Slashdot Mirror
AMD Athlon 64 FX-57 Review
Duane writes "GDHardware.com has the first review of AMD's upcoming Athlon 64 FX-57 CPU clocked at 2.8GHz. They benchmark it against Intel's current fastest 3.8GHz P4 and the Athlon 64 X2." From the article: "Clocked at 2.8GHz, the FX-57 continues the 'San Diego' core AMD released with the FX-55, but is stepped up a paltry 200MHz faster. What's interesting is that while 200MHz on the Intel side of things doesn't always mean that great of a performance gain, not so with AMD."
no u didnt
Now my compy's not fast enough anymore. And I'm on a budget too. Damnit!
[page 1]
AMD continues to raise the bar in performance - both in dual core with its recent X2 chip and now once again in the single core design with its pending FX-57 launch due on June 27th.
[page 2]
The FX-57 is armed with a total of 1152KB of cache (128KB L1 and 1024KB L2) which greatly speeds up commonly called data cues and is a great sized buffer between the CPU and system RAM.
[conclusion]
However, at this point in the game we'd have a hard time giving a full recommendation to anyone to spend close to or over $1000 on a chip that isn't dual core
I realize the price will go down over time, but seriously, who is going to buy this chip? Ok, I know some gamers with too much money on their hands will buy it, but it's still going to be surpassed when the dual cores start gaining ground, especially in gaming (think Christmas '05). Until I saw the pricetag I thought this might be an option for my next build, but not anymore. There are other options, at much lower prices.
-William Brendel
the FX-53 and FX-55 will still make great gaming CPUs from now until the next year or so
Why the hell does it have to be this way?
Umm... Actually, thanks to speeds of other devices (RAM, etc.) not keeping up with CPUs, it doesn't scale linearly. If the FSB goes up, then it does, but multiplier increases don't do as much - only the CPU is faster, not the rest of the system.
No, and why would it? The original pentium core only went up to around 233Mhz. The other Pentiums that hit the Ghz range are the P3 and PM.
They both have a higher IPC than the P4, so no a 1Ghz P4 is not the same as a 1Ghz P3... [it's slower].
In the AMD world they're not always the same either. A 1Ghz AMD64 would be faster in most cases than a 1Ghz AMD32 [e.g. Barton] because of the extra registers and more decode/execute resources [e.g. larger instruction scheduler, more DirectPath opcodes, etc].
Tom
Someday, I'll have a real sig.
Yes. The chip itself will operate exactly 3.8 times faster than a 1GHz chip of the same architecture. Note that there was no 1GHz P4 chip nor was there a 1GHz Athlon 64.
What you don't understand is that different architectures lead to different performance characteristics. This results in a similarly clocked AMD chip outperforming it's Intel rival.
Also, many other systems affect how fast your program runs -- it's not just processor speed.
Please mail to me at matt@purpletentacle.co.uk
If you compare the same chip running at 3.8 GHz it should be ~3.8 times faster than when running at 1 Ghz. However, if you are comparing a pentium IV running at 3.8 Ghz to a pentium III runing at 1 Ghz that probably wouldn't hold true.
I think the pentium III was more efficient per clock cycle than the current pentium IVs.
safe to say I won't be visiting GDHardware for reviews again, not that I'd heard of them before.
"The real target audience for the FX-57 is going to be the ultra-gamer who insists on nothing but the absolute fastest gaming CPU money can buy. It simply crushes everything in its path in game performance and handles most of today's common applications with power to spare."
Please show me an ultra gamer that plays on cutting edge hardware at only 640x480. I guess it was the only test they could find where the the FX beat the X2
Worst Review Ever!
No way he got a 20% increase in performace with only a 7% boost in CPU speed when overclocking the chip to 3ghz.
If he did, why didnt he include the overclock in the benchmarks?
I bet the whole article is fabricated. Wait for Anandtech for a real review.
George Bush + Linux = "I will not let information get in the way of the fight against Windows"
A lot more - in some cases, 2x as efficient. The Pentium M at 2GHz is as fast as the 3.x P4s, if not faster - it's based on the P3 architecture.
... for that price.
From what I've read, while Intel can keep cranking up the core speed of their chips, all those clock cycles are wasted if it spends most of its time waiting around for memory. The northbridge on Intel motherboards is now their biggest bottleneck. So at least part of the reason AMD can get better throughput at a lower clockrate is that it eliminates the northbridge altogether, puts the memory controller on the CPU, and ties everything else together using their insanely fast "HyperTransport" system bus. Any engineers who know more about it care to comment?
if you are comparing a pentium IV running at 3.8 Ghz to a pentium III runing at 1 Ghz that probably wouldn't hold true.
Correct. Everyone knows the pentium III would be faster...
Seven puppies were harmed during the making of this post.
Damn you Intel and AMD, always teasing me with the absolute most bleeding edge hardware that I CAN'T AFFORD. Come on, let's work on bringing down prices as well as bringing up performance.
Some of them say "Lower is better" some of them say "FPS" some of them just don't say anything. It makes it hard to gauge if higher is better or lower is better. I mean, some things are obvious like 3dmark 2005 results, but then it says "4D rendering" what the heck is that? Is it measuring FPS?
Agh, eeh gads!
Who the heck modded this informative? This is a completely juvenille misinterpetation of how CPUs work.
IIRC the Athlon 64 can "only" dispatch 3 'muops' per cycle to its execution units, which themselves are broken-down x86 instructions. The P4 is similar.
Secondly, the mu-ops must be in a certain sequence if you're ever going to dispatch more than 1 per cycle.
Thirdly, in order to keep the dispatcher dispatching, you must keep it busy with new operations and data to execute. Which means you need a good memory architecture and cache.
And: "The size of the pipes is important (you may be doing more per cycle, but the cycles take longer)." I can't fully understand what this sentence is trying to say, but it's probably wrong.
Put it all together, and IIRC the Athlon 64 can barely beat 1 x86 instruction per cycle under optimal conditions. It won't get close to 9.
How the fuck did that get moderated up. It makes no fucking sense and is completely inaccurate (and yes, IAACompEng).
Athlons have higher IPC (instructions per clock) than a P4. Why? The length of the pipeline. Athlon 64s have a SINGLE pipeline, with a length of about 15 (aka "a 15 stage pipeline"). A P4-prescott (90nm version) has a 31 stage pipeline. The P4 northwood had a 20 stage pipeline (note that those are for integer instructions, floating point operations have more stages through the FPU). A64s do not have 9 pipelines, nigh the P4 have 6. And neither get anywhere near the ops/clock you claim. They do have parallel execution units however, and maybe thats where you get your numbers from, but even then they're still not right.
So it takes an integer operation 15 or so cycles to be complete in an Athlon, and 30 cycles in a P4. Thus the higher IPC. Other things also influence performance are cache hit ratio, branch prediction. And thats the reason why the prescott didnt fall on its face-more cache as well as better Branch Prediction Unit (BPU). A lot of improvements went into the 90nm prescott to keep IPC close to what the P4-northwood had. There were some articles at Anandtech when it first came out, comparing it to the northwood.
To parent: Go read some Ars Technica articles about how CPUs are organized before you talk out of your ass about stuff you dont know.
The Doormat
If you're not outraged, then you're not paying attention.
This seems to imply Athlon scales better than linearly (?!) How does that work?
The toad can't burp - and for some reason can't fart either, so it swells up and eventually explodes. --Anonymous Coward
WRONG! Thank goodness I had me some mod points today. This guy has *NO* idea what he's talking about. The P4 has a longer pipeline than the Athlon (and Athlon64) resulting in a higher performance penalty on missed branch predictions and cache misses. This is what causes the performance hit (well, technically there are more reasons than the longer pipeline, but that's a topic for another thread).
I have to question if you understand the word pipeline. A decoder and its associated execution unit are the same pipeline, not two separate ones. Athlons have three pipes.
What that means, is that an Athlon performs ~9 operations per cycle, or 9 * 2.8 Ghz = ~25.2 billion instructions per second, and the intel would do 6 * 3.8Ghz = ~22.8. Those are very, very rough estimates.
Again, completely wrong. AMD's architecture can at best issue 3 instructions per cycle (and this depends on there being 3 instructions present in a 16 byte fetch window, plus various other factors such as branching into the middle of the window and predecode bits being correct), and they can only retire 3 instructions per cycle. As a result the best possible IPC is 3. In reality, a good application will exceed an IPC of 1.
So basically, the Ghz of the processor doesn't mean a whole lot. There's a lot more to take into consideration.
This at least is correct. You might want to read this article for a start, although it really only covers the processor's front end:
Arstechnica processor overview
Actually it can be argued that in "optimal conditions" the ipc should be around 3. While a typical x86 operation (say, add reg,mem) isn't a single micro-op on P4, on Athlon it is.
So it takes an integer operation 15 or so cycles to be complete in an Athlon, and 30 cycles in a P4.
I want to add that a long pipeline isn't as bad as you make it seem. Assuming the branch prediction and cache are working effectively (and there aren't too many data hazards, etc), there will be several instructions in the same pipeline at the same time in different stages.
One integer operation may take 30 cycles on a P4 and 15 on an Athlon. But one million integer operations might approach 1 integer operation per cycle on both processors. This is under very ideal circumstances, and realistically there will always be fewer instructions in the pipeline than there are stages.
Social scientists are inspired by theories; scientists are humbled by facts.
So it takes an integer operation 15 or so cycles to be complete in an Athlon, and 30 cycles in a P4.
You'd better go read again Ars Technica, and understand it this time. Or better, buy and read a book on computer architecture.
Imagine an assembly line that's making cars. You don't wait until the first car gets off the end of the line before you start making the next one!
Correct, as far as it goes.
/bubble penalties.
However: it's not the pipeline length causing "15 cycles versus 30 cycles" that will actually harm performance. It's pipeline STALLS what kill performance--in a perfect world, for example, a hypothetical 10,000-stage single-pipeline processor running at 1 GHz would retire 1 BILLION instructions per second, albeit with a 10,000 clock initial pipeline fill upon powerup.
Do something that causes the pipeline to need to be flushed and refilled, however, and you just lost 10,0000 clocks.
This is where the P4 has problems relative to the Athlon: keeping it's pipeline filled, and the subsequent pipeline flush
Note that there's lots more to this discussion than I wrote here (can you say branch predictors, trace caches, lookaside buffers, etc.), but ultimately all that stuff has to do with KEEPING THE PIPELINE FILLED, and what happens when you don't.
- The race is not [always] to the swift, nor the battle to the strong. -
you can also have that many instructions in the pipeline at once.
No shit sherlock, thats the definition of a pipeline.
The Doormat
If you're not outraged, then you're not paying attention.
If you know what it is, then why did you misuse it?
How the fuck did that get moderated up. It makes no fucking sense and is completely inaccurate (and yes, IAA REAL CompEng) see the pipeline as a queue: if the queue is long, the first person will take more time to go through it, but the second is just behind the first and it doesn't matter the length of the queue pipelines are GOOD it's pipeline STALLS (or BUBBLES) that are bad - when you need to flush the pipeline and reload it with instructions the first instructions takes 30 cycles, but the others take 1 cycle if everything goes right because there already in the pipeline To parent: Go read some real books about how CPUs are organized before you talk out of your ass about stuff you dont know.
"Note that there was no 1GHz P4 chip nor was there a 1GHz Athlon 64."
:)
With Cool'n'Quiet newer Athlons can underclock themselves to 1 ghz. Mine does this.
I rarely criticize things I don't care about.
The Z80 has decodes and executes in single stage and instructions take slightly higher 2 cycles to complete on average. By your logic Z80 must have 30 times the IPC of P4. Therefore P4 must have about 0.01 IPC. Illuminating although completly untrue.
Gentlemen, you can't fight in here, this is the War Room!
Take it easy man... not everybody still lives in their parents house, and not everybody has all the time in the world to eat doritos and learn about CPU's.
The only way to tell the difference between a hamster and a gerbil is that the hamster has more white meat.
Pentium 3 chips are quite a lot faster as pentium 4 chips at the same clockrate. An 1Ghz p3 was about as fast as an 1.4Ghz p4, iirc. Though that was the first model p4, they have optimized it a bit so it will be less in favor of the p3.
Why is it that the dodgiest news/review sites so often get written up here?
Me (Blog)
I actually thought this was funny, but we all know that the moderators are out of touch with reality.
You're asking the wrong question. Even if no one buys this chip, the chip is still worthwhile to have on the market.
A few years ago Wendy's found that almost no one was buying their triple cheeseburgers, so they took triples off the menu. When they did this, they found that sales of their double cheeseburgers dropped to almost nothing. The problem, as they discovered later, was that the presence of triple cheeseburgers on the menu helped to legitimize the double cheeseburgers as mainstream items. Without triple cheeseburgers, the double cheeseburgers became the high end item and mainstream buyers went for the singles instead.
Since profit margins on double cheeseburgers are higher, the chain was forced to bring back triple cheeseburgers, even though triples weren't selling at all, because the sales of their double cheeseburgers depended on having triples on the menu.
Point is, although this is a fast food example, the same thing applies to the computer industry. You HAVE to have a high end item available if you are to have any hope of positioning the more profitable midrange items as mainstream.
As one can clearly asses from our benchmarks...
And that is not correct either. The chip doesn't scale that way with clock speed.
;)
A 3800+ AMD chip will perform, roughly, 3.8 times as well as a 1 GHz Athlon. This is not the true in all cases - it will perform better in some situations, worse in others.
Optimized architecture also means, that the 800 MHz Athlon 64 FX (underclocked by Cool'n'Quiet) could still outperform a 1 GHz Athlon, hence giving it a performance rating of more than 1000+.
While you've been moderated quite "informative", your comment isn't really that - it's partly right, just like it's partly right that the earth is somewhat flat
We do not live in the 21st century. We live in the 20 second century.
You can find the new San Diego core FX benched at FX57 and FX59(3GHz) speeds here.
ignorance is bliss. googlefiberatx.com
Intel generally leads AMD in memory bandwidth, at least without any overclocking.
Intel was doing 6.4GB/s (dual channel PC3200 RAM) when AMD was at 2.7GB/sec. (single channel PC2700).
Also, note that memory accesses don't go over HyperTransport on an Athlon. The memory controller is built into the CPU. This is nice for latency, but bad because it means that Athlon users are stuck with whatever memory technology AMD has selected. At the moment, that means Athlon systems are stuck with DDR right now even as DDR2 prices fall below DDR prices.
Also, HyperTransport isn't all that insanely fast. Amongst other thinks, clocking cycles are part of the "GHz" rating on HT, and so the bandwidth is lower than it might seem.
http://lkml.org/lkml/2005/8/20/95
Isn't that always how it is? I mean, AMD and Intel mostly succeed because they can ship chips in volume and at low prices. Most actual innovation takes place at Sun, IBM, or other server-oriented companies if I'm not mistaken. The Alpha is just a particularly poignant example.
> This seems to imply Athlon scales better than linearly (?!) How
> does that work?
It shows how much the athlon is dependent on bus speeds. it wasn't only the 2.8GHz to 3.0GHz jump (seven percent) that helped it boost speed, it was the much higher percentage jump in bus speed that helped. The athlon 64 was capable of using all that extra bus speed - at 2.8GHz it was being RESTRAINED by the relatively slower bus.
Bullshit, Intels' Dothan (Pentium-M) on an Asus mobo will smoke an FX-57 at less that half the price. Dothans currently hold all the 3DMark records and SuperPi. Check out the scores for yourself.
I e-mailed the author several days ago that leaving out Dothan benches made his review and conclusions worthless. He hasn't e-mailed back. I can't say for sure, but this article sure seemed like AMD fanboyism.
Slashdot "libertarians": Small government for me, big government for those I disagree with. -1, I disagree with you
I don't think that nigh means what you think it means.
No gods, no demons, and no masters. Secular Humanism!
OK, so yet another socket 939 part. The artificial "let's separate our market segments" marketing department stunt that otherwise is the same as socket 940.
When will we see a socket 940 equivalent part? Opterons always seem to lag behind the FX. Seems like socket 940 owners are getting screwed by AMD. Heck, even FX 51 and 53 were socket 940 when they came out so why can't AMD support 940 as well?
In fact, even Opteron 1xx is moving to socket 939! Why? This is rediculous. (And don't give me the marketing spiel of "socket 939 can use cheaper unbuffered RAM" because that is a function of the CPU's memory controller, not the socket!)
It gets modded up because people with moderator points use thing as quickly as possible, with as little scrolling as possible.
Not just that. There are all sorts of bottlenecks for processor performance. The new FX-57 has an improved memory controller, that probably helps some. Also, the P4 is at this point limited by its gigantic pipeline, which is probably hurting it more than clock speed is helping it.
When Intel came out with the 800MHz FSB (6.4GB/s), they handily beat AMD on benchmarks. AMD didn't have the internal memory controller at the time. Plus they had a slow FSB. Plus they were in that awkward time before NVidia jumped on the bandwagon, and so the chipsets for AMD were terrible, most of them bad VIA performers.
AMD may have the upper hand in many benchmarks right now (I guess you don't look at video compression), but it hasn't been that way for long. AMD's most recent rise above Intel really started with Intel's 3.4GHz offerings. In the period before that (the early 875P chipset days, and the aforementioned 800FSB and 3.0GHz range), Intel was beating AMD handily, although not at nearly equivalent prices. This is especially true on memory-intensive benchmarks. It was quite a step forward for a company that at that time was just emerging from the dark days of RDRAM stupidity.
Also note to say that the hypertransport bus is why AMD's dual cores run faster than Intel's dual cores is pure speculation. Do you have any real reasons to put behind that or just assertions?
Even if Intel does use the standard FSB for inter-processor communcations, their FSB is currently faster than AMD's HT. The bandwidth of the HT on AMD chips is 8.0GB/sec. The bandwidth of the FSB on Intel's fastest processors is 10.7GB/sec.
I'm not dumping on AMD. I like AMD. But the amount of misinformed speculation and assertions as to why they are doing well is astounding.
Also, I find the "troll" moderation on my post above insulting. It's seems silly to me, but disregarding that, I find it ridiculous that people automatically moderate "troll" apparently just for speaking any nice things about Intel.
http://lkml.org/lkml/2005/8/20/95
But then again you use the longer latency (30 cycles for an instruction) of the P4 (Netburst) architecture to indicate why it has lower bandwidth (instructions per clock).
Latency and bandwidth are not the same thing, cannot be used interchangably and cannot be transformed one into the other with simple math.
Even with a 30 cycle latency the P4 could easily execute (complete) 1 instruction per clock. And actually, it probably executes (completed) more than 1 instruction per clock.
The reason Athlon has higher IPC is because it has IPC. I know that sounds stupid, but it is the nearest we can come to explaining it. It was designed to do more in every clock, so that it could be faster without going to a higher frequency (lower cycle time). P4 (Netburst) was designed to actually do less per clock so that they could reduce the cycle time and raise the frequency. That's the differences, you could nibble around the edges so more, but that's about as close as you can get without getting into the real specifics of the design.
You're a pretty crappy Computer Engineer from what I can tell.
http://lkml.org/lkml/2005/8/20/95
As if CPU benchmark tests weren't vague and subjective enough, we get a bunch of random graphs without LABELS. Jesus. I mean, you can figure out what most of them probably are but I'd rather just READ it. I mean I hate reading those things enough when they don't mention how they do means or what they normalize to, but just meaningless numbers? Yeah, no thanks boss.
That's somewhat odd, even for /.
Bullshit, Intels' Dothan (Pentium-M) on an Asus mobo will smoke an FX-57 at less that half the price. Dothans currently hold all the 3DMark records and SuperPi
? i=2382&p=7
The article said that the FX-57 crushes everything in its path in game performance. You bring up the Dothan. The Dothan will not "smoke" a FX-57. While it is a fast chip, it is not a desktop replacement chip and lacks the power that the high-end Athlons have in games.
http://www.anandtech.com/cpuchipsets/showdoc.aspx
Nigh - isn't that like when the sun goes down? You know, the opposite of da.
That was one of the better insights I've seen in my lifetime :P.
:).
What I like about these types of insights is that they are simple on the face of it because it is so common sense that you say 'of course' inretrospect, like most of the great ideas in the world
My point was that Intel has not been behind AMD in memory bandwidth in recent memory. At most times, including right now, they are ahead of AMD in memory bandwidth. The parent (now super parent) poster was wrong in saying AMD was ahead.
4 49.html.
Intel reached the memory bandwidth levels AMD is at right now almost 3 years ago with the 3.0GHz/800FSB Pentium 4.
Being stuck with DDR isn't a problem as far as performance. But right now memory (esp. Taiwanese) vendors are dropping their prices on DDR2 trying to accelerate the switch to DDR2 from DDR. This is presumably to get out from under the license fees they pay to Rambus for DDR. But regardless of the reasons, as DDR2 drops in price below DDR, many Athlon users are going to wish they could use DDR2.
As to your comments that memory manufacturers say DDR2 prices aren't going to drop, I could find nothing like that at all. Most news sources say DDR2 prices will drop below DDR prices in the 2nd half of the year. More specific news says things like I mentioned above. http://www.tomshardware.com/hardnews/20050609_065
As fast as having your own onboard memory controller is, it does stifle innovation as far as what memory can be used on motherboards completely. So you had better be SURE your bet is right. I'm not 100% sure AMD's is.
I dunno about Intel copying AMD's plans. I haven't heard anything of it. To do so requires adding at least 160 pins to the CPU package. And it means you can't do multi-chip multi-processing.
As to your comments that this means that I/O traffic doesn't tie up the FSB, you are incorrect. When you do I/O, the data doesn't go directly to the CPU (into the CPU registers), it goes into RAM. And AMD has the memory controller on the CPU, so that means that when you are moving data from the disk to the RAM, the data on an AMD has to go into the CPU on the HT bus, and out on the FSB (RAM) pins. So I/Os still tie up the FSB.
On an Intel, the data never even goes to the CPU, it comes in the south bridge (ATA, including SATA, most other stuff) or directly into the north bridge (GigE), and then goes out on the RAM pins (the magic of DMA). So there's no more or less competition for memory bandwidth in an Intel than on an AMD.
Essentially, AMD just moved the northbridge into the CPU. Why this decreases the memory latency, I'm not sure. I'm not saying it doesn't either, as the numbers seem to indicate it does.
I dunno if Intel is going to copy AMD. Right now, Intel is busy moving the GPU onto the northbridge to save money (esp. in laptops). That means Intel probably isn't going to move the northbridge onto the CPU, at least not on all systems.
http://lkml.org/lkml/2005/8/20/95
All I was doing was responding to the original question that a processor that is X times faster is EXACTLY X times faster for what it can do. You're right -- that's not particularly informative. I was trying to point out that the question wasn't very informed.
$1000, I prefer to buy a $100-200 cpu and spend the rest "underclocking" myself with frosty beverages. It also helps reality look more like computer graphics too.
Actually, if you really know your CPU architecture, you can easily get a five-digit salary, and even six-digit salaries are not unheard of. And these jobs aren't going to India anytime soon. Granted, it's extremely unlikely that such people would show up on slashdot, but it's not like CPU architecture is some pointless nerd thing that no one has any use for. Even a low-level software developer could use a general knowledge of CPU principles to his advantage.
I appreciate any corrections you can give me. I don't like to be wrong multiple times. You can help me be right more often by pointing out where I am wrong and helping me get the correct info.
http://lkml.org/lkml/2005/8/20/95
Based on what you reported, the entire Wendy's chain should be nuked for furthering the cause of obesity in the USA.
You can easily get a 5 figure salary flipping burgers. Hell, with working two jobs, most anyone can easily make a 6 figure salary.
your mods stay if you post as AC? I didnt know that...
That statement says nothing about being truly a desktop, just the fastest gaming CPU.
My e-mail of 18 June:
Date: Sat, 18 Jun 2005 13:12:19 -0700 (PDT)
From: "--"
Subject: FX-57 review
To: info@gdhardware.com
With all due respect, your review is absolute bullshit:
"The real target audience for the FX-57 is going to be the ultra-gamer who insists on nothing but the absolute fastest gaming CPU money can buy. It simply crushes everything in its path in game performance and handles most of today's common applications with power to spare."
I've got a Dothan running at 3.0 GHz that will SMOKE any AMD out there, unless it is on LN2. Why would you ignore the P-M's on desktops with the numbers they have been putting up?
"We clearly see the FX-57 as a 'Don't you wish you could build a chip this fast Intel?'"
They have; it's called the Dothan, and it is less than half the price of a FX-57, even if you include the CT-479 adapter.
Man, you guys have AMD stock or something?
---------------
Los Angeles
AMD locks you into DDR right now. You cannot use DDR3 on an Athlon 66 or FX any more than you can use DDR2.
Your response is non-responsive.
Furthermore, DDR3 is most definitely going to be more expensive than DDR and DDR2 for the near future, probably prohibitively so.
As to memory prices, we'll just have to see. I do not subscribe to the idea that xbitlabs' reprints of stories in the Taiwanese technical news are any better than Tom's. But for that matter, neither is impressive.
I do strongly feel that DDR2 prices will drop below DDR prices. Intel machines use DDR2 now, the majority of the market will be with DDR2, and that'll make that the most common and cheapest form of memory. I could be wrong of course, since RDRAM never dropped in price (although I predicted that correctly too).
I'm not 100% certain Sciencemark measures memory bandwidth completely accurately. I know it is designed to, but it also accidentally measures bus utilitization and certain cache efficiency parameters (more accurately prefetch parameters) also. I do agree it does give an interesting "real-world" counterpoint to the more common mathematical definitions of memory bandwidth.
Finally, this test uses CAS2 DDR, and CAS3 DDR2. Although CAS3 DDR is the most common type of DDR2, CAS2 is by far NOT the most common type of DDR. CAS2 DDR costs about twice as much as regular CAS2.5 DDR (more than regular DDR2 also). The tester should have used lower-latency DDR2 or regular-latency DDR in order to have a more apples-to-apples test.
Finally, your posts are near unreadable due to poor formatting. I used to have that problem too. You should change the pop-up next to the submit and preview buttons to "plain old text" before you submit your post. Leaving it at HTML means that all your paragraph breaks are eaten. Alternately, you could put some explicit breaks in your posts. Using "plain old text" doesn't limit you much, links and such still work.
http://lkml.org/lkml/2005/8/20/95
"Which is immaterial. You can't just pop in the next best memory technology on either Intel or AMD systems - they both put a cap on what you can use. Even going to the next best P4 FSB to use faster DDR2 causes you to replace your CPU, mobo, and memory, so a claimed advantage or disadvantage here would be mistaken."
That's not my point. I think you missed it again. In order to use new memory types on an Intel system, they just need a new chipset, a North bridge. So that means that P4 for example which once used RDRAM and SDRAM later used DDR and DDR2. If Intel chooses not to support a memory type, a 3rd party (who has a chipset license) can support it for them, and the processor can use it. This is not the case on the AMD chips, since the memory controller is in the chip. In order to adopt new memory, AMD must come out with a new chip, and probably a new socket too (as standards often add pins).
This is a big difference.
You don't have to change your FSB to use DDR2 on an Intel. You might in order to get max speed, but I'm not talking about max speed here. I'm worried about pricing. You can leash DDR2-533 to a 800FSB P4 with some degree of effeciency, certainly enough for most people.
As to the non-responsive thing, I say AMD locks you out of DDR2, you respond that who cares, DDR3 is the thing. Well, AMD locks you out of DDR3 also. Your response is non-responsive, it doesn't answer my complaint, it actually states a new one.
As to the CAS stuff, it appears you're right, CAS3 is premium for DDR2. However, I disagree that CAS2 for DDR is cheap. Even at newegg. It's 50% more than regular CAS3 or CAS2.5. It's premium. But you're correct, both modules in the test were premium modules.
Again, as to no real-world disadvantage, you're still wrong. It removes the flexibility of memory module type support from the chipset designer (of which there are at least 4 companies) and puts it in the hand of AMD alone. As DDR2 drops below the price of DDR in the 2nd half of this year, it will hurt AMD, since as you say, they won't have DDR until next year.
My point isn't that you can upgrade yourself, it is that the change in predominant memory type makes the AMD CPU itself obsolete, while on the Intel side, new mobos come out to support the new style.
That hurts AMD in the marketplace. You're right that the impact on individual users is much smaller.
http://lkml.org/lkml/2005/8/20/95