AMD's Roadmap revealed

NoPants writes "It looks like the aces at Anandtech were able to get their hands on some of AMD's internal roadmaps. Anand has some interesting information including the new upcoming Socket 939 CPU standard as well as AMD's predicted release dates for Athlon 64 4000+ processors. Hopefully this will shed some light on what AMD is trying to do with all the different socket types..."

Comments

Maybe all of this...

[BJZQ8]

Maybe all of this preliminary information will help Intel markitecture their way to the Pentium 6!

Re:Maybe all of this...

[somethinghollow]

I think they wanted to call it Pentium !!!!!! after their last really successful processor, Pentium !!!.

well thats nice

[ZenBased]

but my amd 700mhz proc is still quite fast enough to give me debian, fluxbox, openoffice etc.. ah well there must be enough people out there who cant live without a fast proc

Re:well thats nice

[Zathrus]

Unless you are a gamer or looking to compute Pi to the last digit

Or are wanting to do things like rip MP3s (trivial) or burn DVDs (non-trivial; technically it's the MPEG2 mastering, not the burning, that takes the CPU time). Developers, graphics artists, and most engineering can also use as much CPU as is available. For just a plain old file server you do very well by using the cheapest (in terms of purchase and run cost) that you can get. A webserver probably needs more juice. A database server definitely does. Trivial home use excluded of course. I'm not talking about trivial usages -- they can always be solved easily.

Wake me when a cheap "build your own system" RISC alternative hits the market.

When you realize that the core ISA of all x86 chips is RISC let me know. Not to mention that most of the classic "RISC" designs have deviated far from the "reduced" portion of that moniker. Looked at the Power or PowerPC ISA recently? RISC was created not because a reduced instruction set is inherently better, but because it allowed for a number of technologies such as pipelining, branch prediction, caching, and so forth to be implemented. Every single one of those is in x86 architecture now. Sure, the ISA is still a mess, but it's a better price/performance than anything else out there. All the naysayers have been disproven, time and time again. And yes, when I was a little college student I was horrified at the design of x86. Then I grew up.

Re:well thats nice

[dasmegabyte]

Except that programmers like myself write for the slowest currently marketted PC. We take advantage of the excess speed by increasing search capabilities, performing more intricate analysis, using higher quality fonts, sounds, graphics, etc.

Yes, MOST of what people really need to do can be done on a 500 MHz machine. Shit, most of what people do -- search for information, write email, word process -- can be done on a goddamn Commodore.

It is a fact of life that computers are going to get slowly faster, and people are going to expect these faster computers to have better software. Even if it's mostly superficial, we try to deliver that. Most of the time, though, a faster processor is a boon even to Joe Q. Homeuser. Consider a 3 megapixel camera, delivering photographs in excess of 1.5 megabytes. Time was we'd never THINK of doing graphical operations on that much information. Nowadays, it's so trivial that many photoalbums are processing 10 or more such pictures per second!

Anyway, for easy operations like file serving, running a firewall, serving 100,000 or fewer web pages per day, etc...your best bet is a processor with a fast bus and a slow clockspeed. It'll be cooler and more reliable than some 64 bit god (honestly, who needs 64 bits to send packets?)

Re:offtopic..

[Trigun]

The Slashdot staff created this story with a timestamp from a Mars watch. It's thirty nine minutes off.

Still not convinced

[LookSharp]

There's no question Athlon 64 and AthlonFX are great products. That being said:

*Do they really need to be different products? Opteron is your product for server/high-end workstations, Duron (and now Athlon XP) is low-end... you want Athlon64 to be mainstream, right?

*Is it really a good idea to have the memory controller on the CPU? OK, I buy that it increases performance, but it hasn't lowered mainboard costs and all I've seen it doing is causing a rift between the A64 and AFX product lines, since Athlon64 doesn't have a dual-channel memory controller.

*Why in the world introduce an AthlonFX based on Socket 940, especially at the outrageous price, when you're moving to socket 939 imminently?

I think it would have been more of a slam-dunk as a platform and a "brand" to release Athlon64 as all dual-channel, all Socket 939 (or some standard), and left Opteron as the high-end platform. Any other takers?

Re:Still not convinced

[Hoser McMoose]

The marketing of the Athlon64 FX has become a bit confusing. It kind of made sense for the initial launch to combat a precieved weakness of the design compared to Intel's P4 though. With the P4 you get up to 6.4GB/s of memory bandwidth, while the first Athlon64 would only have 3.2GB/s of memory bandwidth. Now, it turns out that the extra bandwidth doesn't actually buy you much on most applications, but this was seen as a weakness, hence the Athlon64 FX. With Intel bringing out the P4EE to compete with the FX, now AMD might need to keep the chip, even if it isn't a worthwhile product (The P4EE isn't a worthwhile product either).

*Is it really a good idea to have the memory controller on the CPU?

Yes, yes it is a good idea. A VERY good idea in fact. Memory latency has only improved by about one order of magnitude in the past 15 years. Meanwhile everything else in the system has gone up by at least two orders of magnitude. Virtually everything that is being done in CPU design these days is to hide memory latency (larger caches, out-of-order executation, branch prediction, even SMT).

Integrating the memory controller reduces latency by 20-30%. At 2.0GHz this makes a BIG difference (this is the main reason why a 2.0GHz Athlon64 is faster than a 2.2GHz AthlonXP), at 4 or 5GHz the difference will be huge.

but it hasn't lowered mainboard costs

You can buy new Athlon64 motherboards for only $100, only 3 months after the chips release. It took ages for Athlon or P4 motherboards to reach that price point. What's perhaps even more impressive is the dual-processor boards that are only $200. In short, it HAS reduced motherboard costs, whether you've noticed or not. It also means that ALL Athlon64's support ECC, chipkill and a few other nifty reliability features, regardless of how badly VIA screws up their chipset design.

*Why in the world introduce an AthlonFX based on Socket 940, especially at the outrageous price, when you're moving to socket 939 imminently?

The Athlon64 FX was a bit of a last minute decision I believe. They found a marketing weakness and wanted the quickest and easiest solution they could find. The answer? Sell your server chip as an "enthusiast" chip. Intel did exactly the same thing for the same reason with the P4EE.

Also, it's actually VERY normal to switch sockets soon after releasing a new processor. Intel's upcoming Prescott will use Socket 478 for only about 6 months before switching to socket 775. The original P4 used Socket 423 for a very short time before switching to socket 478. The original Athlon used Slot A for a year or so before switching to Socket A. The PIII came out in Slot 1 form but then switched to Socket 370 about a year later. The Celeron followed the same path a couple years before.

I think it would have been more of a slam-dunk as a platform and a "brand" to release Athlon64 as all dual-channel, all Socket 939 (or some standard), and left Opteron as the high-end platform. Any other takers?

In retrospec that might seem like a good idea, hindsight is 20-20 after all. However the original split of ALL Athlon64 chips being socket 754 and ALL Opteron's being socket 940 seemed to make the most sense when AMD was desigining them. It wasn't until market conditions changed and a new perceived weakness was discovered that AMD felt they need a consumer chip with a 128-bit wide memory bus. By that time the chip was already late to market and designing a new socket would have added more delay to the equation.

There's also the question of budget chips. AMD hopes to move their entire product line to the Athlon64/Opteron platform by the end of 2004. That means they need a budget chip, and socket 939 with it's 128-bit wide memory bus is problematic for that. Hence the continued existance of Socket 754 and the AthlonXP for that platform.

Re:Grhh...

[binaryDigit]

AMD's new stuff has been pretty impressive, but it really bothers me when they pull this type of stuff: AMD Athlon 64 3700+ 2.4GHz 1MB Q2 '04 AMD Athlon 64 3400+ 2.4GHz 512KB Q2 '04

What's the problem. They're saying that having the smaller cache gives you less performance. Are you upset that they happen to have the same clock speed? I assume you'd prefer nomenclature more on the order of "AMD Athlon 64 2.4/512 and 2.4/1024"? In many ways they way they are currently doing it is more descriptive to the average buyer. No guessing as to how much performance you're giving up by going with the smaller cache ,or how much you're gaining by going with the bigger (performance benchmark inflation not withstanding).

Global Warming

[Anonymous Coward]

Wow if these new chips radiate at a proportionate level to AMDs current offering we could all be wearing shorts in Antartica by new year 05.

Seriously tho. I think AMD ought to work some better thermal performance into its cpu range. A low cost, low temperature, high performance CPU is what is required in the market.

Re:Global Warming

[hattig]

This whole AMD is hotter than Intel thing was true when Intel's flagship processor was the cool (relatively) PIII.

The P4 generates more heat than the Athlon (any variant) for the same performance.

It is such an old, and incorrect joke it isn't even funny anymore.

Re:Grhh...

[TrueBuckeye]

Not to mention the fact that it lowers costs to them, and thus to the consumer, to do this. From what I've heard, they are able to take cpu's with some bad cache, which isn't uncommon, disable that non-functioning section, and then sell the cpu as a 512k cache cpu rather than wasting the entire chip. Lower performance, lower cost, but less waste. This is a far cry from the world of the Intel 486 sx vs dx with the math co-processor fiasco.

Socket, shmocket ... I want RAM!

[ackthpt]

The only reason I'm even considering one of these gems is so I can cram more memory into a system for video work. All the boards I've seen for Athlon 64 max at 3Gb. The SK8* boards for the Athlon FX will take, IIRC 8Gb. Where's the boards I can cram 32 or more into?

Re:Socket, shmocket ... I want RAM!

[BigDumbAnimal]

Call Sun.

socket types

[ynohoo]

Hopefully this will shed some light on what AMD is trying to do with all the different socket types..

Making us buy more motherboards, of course!

Dual Processors?

[Krieger]

What the article does not cover, is when we will be able to purchase non-Opteron Dual processors. Since they are inherently capable, it would be nice to know when we'll be able to build a performance (non-ECC) dual desktop.

G5 looks like ramping up faster

[beelsebob]

By the look of the figures the Athlon 64 is margionally faster than the G5 clock for clock... (the 2.2 Ghz beating the 2 Ghz G5 convinvingly and the 2Ghz ones locked in a tight battle). It looks a lot like AMD are gonna have to ramp up faster though, because IBM are gonna have 3Ghz G5s by Q3 this year, and AMD are only saying 2.6Ghz by Q4. Bob

Re:G5 looks like ramping up faster

[Junks Jerzey]

IBM hasn't hit the power/heat problems that Intel has, primarily because the PPC is a more streamlined processor. It's still very complex, mind you, but the x86 line is so complex, with all of the legacy support and CISC to RISC conversion and wacky nonsense like MMX *and* SSE *and* SSE2 all at the same time. Intel is already talking 150 watts for processors to be released this year. It is quite likely that the PowerPC line is going to pass Intel in the next 8 to 12 months.

Top 5 things you will do with your Athlon64...

[Realistic_Dragon]

Run Doom 3.
Compute Pi to 5.497558e+11 bit precision during your lunchbreak.
Store this value of Pi in RAM.
Install Kazaa and not notice the spyware slowdown.
Use the faster page loading times to get FP more often.

Socket hell

[freidog]

So 940 gets moved to opteron only
939 encompasses both Athlon64/FX chips, starting in Q2.
754 is relegated to the next gen AthlonXPs (with the on die memory controller, but only 32 bit)
462 dies a slow death.

Why can't every CPU made just fit on Socket 7... :)

Re:Grhh...

[glsunder]

The ratings have been mostly accurate when comparing AMD cpus, and they're useful when talking to non-computer people rather than explaining things like cache size, locality, etc. With the Athlon64, it should be even more accurate since they now control the memory controller. If the performance rating can make it easier for the average person to compare systems accurately and pick the better value for them, then I'd say it's a good thing.

Also, normal users don't care about 10% performance differences. I've found a good way to relate to this is to think about old computers. Do you really think of the difference between a P133 and a P166? No, they're both from the p5 generation, with a L2 caching up to 64MB or 128MB depending on chipset and the cpu performance is basically equal (slow). If a p166 will do the job you need nowdays, a p133 probably will too, and it's doubtful that you'll notice much difference.

Re:Real Mhz on the 4000 chip?

[DanglingPointer2004]

False advertising? It's called marketing. They never said "Our 3200+ is equivalent to Intel's P4 3.2C". Find that on the AMD site and you are making sense. Also, compare a duron chip to a celeron chip that are "rated the same". There is no comparison, but then, neither company specifically said "these two chips are the same speed", so you can't really complain. You can't blindly trust a number to tell you how well a processor will perform, there's a lot more to it than MHz and GHz. As for Tom's Hardware, I would look at who pays their bills before counting on their "benchmarks" too closely.

Intel is the Global Warming threat

[Brian Stretch]

Running at full CPU load, an Athlon 64 3200+ uses less power than an Intel P4 3.2GHz. Furthermore, with AMD's Cool and Quiet power management enabled the Athlon 64 CPU slows down to 800MHz and drops to 1.275v when you don't need much CPU performence, ie, while I'm typing this message. ASUS has a nifty little program that displays the current CPU speed and core voltage on my desktop.

AMD CPU power requirements are expected to drop substantially when they switch to 90nm in the second half of this year. OTOH, Intel's prototype 90nm Tejas CPU burns up 150 watts .

AMD chips haven't used more electricity than Intel chips for years. Pay attention.

BTW, Athlon 64 notebooks are out. $1,550 for a widescreen 64-bit notebook! I'm going to stick with my Athlon 64 desktop, at least until I come up with an excuse to buy a portable. Really, I am...

Athlon Thunderbird

[Chicks_Hate_Me]

Correct me if I'm wrong, but I was told the PR xxxx+ was in comparison of what a Athlon Thunderbird clock speed would be.

Re:confuse much?

[Loki_1929]

"There several instances of the same + rating with different specs in the same line. "I got a 2600+" "Which one?""

The idea is that you don't have to know what the exact specs are. What AMD is trying to do is show that CPU performance is relative. Within a few percentage points, every 2600+ CPU will perform equally, regardless of its core, amount of cache, socket, memory lines, clock frequency, etc. That's the whole point - you don't have to know. If you're "in the know", you can look at individual 2600+ CPUs to see which one has more of what will help you in the specific applications you use. If you're a general user, then 2600+ is the only thing you need to know. How many people buy a 3.06GHz P4 instead of a 3.00GHz P4 because they think the former is faster? The clock frequency alone belies the fact that the higher FSB on the latter CPU will actually make it perform far better on nearly every application. AMD is trying to hand you the whole package in a single number to simply the buying process for everyday people.

The fact is, neither AMD nor Intel are telling the whole performance truth, nor could they do so. The only way to do that is to educate consumers about CPU mechanics, latency, IPC, L1/L2 cache, cache hits and misses, branch prediction, pipeline stages, and so on. The average consumer (hell, the average geek) can't understand half of these things. Thus, Intel has chosen to show the clock frequency of its CPUs, and AMD has chosen to use performance ratings that give consumers a performance index relative to the Athlon's Thunderbird core. Neither system is perfect, but neither system is more imperfect than the other, in my opinion.

"Is AMD hoping nobody will know what they're buying?"

AMD is hoping that those who need to know, will know, and that the rest who buy the "bigger number" will at least have an idea of what to get.