AMD to Adopt DDR2 Next Year

Hack Jandy writes "According to Anandtech, AMD has already developed a new processor lineup for Athlon 64 processors with DDR2. The article states that internal AMD roadmaps indicate the processors should debut early next year and will require a new 1207 pin socket."

Comments

DDR2

[Anonymous Coward]

So we have to dance to get the damn thing to work?

Re:DDR2

[Anonymous Coward]

only if u can do "da robot"

Re:DDR2

[Anonymous Coward]

So we have to dance to get the damn thing to work?

Hell, I'll turn cartwheels for it as long as I can get > 90 fps on Ummah Defense I...

Re:DDR2

[Anonymous Coward]

The Intel marketing people in colourful bunny suits has been doing their song & dances for years. It is time for AMD to catch up.

Re:DDR2

[Breakfast Pants]

At least you don't have to purchase any bongos.

Re:DDR2

[Anonymous Coward]

The more I stomp my feet, shake my ass, and sweat, the better it works? Hell, I'm going to Fry's!

Re:DDR2

[Anonymous Coward]

I was thinking "oh no...." before I even started reading the comments. At least your joke was in good taste.

Re:DDR2

[stone2020]

If you read the Intel lawsuit, AMD claims Intel is holding up the ratification of the DDR3 standard. I am surprised AMD doesn't go with DDR500 or 600 for now and go with DDR3 later on since Intel made another memory mistake going with DDR2 (i.e. RAMBUS).

Socket 1207

[fr0dicus]

Will pincounts be the new megahurtz?

Re:Socket 1207

[dimator]

1207 sounds like a prime number, but it's not.

(That's your daily dose of useless information.)

Re:Socket 1207

[Basje]

They need the extra number for the integrated PCI Express controller, integrated on the chip. Chips without that controller will have less pins.

I think they will try to keep the number of pins down: more pins is more expensive to manufacture and transport.

Re:Socket 1207

[ettlz]

Aren't large numbers of pins in the socket dedicated to supplying power to the chip?

Re:Socket 1207

[Petersson]

Will pincounts be the new megahurtz?

My wallet hurtz... But I've discovered I don't need it, since after almost twenty years of gaming I've quitted and I use my Athlon XP system mostly as DVB PVR, video player, some video editing and dvd burning. What I ask for is better OS, not hardware.

BTW since all most of the controllers are on CPU, I expect motherboard prices will decrease since there is not much to remain on them. In extreme, why spend money on the processor pins and the socket itself? Why not solder the CPU to motherboard, like in the old times of some 386 boards? With adequate hole in mobo, we can cool the CPU from both sides.

Re:Socket 1207

[Anonymous Coward]

What I ask for is better OS, not hardware.

Stop using NT then.

Re:Socket 1207

[VitaminB52]

Yep - it's about time for a new CPU logo: Hedgehog inside :) .

Re:Socket 1207

[cbreaker]

"BTW since all most of the controllers are on CPU, I expect motherboard prices will decrease since there is not much to remain on them."

There's a lot of stuff on the boards and there's a lot of engineering that goes into making them. Just because the memory controller and PCI express controller are on the CPU doesn't really make the mainboard any less complex.

Don't expect prices to drop at all. And about the solder to the board thing - hell no. There's very rarely ever a defect with a CPU, but mainboards attach directly to all sorts of equipment and power. It's a buffer zone between the cpu and everything else. And mainboards aren't infallable - they die.

Plus, I've been able to upgrade JUST the mainboard, or JUST the CPU in the past to get real big jumps in performance or features without spending a lot of cash. If everything was integrated you could never do that.

Re:Socket 1207

[carl0ski]

you clearly never had a pre 586 cpu
i dont think he is not talking anything about the memory controller

IO Hardware control was the reposibily of extra hardware

network cards handled all io traffic

sound cards were very complex and the cpu did

absolutely nothing to help

A prime example is the Wintel modem

Linux doesnt support them well as drivers/software pass all the work to the processor

more expensive modems dont use the cpu at all extenal modems generally.

Example my tv tuner PCI used to work flawless on my old pentium 100

however with my current much newer 1ghz machine it has a cpu usage and slows (drops heaps frames) when i'm loading programs.

now everything has been offloaded for the cpu to do

mainly at the time it was because they were fast enough and the io components add extra cost.

You still buy IO cards but are they compatible with new boards?

Re:Socket 1207

[StarWreck]

An integrated PCI Express controller would solve the only down-side of AMD's current use of an onboard memory controller. That downside is added latency when using system memory for the graphics subset.

In AthlonXP and Pentium 4 architecture, data from the graphics card would only have to pass through the North Bridge to get to the memory. However, in current Athlon64 architecture it has to pass through both the North Bridge and the memory controller built into the CPU. This slows it down a bit.

Adding an integrated PCIe controller to the Athlon64 would get rid of this slow-down once and for all... however, this problem has not really been much of a problem because graphics cards have so much memory nowadays that they hardly ever have to access system memory anymore.

Re:Socket 1207

[Bullfish]

You think that mobo prices will fall because they can integrate less onto the chipset? Har-Har! You probably believed in trickle-down economics.

It the real world, it doesn't work that way. Look at cell phones, they don't cut the cost as they become cheaper to make, they just give you new features you don't need and that cost you money and keep the price the same (or raise it a bit).

Re:Socket 1207

[Anonymous Coward]

Right, it's the product of two primes, 17 and 71. Which means we could use it for RSA encryption. Really, really weak RSA encryption.

Re:Socket 1207

[Neoprofin]

Beyond just the ability to upgrade, the other reason that seperate components for PCs have become so popular is that these days if you fry a mobo it's nice to only have to shell out $130 for that and not $130 for a mobo plus the cost of a new processor. Usually processors don't fry, but I know a lot of people who have killed boards.

Re:Socket 1207

[Anonymous Coward]

Why in fuck's name would you format like that, you fucknig stillbirth?

Re:Socket 1207

[Matimus]

Motherboards are already one of the cheapest components, this move COULD reduce the price more, but it's probably not going to save anybody a whole lot. Unless you go into the server side of things, high end motherboards usually cost less than $200, and most pretty good motherboards cost right about $100. Where as chip prices range anywhere from ~$100 - $1000+. I just built a new computer with a motherboard that I won at an event. I'm happy with my new computer, but I quickly found out that a free motherboard really wasn't a good justification. Almost every component cost more than the motherboard I had won.

Re:Socket 1207

[poot_rootbeer]

Why not solder the CPU to motherboard, like in the old times of some 386 boards?

Good question. The kneejerk answer would be something about modularity and allowing people to upgrade their CPU while keeping their mobo, or vice versa, but is that really worth all the costs of socket design and engineering?

I mean, we have $100 motherboards and $300 CPU's these days. It's not a financial hardship to upgrade both at the same time. And the point of interchangeable CPU's is negated anyway when each new model that comes out uses a different socket design. I don't think I've had two successive CPUs that used the same pinout since upgrading my 486SX to a DX in 1993.

Re:Socket 1207

[Anonymous Coward]

He is unsucks essfully attempting to compensate for the minscule size of his penis.

If "he" is actually a "she", she is unsucks essfully attempting to compensate for the cavernous size of her vagina.

If "he" or "she" is actually both or neither, then I have no idea why he/she/it formatted like that ([size of anus? - ed]); it's one of the great mysteries of our time, second only to why Clinton and Bush were both re-elected.

Re:Socket 1207

[ciroknight]

I honestly don't see why the rest of the industry isn't crying foul; with AMD integrating all of the components, they're essentially taking jobs away from chipset manufacturers. Of course, chipsets can still be used for things like USB and such, but how long until AMD realizes the industry dream of Computer-on-chip?

But, this is a distant thing, and as long as Intel still exists, I'm not worried about AMD becoming a monopoly and completely locking us into our system's components. But I'm sure if Intel made these steps, everyone would be crying foul.

Re:Socket 1207

[fbjon]

They can also give new features that are cool.

Re:Socket 1207

[cbreaker]

I clearly what?

My first machine was a C64, then a 8086, then an 8088, then a 386sx (skipped the 286, although the 386sx is a 386DX on a 286 board) then a 486SX 2, then a Pentium 75, to the DX4 120 (AMD), then a Pentium 133, then a Pentium 166MMX (the Miracle 166 chip ran great at 300Mhz) and so on..

The 386sx boards were SO simple that the biggest surface area on the board were the ISA slots.

So you're going to tell me that everything is loaded onto the CPU now - what AMD and Intel CPU's have built in Ethernet, Sound, Video, Serial, USB, PCI? Ohh wait - NONE OF THEM. It's all duties of the mainboard now, and the more shit that gets crammed into these things the more expensive they will be.

For a long time you could get a really nice mainboard for less then $100. Now, if you want a really nice board you'll be dishing out $200. Is it because there's less demand? No. Is it because the parts are more expensive to make now? Nope. It's because they do a lot more now in standard configurations and they require a lot of engineering.

Take a look at the video on Toms Hardware about the Gigabyte factory tour.

Your observations are ignorant of the technology. Winmodems? All they offloaded was the encoding software to make them cheaper to produce - you still need the rest of the modem parts. TV Tuner? You were using an overlay on the old card, and your new one probably doesn't. Overlays are great for slow machines but you can't do shit with the video. Bus-fed video allows you to record, edit, etc.

The "IO cards" now are the same as they were in the past but because the rest of the computer has advanced enough, many of them send their data through the processor allowing you to actually DO something with it. And just because most mainboards now integrate everything onto the board instead of forcing you to buy expansion cards for evertything, don't think for a minute they aren't expansion cards. They still "plug" into the PCI bus. Meanwhile, the CPU is still just that- a CPU. AMD recently took the memory controller and put it on the CPU, and some bus functions. But Intel doesn't, and the rest of the machine is business as usuall - no different then an 8086 in basic design.

Re:Socket 1207

[toddestan]

With adequate hole in mobo, we can cool the CPU from both sides.

Or perhaps we could just put another slot in the motherboard. Kind of like a PCI slot, or maybe a DIMM slot. Then we can put the CPU on it's own riser card, which would allow the CPU to be cooled from both sides, while retaining the ability to easily swap CPUs. Since it would be an AMD chip, I'm thinking maybe it could be called "Slot A". Yeah, that's a great idea.... oh wait....

Re:Socket 1207

[Anonymous Coward]

having all these controllers on CPU will improve performance and it will also make it easier to create drivers for linux.

Re:Socket 1207

[sumdumass]

He wasn't saying that the processor had the features built into them. His statment wich is corect is that the processing and functions of the componants have been offloaded to the processor and system memory instead of the actual device taking care of it. In the older systems, the proccesor couldn't handle the increased proccesing load but in newer systems they can. Manufacturers hav opted to do this to save expence. Look at the minimum system requirement on a network card or modem, some will say intel compatible system while others say P233MHz or somethign even higher. Thats not because somethign in the bus has changed dramaticaly. It is because it is using the processor to perform functions. Another way to tell is because some require a windows operating system. network, and modem(at least)functions shouldn't change form one card to another and all be implemented in network stack or ndis wrapper. Nothign has changed to the point that this is so different except the offloading of proccessing.

Network cards and modems are particularly notorious for this. All the actual chips do anymore is just process the dsp conversion and the rest is done with the system hardware. An old 3com nic will in most cases outperform most onboard network cards bacuase of this. Combine this with some proccesor or memory intensive application and you can physicaly see the difference. Try loading a quickbooks file around 100 meg or more from a file server with the bult in nick that sufs the porn site just fine then drop a used ($20) 3com 3c905 series card in and watch the difference.

This difference can usualy be seen with other nics too. If you get a real nic and not one of the cheapies that work well enough for home use. If you replace the on board nic with the same type of pci addon card there is generaly a different.

Re:Socket 1207

[cbreaker]

You're right, and I definately understand that.

But offloading processing to the CPU isn't really making anything more simple at all from a hardware perspective, like he suggests. Whether or not the NIC has a dedicated processor instead of a simple DSP is moot; it still has to hook into the bus, power, etc much in the way that dedicated processor would. You still need the supporting hardware, even if it's more simple internally.

So the mainboards aren't going to get cheaper as he suggests. If anything, they've been getting a lot more expensive because all the companies are really cramming a lot of stuff on them now. So even if the integrated NIC, modem, USB controllers, and SATA raid controllers are all CPU hogs, it doesn't matter because the boards are still very complicated and expensive to produce.

Ohh, ps. If you're planning on using Windows XP x64 edition, you can't use the 3c905a/b/c. Unsupported. 3Com has no drivers and does not plan to make any. It sucks, because I have about 20 3c905b's that I used to give to friends when they'd bring their computers over (so that 25GB copy wouldn't take fifteen hours.) They're good cards. This is just yet another example of why OSS operating systems are a good frickin' idea.

Oh well, I still have plenty of machines that will work with them.

Re:Socket 1207

[sumdumass]

Well it seems that we are in agreement more then i first thought.

On the other hand though, If the nics and all had thier own processing power then the already established system bus would be all that is needed for comunications to that device. I suppose that they would find some way to make it better while still complying to the specs and that would increase the prices. I see your point and while i think it is very valid, i just don't think the complexities you point to would remain outside of comunicating with the processors. Maybe i'm just wrong on that.

Yes it does suck about the Windows XP x64 and 3c905 cards. I'm not sold on the newer 3com cards either. i don't think they live up to the older 3c905 series when going 10/100. I have never got a reliable gigabyte network going were i could sustain a true 1000 speed so i kind of think those are a little useless at the moment (too many things interfere).

I also concure on the OSS operating systems. I don't think it would be too dificult for someone to make a opensource driver for the 3com nics and XP x86 but i'm not advanced enough to try at the moment.

Re:Socket 1207

[cbreaker]

" I have never got a reliable gigabyte network going were i could sustain a true 1000 speed"

Yea. Although you can use the best Cat 5e or Cat 6 cable out there, with the consumer-grade gigabit over copper they put in machines and the $90 switches at CompUSA, it's pretty bad.

We've got an Extreme 7i at the office for our non-fiber servers (Fiber is still the only route if you truely need gigabit with no nonsense) and it performs very well. But the switch is several thousand dollars and the Intel NICs in the servers are at least $500 a pop - nevermind the 4-port NICs. Cheaper then fiber, much more realistic speeds then consumer grade stuff, but still not cheap by any measure.

I'm waiting to find some decent gigabit cards like the 3c905's were for 100mbit. And at a decent price.

As far as someone creating a 3rd party 3Com driver for x64, maybe it wouldn't be impossible porting a linux kernel module to Windows but it wouldn't exactly be a direct port.. And OSS drivers for Windows are - well, I've never seen any.

But there's plenty of support for the cards in Linux/AMD64. Your hardware doesn't automatically become obsolete just because the vendor or Microsoft decides you're not worth supporting. (obviously, support for things can and do become depricated in Linux too, but not as arbitrarily, and since the code is available it makes it easy for some decent kernel hacker to update anything to work with the new version.)

But back to the topic;

"If the nics and all had thier own processing power then the already established system bus would be all that is needed for comunications to that device."

On the server NIC's we have, there is still only one chip on the things. They are both the DSP and the processor in the same chip. While the boards look moderately more complex then a $5 linksys 10/100 board, I think it's mostly because the server NIC doesn't integrate all those little capacitors and stuff in order to produce a cleaner signal.

I think the whole reason that many controllers now utilize the main CPU to do processing is simply because it's cheaper, and it's possible. Like you said, the CPU was never fast enough to handle these tasks in the past. It's cheaper because the controller card chips are much more simple. But I still maintain that they don't necessarily make the physical side of things less complex =)

Aarrrrgh....

[Solder Fumes]

And here I was, thinking Socket 939 was going to be good for a LONG time, and bought a new motherboard....

Oh well, it's not like motherboards are the most expensive part of a computer.

Re:Aarrrrgh....

[Mgs0008b221]

Ah, the life of a helpless consumer. ;)

Re:Aarrrrgh....

[NerveGas]

Personally, I never buy a motherboard with future CPU upgrades in mind. It's just not worth it, upgrading your CPU within the same general architecture rarely gives you much real-world performance.

The real performance boosts come from radical architectural changes - new memory subsystems, new processer types, new interconnect, etc. - and for any of those, you're going to need a new motherboard, period.

steve

Re:Aarrrrgh....

[Anonymous Coward]

Very true.

But still. I was hoping that with the new era of dual core CPUs, I would for the first time actually upgrade the CPU without changing MB at the same time.

Re:Aarrrrgh....

[Anonymous Coward]

Personally, I never buy a motherboard with future CPU upgrades in mind. It's just not worth it, upgrading your CPU within the same general architecture rarely gives you much real-world performance.

But this was THE major attraction of Socket 939 to me - you can whack a whole new type of CPU (one with two cores) into your existing board and have it work fine, possibly just with a BIOS upgrade if needed. That's a far more important archictural change than DDR2 or SATA or some other standard.

Compare this with Intel, pushing new chipsets and sockets and bus speeds and memory types at every damn opportunity. Socket 939 was/is actually a really good investment.

Re:Aarrrrgh....

[ksheff]

swallowing product and crapping cash

Re:Aarrrrgh....

[gilesjuk]

Always good to upgrade every year or so. Keeps things running smoothly. Either that or clean your hardware periodically in PCB cleaner :)

Re:Aarrrrgh....

[Ours]

I gave up on the upgrade this and that on my PC. Now I just sell my PC and get a new one every 2.5 years. I tend to keep the sound card, network card and DVD-Writter but recently I'm not even bothering with that. During the 2.5 years I'll never buy extra RAM, change CPU and anything else that I won't keep when I'll sell it. So now it tend to just add hard-disks and external peripherals but nothing else. I save the cash for the next purchase.

Re:Aarrrrgh....

[Sinus0idal]

Even 2.5yrs isn't strictly necessary unless you are a)a major gamer or b)doing cpu intensive work.

I recently completed a CS degree with only a 1GHz machine to my disposal and only recently have I upgraded. It played all the games I needed fine.. GTA, UT just not any of the latest whiz bang stuff. I had that machine for a good 5 years I believe.

Re:Aarrrrgh....

[Cyn]

Well buying a dual processor motherboard and sticking a Ghz proc in it, then 2 years later tossing in 2 2800 bartons - sure made a huge difference.

Still, you're absolutely right about the subsystem and bus speeds, those are where you can squeeze some overall system performance.

Re:Aarrrrgh....

[NutscrapeSucks]

You are still talking about upgrading a system that's less than one year old. That has nothing to do with "investments" and everything to do with your hobbies. (Although, unlike most Intel stuff, Socket 939 was pretty much designed only for the "enthusiast" market.)

Re:Aarrrrgh....

[pato101]

probably but I'm in the same case as the parent poster, and I wish to upgrade my AMD64 to dualcore one perhaps next year. That would be a big improvement for my architecture. That's why I chose the 939 socket. Yes, an anonymous covard has the same point, don't mod me up, don't mod me down.

Re:Aarrrrgh....

[Anonymous Coward]

It's not like Socket 939 will stop being good... in fact, its fricken awesome.

Re:Aarrrrgh....

[Anonymous Coward]

Probably true but not considering that the Athlon 64 X2 (two cores) is designed for socket 939..

Re:Aarrrrgh....

[maraist]

The consolidation to 939 was probably nothing to do with having a "foreward looking standard" as having fewer current generation parts to test.. Every quarter, there is a new technolog, either driven by the video card industry or the CPU industry. The common denominator is the motherboard.. So in essence, choice of a MB is more critical than the other periferals like CPU / vid-card, because it limits your choices.. Now imagine that you are a MB manufacturer and there are two major CPU makers, and EACH maker requires 6 different configurations. THEN, you have 4 different classes of chipset makers (SiS, Via, AMD/Intel(native), nVidia, and soon ATi). To provide a complete suite of choices for the customer (so nobody can say, hey, ASUS, I like you, but I needed this particular configuration).. If as an MB maker, you picked the wrong horse (ATi, v.s. nVidia for your high end chipset), and your competition chooses the opposite, then there is a generation of customers that have cozied up to another name-brand.

As MB makers produces the dozens of possible combinations, they have to test these guys, and invariably, some have problems.. That's why you have different versions of the same motherboard (though they never advertise which version you're purchasing). You generally never want a rev1, but by the time you get to a stable rev, the next generation is out with dramatic enhancements.

Thus, in a given period of time, it is in the interest of CPU makers, motherboard makers, and chipset makers alike to limit the number of possible combinations of motherboards needed. If any one component is unstable, then the user experience will be negative.

BUT, as we generally require new types of features (such as DDR2 v.s. DDR1, PCI-x v.s. AGP), you absolutely require a new MB from time to time.. But, you want this new generation of parts to have as few combinations.

Re:Aarrrrgh....

[NerveGas]

You still can stick in a dual-core CPU, and you'll be just fine. You just can't switch from a DDR to a DDR2 memory subsystem.

Think about it - not only will a DDR2 memory controller pinout be incompatible with a DDR memory controller, the memory sockets on the board aren't the same, either. There's just no reasonable way of doing it - but you can do just what you mentioned, go to a dual-core chip with DDR memory.

steve

Re:Aarrrrgh....

[toddestan]

I gave up on the upgrade this and that on my PC. Now I just sell my PC and get a new one every 2.5 years. I tend to keep the sound card, network card and DVD-Writter but recently I'm not even bothering with that. During the 2.5 years I'll never buy extra RAM, change CPU and anything else that I won't keep when I'll sell it. So now it tend to just add hard-disks and external peripherals but nothing else. I save the cash for the next purchase.

I thought the same way, but now over 3 years since I built this Socket A system (Athlon XP 2000), it's still pretty fast, though I wish it was a bit faster. Now, I could build another computer - but I'm kind of hesitant right now and would rather wait until the whole 64 bit/dual core thing shakes out - or atleast comes down in price. So I'm considering just buying a 3200 Barton (for a bit over $100) and sticking with what I got for a little while longer.

Re:Aarrrrgh....

[corngrower]

Just drill another hole in that 939 socket. That shouldn't be too hard to do. What's one pin left unconnected anyway. I doubt if you'ld ever notice it wasn't connected.

Yeah, They've upped the power on the socket too, but what the hey, run a few more 12 guage wires to the socket from the power supply. That ought to do it. Up the voltage on that CPU cooling fan a bit as well. That will make the blades turn faster. That should be about all you need to dissipate a few more watts.

Re:Aarrrrgh....

[mordejai]

I agree. Of all the things you can upgrade, the CPU is the less probable.

For example, I bought a PIII in 1999. Since then, I upgraded the HDD once (8+30GB), the video card once (8->32MB), the optical drive twice (CD->CDRW->DVDRW), and the memory twice (128->256->512).

Now, I'm gonna buy an Athlon64. And I'll go for a 939, not for the possible X2 upgrade, but because 754 boards have only 3 memory slots. And, in a Windows World, I'm pretty sure THAT is what's gonna need an upgrade sooner.

What next 10000 pin processor?

[surajbarkale]

They better start using RF for processor to peripheral communication (using waveguides maybe).

Re:What next 10000 pin processor?

[VitaminB52]

They better start using RF for processor to peripheral communication (using waveguides maybe).

On-chip lasers are better suited for this task. Shorter wavelength than RF --> larger bandwidth.
Intel is investing some R&D money in this technology.

Socket A

[tor528]

What happened to the good old days, when pin counts lasted years and years?

Re:Socket A

[damsa]

It's probably because of the law suits, and counter suits between Intel, Amd and Via. No pin compatibility between anything. If AMD wins the anti trust suit, you might see pin compatible architechture, but I wouldn't count on it.

Re:Socket A

[orz]

They integrated the memory controller on to the CPU. Now, every time they switch memory technology (DDR -> DDR2 in this case), they have to switch CPU sockets also.

Re:Socket A

[smittyoneeach]

Well, the earth rotated, and orbited the sun, and, um, here we are.
Here is your morning dose of Solomon:

Say not thou, What is [the cause] that the former days were better than these? for thou dost not enquire wisely concerning this.
ECCL 7:10

Re:Socket A

[zakezuke]

What happened to the good old days, when pin counts lasted years and years?

Ummm... we have socket 754 which is pretty much supports the older athlons as well as the Sempron and Athlon 64s. This is common on those sub $400 pcs that you find on the retail circuit.

It's sort of like the old super socket 7 of old. Nice the fact that AMD offered CPUS as fast as 450 and 550 IIRC, even a tad higher than 450mhz in the amd k6-3 mobile if you were lucky enough to find them. While not nessicarly the best upgrade choice they are not only an option for the budget minded but most importantly those last generation high end chips either hold their value or increase in value. The socket-7 run would have been limited to 233/266mhz had it not been for AMD.

But this is all accidemic as well... Socket-a has been around for a very long time... offering speeds as low as the 600mhz... as high as 3.2ghz AFAIA. More over they are still in production.

So what happened to the good old days when pin counts lasted for years and years? They are still here, and in fact improved thanks to AMD so long as you ignore Slot-A. While I would strongly reccomend going 939 if you can, 754/slot-a is still an option even for those who gotta have 64bit CPUs. Just like the end of the 21st century when you "could" go slot-1 or stay with (super) socket 7 a while longer, or hell even 72pin simms if you really wanted to.

Re:Socket A

[Biogenesis]

Remember that before socket A there was slut A, a CPU package that lasted about as long as 754. I was hoping that 754 was the new slot A and that 939 would last a long time too though :(. Maybe 12xx whatever will last a long time, then the next *3* sockets after that will be short lived follow by a long lived 4th, then the next 4 sockets will be short lived...you got the idea 3 lines back didn't you?

Re:Socket A

[Anonymous Coward]

before socket A there was slut A

Jokes are just too easy here....

Re:Socket A

[Slack3r78]

Actually, as others have pointed out, current speculation is that the extra pins are driving an on die PCIe controller, not particularly because of the memory switch.

Also, wasn't the DDR3 spec design by JEDEC to be pin compatible with DDR2? Meaning that AMD probably wouldn't need to change anything outside the CPU itself when DDR3 rolls around in a few years.

Re:Socket A

[Shanep]

Remember that before socket A there was slut A, a CPU package that lasted about as long as 754.

I thought slut A would only do things for you for a little while and then move on to the next guy?

Re:Socket A

[Shanep]

I forgot to mention that at least slut A always comes with at least 3 ports. The others come with a maximum of 2, but usually only come with 1 unreliable port which is hard to get working. These common units tend to be expensive to run and also come with an output only port which is very noisy.

Re:Socket A

[Massacrifice]

More over they are still in production.

Socket A CPUs are not produced anymore. AMD stopped production last spring. I was looking for the picture wher they have all the engineers rounded for the final goodbye, but all I could find was this :

http://www.theinquirer.net/?article=21377

Ok, I've found the damn pic. Here :

http://www.theinquirer.net/?article=24544

Re:Socket A

[Anonymous Coward]

hahahahahaha!

Re:Socket A

[AxemRed]

Pin counts last long enough for me... I have a 939 with a Athlon 64 3000+. Right now, I could get a Athlon 64 X2 4600+ for the same board. Of course, I'll have to wait for the price to go down.

Re:Socket A

[sznupi]

754 has nothing to do with older athlons and socket A never hit 3,2GHz, 3200+ was only PR rating (cpu run at 2,2GHz) (and I assume that mentioning slot-a second time ("you can, 754/slot-a is still") was something almost in line with typo...) Although I agree that 754 will probably live a while...also thanks to mobile CPUs...

Re:Socket A

[sznupi]

754 will live, with mobile CPUs. I wouldn't be surprised if longer than 939...

Re:Socket A

[not-enough-info]

Just watchout where you get that arctic silver. It can be a real pain to clean up. Though I'm sure for most /.ers they're not gonna care as they'll just replace the whole system in a few months anyways.

Re:Socket A

[toddestan]

It's sort of like the old super socket 7 of old. Nice the fact that AMD offered CPUS as fast as 450 and 550 IIRC, even a tad higher than 450mhz in the amd k6-3 mobile if you were lucky enough to find them. While not nessicarly the best upgrade choice they are not only an option for the budget minded but most importantly those last generation high end chips either hold their value or increase in value. The socket-7 run would have been limited to 233/266mhz had it not been for AMD.

I remember I spent a lot of money back in the day on a K6-3 450 and a Super 7 board. I was thinking "Yeah, Socket 7 has been around a while, so it should hang around a while longer". Then like a month later AMD announced the Athlon and Slot A and I was like "Doh!". Then some other things changed the voltages for AGP cards and it really made that system an expensive dead end. But on the other hand, the system is solid and is still in use today - one of the most reliable computers I have ever built.

Re:Socket A

[orz]

The article (unlike the submitter) primarily talks about a 940 pin socket M2. No extra pins, just not compatible with AMDs existing socket 939 or socket 940.

1207 and Virtualization

[StarHeart]

The blurb mentions 1207, but the article only talks about M2(940). I have read mention of 1207 in relation to chips with the PCIe controller onboard. But not signs of tha in this roadmap.

This roadmap seems to suggest at least that virtualization will only come in chips with the M2 socket. I will be disappointed if that is true. I had planned to upgrade to dual core chip with virtualization, but keep my 939 board. Maybe by then I will be looking to upgrade to PCIe and won't care. I have an AGP board now.

Re:1207 and Virtualization

[orz]

The article does mention a 1207-pin "Socket F" in a footnote at the end. But yeah, the submitter got that detail wrong.

Re:1207 and Virtualization

[oldwolf13]

I'm looking at going to AMD64 with socket939, and was looking for a board with both AGP and PCIe, but can't find one.

I have a Radeon 9800 Pro which cost me a fortune at the time and I'm not ready to ditch it for a lesser PCIe card (since I can't afford a comparable PCIe that I can afford).

Any ideas why nobody supports AGP, PCI, and PCIe on the same board? I'm tired of having to buy all new everything basically to upgrade... what's the point now... I'm probably going to have to replace my RAM as well.

As to whoever posted about always having to upgrade motherboard... well I went from like a 1.3 Duron to a 2500+ and that was a hell of an improvement. Stayed with the same motherboard.

Re:1207 and Virtualization

[cartoon]

Well, look no further:

http://www.anandtech.com/mb/showdoc.aspx?i=2471

It's a full AGP x8 implementation, not a pokey AGP-thru-PCI or something like that. In fact, at present AGP is faster than PCIe on it... But read the review for yourself.

Re:1207 and Virtualization

[m0ng0l]

But there are no boards that currently use this chipset....

Re:1207 and Virtualization

[TooncesTheCat]

Why are people buying PCIe motherboards. Games right now dont even use 1/2 of the banwidth on a AGP 8x slot. It is just another case of them pushing something you do not need on you. If the market would stagnate for a while and let the developers use the 8x bandwidth to its full potential then we could think about PCIe. But as of right now their is no point on getting a PCIe card. But the thing is they are pushing really hard to make PCIe the next video slot solution. So that is why newer video cards are coming out with PCIe first then AGP later. Fuck Nvidia :

Re:1207 and Virtualization

[Giggles Of Doom]

I miss the good old days of Socket A. That thing lasted forever. Now both AMD and Intel seem to come out with a new socket every other year.

Re:1207 and Virtualization

[kesuki]

Games right now dont even use 1/2 of the banwidth on a AGP 8x slot.

Right because we all know a $400 GPU that has 3x the transistors of the main CPU can only be used -- For Games. PCIe was designed to take care of the 'limitations' of AGP, manyly that is was a one way street, ramp up the ability to push data to the card, and ignore the ability to send data back on the return trip.

PCIe Is required for card to card intercommunication, AGP simply doesn't have ability for the cards to send enough data back down the bus for 2 cards to work in parallel on the same graphical computation.

Considering how powerful and advanced modern GPUs are a bus system that allows them to return data to the main GPU is very important, because it allows 'other' programs to use them, programs that aren't 'primarily' concerned with outputting to the 'screen' but rather to a file, such as 3d-rendering software, image manipulation programs, mpeg-4 compressors, even audio effect software can utilize the GPU, because while it's designed for calculating graphics certain audio computations can be run on them faster than on a general purpouse cpu.

PCIe is better, it makes GPUs more valuable to non-gamers, and frankly gamers while a lucrative market segment are not the 'whole' picture. Not by a long shot. Many graphic professionals pay $1,000 or more for a 3-d card that can perform the rendering work in hours that a normal CPU would take weeks to compute.. So since PCIe works for gamers too, it would be more profitable for the graphic card makers if PCIe was the prevalent and dominate technology and they didn't have to build 'high end' agp cards anymore..

Re:1207 and Virtualization

[some guy I know]

a 1207-pin "Socket F"

This is silly.
Pin count should be a power of 2 (or power of 2 - 1 to provide a "key").
So there should be 2048 (2047) pins on the next CPU, say, a 32x64-pin rectangle.
Added advantage: People will attempt to insert a rectangular package the wrong way only half the time, whereas they will attempt to insert a square package the wrong way 3/4 of the time, meaning a 16.6666666666% (approx.) decrease in wrong-way-package-insertion attemptationisticalizations.
Unfortunately, this advantage will go away when the pin count is upped again to 4096 (4095).

Re:1207 and Virtualization

[mattyrobinson69]

or you could put a few dummy pins in, to prevent r-tards from not putting the chips in the wrong way

PCIe too

[Max Romantschuk]

According to the Inquirer AMD plans to integrate PCI Express as well. This would be very nice indeed, but I guess it's not exactly press release grade information at this point.

Not yet.

[aaron_ds]

1207 pins, pffft!

I'll hold out for the 1337 pin AMDs.

I for one welcome our elite cpu overlords.

Re:Not yet.

[sargosis]

i'd buy one and i'm an intel guy

Transitional moment

[Anonymous Coward]

I really am a geek now.

I read that as AMD would be adopting Dance Dance Revolution 2. XD

Re:Transitional moment

[iwan-nl]

If you think ddr2 the game is more geeky than ddr2 the memory, we must have a really different definition of geekness.

Re:Transitional moment

[Actuator Man]

I thought they bought East Germany.

Re:Transitional moment

[Anonymous Coward]

You got the word "geek" confused for the word "dork". Don't worry, dorks do that.

Re:Transitional moment

[kelzer]

If you think ddr2 the game is more geeky than ddr2 the memory, we must have a really different definition of geekness.

I'd say an ubergeek is someone who knows both definitions.

Congratulations!

Of course they changed the socket... again...

[OlivierB]

They make a living out of this!

I am an Apple kind of guy.
When I switched a couple of years ago, the thing I was most upset about was the inability to upgrade my system myself.
I was afraid that with Macs I would be locked in the hardware and would have to upgrade the whole machine when I needed an upgrade. Well that's true: if you want to upgrade the CPU on your Mac you have to change your machine (Ok you could maybe buy some "overdrive" for your Mac).
Well on x86 it's the same thing!
Theoritically you could swap out your processor for a faster one, but the average production life of a CPU socket is LESS than the average time you use a CPU before thinking about upgrading it.

So on x86 when you think about upgrading that 2 year old CPU to something new, well the pin layout has changed and you need to buy a new motherboard, with new type of Ram, and now new components (SATA, PCI-X etc...)
Although you could change all these components idividually, you must admit just changing the whole machine is often a better deal.

I highly suspect intel has a built-in incentive to do so as they produce chipsets for the motherboards, and most of the chips in the new parts involved when "upgrading".

Upgrading no longer exists, it should be called "changing-my-machine".

Re:Of course they changed the socket... again...

[TERdON]

I usually try to buy cheap second hand stuff to do the upgrades with. Cheaper (duh) and I don't have to worry about production still going on (unless I'm looking for ancient parts).

Re:Of course they changed the socket... again...

[OlivierB]

While you are probably getting a good deal out of this it cannot be done by everybody:

-- as we've seen a few days ago, people will often give, store or throw away computers. So you won't be getting a Cpu from them.

-- Also, the person who sells you his CPU, if his machine was still functional before the sale, how will he make his machine work now? Get another CPU :-)? It's a catch 22. In days where people are trying to extend the life of machines that isn't very practical.

-- Say you bought your machine right before they announce a switch: They probably won't release much faster CPUs until the switch. So what's your incentive to buy a 2nd hand CPU for a ~20% performance increase?

As much as I like your solution, it isn't mainstream. Also please consider that as they change PIN layouts all the time, your potential market for a compatible CPU will be further reduced

Re:Of course they changed the socket... again...

[eddy]

So on x86 when you think about upgrading that 2 year old CPU to something new

Stop blathering. On Socket-A I went "Duron 800", "Athlon 1333" and then "AthlonXP 2400+". That's three processor upgrades on one platform and the VERY SAME MOTHERBOARD.

Just recently I bought a new Socket-A MB (they're dirt cheap) and 2GB of DDR (which is similarly dirt cheap), so the last CPU has seen a MB upgrade too.

And do you know what? I play modern games on that sucker.

(I'll go S939 soon with a nice Venice and get a real use for all that memory)

Re:Of course they changed the socket... again...

[itsmekirby]

It is pushing it quite a bit to say x86 involves upgrading your entire system as well.

The case, power supply, hard drives, and optical drives can generally stay, even when an entire processor-motherboard-video card-memory system has to be changed. PATA is still supported on pretty much all new motherboards in addition to SATA. Those parts are about $150-$400 which don't have to be upgraded. It's hardly low enough to consider insignificant when comparing a mac upgrade and a pc upgrade.

Re:Of course they changed the socket... again...

[richlv]

this is happening through whole industry (and all other industries as well) - artificially reduced life span for everything.

it is possible to design compatible, long lasting technologies - but that would reduce profits. there is no long term thinking for consumer goods - of course, that's bad for environment, but who cares if only short term profit is ok ?

i've been thinking about a new computer - but when i try to select some fundamental parts to build around, i start to doubt. what disks shoulod i choose ? sata ? how long will this standard last ? what type of memory ? what motherboard factor ? what type of extension cards (ethernet, videi, sound etc) ?

for me also a big problem is interoperability - previously it was easy to test components by swapping them - change video, network, psu and check that mb again. also getting some system on feet fast was a lot easier because almost all parts were identical.

now there's a problem of finding the correct keyboard (because when i find an at, i need ps/2 and vice versa), finding correct ram is almost impossible - soon video and other cards will be incompatible with existing sets.

what about dvd* ? how long will which standard last (as can be seen, soon there will be another...) ? which of them will be the best bet for backwards compatibility ?

i plan to buy a new computer after a couple of months, but i am heavily puzzled wether that is a good idea (i'm already sure that's a bad investment ;) ).

Re:Of course they changed the socket... again...

[TheIndividual]

The trick is not to buy the most recent CPUs.

For example I bought an AthlonXP 2000 when I already knew that 3000's will be available some time in the future.

Saved a lot of money and ended up having a spare CPU when I switched to the faster one (on the same board).

Re:Of course they changed the socket... again...

[OlivierB]

Generally, If you want to see improvements you will need to upgrade other components.
While I agree PATA is here to stay for the next couple of years at least, you will be forced to upgrade one day or another, just as for hte motherboards.

I'm not arguing that Macs are cheaper; they are not. I didn't switch for cost and performance alone; I also factored ease of use etc.. For me it turns out to be the better "deal" (not in a dollar way).

Keep in mind as well, that Macs don't depreciate as fast as PCs. So when I am "forced" to upgrade my whole machine, I might end-up paying sometimes less than what it would be for a new PC:
Remember, when you already have a machine (or car for example):
Cost = New machine price - Old machine resale price
With Macs the latter is usually higher.

I also think that machines should be balanced and that Hard Drive and Graphics card upgrades are about as much as I will do because changing CPUs without improving memory bandwith, latency, HD I/O has little real performance improvements

Re:Of course they changed the socket... again...

[jp10558]

Ummm... Ok... I guess it does depend on how often you upgrade, but I went from my AGP TNT2 Ultra in 99 through my FX5600 to my current 6600GT all with the same basic motherboard slot. Sure I'll have to change slots the next upgrade in two years, but I'll have had 3 revisions on the same slot.

My DDR RAM has worked from 2003-present, and likely into late 2006 when the DDR2 starts to reach real world consumer price levels. That's 3 different motherboards for me and 2 different CPU Archetectures.

The only thing that has really been changing so fast IME is CPU to Motherboard. It's not really worth planning to upgrade your CPU without also upgrading your motherboard.

That all said, this next upgrade will be a bitch as it seems every component hit an upgrade cycle at the same time here (I hope this doesn't become the norm, or a lot of the x86 cost benefit for us build it yourselfers goes out the window as you said) and I'll need to look at PCIe, Socket 939 (or this new one), DDR2 and SATA. I'll need a good $2k for the upgrade vs the $300-- $500 I'm used too. Ouch.

Re:Of course they changed the socket... again...

[Anonymous Coward]

It sounds like you're upgrading CPUs too often. Do you really see that much of a difference between Athlon 1.3GHz and XP 2400+?

Personally when I use CPU frequency scaling on my laptop to switch from 1.4GHz to 600MHz and do a few C compiles, I hardly notice a difference there.

Re:Of course they changed the socket... again...

[Anonymous Coward]

You don't need SATA - you won't see any noticable performance boost in terms of throughput or CPU utilization. For some reason, all over this board people seem not to realize that. Its a good upgrade to make if you plan to buy a new drive, but otherwise don't waste your money when you can reuse your old one. You'll get no gain, unless you count the extra porn you can store. :)

Of course, your right about the rest of the cycle. Thank god PCIe is the new (stable) standard, so we won't be hoping around... though ram and sockets seem to change every other year. The perfect time to upgrade is when DDR2 and dual core become standard, because a new rig should have a lot of life in it.

Re:Of course they changed the socket... again...

[iPod is UNIX]

I learned early on never to take advice from an Apple user, they just make arguments for Apples current product line adjusting them as Apple changes directions. I'm going to present some of the most mindbaffling arguments from the Apple community that you may check with other sources and find out they are pretty much right on.

Apple products and Apple users arguments:

• The Newton, try to convince the Apple user this never was a very good PDA and by todays standard is totaly "out there", 8" x 5" x 1" inches and about a pound without batteries, for reference a palm is about 5" x 3" x 0.3" and about 0.3 pounds. The newton is still by many Apple users the PDA to have. Now ask the same Apple user why the iPod is much better then a Creative Zen. The Zen is to heavy, by 0.1 pounds.
• The time around 2000 when Apple users where still making arguments for cooperative multitasking which to the rest of the industry was pathetic and laughable. As Mac OS late but surely got preemptive multitasking by copying it from FreeBSD the Apple users finaly could lay the cooperative multitasking argument to rest. Not because preemptive was better (which it of course are) but because they had it too.
• The early stages of OS X (which really where an open beta), slow kernel, slow UI and not even easy to use. To the Apple users was of course the best thing. In reality it was so bad Apple don't even offer security patches for those machines even though they are just a few years old.
• The G4 cube. A bastardised computer, impossible to use. You needed to stand up to load a cd in the tray (top loaded). You had to turn the computer upside down to connect peripherals (all connectors was at the bottom of the case?!?). It had heat troubles taking down most of them. Of course by the Apple user touted as a marvelous piece of equipment and even today by many Apple users seen as the height of Apple design and innovation.
• The Mac Mini, we haven't seen the last of this yet I'm afraid. Of course by the Mac users seen as the future of Macs. Reality: Apple are in 2005 selling computers with 1.25ghz CPU and 4200RPM drive for $499, this excludes keyboard mouse and monitor and includes not even enough RAM to run the included operating system. If you could buy a similar spec PC (which you can't because there are no that slow) you would get at least keyboard, mouse and monitor. It will probably not take long before a hoard of not very happy Mac mini users put these to rest when they find out you can't even run todays software reasonably on a new computer, and tommorows will be next to impossible. The argument from the Mac crowd is that if you buy a Mac mini to play games you are stupid. Is there any other software for the Mac mini I must be stupid to try running?
• Unix, first let me explain that OS X is not a certified Unix. Unix is a trademark hold by Open Group and Apple is using the trademark without permission. Certified Unixes includes Solaris, True 64 HP-UX and other Big leage names. To an Apple user Unix has always been something weird and strange and generaly bad, the usual "not invented by Apple syndrome". Now the Apple user tells you he has a Unix too and Unix by now is the greatest thing thing sliced bread. A real life story was the Apple user who told me "All modern science is based on Unix", that tells you how much the typical Mac user knows what is under the hood of their computer. They tell you Apple is the largest supplier of Unix world wide. Of course OS X doesn't even remotely classifies as Unix and recent test has shown it is at least 10 times slower then Solaris on simple database serving. This of course gives Unix a bad reputation so you can imagine Open Group being more than upset (they have of course sued Apple over infringement). Real Unixes also has 8-

Re:Of course they changed the socket... again...

[Tim C]

So on x86 when you think about upgrading that 2 year old CPU to something new, well the pin layout has changed and you need to buy a new motherboard, with new type of Ram, and now new components (SATA, PCI-X etc...)

I've had the same basic type of RAM in my machine for at least 4 years, and the same basic type of hard drive and optical drive for 7. I've had the same basic type of graphics card for at least 5 years. In that time I'm on at least my fourth new motherboard and processor.

Yes, new technologies are developed that replace the old ones - that's called progress. Or would you prefer that we were all still using 8bit machines like we were 20 years ago?

Besides which, isn't the memory controller for AMD systems integrated into the CPU? It stands to reason then - new RAM, new CPU, new CPU with different features, new pin out...

Re:Of course they changed the socket... again...

[Rich0]

Although you could change all these components idividually, you must admit just changing the whole machine is often a better deal.

However, keep in mind that a computer is only as fast as its weakest link. If the CPU can go up to 4GHz, but the memory bus is so slow that the CPU sits around all day waiting for data, then you might as well save money on that CPU.

Hence, motherboard technology and CPU technology tend to improve hand-in-hand. Ditto for memory speed.

Even if you could double your CPU performance on the same motherboard, would you want to? The only exception would be if you started out with a well-below-cutting-edge CPU to begin with - in which case you could upgrade to what was top-of-the-line at the time and your motherboard would probably still be fine.

In any case, I still find it cheaper to upgrade my own PCs. Why?

1. Keep the optical drives, case, floppy drive, video card (if you're a non-gamer), etc. (Maybe even the power supply if you have a decent one.) That right there is a few hundred dollars saved over anything Dell will sell you.

2. Get a decent motherboard. Performance tests have shown that spending an extra $10 on a motherboard can often improve performance SUBSTANTIALLY (moreso that spending it on CPU/RAM). Big brands know that consumers don't ask about motherboards, and they don't look at benchmarks, so they save the $10. Don't make the same mistake. Likewise, a decent motherboard will often give you technology like firewire/extra USB/etc for literally $10 more. That saves quite a bit when you don't have to run out to walmart for a USB hub.

3. Be able to overclock if you want to - in general a retail motherboard/BIOS gives you far more control over the computer. Don't be a ricer, but feel free to experiment with timings that improve performance on your hardware without wearing things out.

4. Get the components you want. Thinking about dabling with linux? Well, check the compatibility lists and you don't end up with that motherboard which was popular with Dell or Compaq or whoever which has some random glitch with linux. ACPI is nototious for being broken, and if you are using features like powernow you want a system that isn't glitchy. Check out IRC for people groaning...

Honestly, I was surprised that I actually saved money building a PC myself - Dell has access to all kinds of deals that I don't have access to. On the other hand, Dell only sells "genuine" Intel, which handicaps them. Plus, Dell forces you to toss your perfectly good case, floppy, opticals, etc. Oh, you also get penalized for not buying a monitor, and I hope you like your growing collection of keyboards/mice/etc... You may not end up being far ahead on price, but often your performance ends up being better.

Just some random thoughts...

Re:Of course they changed the socket... again...

[tomstdenis]

Just add a bit...

The PPC [G3] is about as efficient for bignum math as the AMD Athlon-XP and the P4 ALU. See this chart for instance.

I think the G4 maybe be slightly better [the instruction set remains the same w.r.t. bignum math] but still same ballpark.

The problem with that design really is that while it has a good RISC ISA and lots of registers it's simply not meant for math. You have to execute two 4-6 cycle multiplies to get one 32x32=>64 product whereas other cpus can get the full product in 6 cycles AT A HIGHER FREQUENCY.

Where the PPC is nice though... is in embedded work. It's not quite as efficient as the ARM but can be clocked higher. So it's a good tradeoff between something like an ARM which gets enough MIPS and way low power and an Athlon which gets very high MIPS but takes a lot of power [by comparison].

All in all though I think the PPC can hold it's own against the x86 offerings. It just doesn't scale as nicely.

I don't know about the G5 though [I don't have access to one]...

Tom

Re:Of course they changed the socket... again...

[CaymanIslandCarpedie]

Keep in mind as well, that Macs don't depreciate as fast as PCs WHAT!!!! ;-) Not saying it good/smart/easy but a pretty scary percentage of PC users are still running Win95 and getting along just fine (at least in thier eyes) with todays applications. How many Apple users are still using the Power Mac 7500 and what modern applications still run on it? Hell, Apple COMPLETELY changes its entire architecture more often the many users upgrade thier PC OS ;-)

Re:Of course they changed the socket... again...

[dreemernj]

iPod is UNIX (893801): +5 Karma points in dreemernj's book of life.

Re:Of course they changed the socket... again...

[HairyCanary]

You started with a Duron, which means you started low. If everyone did that, sure we could upgrade several times. And most motherboards that supported old Athlons (like the 1333) did not support the faster XP procs even with a BIOS upgrade.

So basically all you're saying is that -your- purchasing style involves upgrades. The rest of us, who tend to buy at the sweet spot or upper end, not the low end, have to upgrade the motherboard and usually the memory when we upgrade the CPU.

Re:Of course they changed the socket... again...

[VolciMaster]

You know, I don't mind people arguing for or against certain things. I do mind when I need to reread almost every sentence to figure out what was said. If you're going to write something long and involved like this, please use a spelling and grammar check. I routinely write long email messages, blog entries, and Slashdot posts, but always go through them with at least a spell checker before I click 'send' or 'submit'.

I'd like to point out some minor points that you may do well to revise in your thinking, too.

• In just a few years Apple has changed computer architecture 2 times, and completely rewritten the OS 1 time. - the transition to PowerPC happened quite a while ago, being introduced in 1997. Eight years and running is quite a while to stick with one architecture. Has x86 been used longer? Yes. But, unlike Apple's use of the Motorola 680x0 line (32-bit from the start), they kept reengineering it to move to eventual 32-bit support, which is a primary reason for Macs costing more for a long time.
• To many people it just doesn't look professinal the way windows 2000 or Gnome do - well, I suppose this depends on what you think 'professional' looks like. If you're a Windows user, I imagine that your understanding of 'professional look' would be Win3.1, 2k, XP, etc. If you're a Linux user, Gnome, KDE, etc. If you're a Mac user, there's not a whole lot you can change about the interface (just as there isn't a huge amount you can in Windows out of the box (and yes, I know all about the Accessibility control panel)), but you're understanding of 'professional' is what you use. Professional media types think of Mac OS X as professional looking, and so do I.
• As to the G4 Cube and Mac mini - I happen to think they were (and are) both very good examples of design. The G4 Cube had cooling issues, and probably should have been more of a design showpiece than a production computer (or have added a large, slow-turning fan). The Mac mini should be compared with low-end laptops, in my mind. Yes, laptops come with the screen, keyboard, and mouse built-in, but it also means that they can't be replaced as easily.
• The Newton, try to convince the Apple user this never was a very good PDA and by todays standard - ok, there's a problem right away: "by today's standard". You can't compare the Newton of 1993-99 with today's standard. It doesn't work. Compare the Newton to other PDAs of 12-6 years ago. For an early example of a budding industry that hadn't figured out what it was doing yet, the Newton was good. Is it as good as something available today? No. Was it better than most of its contemporaries, I never used one, so can't say from personal experience, but it was well known, and fairly successfully marketed for over 6 years. Perhaps Apple should have done more to market the Newton as an add-on to your computer rather than a replacement for the computer, much like they are now marketing the iPod as an accessory, not a replacement for much beyond portable CD and tape players.
• Now ask the same Apple user why the iPod is much better then a Creative Zen. The Zen is to heavy, by 0.1 pounds. - I like the look and operation of the iPod. There are other players out there that play more formats, have bigger drives, or are smaller (though not many smaller than the Shuffle). However, as a total package, of the MP3 player (which is the only format people care about that I've talked to), interface (that thumbwheel is really nice, and very intuitive), and the iTunes store (of the online music stores, the easiest to use, in my opinion), the iPod pretty much earns a best overall title in my book, and lots of other people's books, too.
• The iPod Shuffle - does exactly what it's supposed to, and is relatively cheap. Yes, I would like it to have an SD or similar slot that

Re:Of course they changed the socket... again...

[Shanep]

Wow, is this a new troll?

Regardless. Troll you need to learn to spell, realise that zealots and reasonable people are in all camps and that you should replace "Apple user" with "Apple zealot" and then realise that you could replace "Apple" with any other name like Linux, Wintel, etc.

The Newton had the best handwriting recognition I have ever seen, even for cursive. Practically speaking it was too bulky and heavy. Every Mac user I know avoided OSX until it was thrust upon them when they bought a new Mac. So it was hardly suddenly the bees-knees in their eyes. The Mac mini is a cheap Mac for basic computing that introduces people to OSX. The real legally acceptable meaning of UNIX is not very meaningful when people just want to get stuff done. I use Solaris, OSX and the BSD's, but at the end of the day I prefer to use FreeBSD on x86 to get stuff done fast. Oh, FreeBSD is not UNIX (R) so therefore FreeBSD is suddenly much less useful to me. When UNIX-like OS users refer to their working environment as "UNIX", they know what each other means in the practical sense. I like PPC. I wish it would ramp up as fast or faster than x86. I am disappointed. I will continue to use Apple products where they shine.

The OS X GUI... To many, many people the OS X GUI is just to much to bare. To many people it just doesn't look professinal the way windows 2000 or Gnome do

Okay, now I know you are completely off the planet.

The Mac user just belly up more cash not realising Apple is doing this things because they have no long term plan for their computers or software

Actually, the truth is the complete opposite. Apple made drastic changes to go back to the beginning and lay long-term foundations. Microsoft on the other hand has massive inertia in massive code bases. Their software is completely covered in bandages because of a lack for long term planning. Apple is now at a stage where they pretty much have a good foundation OS for the future, which runs on an architecture which will keep them competitive in the long term (x86). If Apple introduces a Wintel compatibility layer when they switch to x86, Microsoft might suddenly find themselves very much up "shit creek" without a paddle.

PS, the iPod Shuffle is excellent for exercise. I am actually considering complementing my iRiver with a Shuffle because it is light, durable and easy to keep rainproof. Which is perfect for that niche and cheap enough to fit it.

If you are going to troll effectively, please improve your spelling, grammar and arguments.

Re:Of course they changed the socket... again...

[Anonymous Coward]

You needed to stand up to load a cd in the tray (top loaded). How short are you that you had to stand up to load the cd?

Re:Of course they changed the socket... again...

[g0at]

The early stages of OS X (which really where an open beta), slow kernel, slow UI and not even easy to use. To the Apple users was of course the best thing.

Well, that's because we were sick and tired of our systems locking up and crashing all the time. Pre-emptive multi and memory protection were way waaaaaay overdue. (Before the Mac I had an Amiga in 1987 so I was already used to a decade and a half of pre-emptiveness.)

Certainly though, I agree, 10.0 and 10.1 were no hell compared to where we're at today.

-b

Re:Of course they changed the socket... again...

[Nailer]

If you could buy a similar spec PC (which you can't because there are no that slow)

Please provide a link to a computer the same size (or even a little bigger) as a Mac mini that's as powerful.

Last I checked, there was a 700Mhz Pentium I, and that was about it. Just in case somebody feels like mentioning shuttle, most shuttle cases are a about 5-6x the size of a Mini.

No doubt in the next year you'll see a lot more small integrated boxes in x86 space. In the meantime, I'll be sitting in this couch, with my wireless keyboard and mouse, staring at my 4 meters of screen running OpenOffice, Firefox, XChat, Evo, vpnc, and every other app I need to get work done / geek out (yeah, I use Linux on it - Fedora 4 to be precise) from the tiny box in my stereo cabinet.

Re:Of course they changed the socket... again...

[Anonymous Coward]

you'll be alright.. don't hate the player, hate the game :)

Re:Of course they changed the socket... again...

[Neoprofin]

Nope. SATA is nice, but all mobos still suppport IDE, and most new ones support both. PCI-X is here, but PCI and AGP is still more common and has been around for ever. Just because new technologies exist doesn't mean that they're the defacto standard and the only choice, don't be a drama queen.

Re:Of course they changed the socket... again...

[dmaxwell]

You're mostly bang-on. I will point out that OS X can be quite happy in 256MB ram IF you won't be running Classic. I'll grant that OS X really likes ram and gets quite the increase in responsiveness if you go 1GB ram or more. It's just that an OS X box not running Classic is quite usable with 256MB.

Re:Of course they changed the socket... again...

[Phydoux]

As I was reading your post, I began to have nightmares about reading John C. Dvorak articles!

Up until now, I thought that he was the only one who wrote like that.

Re:Of course they changed the socket... again...

[Rudeboy777]

Can I ask what board that is? My circa-early-01 1200 MHz Thunderbird system came with a board that claimed not to support higher than 1.4 GHz

Re:Of course they changed the socket... again...

[poot_rootbeer]

Apple products and Apple users arguments [snip]

Did Steve Jobs give you a wedgie or something? I don't understand where this outburst of anti-Apple invective came from, especially since this thread is about AMD processors.

Re:Of course they changed the socket... again...

[trintron]

When comparing Mini to the Shuttle, you can actually upgrade parts if nessesary, sarcasm: and won't be so owned by Apple. :O

Re:Of course they changed the socket... again...

[larkost]

1) The Newton was the first PDA, and the first ones were the size of a current Palm/PocketPC. You are thinking of the later Newton 2K series. And for the markets that were actually buying these devices at the time (Doctors and Industry) they were the right product. Those markets were just not big enough.

2) For the computers it was designed for Cooperative Multitasking was the correct choice. Preemptive takes too much processor power for those older computers. Remember, Apple started doing this when there was no such thing as multitasking on the PC. Apple was stuck with the choice for a while, but has now gotten out of it.

3) You are overstating it, and it is not the machines that have been left behind, but the OS.

4) I own a Cube, and still use it daily, and you are just plain wrong about it. I have not had any heat problems, and how often do you change your peripherals.

5) You are mistaking MHz for speed. You can learn why this is a mistake many places. You are also forgetting that the mini (or Macintosh in general) is not being about as cheap as possible. If you compare it against comparable computers you will find that it is very competitive.

6) No, MacOS X is not a certified UNIX OS. The only OS currently available that is (that I am aware of) is SCO UNIX. Linux is not, the BSD's are not, Solaris is not. What is your point?

MacOS X is a OS in the tradition of unix (notice the lower case name), and is mostly POSIX compliant (but not certified as such). That is what really matters.

7) PPC is a great chip, but there is not enough research going into it to keep it competitive. In 2 years, when Apple's transition is complete there will be no PPC chip that is competitive. That is what came out of negotiations between Apple and IBM (after the relationship with Motorola/FreeScale suffered a similar fate).

Oh... and Altivec is significantly better than SSE2 (which is significantly better than SSE, which is significantly better than MMX). No one who has done real research in the field disputes this. But Apple is not going to be using chips with SSE2, they are going to be using SSE2's successor, the one that Intel is promising will be much more Altivec-like.

8) Most desktops are going towards XP, and everything points to Longhorn being even more of that treatment. In comparison Aqua looks very professional. And I think you are just wrong even with the comparison to Win2K.

9) On software longevity you simply don't know what you are talking about. I have programs from system 6 that still run in MacOS X's Classic layer. There we are talking about a contemporary with DOS... and at the end of its life Win3.1. Do you have any Win 3.1 software that still runs?

There are lots of programs that don't work, but most of those touch hardware, and the same rules apply to Windows as they do Mac.

Oh... and your time-scale is way off. The MacOS X transition is now 5 years old, and the PPC transition was nearly a decade ago. That is not "a few years".

Mac's also enjoy a much longer average lifespan (usability, and dependability) than PC's with 5 years being close to average for corporate usage. 3 years is pushing it on the PC side.

Re:Of course they changed the socket... again...

[sznupi]

I was able to do something like this with Duron 600 :> Bought it when it was sweet spot, just when they come out. Was working good for almost 2 years...then mobo burned and I got as a replacement one that could go up to 2400+ (Stuck in 1700+, D600 also burned; shop wanted to five me D850 (slowest at the time), so I only paid the difference). And now I have one year old SiS746FX mobo with DDR...still with that 3,5 half years old XP1700...sometimes upgrades work...

Re:Of course they changed the socket... again...

[loose_cannon_gamer]

I have to agree with the parent against the grandparent on this one. I bought an Asus CUV4X (original) in 2000, originally with a P3 celeron 600. Then it became a P3 1Ghz. Then it went to my brother, with a celeron 600. Then a 733. Then back again to a P3 1Ghz. Now it has turned back into the celeron 600 (as a low heat always-on server, back in my house). That's three different processors, in 6 configurations, and I'll swap back in the P3 1.0 come cooler weather, all in one motherboard.

For those curious, in fact that Asus motherboard has outlasted several cheaper socket 370 motherboards which were purchased later than it to replace it.

It just goes to show that in some areas, paying for quality is worth it, and I believe motherboards are one of those areas. If you build a system with solid hardware, to fill a role (other than "highest frames per second on latest first person shooter game Q among all my friends"), there's no reason to get all excited when technology marches on and leaves you with "outdated" but perfectly functional hardware.

And a note for hardware designing EE/CE/CS types -- I now have a house full of Asus hardware because of that one motherboard. Think about it.

Re:Of course they changed the socket... again...

[ciroknight]

Try that now, buy a machine, and in a year buy a new processor and try to stick it in the machine. You won't be able to do it. Then try it again in another year. Once again, you won't be able to do it.

Re:Of course they changed the socket... again...

[VolciMaster]

yes, I know he's posted this several times elsewhere on /., but I felt someone should respond to this in a calm, considered fashion.

Re:Of course they changed the socket... again...

[LordSah]

done

Re:Of course they changed the socket... again...

[toddestan]

Last I checked, there was a 700Mhz Pentium I, and that was about it.

Interesting, as I always thought the Pentium I ended at 200Mhz. Though the MMX version went to 233Mhz, and I believe there was a mobile MMX chip that went to 266Mhz. But hardly to 700Mhz.

Re:Of course they changed the socket... again...

[Nailer]

Me: Please provide a link to a computer the same size (or even a little bigger) as a Mac mini that's as powerful.

You: done

But not really.
The Mac Mini is 16.5 x 16.5 x 5cm
Volume of 1361.25

The box you linked to is 25.4 x 14.78 x 7.09cm
Volume of 2661.67

Which is nearly twice the size of the Mac Mini

Re:Of course they changed the socket... again...

[Anonymous Coward]

2) For the computers it was designed for Cooperative Multitasking was the correct choice. Preemptive takes too much processor power for those older computers. Remember, Apple started doing this when there was no such thing as multitasking on the PC. Apple was stuck with the choice for a while, but has now gotten out of it.

Not correct. In 1985, the Amiga 1000 with a 7.16 MHz 68000 processor and 256k RAM ran a preemptive multitasking OS. No performance issues there.

Re:Of course they changed the socket... again...

[corngrower]

Way back when, in the days of the Apple II. Apple had an open architecture. You could add in your own boards and peripherals to the system. Then they made the mistake of making their architeture closed. They lost market share big time because of this. Big time. What's Apple's market share these days?

Re:Of course they changed the socket... again...

[eddy]

>Try that now, buy a machine, and in a year buy a new processor and try to stick it in the machine. You won't be able to do it.

Wanna bet? If I buy Socket-939 today, in a year I will easily be able to get a nice dual-core S939 CPU and plug in. There's a lower-grade dual-core that hasn't even been officially announced yet, much less released. Of course, in a year I wouldn't go for that when I can change my Venice 3200+ into a nice 1MB cache dual-core 4800+ or something even better.

Just for the record, let us remember that the last Socket-A CPUs left AMD two months or so ago (there was a picture at TheInq from the last batch shipped). Likewise, S939 will be with us at least another year.

You know, just because you might have to plan for it, doesn't mean it's "impossible".

Re:Of course they changed the socket... again...

[eddy]

>Can I ask what board that is?

It was a revision 1 Abit KT7A-RAID. It doesn't "officially" support the XP-series of CPUs either, but it'll work if you know what you're doing. You have to set some specific settings in the BIOS to get it stable, but it really is stable. The XP2400+ was the last CPU you could just plonk in there and run, since it was the last one to run at 266MHz bus (there might be some higher speed grade mobile part too, but let's not go there).

Many older boards support this particular CPU even though the the official word is "Max 1.4GHz Thunderbird". Best way to know if it can be done and what's required is to use Google Groups.

Of course, getting hold of a cheap 2400+ XP probably isn't easy nowadays, so it's probably not a worthvile upgrade unless you can find a used one for cheap, in which case it can be a great upgrade.

Re:Just had to say it...

[Anonymous Coward]

Please refer to this event as a "deferred success" in the future.

What rate will it begin at?

[Anonymous Coward]

http://en.wikipedia.org/wiki/DDR-2

The article states that 400MHz DDR is better than 400MHz DDR2, but there will never be standard DDR made better than 400MHz.

I wonder what rate DDR2 for the new AMD's?

The standard goes up to 1000MHz. Add dual channel to that and damn

Re:What rate will it begin at?

[orz]

The article mentions DDR2-667

Re:What rate will it begin at?

[TheRaven64]

Exactly the same thing happened with DDR - it has a larger latency than SDR. In practice, this is largely mitigated by large caches - any time you go out of cache you take a significant hit, and it becomes a matter of being able to load a replacement cache line as fast as possible - which is where the faster speed of DDR2 (and previously DDR) help. The first memory access is slower than it was, but the next one is faster because by that time it's already in the cache.

First Prime Factorization Post

[2*2*3*75011]

1207 = 17*71. I wonder why this beautiful factorization isn't mentioned in the article.

Re:First Prime Factorization Post

[Anonymous Coward]

By your reasoning, the AMD geode LX800 takes the crown:

481 pins = 13*37

OMG, 1t are t3h l33t CPU!!1!!11one11

http://www.linuxdevices.com/files/misc/amd-lx800.j pg

Re:First Prime Factorization Post

[fbartho]

Soooo... is that where your username comes from?

Re:First Prime Factorization Post

[Anonymous Coward]

multiply the primes that are in his username and the total is his user ID.

Re:First Prime Factorization Post

[MexicanMenace]

I was a math major for a while. It's when I realized I couldn't stand being around people like this that I jumped ship. :P

Re:First Prime Factorization Post

[VoidWraith]

I'm still amazed you got your UID to match your handle.

Re:First Prime Factorization Post

[MyLongNickName]

And, for some reason, you still come to Slashdot...

It doesn't matter that they change the socket.

[Ihlosi]

Really, it doesn't. Since the memory controller is integrated in the CPU, there is no way to make one of these things run in todays DDR1 mainboards, regardless of the pin count. And since DDR2 has a different pin count than DDR1, of course the pin count of the memory contoller has to change, hence the pin count of the CPU changes.

Anyone complaining about "yet another socket" apparently hasn't understood this.

Re:It doesn't matter that they change the socket.

[Jeff DeMaagd]

It is a new socket, even if practical considerations dictate it.

I am curious, has DDR2 been shown to be worthwhile? I sure hope so.

Cool

[Anonymous Coward]

Nice to always know technology is progressing..i love AMD.

Posted anyonymously because i forgot my pw.

Offtopic

[Anonymous Coward]

This comment is offtopic. If you have moderator points, please apply the appropriate moderation to it.

Re:Offtopic

[name773]

you're just trying to get them to waste mod points ...

(good technique)

Re:Offtopic

[TheRaven64]

This comment is informative[1]. If you have moderator points, please apply the appropriate moderation to it.

[1]in a self-referential kind of way.

Re:Offtopic

[SmittyTheBold]

Even when they're given explicit directions, the mods get it wrong.

Ever increasing pinnage

[Timberwolf0122]

I think this is going to cause a proble later on as we are either going to need biger chips or make the pins out of C60 as they are going to start to get a bit bendy.

I am quite excited about the DDR aspect on a dual core as this should help eleviate the bottle neck of two processors on one bus.

great

[axonal]

Great, more fat kids falling off the dance floors.

Re:great

[Cantide]

I tried to quit! I need a 12 step program...

Re:great

[not-enough-info]

Hey, if there's a chance to see that one-legged DDR guy dances his foot off again, then I'm all for it.

I rule

[Anonymous Coward]

It comes down to how much performance we get by with DDR2-667. Although I am of the opinion that it's not going to matter until we get socket 1337 chips.

damn.

[chrisxkelley]

okay, i go to slashdot for my nightly fix and at a glance, this article seems to say that new ddr machines will now be stocked with amd processors. i was like "Shit yes! now those dudes in that malls who rule the ddr machines will have to compete with AMD fastness!" theeen i read it. damn article ruined my evening. another socket change. psh

Shocked I tell you!

[Groo Wanderer]

Now, can someone tell me why this is news again? I realize Slashdot is behind Anandtech, but dear ghod, AMD even announced this a while back. Listen to the last analyst day broadcast, when a company anounces something, 2 months later it is not really hot news anymore.

For the people out there complaining about pin counts, look at Intel. They went from 423 to 478 to 775 in the life of the P4. AMD has only had two sockets, 754 and 939 in this span. If you want to drag servers in to it, add 940 on the AMD side, and 603, 604 and 772 on the Intel side. Since we are bringing in other chips, Intel also has 479 and 480 on mobile, AMD has the same 754 there.

AMD has a lot more stability on the socket side, and the move to the new sockets is strictly due to added functionality, DDR2 in this case, PCIe and FBD on the server side, eventually.

-Charlie

P.S. The person who approved this topic/story should be slapped with a half-thawed cod fish until they get a clue.

Re:Shocked I tell you!

[Anonymous Coward]

AMD has a lot more stability on the socket side, and the move to the new sockets is strictly due to added functionality, DDR2 in this case, PCIe and FBD on the server side, eventually.

PCIe is already on the 939 boards (ASUS A8N-SLI box sitting right next to my desk here, for example).

Re:Shocked I tell you!

[Dachannien]

Well, this was going to be a dupe, but whoever was in charge of posting the original article forgot to do it.

Re:Shocked I tell you!

[oldosadmin]

Heh, you must be new... wait... no... your UID is lower than mine...

You must just be unobservant.

Re:Shocked I tell you!

[ksheff]

or just not give a damn anymore.

Re:Shocked I tell you!

[Anonymous Coward]

Now, can someone tell me why this is news again?

Because a reputable publication commented on it. Your rumor mill blog mentioned something seven months ago. Let's look at some other noteable entries by the great Charlie Demerian:

A personal favorite of mine. Let's quote:

• Any scholar of the great ninja vs pirate wars will immediately understand that pandas and cows are bitter enemies fighting for the affections of the zebra overlords. Do not make the mistake of mentioning cows or Gateway with a handful of M&M while in San Fran next spring, you have been warned.

You actually get PAID to write?

Some other noteables:
Curse of Everywhere Girl hits IDF
Celebrities spotted at E3
Inquirer equivalent of a slashvertisement
Another
ATI offers succour to notebook users

Are you the nerd equivalent of joan rivers?

Dominoe effect

[eddy]

I forgot the best part; my old Socket A MB (KT7A-RAID) ran on SDRAM of course, so when I upgraded that MB, I moved 512MB of SDRAM to my old linux server (K6-2 based FIC-503+), which up till that point was running on 80MB of EDO.

Now that's what I call an [cheap ass] upgrade.

Re:Dominoe effect

[Molochi]

If you can upgrade this way it is great. I see more instances of incompatability between old versions of motherboards with a given socket design than compatability. Mostly it seems that old boards can't supply the wattage needed by a new processor reliably (lower voltage higher amperage) even assuming they do support a bump in FSB speed.

KT7A was a good board for this but the KT7 (the board available when duron 800s came out) was not. I went with a A7V133 (another KT133A mobo) around this time as my older A7V didn't support a 133FSB and missed out on Athlon XP compatability. The board just wasn't stable with them. The next revision (I think it was 1.05. or something) worked fine with AXP but that board revision was only a couple of months younger than the CPU it supported.

However, it's pretty sweet when you can pull off several generations of upgrades. I had the best luck upgrading slot1 i440BX boards. There was a lot of effort put into keeping these boards upgradable by 3rd parties and it was pretty cheap to do so. A slotket adapter was cheap and many supplied addl power to keep newer cpus stable. Later S370 boards had the promise of upgradability but suffered from most of the same upgrading issues encountered when trying to upgrade SocketA systems.

Well they were going to move to dd2 anyway

[hobotron]

Im just glad they waiting this long, they let Intel take the first strides, that were almost backward (DDR2 performace was not better than DDR at launch, and im wondering if is now, also DDR2 was MUCH more expenisve than DDR).

Also to those with socket 939 boards, (like me), Id just do what I have always done, throw the cheapest processor that you can find in the system when you buy it, then wait till the socket line is at an end, then buy the last and fastest processor made for it when the price is lower because everyone else is upgrading their entire system to the new socket. You get the maximum performance gain with the lowest cost.

AMD's use of DDR2 wont bring revoultionary performance, but it will bring DDR2 prices down, and help stabalize the DDR2 market into hopefully something that can bring about some of the promises we were given at its launch.

DDR2 ?

[Arthur B.]

That is great news for Reiser4

(SELECT /.ers WHERE CLUE > 0)

Re:DDR2 ?

[Anonymous Coward]

Records found: 0

Re:DDR2 ?

[Arthur B.]

Reiser 4 uses dancing trees I know it was far fetched

Read and learn :)

[Groo Wanderer]

http://www.theinquirer.net/?article=24756

I agree with the author's views 100% here, mainly because I wrote it. :)

-Charlie

Re:Read and learn :)

[joib]

There was quite a bit of discussion about this over at aces hardware. The consensus seems to be that it isn't about reduced latency (shave off a few dozen ns for I/O, who cares), but about cost for low end (no need for a HT PCIe bridge in the chipset for single socket systems) and on the higher end freeing the third HT link for coherency traffic on multisocket systems.

Pushing old hardware...

[Gopal.V]

> So on x86 when you think about upgrading that 2 year old CPU to something new, well the pin layout has changed and you need to buy a new motherboard, with new type of Ram, and now new components (SATA, PCI-X etc...)

You have to realize that paying for a new box out of the shop is not always in the means of a student out there. I started out my computer adventures with a P 3 450 Mhz on a VIA board with 32 MB Ram and an 8 Gb HDD. The very same machine currently has the same CPU , 320 Mb of RAM, an Asus , 48 GB hdd, USB2.0 adapter and an OpenGL video card (of course it is now is being used by my sister to write word documents).

For a well paid geek, upgrading might not make much sense. But for a student with a slow & steady stream of cash, such mild upgrades do work nicely.

DDR? AMD? ITG?

[Cantide]

I thought DDR2 was made by Roxor...

Netcraft confirms: Socket 939 is dying

[Anonymous Coward]

It is official -- Netcraft confirms: Socket 939 is dying

One more crippling bombshell hit the already beleaguered Socket 939 community when IDC confirmed that Socket 939 market share has dropped yet again, now down to less than a fraction of 1 percent of all sockets. Coming on the heels of a recent Netcraft survey which plainly states that Socket 939 has lost more market share, this news serves to reinforce what we've known all along.

You don't need to be a Kreskin to predict Socket 939's future. The hand writing is on the wall: Socket 939 faces a bleak future. In fact there won't be any future at all for Socket 939 because Socket 939 is dying. Things are looking very bad for Socket 939. As many of us are already aware, Socket 939 continues to lose market share. Red ink flows like a river of blood.

Let's keep to the facts and look at the numbers.

All major surveys show that Socket 939 has steadily declined in market share. Socket 939 is very sick and its long term survival prospects are very dim. If Socket 939 is to survive at all it will be among socket dilettante dabblers. Socket 939 continues to decay. Nothing short of a miracle could save it at this point in time. For all practical purposes, Socket 939 is dead.

Fact: Socket 939 is dying

So what is this?

[Sycraft-fu]

Some kind of filmsy rationalization for Apple's impending switch to x86? Give it a rest. As others have noted, you do not have to upgrade entire systems. I currently have an Intel 478 motherboard, at the time I bought it, top of the line. You do the research on that, you find it was over a year ago. The processor is 2.4ghz. Now I can increase that by 50% on the same socket, which I'll do when the time comes I feel my processor is the limitation in my system. I imagine that'll be at least 6 more months, probably a year. After that, based on past experience, it'll be at least 12 months, probably 18-24, before I decide again the processor is what needs replaceing and scrap the board and do a major upgrade.

Well, ok, I should be honest I won't scrap it, I'll move it to another system and use it, probably one of my web servers, but at any rate.

Now you are crowing on about SATA and PCIe (you said PCI-X, you are mistaken, PCI-X is 133mhz 64-bit PCI which Apple has used for awhile, despite it being worthless for all but high end SCSI cards, PCIe is PCI-Express which is what is now being sold). Ok, great, now go and review when PCI and better yet PATA came in to existance. We are at a time when two major periphal busses are being updated. Nothing bad about that, but please don't pretend it's common. I got my first PC in 1993 and it had PATA. I didn't stop using PATA drive until 2005, and I still support a few thousand systems that use PATA. PCI is a little more receant, but still an old techolongy in computer terms.

So please, let's cut the shit. If you like Apple's products, that's fine, there's a lot to like about them. However let's not start some bullshit game like you never upgrade an x86 CPU. I have not ever owned an x86 system that I didn't upgrade the CPU in at least once. It started with a 486 that I got a Pentium Overdrive for, and now I have a P4 that I'm looking at upgrading in the next 6 months.

Upgrades to PCs, video card, harddrives, CPUs, and so on are common, at least among enthusistst. This isn't a shot at Apple, but please don't try to pro your platform up with bullshit.

Re:So what is this?

[OlivierB]

Whooooohhh
Cool it buddy!. I 'm not rationalizing Apple's move to x86. I'm not saying it's cheaper and I'm not suggesting people should move to Apple. It's all about different needs for different people. Please don't put me in the same basket as apple fanboys

My point is that I am very sorry to see that commodity hardware is becoming less "commoditized" because of all the different standards: SATA, PATA, x86, x86-64, DDR2, Rambus etc...
So switching to a Mac I was giving up less than what I had imagined.
I had a different experience from yours: you know I did have other computers before I made it to Apple and yes they were great. However with my average upgrade cycle of 24-36 months I have never been able to leverage my motherboard or CPUs in an upgrade.
So you bought more wisely than I did, or were lucky to embrace an architecture at the beggining of a life cycle: well done.

I never buy top-end stuff: I find there is most value in mid-range. The problems you get with that are those I listed.

I still have a couple of x86 PCs built from spare parts I had. I use them as home servers and they will continue to live as such because I do not need 3Ghz for SMB and light web traffic. When they die I will be forced to upgrade the motherboard and CPU (and most possibly the ram too).

My mention of Apple was just to say that MOST of the people on x86 will go ahead and change the whole machine because of incompatibility issues.
You fit in the 5% tail.

Re:So what is this?

[Anonymous Coward]

You're way out of it, and I agree with you. It should be noted that PATA has been upgraded quite a number of times to increase bandwidth, storage capacity, or change the IO mechanism (e.g. PIO -> DMA). Its not a big deal, but was kind of annoying when you needed TSR loaders to extend the BIOS, new PATA cards, etc.

Also most users (even us techies) often don't upgrade our CPUs. Back in the day we'd have a dozen computers and tons of spare parts, because a CPU upgrade was expensive and left you with a useless part. Its far more common to upgrade your mainboard, ram, and cpu in one shot. Back then if we couldn't use our old parts, we found family or friends who could. These days, computers are so common its sadly just trash.

I just upgraded my system from a 1ghz Athlon to a 3200+, and have no idea how to unload my old cpu/board/video (ram went to another PC). I no longer keep a garage full of monitors, kb/mice, cases, spare parts (I miss Number9 Imagine series), and old PCs relics I loved (TRS80, Tandy1000, Apollo, and MFM drives).

Its easier to upgrade the core platform and incrementally add new hdd or peripherals as needed. There's just no more pressing bottlenecks for us to dabble with SCSI or exciting upgrades like tape or cdr backups. RAM is cheap and fast, and CPUs are fast enough for most tasks. And video was never a priority for old school techies.

What you should say is that all those individual upgrades are common for gamers, who are the last segment willing to spend the money and time for such small incremental improvements. For the rest of us, it just isn't that common.

That's because...

[Phil John]

...compiling is, by and large, I/O bound, not CPU bound. Of course, modern optimizing compilers use more CPU than their older brethren, but the initial statement still holds true.

Re:That's because...

[Anonymous Coward]

Yes, and the point is that not very many things in day to day usage are CPU bound, and more often they are I/O bound. Unless you are crunching numbers, a faster CPU doesn't always make a huge difference.

Is switching from 1.4GHz to 1.7GHz or whatever the XP2400+ is clocked at really worth it? Probably not.

Re:That's because...

[Anonymous Coward]

Yes, and the point is that not very many things in day to day usage are CPU bound

If you read the OP you'd see that he play games. CPU is important in gaming.

Is switching from 1.4GHz to 1.7GHz or whatever the XP2400+ is clocked at really worth it?

The XP2400 runs at 2GHz, and the previous CPU was 1.33GHz, so a 666Mhz upgrade. And you get SSE to boot (which pretty much is a gaming requirement today), and the new CPU draws less current and generate less heat.

As you recall, the 2400+ rating compares to the thunderbird generation, so the AthlonXP 2400+ Thoroughbred should perform like a Thunderbird at 2.4GHz!

in other news

[Demoknight]

intel announces chipsets with ddr3 support available before amd gets ddr2

Re:in other news

[Neoprofin]

Just like they beat AMD to dual cores and quality retail 64-bit?

Is DDR2 worth waiting for?

[Tim C]

I currently have an old 2.4GHz P4 (the B revision, iirc), and am thinking of upgrading. I'd been drooling over the Athlon X2, mainly for the cool factor I admit (as I mostly do gaming on this box, with a little C# and Java development now and again). By "gaming" I mean Half Life 2, Doom 3, UT2k4, that sort of thing.

I have an AGP GeForce 6800GT, and would like to pair it with a suitable processor and RAM. So, the question is, is it worth holding out for DDR2, or should I just upgrade now? What are the real-life performance benefits of DDR2 over "normal" DDR? For that matter, will these systems even support AGP? (I'm going to have enough trouble convincing my gf to let me upgrade as it is, without having to replace a brand new graphics card at the same time...)

Re:Is DDR2 worth waiting for?

[Anonymous Coward]

If you want to use your AGP-card then 939 is the way to go since there will be almost 100% PCI-express graphics on the new sockets, and since you can reuse your ram with good performance (DDR2 in itself won't give you much extra) go for broke right now

Re:Is DDR2 worth waiting for?

[elchuppa]

a lesson I learned a long time ago: Girlfriend does not have to know what goes into the box. It's the magical part of your life where you can invest all the cash you want without her knowing the difference. Assuming that is, that you have the good sense to own a case without a window into it.

Besides, it's 'girlfriend', not 'wife'. She shouldn't have a say in the matter (yes I'm not that naive, I know that's not the way it is, but it's important that we keep in mind the way it should be).

Re:Is DDR2 worth waiting for?

[HerculesMO]

You can bet every motherboard made for that new socket will be only PCI-E.

Re:Is DDR2 worth waiting for?

[epaton]

anantech recently did an article which showed even when they upped to 500mhz ddr and it took doom3 and encodeing a dvd on a dual core cpu to even touch that bandwidth, you would be better off with the lower latency

socket 754

[Anonymous Coward]

Hmm, having just bought a socket 754 motherboard, I was hoping that amd would continue at least the sempron line on that socket, allowing me to upgrade in the future. But when looking at their desktop roadmap, it appeared socket 754 was going the way of the dodo. But on their mobile roadmap, it looks like socket 754 still has a long way to go. Woo Athlon 64 4000+ here I come.

To the person who modded me to troll

[Groo Wanderer]

I would say I told you so, but I did.

http://www.theinquirer.net/?article=20749

Please notice the similarities between the 7+ month old article and the current one. *YAWN*

-Charlie

Re:To the person who modded me to troll

[Anonymous Coward]

Oh you mean that "article" that you wrote? I'm sorry, that trash magazine you work for doesn't even compare to a real publication like Anandtech. The inquirer needs a new motto "we get it right about a third of the time".

.. that and they could probably use some new competent writers.

No reason to *buy* DDR2

[TwoTailedFox]

DDR2 looks to me like RDRAM. High-Latency, High-Frequency, eclipsed in performance at the same frequency when compared to vanilla DDR.

Keep your eyes peeled for DDR3 next year, though.

A little scary

[IPFreely]

The first thing I thought of after reading that is "Is this safe?"

With HT and Memory, the CPU is only connected directly to things on the main board. Adding PCIe to the mix means the CPU is now directly connected to external devices (PCIe slots). Having spiked a MB myself through careless handling of addin cards, I imagine it would be very easy to scrap your CPU by doing the same thing. And it wouldn't even have to be the users fault. A badly designed card could do it. If they put in some sort of surge protection on the MB around the CPU, would that just increase the latency they are trying so hard to cut?

As mentuioned elsewhere in this forum, it may be more about cost than performance, in which case adding surge protection may be acceptable for performance but would also add its own price.

Plus one for performance, minus one for safety.

Dance Dance Revolution? (offtopic)

[b0rk b0rk b0rk]

Was anyone else initially thinking "DDR? Wha? Dancing chips, now?"

Phew

[Trogre]

For a second I'd read that as "AMD to Adopt DRM2 Next Year".

Here I was preparing this elaborate rant on how DRM (and this new DRM2) is taking away our freedoms and how I'll never buy AMD again.

Oh well, go AMD!

musical sokects

[Truekaiser]

amd is going to kill them selfs if they continue to play musical sokect's.

Re:musical sokects

[springbox]

amd is going to kill them selfs if they continue to play musical sokect's.

Just like Intel did, right?

Re:musical sokects

[Anonymous Coward]

You're gonna kill yourself if you keep going with that vodka.

Re:musical sokects

[Truekaiser]

There is a Differnce
intel only has two differnt sockets out at the moment for their desktop market.
socket 478 and socket T(775)
socket 478 has already been officaly discontinued by intel though two major motherboard manufactures still produce motherboards with it(foxconn and ecs)for the moment but they are going to drop those lines very soon.
on the other hand amd with the introduction of this new socket will have a total of 4(754, 939, 940, and M2)differnt sockets out still being made with their respective cpu's.
Confuseing your customer is the best way to drive them to their compitetion.

ps: i did not mention socketA(462) because amd stoped making the cpus in june and by the end of this month most motherboard manufactures will have stoped making socket A motherboards. the ones you can get now are the channel buffer stock they make to counter shortfalls in production.

Re:musical sokects

[Dr_b_]

What about the Xeon socket 604 and the Itanium socket mPGA700? I'll tell ya whats confusing about using the same socket, is buying a mobo, and thinking a prescott cpu will work in it. The pin counts are the same, like with 478, but it won't work because of the voltages. Diff sockets at least are good in that regard, you can't plug the wrong or incompatible cpu into it. Isn't it fairly easy enough then to buy a mobo and shop for a cpu with the correct pins, instead of trying to figure out if its got the right microcode or voltage?

Re:musical sokects

[Truekaiser]

i said desktop sockets.
as for voltages, it is only confuseing when they don't document or openly print the differnces.

Not a bad idea actually

[goldcd]

Every time I want an upgrade I seem to end up shelling out for a new motherboard anyway - bundle them together and I get to save a few pence.

Oh, so now we're getting into an oldest...

[dpilot]

So now it's become an oldest/slowest computer contest? I see your 1GHz and I lower you to 380MHz.

My main machine at home that my wife and I use is a K6-3 running at 380 (4x95) MHz, and our server is a 300MHz Celeron-A. The kids share an Athlon XP 2600. We recently retired our Cyrix P150+ server.

Re:Oh, so now we're getting into an oldest...

[CastrTroy]

I'm running a PII 266 at home. It does kind of chug along, and Linux takes forever to boot. Oh and compiling is a nightmare. But for most everyday tasks it performs rather well. I've been meaning to upgrade for about a year, but it isn't a major priority. I am waiting until the one I want to buy drops below a certain price. And that one will probably last me another 8 years.

Re:Oh, so now we're getting into an oldest...

[TobyWong]

I have a rusty abacus at home that I use for my banking. My wife wanted one too after she saw me flying through my interest calculations but I told her "we don't waste money on newfangled technology" and bought her a pile of shiny beads to do her counting with instead.

Re:Oh, so now we're getting into an oldest...

[Articuno]

I have a Pentium 166! And a 286 too!! So, I win ? (err.. maybe "lose" is a better word in this case...) I, for one, welcome our old computers overlords!

Re:Oh, so now we're getting into an oldest...

[Anonymous Coward]

I still have an IBM XT. 5 1/4 floppy, no hard disk, green monochrome display. Those were the good old days.

Re:Oh, so now we're getting into an oldest...

[carl0ski]

have you tried the muliplier of 6 on that k6-3?


some k6-2+ and k6-3

let you set the muliplier to 2x and

it assumes it is 6x

very handy i once had a k6-2 at 6x83 on a p1 board maxed out :)

it retired 2 years ago

Re:Oh, so now we're getting into an oldest...

[dpilot]

No, but I have run it with 4.5 x 95. But this machine tends to have thermal problems, so I run it conservatively. When it only (?) had 64MB, I ran the bus at 100MHz, but it would never run more than 1 stick reliably at that frequency, so I backed off to 95MHz and have been mostly happy. I still need to run the AC in order to "emerge -uD world" or build a kernel, in the Summer.

Re:Oh, so now we're getting into an oldest...

[dpilot]

But are you actively using it?

Re:Oh, so now we're getting into an oldest...

[stanmann]

The XT came with a Hard drive.
[/pedant]

Re:Oh, so now we're getting into an oldest...

[Gr8Apes]

I have a P133, a P200 MMX, and a PPro 180. I keep the P133 around for old games, and a recipe program my wife still uses occasionally. (I really should get that info migrated).

As for long term "modern" computers, I bought a P4 2.4 GHz a few years ago, recently upgraded the RAM to 2GB, and it still handles email just fine, thank you. :) The RAM upgrade was for picture editing and scanning, I actually could not scan a picture because I didn't have enough RAM @ 1GB!!!

I'll probably yank the MB/CPU combo and drop in a new AMD 64 X2 in about another 6-12 months, when they drop to ~$200 for the pair. (My RAM is compatible with those MBs. Although I'd really like a dual Opteron setup, I just really can't justify it for reading email and a few other little tasks I do at home).

Re:Oh, so now we're getting into an oldest...

[Sinus0idal]

lol no defo not a slowest computer contest or I would have entered something better than a 1GHz thunderbird. I mean my server is a 90mhz Digital DECpc and it does all I need it for (mind you wouldn't want to see it take a slashdotting).

Was just pointing out that a 2.5 year cycle is unnecessarily short.

Re:Oh, so now we're getting into an oldest...

[AvitarX]

The RAM upgrade was for picture editing and scanning, I actually could not scan a picture because I didn't have enough RAM @ 1GB!!!

what were you doing?

a full page at 1200 dpi is less then 400MBs.

Re:Oh, so now we're getting into an oldest...

[Retric]

2.5yr is based around best ROI. AKA Buying a 400$ pc every 2.5 years is better than buying an 800$ pc every 5 years.

Personally I tend to buy cheep stuff, over clock it and replace it when it breaks. The difference between a cheep CPU and the best CPU you can find now tends to be a lot less than the difference between the best CPU now and a cheep CPU you can find in 2.5 years. (This does seem to be slowing down some for RAM and CPU's but it still works for graphics cards.) It even works for those who find a 500$ system overkill now you can always eBay for a 2-year-old machine that costs 150$.

Re:Oh, so now we're getting into an oldest...

[JohnnyLocust]

Do you have the older 7 pin abacus, or the 13 pin floating point unit model?

Re:Oh, so now we're getting into an oldest...

[Anonymous Coward]

I wonder if he can do online banking on that abacus.

Re:Oh, so now we're getting into an oldest...

[Gr8Apes]

scanning @ 2400 dpi. The scanning program wasn't the only thing running, though I supposedly had 700MB free. I run with a very small virtual memory space though.

Re:Oh, so now we're getting into an oldest...

[AvitarX]

My question is more along the lines of where can you output that kind of quality, and what original had that kind of detail.

The only thing I can think of that would require (or even benifit) from 2400 dpi over 1200 dpi at that size is showing an extreme detail to analize how ink soaks into paper, but even then it seems excessive.

Re:Oh, so now we're getting into an oldest...

[Gr8Apes]

It was an experiment in scanning. Printing at that res would, as you correctly state, most likely be an experiment in ink soak-through. :)

Known troll: check the parent's history! [NT]

[Anonymous Coward]

Nice Teeth

Socket M2

[hattig]

Stupid article, Socket M2 is the dual-channel DDR consumer socket, it has 940 pins.

Socket 1207 is the next Opteron socket, with PCI Express and possibly 4 HT links on board to enable even greater scaling up to 16 and 32 processor systems.

Socket S1 is the new mobile socket, and it has around 640 pins, and supposedly supports dual channel DDR2, along with a single HT link I imagine.

Slashdot requires you to wait between each successful posting of a comment to allow everyone a fair chance at posting a comment.

It's been 8 minutes since you last successfully posted a comment

Re:Socket M2

[hattig]

I meant M2 is the dual-channel DDR2 consumer socket, bah.

S1 - One HT Link?
M2 - Two HT Links?

1207 might use FB-DIMMs for memory instead of DDR2.

huh?

[fbartho]

So... you ungeeked yourself by not understanding something was referring to ram, and you "regeeked" yourself by thinking of a gamecontrol pad/arcade game that is completely mainstream, and not at all restricted to geeks?

*blinks* ok. I guess that depends on what the meaning of "am a" is now, doesn't it?

what a troll

[Anonymous Coward]

It doesn't take paragraphs to dispel these myths.

* You're comparing the Newton to "today's standards" when it is no longer made, rarely used, and is 12 years old.

* Cooperative multitasking was was all the mac had at the time, so marketing had to come up with something. Of course it was crap.

* Apple didn't have a choice but to fast track OS X. See above problem. Who cares?

* G4 Cube was a canceled project and therefore an admitted failure.

* The Mini serves one purpose, and that is to allow people to try out the Mac computing experience for only $500. It is not a "serious" computer for anything.

* Linux is not UNIX either. Nobody seems to care.

* It's not Apple's fault that Motorola and IBM allowed PPC to fall by the wayside.

* You opinion is that professional is another term for "ugly." This isn't really a point. Windows and GNOME have the same clutter problem.

* Is it flip-flopping or staying as competitive as possible by choosing the best available technology? I see you're already biased.

You must not experience too many honest mac users that will tell you things like:
* Safari is a leaky-ass web browser
* Dashboard is just some fancyness to push a very advanced set of programmer APIs to the masses.
* The Opteron kicks the shit out of the G5
* Apple should have gone with the Opteron (but we'll consider forgiving them if they add the chips to the lineup when the supply issues are abated)
* The iPod Shuffle is for cheapskates who want iTunes and the Music Store and it needs a display in the next revision. Nobody really cares about FM, otherwise they wouldn't tote their own music around.

Sorry you've had the zealots lie to you. Most of us believe that Apple offers the best tradeoffs, despite any flaws, real or "perceived." You should try hanging out with less asinine Apple users.

Date - 2010 AD

[dbucowboy]

"AMD is releasing the new 250 Ghz, 128 bit processor today and will require a 1,207,201 pin socket."

Pin count and price

[marat]

It's stunning how far technology has advanced - back in i4004 times pin counts were severely limited by price issues: pinning out one line cost about $1. Because of this bus width of the first microprocessor was just 4 bits. But now we build processors with more than thousand pins and a retail price of just... let me check... or, wait!!!

Re:Pin count and price

[chawly]

And I, for one, welcome our pin counting overlords. Is a 1207 pin socket going to hurt my sex life ? Somehow I don't think so. Might hurt my pocket-book though - if I bought one. It might hurt my ego even more - if I bought one and obtained a negative performance change. It would make me cry big salt tears if I bought one and somebody said something about division by four. Yep, how many pins on each side of this miracle ? How many sides ? Is 1207 a prime number, or are we being primed for the "world stupidy prize" ? Just a question .... sorry about it. Did I just invent the hyp.... othetica question ? Or is this only hype ? Seems so to me - if fact it seems worse than the World Wrestling Federation.

I have to agree with you on many points

[Anonymous Coward]

I've been a mac user since 1990, but I'm beginning to like the GNU/Linux/X/Gnome platform better. I'm really digging the Gnome 2.10 interface in Ubuntu 5.04. I changed the theme from Human to Glider though. The brown was really getting to me. I'm starting to like this a lot better than OS X 10.4.
I think my next computer is going to be an AMD 64 X2 once they add the hardware virtualization.

Socket 1213

[diesel_jackass]

They should add the extra 6 pins they'd need. (more == better, right?)

Then they could market it with:
This one goes to 10.
This one goes to 11.
BUT this one goes all the way to 1213!

Latency to clock ration

[charnov]

Other than the obvious reason (new socket, etc.) the biggest reason they waited is that DDR2's latencies were sky high which kills AMDs model. Also, because of the lower clockspeed, AMDs were not starved for memory bandwidth. With the dual cores and the new DDR2s greatly reduced latency it makes sense to start the switch.

Re:Is DDR2 worth waiting for? - No AGP

[flyingace]

Agree. Apparently Longhorn will use the video ram in both directions. That is Video Card -> CPU and CPU -> Video Ram. The former is pretty fast with AGP and not the later. If you want the later to work you are better off with PCI-E. Hence Longhorn will force the adoption of PCI-E.

You have to wonder...

[Belial6]

You have to wonder if adding the extra 130 pins (even if they were not connected to anything internal) would increase or decrease their sales.

My rule of thumb.

[dmaxwell]

I build myself a new machine every two or three years. When I'm ordering the parts, I get what was top of the line one year before that date. If you go newer than that then prices spike upward dramatically for very little increase in performance. I find that such a machine lasts three years if I'm keeping up with the latest and greatest in software. I consistently spend between $500 and $600 doing this. That isn't bad spread out over two or three years.

I also don't count on using much from my previous machine; usually just the hard drive and optical drives (which are used until failure or a format is forced on me). I have a floppy drive that goes clear back to my 486 days but everything else has been churned out. The floppy only gets used for firmware upgrades and will likely be chucked the next time I build.

AMD Tech Tour Q and A with engineer

[valhallaprime]

Well, I am freshly back from the AMD tech tour event in East Brunswick last night, and this specific question came during the Q and A with "TEH EXPERT5" - The question of DDR2 support.

The actual engineer on staff at the event answered it, and stated flat out that there was no performance gain until at least DDR2-667, and that alone "was only about 5% or so faster than DDR400 running in dual channel mode". He even went so far as to say that "DDR2-533, with it's increased latency over DDR400, has a negative impact of OVER 5%", and makes no sense to jump to. This was because of the efficiency already inherent in the HyperTransport bus, according to him.

He talked for about 5 minutes on the issue, and the gist of it was that until DDR2-667 specifically started to become more affordable, the incremental speed boost didn't make any sense for anyone, including and users and AMD Proc Support.

Incidentally, he also mentioned that DDR2 would (of course) require significant redesign in the built-in memory controller of the 939 chips, unless registered memory was used. This sorta implies in a friday morning-drove-all-night-from-NJ way that the current 939's would not support DDR2 if there were to be 939 mobo's with DDR2 support.

Mod Parent Down

[Anonymous Coward]

So I do not have to see this much whitespace wasted.

Headline: "AMD follows Leader Intel on Memory"....

[Glasswire]

The folks that normally crow about Intel being a technological follower of AMD should graciously acknowledge that Intel was ahead of AMD on DDR2...

DDR2

[Deliveranc3]

The worst thing about this is that AMD customers were supposed to be spared the bullshit that is DDR2...

DDR2 IS SLOWER and was designed to be cheaper but that never happened.

DDR is still in production, so why are they not simply waiting for DDR3?

This boggles the mind! Does anyone have ANYTHING favorable to say about DDR2?