I want (or would like) aren't available yet. Mainly I'm waiting on the NV30 to be released. A 333 MHz bus Athlon would be nice (although there's a slim chance of returning to a P4 -- I'm still vaguely looking for someone doing benches comparing an RDRAM and DDR setup) and Serial ATA would be nice. USB2.0 and Firewire are everywhere now, and all the new MBs for Athlons will do thermal die checks.

Oh, why do I want a new system? UT2k3 and Doom3 of course. Duh.

Re:Not a troll, just a question ...

[DarkRabbit]

I know you are going to get slammed by the 'Bill Gates said 640k was enough' crowd so I thought you would like to know that Doom 3 isn't going to require a Freon cooled GeForce 3000 XP+ SupraGamer to be playable ...

Doom III playable on current hardware says John Carmack in Interview with GameSpy

-DR

Re:Not a troll, just a question ...

[archen]

Mozilla

yes I'm serious. At times it's still the only application that can actually make Winamp skip. (That's with an athlon 1700)

Re:Not a troll, just a question ...

[balloonhead]

For most users, the answer is no, you don't need any faster. Today's 'obsolete' chips (like my AMD XP 1500+) are much faster than what I need. For a few power users (scientific stuff, large app compiling, rendering, SETI...) these will offer some benefit. For a few 'bleeding edge' enthusiasts, they are desirable. For the rich, they may as well have them. Everyone else - 800MHz is probably enough; it's about the bare minimum for decent DVD playback (you can scape by on a bit less but that's a reasonable minimum.)

Remember also that today's 'normal user' chips are yesterday's fastest - if chip makers thought the way you did, then we might have stopped with only a few hundred MHz (and of course 740K RAM).

The other benefit is in coding - today's languages are designed with bloat in mind. That's why C++ is better than C (war! war!) - C is more efficient, faster, and therefore better... but C++ on a faster processor is just as fast (or faster), who cares about efficiency, and easier to program and maintain (the whole purpose of OO programming) - the point is, let the RAM and processor take care of the dirty work, and give us the apps to play with. The bloat simply doesn't matter any more. Obviously this is oversimplifying somewhat (peace! peace!) but the principles hold true... just think, if processors were fast enough, and computers were powerful enough, then programming languages could be so powerful (read - idiot friendly) that all you would need to do is fire up your 'Microsoft NaturalLanguage++(h4X0r 3d¦7¦0n)' and type in "/\/\4k3 4 l33t g4/\/\3 4 8¦7 l¦k3 'quake 8'" (make a l33t game a bit like Quake 8 - I'm not very fluent in h4x0r I'm afraid) and it would do all the dirty work for you.

In the short term it's nice when things do start a little bit faster, but for most intents and purposes it doesn't matter. In the long term, when your house central server is automagically ordering your groceries, booking the car in to get a service, paying your bills, playing you two different DVDs in different rooms and some elevator music in the hall, running SETI and PersonalGenomeDecoder (OK, I made that one up) in the background, and the kids are deathmatching with a video-link up to their friends (who live in big plastic bubbles on the moon), and sending all this info and more back to microsoft; only then will you appreciate what these slight incremental power-ups might do in the long term.

Re:Not a troll, just a question ...

[StoryMan]

Did you read the last part of his message?

He *is* programming. Inform is an OO programming language for creating Zmachine text-adventure games -- like the old Infocom games.

Check out the Inform website.

Re:Not a troll, just a question ...

[Christopher Thomas]

That said, I'm curious about what people are using these super-fast processors for.

Well, I was *going* to use it to try materials/chemistry simulations based on brute force approximate solutions of Schrodinger's Equation...

But, in practice, it's been for gaming. Tribes 2 ran like a slide show until I upgraded both my processor and my video card. It's also nice being able to play Amiga games under WinUAE without the sound skipping.

Re:Not a troll, just a question ...

[friedmud]

Exactly!!

Just slap Gentoo Linux on the system and you will be using every MHz you got to create the ultimate operating sytem.

I personally just have a 1.2GHz with 512 SDRAM and I am severely wanting to upgrade right now - but I think I am going to wait until next summer so I can get a Hammer... mmmmmmmmmmmmmmmmmmmmmmmm 64-bit Linux! Delicious!

Derek

Re:Not a troll, just a question ...

[aussersterne]

Someone already mentioned video editing and DVD mastering. This is job #1 for my fastest single-CPU setup, which is always OC'ed and which cannot possibly get fast enough. I'm a small operation with only a few machines. It is no fun to have to stop working and wait hours for a chunk of video to encode.

ALSO, there's a lot of frame-by-frame work going on here with things like Photoshop/GIMP (yup, use both) and that's where my current dual Xeon setup comes in. Here again -- for some of the more complex filters or transformations (i.e. perspective transformations and so on) on many successive images or frames... you run out of CPU very quickly. I/O is actually somewhat easy right now... a couple of mid-range drives in a RAID can easily deliver data as fast as I need it or write it as fast as I need to store it. I'm always waiting on CPU.

Re:Not a troll, just a question ...

[davros74]

You're thinking too much along the lines of a home-user running Windows. There's a very real presence of Intel (and unfortunately, not so much AMD) CPUs in low cost workstations for business use. Linux is certainly part of the picture as well, as almost all EDA vendors have or are releasing Linux versions of their tools.

What do I use my 80x86 cpu for? Well, I work in a hardware engineering group which does ASIC and FPGA design. We have a CPU farm of about 30 machines with Intel P4s running RedHat 7.2. (May see AMD Hammer chips in the future - we are excited about this possibility). We run everything from RTL and gate level VHDL and Verilog simulations, to chip synthesis, to test insertion and fault grading simulations. One of the last chips I worked on required such a large set of ATPG vectors (and the design was just so huge), that it required breaking the test vectors into ten groups, and even then, just one file (10% of the total) required an 8GB Sun box to convert the vectors to the fabs tester format, and the gate-level simulation took 10 days. PER FILE. Yes, that was total of 100 CPU days of simulation time for one chip just for ATPG vectors. And these were running on 1.8GHz Pentium 4s with 3GB of RAM. Not surprisingly, leading edge tools in this field are starting to look at distributed simulation over high-speed backplanes (read: not ethernet).

Tomorrow's technology is designed and verified on today's hardware. Every generational step in every sector of industry leapfrogs like this. You can't design next year's high performance video card using 80286s. Definitely for ASIC/FPGA design, there isn't a system fast enough for how quickly we (the engineers) would like a simulation to finish in. Being able to run more simulations overall means a better design out the door. More stuff caught up front. The faster a simulation runs, the quicker it will finish, meaning we can get by with fewer high-priced licenses for our EDA tools. (Licenses are usually in the tens or hundreds of thousands of dollars EACH).

You can never run enough tests before a product is done. How many tests/simulations are run, depends on how long they take to run. Give me the fastest CPU you got, decked out with the most physical RAM it can handle. (Sadly, the 32bit limit on current 80x86 platforms is hurting us badly - go x86-64! go AMD! Capture the workstation market!)

Re:Not a troll, just a question ...

[mcelrath]

Again, video.

My computer is admittedly aging. I have a Matrox Marvel G200 video card and I record TV shows direct to disk, edit out the commercials and then re-encode them to mpeg2 or divx to burn on CD. At this moment I've got almost all of Farscape. Once upon a time I recorded all of Babylon 5 to videotape. Call me an eccentric collector of sci-fi. ;)

These newer Athlons/P4's should be able to record straight to MPEG4/DivX without an intermediate step which would seriously reduce the disk space and time required. My current computer takes about a day to encode a 1-hour show to mpeg2.

However I'm waiting for the dual Opteron systems to come out. That way I can also use my computer while it's recording/encoding.

Yeah, I could get a PVR, but then I would lose the ability to edit out commercials, and do you know any PVR's which use MPEG4 and can transfer shows off so I can burn them on CD?

I can't stand to watch regular TV anymore. About 3 seconds into the commercial I get bored and want to go do something else. Besides I don't like them telling me when I have to watch their show. I have better things to do. I'll watch it on my own time thank you. ;)

-- Bob

Re:Not a troll, just a question ...

[dpilot]

Speech recognition?

Not personally involved, but close to a year ago I seem to remember reading that good speech recognition with the fastest Pentia (Don't remember if it was PIII or P4.) and K7 was still "slightly slower than realtime."

Seems to me that that situation may have changed, that we may now have faster-than-realtime speech recognition. Maybe now the system can figure out what I said, and then have time to do it, too.

Re:Not a troll, just a question ...

[afidel]

funny, I run it on a P2-300 and don't have any such problems, maybe something is wrong with your soundcard and/or motherboard (many Via chipsets have pci bugs that don't agree with many soundcards).

Re:Not a troll, just a question ...

[geekoid]

video editting, big number crunching, keeping the house warm.

Re:Not a troll, just a question ...

[tshak]

I second video editing. With high speed disk drives (10-15K RPM, 3.5ms access time, etc.) my bottleneck is always Premier eating up my CPU. I do not do video editing professionally, just for fun. I want to render in real time, but I'm actually rendering at over 10times realtime if there's a lot of filters and/or effects.

Re:Not a troll, just a question ...

[archen]

Ever veiw pages with heavy Chinese or Japanese? Win2k reports Mozilla will spike up to 99% CPU time on some Japanese pages, and essentially stay there until you get off the page. I think Mozilla 1.1 fixed all that, because I haven't had a problem with that for a while now. I still had to put up with that behavior for months.

Re:Not a troll, just a question ...

[Billly Gates]

Easy. Minimal system requirements for Netscape 7

Re:Not a troll, just a question ...

[joebp]

OpenOffice 1.01 still opens pretty quickly.

OpenOffice.org 1.0.1 compiles in 6ish hours for me. Hence, I need more speed ;)

Re:hm?

[TheWickedKingJeremy]

When will the CPU race finish? Will it ever?

You didn't get the memo? It ends June 17th 2004... From then on all technology will be at a stand-still, and most of us will find new jobs involving drills...

Re:hello

[scotch]

processor : 0
vendor_id : GenuineIntel
cpu family : 9
model : 11
model name : Pentium XI (SmeltTown)
stepping : 61
cpu MHz : 40094.670
cache size : 4096 KB
fdiv_bug : no
hlt_bug : no
f00f_bug : no
coma_bug : no
fpu : yes
fpu_exception : yes
cpuid level : 2
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 mmx fxsr sse
bogomips : 7605.97

wake me when FSB is 400MHz (800MHz in amdlingo)

[johnjones]

they have the FSB running at 400MHz and call it 800MHz

thats surposed come come out with the chipset and Opteron (or whatever marketing call it, the K8 )

Intel be worried very worried

regards

John Jones

I want a Hammer!!!!! :(

[exhilaration]

Who cares about the "2100+" or "2900+"? I've been holding off upgrading for months waiting for AMD to release Hammer.

:( *sniff* so close, yet so far away... Hammer..

Product Naming

[Aknaton]

Am I the only one who really, really disklikes the naming scheme of these processors? Although I know that clock speed does not always reflect performance, I would still rather see CPU names that include it.

Re:Product Naming

[liquidsin]

Why? They tell you the speed of the processor. Why does the name have to tell you the MHz? Do you go around telling people what kind of car you drive, and tacking the horsepower on to the end of that? "Yeah, that's a '98 Chevy Cavalier 2.2L 110 HP".

Re:Product Naming

[kaladorn]

Good point. Especially when even for cars, one would have to ask: At what RPM is the horsepower obtained? And what's the torque curve look like? And what's the curb weight of the car? etc.

Raw statistics are meaningless without an understanding of the entire context. Further, they actually enter the category of outright misleading.

And for the record, a moderately good car with a great driver is far more impressive than a great car with a useless driver... :)

No, FSB isn't really much of an issue *yet*.

[Kjella]

Tests run by overclockers show something like a 1.8% increase in performance for an AMD 2000+, not exactly much to write home about. However, like with rambus when it was first introduced, it's reasonable to assume that ir will become more important with higher clock speeds. If you believe that this or going to 400MHz FSB will drastically change anything, do some research. The Athlon is not designed like the PIV, and doesn't benefit from it in the same way.

Kjella

Re:why?

[Anonymous Coward]

Do you collaborate with the rest of the people that say the same thing on every article about faster processors? Or did you just copy/paste a post from the last thread?

Porno.

[autopr0n]

The custom viewer app I wrote to moderate Autopr0n.com is still pretty slow on my duron 1.2ghz. It basicaly renders all the .jpg on a page into a back buffer so youc an flip through them quickly. It can take up to 10 seconds to decompress all the pics.

So, it would be nice to get as fast a computer as I can get my hands on :)

FSB we don't need to stinking FSB

[spiro_killglance]

The clawhammer and operatons don't have a front
side bus, anymore, the memory controller is
on the chip

Re:FSB we don't need to stinking FSB

[barawn]

The FSB of a Hammer chip is the bus linking the chip to the North Bridge, which no longer includes the memory: it still has the AGP port, and needs to get to the PCI bus somehow. In the Hammer system, the FSB is actually a 32-bit HyperTransport link, running at 400 MHz DDR, so 800MHz effective, for a combined bandwidth of 6.4 GB/s. So yes, Hammers still have an FSB.

Re:FSB we don't need to stinking FSB

[hattig]

No, the northbridge is on the hammer chip itself. the FSB is on the chip itself, between the core and the switch fabric that is on the chip that connects up the CPU core, the memory controller(s) and the HyperTransport interface(s).

HT is just a system level interconnect.

Running at 800MHz DDR (1600MTransfers/s), 16-bits in each direction for 3.2GB/s in each second in the case of Hammer.

So no, Hammers do not have a FSB as such.

Re:FSB we don't need to stinking FSB

[barawn]

Depends what you define as a northbridge, and what you define as a FSB. The bus type (EV6, HyperTransport, whatever) is just a name for the signaling and protocol - the name of the bus itself can still be "Front Side Bus".

The "traditional" northbridge had a memory controller and an AGP controller, as well as a PCI controller. The PCI controller got moved completely off the North Bridge to the South Bridge and replaced with a proprietary interconnect in a lot of modern chips. The memory controller was moved on die, but the AGP controller is still off-die, and thus needs a chip for it. This chip could be called the "north bridge". It's just a name - AMD calls it the "HyperTransport AGP 3.0 Graphics Tunnel" (which doesn't really make much sense, as it also has a HyperTransport link to a south bridge - how does THAT relate to graphics?) but it's still a North Bridge, just without the memory controller.

There are two HT links on the system, which is why it makes sense to call it a "north bridge" and a "south bridge": there's a HT link from the CPU to the North Bridge (the AMD 8151) and a HT link from the North Bridge to the South Bridge (the AMD 8111).

So, yes, they do have a FSB, unless you want to call it something else: "highspeed HT link" and "lowspeed HT link" (for the North Bridge-South Bridge interconnect) maybe? Got me. It doesn't matter. The FSB has always been the high speed link out of the processor to a bridge chip, which then has a low speed link to another bridge chip which has all the PCI, LPC, ethernet, all that crap. Hammer doesn't change that, it just removes the memory controller from the North Bridge.

Re:FSB we don't need to stinking FSB

[hattig]

It is called an AGP tunnel because "tunnel" is a HT term for a device which has an upstream HT connection, and a downstream HT connection.

The FSB has traditionally carried CPU signals to the device that has the memory controller. As the memory controller is on the die of the Hammer, there is no FSB off the chip, just a high speed interconnect to connect up further processors or I/O devices.

The Hammer core does have a FSB. It runs at core speed and connects to the on-die switch that connects up the core, HT links and memory controllers.

HyperTransport is a point-to-point link, not a bus. Maybe you could call it a Front Side Interconnect, or how about Processor Interconnect, because Opteron's will have 3 HT links - and 3 FSB's on a processor sounds a bit silly...

Re:FSB we don't need to stinking FSB

[barawn]

It'd be more appropriate to call it a "North Bridge with AGP8X and HT Tunnel", as there's definitely no requirement that the north bridge connects to the south bridge using an HT link. I think one of VIA's chipsets is doing that so they can use an old south bridge... Anyway, it'll be most appropriate to ditch the "North/South bridge" concepts if they ever switch to one die for the PCI bus+AGP port+everything else. Then it'll just be a HyperTransport system hub. Of course they'll also have about 1000+ pins on one chip, but who's counting? :)

However, it really all comes down to what you define as the "FSB", which is tough because Intel just made up the term back with the PII. You can call it the core/memory/I/O interface, in which case, yes, it's internal now. You can also call it the processor's external data bus, in which case it was replaced by the HT link. One block diagram I saw from AMD called it the "system interlink" or something like that.

It's difficult to try to use old acronyms on a new design, especially because Hammer is really quite a striking difference from the old designs. I'm sure AMD will use something like "system interlink" or something like that for the main HyperTransport link. I doubt it, though - people use "bus" for just about any topology now. Wasn't the EV6 bus for the Athlon really a point-to-point link, anyway?

Well, though, if there's one thing we can agree on, though, no one will ever be claiming that a Hammer-based processor is limited by its "system interlink" or whatever. 6.4GB/s is way more than enough for now, especially when all you're doing is shoving data at the PCI bus and the AGP port.

Re:FSB we don't need to stinking FSB

[hattig]

The nVidia Hammer chipset is a single chip design. On-board AGP controller, PCI controller, all the USB, Firewire, IDE and other gizmos as well.

Some SiS Athlon chipsets are single chip as well. Pretty stable as well and well featured, and cheap.

A HT device that only has a HT uplink is known as a "HT Cave".

Old southbridges used to be PCI devices. E.g., VIA 686A/686B as used on the KT133, etc.

1000 pins doesn't seem to be a real problem for BGA devices like chipsets at the moment. AMDs 8131 is around 800-900 pins IIRC.

Re:FSB we don't need to stinking FSB

[Hoser McMoose]

Both the terms "front side bus" and "northbridge" are rather outdated and really don't apply to the Hammer architecture at all.

"Front Side Bus" stemmed from the PPro, which had two memory data buses, one on the "front side", which connected to the memory controller and everything else, and one on the "back side" which connected to the cache. With the Hammer, the cache is integrated so the "back side" bus never leaves the die (same as with the P4 and the AthlonXP) while the memory controller is also integrated, so part of it's "front side" never leaves the chip either, it just has a memory bus. Hypertransport is a chip-to-chip interconnect that is used for the rest of the system. You can call it a "front side bus" if you like, though the term really doesn't make any sense in this context.

"Northbridge" and "southbridge" also no longer make any sense. These terms originated because they were the "north" and the "south" end of a PCI bridge, which gave the processor a way to talk to PCI devices. Of course, the functions that these chips now perform no longer have any PCI at all in them, and the bridge is entirely in the I/O chip (what some people are still calling the "south bridge"). Intel is using their Accelerated Hub Architecture instead of PCI for interconnects, while AMD will be moving to using Hypertransport again as a chip-to-chip interconnect. AMD calls their chips "hypertransport tunnels", which is a somewhat more accurate title then a PCI bridge.

FWIW the only chipsets that I'm aware of which still use actual PCI north and south bridges are the AMD 760MP(X) and the ATI Radeon chipsets. AMD uses 66MHz/64-bit PCI to interconnect their chipsets, which gives them as much bandwidth as any competing technologies. ATI, meanwhile, is using 32-bit/33MHz, which is part of the reason why their Radeon chipset will likely really stink in the desktop market (nice for laptops, useless for desktops.. but I digress)

Re:FSB we don't need to stinking FSB

[barawn]

You're not going to get any argument from me here: "FSB" was a term that stuck around far past when all chips had moved cache on die, and should've just been called "system bus" for a while now, especially when the system bus speed and the memory bus speed were independent, which means that the "system bus" wasn't even really a "memory bus" - it was just a bus to a chip that contained a memory controller. NVIDIA's nForce chipset really showcased that, with the DASP integrated into the chipset. They were really trying to be a secondary processor. Now, with a HyperTransport link out, the best term would be "system interconnect" to satisfy those who can't stand using "bus" for a point to point protocol.

As per the North Bridge/South Bridge distinction, I'll agree that the original idea of the word no longer applies, but no one really has a good set of words for them yet. The "I/O" chip isn't really a pure I/O chip - AGP is an I/O port as well, and it also contains system monitoring information. Yes, it's input and output to the processor, but, well, everything is by a strict definition.

I dunno. "AGP tunnel" and "I/O chip" don't sit well with me. The first name stresses AGP too much, and while it's the main reason for the chip right now, it may not stay that way - in addition, the AGP chip doesn't need to be a tunnel (I think one of VIA's Hammer chipset is still using V.link, since they're using an old southbridge - I think). I think I'd prefer "High Speed Peripheral Interface Chip" and "Low Speed Peripheral Interface Chip" - that's pretty much dead on for the differences between the two chips, and the reason for the separate chips. Yes, there are those who merge the two chips (thus creating a combined Peripheral Interface chip) but many motherboard vendors will want to keep the two chips separate for reuse in multiple platforms.

My ears are bleeding.........

[Ride-My-Rocket]

Please never utter those words again. I actually had to live through that decade -- once was enough. :P

Try Seti@Home

[DrDebug]

Ever get addicted to looking for little green
men? Try Seti@Home. Once you are hooked,
you will want to process a workunit as fast
as you can!

Or, if you want to aid humanity another way,
try Folding@Home, where they 'fold' proteins.
There are a couple of zillion ways to fold a
protein, and figuring them out sooner than
later will definately aid people.

Faster CPU's can only help the cause.

The CORRECT answer is...

[Chris Burke]

They have two clocks, ninety degrees out of phase, and the latch data on the rising and falling edges of both clocks. This provides the four transitions needed for quad-pumping.

As opposed to the other answers, which made no sense.

Have a nice day.

Re:For fuck's sake

[scotch]

Would you PLEASE stop telling everyone what to do?

You were amusing with your cooler-than-thou attitude for about three syllables, but then it got old. I've never posted that joke BEFORE, so for fuck's sake take the stick out of your ass, get off your goddamn high horse, and get over yourself!

Remember the Osborne II

[geoswan]

Might as well wait for the Hammer. The built in memory controller should to wonders for latency. Of course the 64 bit stuff will be a nice future feature to have.

Do you remember the Osborne computer? It was a very popular CP/m computer. Osborne computer grew like crazy. Osborne announced an "Osborne II" computer, and IIRC, sales dried up, as everyone waited with baited breath for the new model. Because revenue shrunk Osborne couldn't afford to finish development of the new model. Then the IBMPC came out, and his target market disappeared.

If too many people hold off purchasing an AMD now, because they want to wait for the newest, whiz-bang thing, then the possibility exists that AMD will not be able to finance the development of the K8 on time, or even that AMD will go bust.

Re:For fuck's sake

[scotch]

Jesus, lighten the fuck up man, that joke may be old, but there is a certain humor in using an old joke at the right time. Everthing that comes out of your brilliant mind is witty and original, so maybe you don't have to resort to timing for your humor. Well for "fuck's sake", have some pitty on those of us who aren't nearly the fine specimen of the human race that you are.

Here's a free clue for you: if you think that by blessing us with your advice to "shut the fuck up" you will in anyway reduce the number of times "hammer time" or any other lame slashdot joke is repeated, or that you will increase the caliber of discussion around here in any way, you are seriously fucking deluded. The only effect your little frothing at the mouth has is to contribute to the general inane posturing that goes on around here.

PS. take the stick out of your ass.
PPS. nice hand waiving to dismiss the objections with the inaccuracies in your first little hissy fit
PPPS. get over yourself, you are not so fucking brilliant that others' stupidity can cause you pain

Re:why?

[geekoid]

That lastest games will not run reasonbly well on your system. By reasonable I mean at a good resolution, with most of the options on.

While technically you are correct, it is above there min requirements, but min requirements seldom run the game in a playable manner.

Probably mistake somewhere.

[TheLink]

If the FSB is at 266 there's little gain using DDR333.

They probably ran the FSB at 266 and the memory at 333. Go look at their other benchmarks at:

http://www17.tomshardware.com/cpu/02q3/020826/p4 _2 800-05.html

You'll see:
Athlon XP 2600+ (2133/133/166MHz).

So FSB at 133. See the bandwidth benchmarks in same article, the 2600+ seems to do about the same as the 2000+ benchmarks in the article you cited, so the memory is clamped to the same limit.

They have a Athlon 2666/166/166, so they have run it at 166/166, but they need to lower the multiplier to do a proper DDR333 vs DDR266 test.

That said,
http://www17.tomshardware.com/cpu/02q3/020821/at hl onxp-04.html
Says CL2.

Whereas:
http://www17.tomshardware.com/cpu/02q3 /020821/athl onxp-20.html
Says CL3.

Whatever it is, something is wrong somewhere.

Here's a better test

[TheLink]

http://www.sharkyextreme.com/hardware/cpu/article. php/3261_1363691__9

Look at the SiSoft Mem BW scores 2400. That looks more like DDR333 with Athlons :).

Don't know what the PCMark numbers mean. Wouldn't be surprised if they mean nothing ;).

Anyway it's a pity they didn't test Quake 3 with the 1700+ overclocked to 2200+/166. Coz I believe Quake 3 is very memory bandwidth intensive.

Whatever it is Tom's Hardware's SiSoft mem bw figures of around 2000 for both DDR266 and DDR333 are very suspect to me.

Re:For fuck's sake

[scotch]

I have to concentrate on work now - it's been fun! Let's do this again sometime, OK?

HAND

PS. shut the fuck up

Deceptive marketing

[AntiNorm]

and will work at 2.17GHz

If it runs at 2.17 GHz, then why the hell are they marketing it as 2.7 GHz? Being an EE, I am well aware of the fact that different architectures -- like IA32 vs AMD -- have different per-clock-cycle performance aspects. Yes, I also know that the customer just sees numbers and thinks 'gee, P4s are running at 2.7 GHz now while Athlons are at 2.17 GHz. P4s must be better then.' But I don't see it as ethical to get around this assumed ignorance by telling what amounts to an outright lie. AMD should instead win customers from Intel by convincing people that their processors are better even at lower clock speeds (which they are, really). If people started to think that AMDs were better at lower clock speeds, AMD's popularity would explode.

I am not being an AMD basher here. I have always been an AMD user, and continue to be one to this day. And contrary to popular belief, there are plenty of computer stores out there that label 2700+ systems as 2700 MHz. Even then, AMD knows damn well that most users think 2700+ means 2700 MHz, and that they don't realize that the s/MHz/+/; is just AMD's way of obscuring the misleading marketing. Fact is, the stores and AMD *are* marketing the systems as 2700 MHz, which they are not.

Re:Deceptive marketing

[Performer Guy]

You said it, you're an EE. Most of AMD's customers aren't and they need to sell chips. Like it or not Intel deliberately created a pipeline with too many stages so they could clock their chip high. People care about performance, AMD is telling people their chip performs like an Intel or better at some clock rate. Otherwise their customers wouldn't get it because they aren't EEs. You clearly know the real clock so there's no problem. Get over it, if AMD marketed like you suggest they might be out of business already.

Re:Deceptive marketing

[Hoser McMoose]

AMD tried that, and their popularity dropped, profits dropped and they were barely able to sell their top-end chip for $100. Apple is still trying it, and still failing miserably (though Apple has the unfortunately downside of actually have very slow chips, regardless of whether you look at their clock speed).

Face it, the few that know enough to bother looking beyond a single number know enough to realize that clock speed is a completely useless number.

Personally I don't really care too much one way or the other. AMD replaced one totally meaningless number with another totally meaningless number. Ideally this sort of thing would just encourage people to actually look beyond the pretty number and try to figure out what it actually means, but both myself and AMD's marketing dept. are well aware that that sort of thing is WELL beyond what 99% of the buying public are going to do. So, model numbers it is, and they worked. AMD's profits and average selling price have increased a lot since the indroduction of their model numbers, even now in a bit of a PC market downturn.

Re:How fast is fast enough?

[Billly Gates]

Just wait for doom3 and ut2003. Oh wait you said productivity. Ah, nevermind.

Re:why?

[Billly Gates]

I believe it was doom1 and doom2 that started the whole pentium craze in the early and mid 1990's. The 3d accelerator phase started after quake1. Todays games are old and made for late 1990's hardware. My old pentiumIII 700 with a geforce2mx can play quakeIII3 at 800 x 600 at 50 fps without a sweat. However from what I read I will be lucky to get even 2 or 3 fps for doom3 or unreal2003, even on a pIV with my video card.

My point is that people are running older software which was made for older systems. In the 1990's software was evolving faster then hardware and this is why many people hated Windows and found Windows3.1 as slow as a dog. I remember only using Windows for the world wide web and used the dos compuserve for everything else. Today its vice versa. Games and office suites mostly used today are old. Of course I am sure OfficeXP would fly on my old machine and its only games or speciality apps that would require a new systems. I bet as cd burning software gets more popular and the latest games come out, that people will once again be upgrading. Infact this might be Microsoft's only hope for OfficeXP migration. Mainly from people buying new pc's altogether.

I plan to buy one soon because I use Gentoo Linux which requires beefy hardware for its package management, as well as run UT2003.

Re:AMD HEAT issues AC is not cheap!

[Hoser McMoose]

Processor myth #1: AMD chips generate more heat then Intel chips.

Check the data sheets sometime, the amount of heat that an AthlonXP and a P4 put out is nearly identical. Both are also only in the 60-70W range, or about the same as your typical light bulb. We all know how turning on a single light in your house cause your AC bills to skyrocket!!!

Re:Power usage?

[Hoser McMoose]

Well, FWIW the Thermal Design Power of the P4 at 2.8GHz is 68.4W. This number is kinda-sorta-but-not-quite the maximum power that the chip will consume (I've ranted about this in other threads :> ). AMd's AthlonXP 2600+ has a maximum real-world power draw of about 68W as well (AMD does document the real exact figure, but I don't have the data sheets in front of me and they don't open properly in Mozilla anyway due to a bug in the Acrobat plug-in... but again, that's for another rant).

So, is AMD going to have to redesign their motherboards to accomodate this new chip? Well, maybe and maybe not. The Athlon 1.4GHz processor is the highest power consuming chip of the Athlon line to date at 70W maximum, and that's probably going to about what an AthlonXP 2700+ will require. However the 2700+ will get that 70W at a lower voltage but higher current. Long story short, it depends on the motherboard. AMD tends to leave this sort of thing more up to the motherboard manufacturer's designs, while Intel kinda forces motherboard manufacturers to do everything the Intel way. So, chances are that some AthlonXP boards will be able to accomodate this chip without change, while others will need a new voltage requlator.

Re:Little Confused...

[Hoser McMoose]

I don't know that the Barton core will use any lower latency, or lower power at all.

The Barton core won't actually be manufacturered by AMD, "Barton" is the code name that is now being used for the AthlonXP processors that UMC will be building for AMD. First off, UMC is still a few months away from shipping chips. Secondly, they might not be able to manage as high clock speeds as AMD manages in their own fabs. AMD's Fab 30 in Dresden (where the Athlons are currently all produced) is quite a high-end fab that is really tuned for high performance/high clock rate chips. UMC's fabs tend to really emphasize low cost/high yields, so they might not get the same sort of bin splits that AMD can get on their own.

The real benefit of the Barton core should be for multiprocessing setups, where the extra cache should really help out a lot. For single processor setups, I doubt that we'll see that big of an improvement over the current Thoroughbred core, probably only about 3-5%, or about the same as the improvement seen when increasing the bus speed from 133/266MHz DDR to 166/333MHz DDR.

Re:why?

[Hoser McMoose]

Well, 10 years ago everyone was saying the EXACT same thing about how we really didn't need these new fandangled Pentiums, and that a 486 was really fast enough for everything, and that in 10 years time we would all be laughing at the people who thought that they needed a 100MHz processor!

The more times change, the more they stay the same. 10 years from now, we'll all be looking back at these AthlonXP 2700+ systems and wondering how we managed to get by with such SLOW processors, because runs REALLY slowly on anything less then a 10GHz chip.

It's been happening for the past 20 years, and I don't see anything today that makes this look different.

This troll is driving me crazy

[fm6]

What's the point of this one? To stamp out urban legends by poisoning the market? To create an urban legend? To drive me crazy?

Oh wait, that must be it. Never mind!