AMD Unveils Barcelona Quad-Core Details

mikemuch writes, "At today's Microprocessor Forum, Intel's Ben Sander laid out architecture details of the number-two CPU maker's upcoming quad-core Opterons. The processors will feature sped-up floating-point operations, improvements to IPC, more memory bandwidth, and improved power management. In his analysis on ExtremeTech, Loyd Case considers that the shift isn't as major as Intel's move from NetBurst to Core 2, but AMD claims that its quad core is true quad core, while Intel's is two dual-cores grafted together."

Comments

Intel's Ben Sandler?

[TheLink]

"Intel's Ben Sander laid out architecture details of the number-two CPU maker's upcoming quad-core Opterons."

Talk about dual processing being grafted on... ;)

Re:Intel's Ben Sandler?

[leonmergen]

I hate your sig.. it's so early.. how dare you.. I hate your sig, need coffee.. aahh

Re:Intel's Ben Sandler?

[mattmacf]

Ok guys, CUT THE SHIT!! It's four in the morning and the last thing I want to do is request a new password just to finish posting a comment and come back to find new comments that prompt me to need to go back to my email and login AGAIN

See, some of us just don't ever logout and everytime I come back to my computer Slashdot is waiting happily for me to return. but you couldn't just let that be, could you? nooooooooooo... every JACKASS WITH AN AGENDA and a COMPLETELY UNFUNNY SIG has to dick me around tonight instead of just letting me post in peace.

My apologies, I seriously need some sleep.

--
Forget your password? Have your password mailed to you by entering your nickname, uid, or email address.

Memory Controllers

[ExploHD]

the memory controllers now support full 48-bit hardware addressing, which theoretically allows for 256 terabytes of physical memory.
256 terabytes should be enough for anybody.

Re:Memory Controllers

[edwardpickman]

12 billion polygons and 25 terrabytes of texture maps in Maya will max that out easy. Why such a low ceiling? They should have gone for at least 512 terrabytes of physical memory. By the time they release Duke Nuke Em Forever it'll exceed that for sure.

Re:Memory Controllers

[aussie_a]

No he didn't. But please, don't let that stop you claiming otherwise.

Re:Memory Controllers

[Kangburra]

To be fair it has to be one of the top ten mis-quotes.

Re:Memory Controllers

[SlashGeO]

Famous last words!!

Re:Memory Controllers

[Anonymous Coward]

Please read parent more carefully.

Clearly "Bill Gate's" is shorthand for "Bill Gate has", not referring to Microsoft's Bill Gates at all, and you can't actually *prove* he didn't delcare that in 1981, nobody knowing what "delcare" means.

The joy and marvel of logic at play!

Re:Memory Controllers

[Xichekolas]

... I for one welcome our Degenerate Overlords!

*ducks*

Re:Memory Controllers

[Anonymous Coward]

Like how he missed the world processing, file/print sharing, directory/identity
services, internet, internet search booms? The reason why they're ahead in those
fields is because they used (questionable) business tactics to claw their way to
the front.

The "computer in every home" thing wasn't a profound prediction: there was a
big dollar sign pointing there from his business model.

Aside from the fact that there isn't really a (PC) computer in every USA home,
if we do generously concede Bill that one prediction, he has missed a great
deal more to really be considered a computer visionary.

A business visionary, perhaps.

Re:Memory Controllers

[Dun Malg]

Of course he said it. Just because he later denies it, doesn't mean he isn't lying.

Right, and just because nobody can come up with a time place, witness, or any other evidence at all of Bill Gates' supposed quote actually coming from Bill Gates' mouth, that doesn't mean he didn't say it. Perhaps a goddamn psychic heard him think it in the shower one morning! Fucking moron.

Re:Memory Controllers

[deficite]

Whoa, fella. Way to rant about something offtopic! If you read carefully (possibly something that a slashdot poster would have problems with) he was commenting on a misspelling. I "delcare" that you are an idiot. (Hmm, big shock that you posted AC)

Re:Memory Controllers

[x2A]

Yeah he missed the whole networking thing, that moves resources out of the local workstation by being able to share them... he was assuming the pc would do more on its own. He's always wanted to tack on extra functionality to their os's, whether they came up with the ideas themselves, or could even pull off doing it themselves, is another matter.

But more to the point, it's urban myth, there's no citation, nobody who will stand by claims that they heard him say it, or anything (if you can prove me wrong, by all means, go ahead, end the debate). Therefore there's no reason to believe it over and other mythical claim, you may as well believe the world is 6000 years old, or any other claims based on "faith" rather than any actual evidence.

wha?

[macadamia_harold]

AMD claims that its quad core is true quad core, while Intel's is two dual-cores grafted together.

So Intel's Ben Sander claims that AMD's claim is that Intel claims that their dual-cores grafted together qualify as quad-core technology? That's not confusing at all.

Re:wha?

[Anonymous Coward]

AMD is limited to claims nowadays!

On snap!

[joe_cot]

"In his analysis on ExtremeTech, Loyd Case considers that the shift isn't as major as Intel's move from NetBurst to Core 2, but AMD claims that its quad core is true quad core, while Intel's is two dual-cores grafted together."
BUUUUUUUUUURNED
Next week: Intel responds by telling us how fat AMD's mother is.

Re:On snap!

[jpardey]

I thought next week intel was going to talk about its use of "Burst Processing" on AMD's mother.

Re:On snap!

[chowdy]

Your mother is so fat, she sat on an A64X2 3800+ and made two A64 3000+'s!

Re:On snap!

[mr_matticus]

Wouldn't "burst processing" be a feature of AMD's mother?

Mine goes to eleven

[Anonymous Coward]

Quad core? Bah! That's only 4.
Wake me up when they have a processor that goes to eleven.

Re:Mine goes to eleven

[masklinn]

8 cores with 8 virtual threads per core (for a total of 64 threads/CPU) enough for you or not?

Re:Mine goes to eleven

[rf0]

#include

Re:Mine goes to eleven

[AcidLacedPenguiN]

why not make 10 faster?

Re:Mine goes to eleven

[jo42]

Eleven? Wake me up when we have 42.

eh?

[dexomn]

I thought Ben Sander worked for AMD.

Re:eh?

[thoryorak]

Ben Sander does work for AMD. It's pretty obvious in the original Extremetech article. Unless AMD is letting Intel people work on their new projects.

Re:eh?

[mr_matticus]

Why not? Microsoft's Windows division lets Apple do their work.

Sorry, couldn't resist. Of course Ben works for AMD. This is just yet another flawed Slashdot summary (are there ANY accurate and well-written summaries on Slashdot these days?).

Intel's response?

[Kangburra]

Intel Talks Up Power Savings

Could I still count???

[RuBLed]

"AMD claims that its quad core is true quad core, while Intel's is two dual-cores grafted together."

So a Siamese twin is not really a true twin because they are two persons grafted together? :)

Re:Could I still count???

[Mikachu]

So a Siamese twin is not really a true twin because they are two persons grafted together? :)

No, actually, it's more like saying that siamese twins are not actually two people in the same body because they're grafted together.

Re:Could I still count???

[Tsagadai]

Actually as opposed to creating siamese twins AMD claims to have created one twin with 4 split personalities.

Re:Could I still count???

[Derosian]

This is like the difference between an AMD processor processing something just as fast at say, 1.8gig, as an intel 2.6gig.

Why? Because Intel will look for shortcuts to a goal. Yes, we have quad core processors, but not telling us the processors were not made for that purpose they will however most likely run faster. While AMD will make a processor for the specific purpose of it being quad core.

Re:Could I still count???

[MiniMike]

No, it's more like comparing Siamese Quadruplets against two sets of Siamese twins stapled together.

There's a nice image to drink your coffee to...

Re:Could I still count???

[MrNaz]

One twin? What have you done with the other you horrible man?!

Re:Could I still count???

[SultanCemil]

Ahhh, if only I still had some mod points. That is one of the funniest posts I've read in a *long* time.

Re:Could I still count???

[dobestpossible]

"Yes, we have quad core processors, but not telling us the processors were not made for that purpose they will however most likely run faster"

  Not Not A double negative or you aren't not illiterate.

Re:Could I still count???

[Chris Burke]

Um, actually, in this case it's more like saying Siamese twins are the only true twins, and taking two non-conjoined twins and sewing them together does not make them Siamese.

Re:Could I still count???

[drinkypoo]

First of all, double negatives are a time-honored and perfectly acceptable staple of the English language. Your english teacher told you they weren't (and that ain't isn't a word, which it most certainly is) because she wanted you to believe that they weren't until you had the maturity to use them correctly. I know that I was a precocious little bastard when I was a kid, always reading books that were well beyond my age, and confusing the shit out of my fellow students. Not all of them of course, there were the few other nerds around, who I naturally hung out with.

Re:Could I still count???

[dobestpossible]

Well if confusing the reader to the point that your point is misunderstood, congratulations, stupidity prevailed. I'm sure you think that using cleverness and wit makes you smart.

Re:Could I still count???

[drinkypoo]

The point is that slashdot is supposed to be news for nerds, not comments for idiots. I don't write to the lowest common denominator; if my target audience (people with two neurons to rub together) can understand it, then my mission has been accomplished. When I'm actually writing to the world, then I assume that my audience is dumber - unable to comprehend double negatives and so forth - and I write at the average american's eighth grade reading level. (Unless they're still in high school, then I aim at about grade four.)

Re:Could I still count???

[dobestpossible]

Der der der....rite at mi grad level i cont fegur aot yur shit.
  There is a reason that English teachers tell their students not to use double negatives. Can you guess why??

(Hint: Necessity)

Re:Could I still count???

[AvitarX]

Do you still think essays need to be 5 paragraphs?

Must the introduction always end with a thesis statement?

Should the conclusion never add new ideas?

These are all things that are tought through highschool at the middle and lower levels, are they all true because of it?

The sentance Yes, we have quad core processors, but not telling us the processors were not made for that purpose they will however most likely run faster is hard to parse, but not because of the double negitive. The problem with the sentance is, Intel appears to be the subject, and the collective they is their employees perhaps? Do they run faster?. Also I am courious weather the poster is opining faster as in more actual work, or faster as in GHz.

Improper usage of a double negitive would be AMD does not design their CPUs to not be dual core instead of AMD does design their CPUs to be dual core. Both sentances have the same meaning but one is hard to read. ...not telling us the processors were not made for that purpose... has no obvious positive statement that is equivelent and therefore not a problem use of double negative. It is possible the end of the sentance was so fucked up you through your frustrations at the wrong point though.

How would you correct the sentance BTW?

PS.

Don't pick apart my grammer, because it is how I encrypt my messages.

Re:Could I still count???

[dobestpossible]

"It is possible the end of the sentance was so fucked up you through your frustrations at the wrong point though." - Or your frustrations arise from the wrong use of "how I encrypt my messages."
  "PS. Don't pick apart my grammer" - What are you going to do about it? Cry like the nerdy bitch U R?
  "How would you correct the sentance BTW?" - For starters, I would either say it is or is not. Who reads your crap and thinks "Oh, the author is so smart the way he encrypted the sentences. I wish I could do that..."
  "The problem with the sentance is, Intel appears to be the subject, and the collective they is their employees perhaps? Do they run faster?" - As if no one gets the jist of a morons ramblings. Oh, that's right, you think it was encrypted. My bad for piecing together the puzzle.
  You started a trend, I heard colleges across the world are going to force essays to be only 5 paragraphs long (no more, no less. Or isn't not more, ain't not less), Introductions don't need a thesis, conclusions should not have any ideas at all, and the entire paper has to be encrypted by the use of double negatives, sentence fragments and run ons, and incorrect punctuations, spelling, and lack coherence throughout.
  P.S. I just wanted to put a post script here for the chocolate anal fudge of it.

Re:Could I still count???

[dobestpossible]

MMMMmmmm...Yummy
Will u roll in a bathtub with me? Bring a friend, I don't care if it's a troll too.
I'll play "Let's get it on" in the background while you nibble on my nipples...yeah that's the spot. Harder baby, harder

Re:Could I still count???

[Nakarti]

No, Siamese twin is not dual-head person, because they are two people grafted together.
They cannot think and act as one, only get very close, but my other personalities and us are all in one brain.

Understand?

How obscure is this?

[MobileTatsu-NJG]

"AMD Unveils Barcelona Quad-Core Details"

It's the processor that runs like a dog with no nose!

Once again...

[tygerstripes]

Firstly, can I just say that stating that "the shift isn't as major as Intel's move from NetBurst to Core 2" is like... er... comparing a decent incremental car improvement with swapping a bicylce for a car. Or something. I'm not saying Core2Duo isn't great tech, but look; Netburst was shit. Everyone knows it. They flogged that horse for far too long, so comparing on the grounds of the proportional improvement is not a useful comment. It's like when the thick kid in school got the "most improved" award, and everyone sat there and went "Well yeah, but what was his alternative?".

As for the quad-core thing, it's the same story all over again. Intel rush out a solder-together-two-chips job to beat the competition to market, and then the actual innovators come out with something coherent that works more efficiently etc.

I'm not saying the AMD will necessarily be better. What I'm saying is I don't care who gets to market 2 months earlier. I want the better chip, and I can live with the mystery for a few weeks.

Although, frankly, I can barely afford to eat having just built a decent Core2Duo rig, so I won't be investing either way just yet...

Re:Once again...

[Calinous]

NetBurst was a good architecture - the only problem with it was total heat, and hotspots inside the processor. This kept it from reaching its expected 10GHz (but was able to run at 6GHz on liquid nitrogen cooling). Now, if an 3GHz P4 is underwhelming, you couldn't say so about a 6GHz one.

Re:Once again...

[tygerstripes]

It wasn't that good. AMD came out with an architecture which, in practical terms, was better designed, while Intel just kept trying to push the envelope with this very hot chip, and steadily lost market share as a result. Core2Duo is fantastic, relatively speaking, but it was a very long time coming...

Re:Once again...

[eebra82]

Firstly, can I just say that stating that "the shift isn't as major as Intel's move from NetBurst to Core 2" is like... er... comparing a decent incremental car improvement with swapping a bicylce for a car. Or something. I'm not saying Core2Duo isn't great tech, but look; Netburst was shit. Everyone knows it.

We all know that already. The point is that AMD needs that kind of jump to get ahead of the competition like it was half a year ago.

I would say that if the writer's point of view is that AMD needs an equally big advancement in tech and it's dead on right. Why? Simply because the lead that AMD had was erased by so much, and AMD will need so much to recover if it wants the top position back.

Re:Once again...

[masklinn]

NetBurst was a good architecture - the only problem with it was total heat, and hotspots inside the processor. This kept it from reaching its expected 10GHz

... and NetBurst was underperforming at the speeds it could reach. In a word, no, it wasn't a good architecture because it only worked well in FantasyDreamLand where heat dissipation doesn't exist.

Re:Once again...

[LaughingCoder]

Netburst was designed for a market that touted clock speed as the performance measure for CPUs. AMD, with a big helping hand from the gamers, changed the game into rewarding true benchmark/performance rather than simple clock speed. I suppose if Intel had managed to achieve 10GHz clocks their performance would have been top notch, though one wonders how long those instruction pipelines would have to be ... and how much power they would have burned.

Now Intel has out-benchmarked AMD, and is attempting to change the rules again to performance-per-watt. This next wave should be interesting to watch.

Re:Once again...

[Waffle Iron]

That's like saying "The Titanic really was an unsinkable ship - the only problem was the poor quality steel and rivets that made it vulnerable to opening huge gashes in the hull during collisions".

Re:Once again...

[JebusIsLord]

Well that's not really completely true though, is it?

I mean, if you read about the K8L architecture, they can throttle the voltage and clock rate for each core independently, and all 4 share a common L3 cache as well. Plus there are additional hypertransport links for inter-core communications.

The Intel solution has none of these, as a direct or indirect result of not being a "true" multicore setup (technologically very similar to a quad-cpu setup, really, let alone two dual core cpus). This doesn't mean they won't still be faster (Core2 is a better design it seems, despite these limitations). And it doesn't make me a "fanboi" for discussing it.

My prediction is that Intel will hold the performance lead, but by a narrower margin.

Re:Once again...

[HoboMaster]

No one's saying it's not a quad-core. The issue is that it isn't going to perform nearly as well as a processor designed for four cores.

"Within that black box of a CPU, it doesn't matter how things are connected."

Depends on what you're looking for. If you just want it to work, then yeah. But if that's all you're looking for, why are you buying a quad-core?

I've got my own suspicions as to your "fanboi"ism.

Re:Once again...

[JebusIsLord]

Well now, I never used the term "true multicore" for either design. Both are "true multicore" in the sense that they both have multiple cores. I never argued anything about that. Read it again! I'm just saying there is some validity to the assertion that the core 2 quad is more cobbled together. They CAN'T share cache between the two halves (You are correct in that the core 2 duo shares L2 cache - didn't know that before), you can't alter voltages and clock rates independently either.

You are wrong about the chip being a black box, though... the Intel design has all cores speaking to each other across the front-side bus, which is expected to hurt the 4-way and cripple the 8-way designs (theoretically, I know). The AMD design also requires the cores to talk across the HT bus, so that's why they added more HT links. Intel doesn't have that option.

Re:Once again...

[dfghjk]

"Well now, I never used the term "true multicore" for either design."

Really? You said:

"The Intel solution has none of these, as a direct or indirect result of not being a "true" multicore setup..."

"I'm just saying there is some validity to the assertion that the core 2 quad is more cobbled together."

No processor design is "cobbled together". One could argue that Intel's methods are superior because they have something that AMD doesn't (MCM). They have 45nm as well of course. Intel's tech enables them to bring quad-core to market sooner, so by that measure Intel's design is superior to AMD's. The custumer doesn't care.

"the Intel design has all cores speaking to each other across the front-side bus, which is expected to hurt the 4-way and cripple the 8-way designs (theoretically, I know). The AMD design also requires the cores to talk across the HT bus, so that's why they added more HT links. Intel doesn't have that option."

Doesn't matter because both designs ultimately share a memory controller. HT is helpful in MP systems where each processor adds a memory controller. In a system with a single CPU and memory subsystem, all cores share the same memory one way or another. AMD's advantage is in larger multi-CPU designs, not multicore.

Re:Once again...

[JebusIsLord]

Hmm, guess I did say that. I take that part back, then, since it is no longer my opinion.

I didn't say Intel was doing shoddy work, i'm saying that their initial quad-core offering is cobbled together. They don't even plan on making it widely available. I'm sure their 45nm quads next year will be excellent chips.

You will see the FSB bandwidth limitation become much more of a burden as more cores are added to the Intel architecture, while AMD's design has the access to the memory controller scale with CPU speed, so there is more breathing room. This is the prediction that many technologists have made. Right now Intel's design is clearly faster, and is a better purchase - I agree. Even their cobbled-together quad is gonna be a hell of a fast chip!

AMD's design has some advantages though, and I'm hoping they pull through. Not because I'm a "fanboi", but because competition in a market is healthy. I'll always buy the winner.

Re:Once again...

[GWBasic]

As for the quad-core thing, it's the same story all over again. Intel rush out a solder-together-two-chips job to beat the competition to market, and then the actual innovators come out with something coherent that works more efficiently etc.

I wouldn't describe it like that. Take a look at this: Intel Pledges 80 Cores in Five Years.

Note to AMD: We don't care

[cperciva]

AMD claims that its quad core is true quad core, while Intel's is two dual-cores grafted together

Note to AMD: We don't care about the implementation details. We care about performance, cost, and power consumption; the clock speed, cache sizes, and how cores talk to each other is irrelevant.

For all I care, Intel's "quad core" processor could be using a team of psychic circus midgets.

Re:Note to AMD: We don't care

[ilmdba]

We don't care about the implementation details.

I think you mean you don't care. The rest of us that use processors by the thousands, will be over here caring about the implimentation details.

Re:Note to AMD: We don't care

[Anonymous Coward]

Some of us do care. Some for work, some for fun. AMD's "designed as quad-core" approach has some notable consequences, especially in the cache layout that (on paper, of course) seems very well suited to virtualization -- much more so than the Intel solution in TFA.

AMD: a shared L3 feeding core-specific L2 caches. Intel: each core-pair sharing a L2 cache. AMD's approach better avoids threads competing for the same data (thanks to copying it from L3 to every L2 that needs it), while keeping access latencies more uniform and predictable (thus better optimizable).

Other AMD enhancements look more like catch-up to Core 2: SSE [and it's "Extensions", dammit, not "Enhancements"] paths from 64bit to 128bit, more advanced memory handling (out-of-order loads versus Intel's disambiguation et al.), more instructions per clock by beefier decoding (more x86 ops through fast path instead of microcode) and more "free" ops (where Intel added way more discrete execution units from Core to Core 2).

If AMD's quad manages to be better due to better memory bandwidth and latency (in practice), then they were quite right about "true quad-core" :)

Re:Note to AMD: We don't care

[tygerstripes]

Is that meant to be sarcastic?

You don't care because you don't understand. Performance, cost and power consumption are directly affected by such things as clock-speed, cache, core integration, architecture etc, and different aspects offer different advantages for different uses.

If it were that easy to put a reliable figure on Performance, the Megahurtz shambles would never have happened.

Re:Note to AMD: We don't care

[Jeppe Salvesen]

You are badly confused.

The ends is a well-performing processor that doesn't produce too much heat or cost too much.

The means is technology. The implementation details may be fascinating, but what matters is benchmarks vs total power usage.

Re:Note to AMD: We don't care

[mabinogi]

> You don't care because you don't understand. Performance, cost and power consumption are directly affected by such things as clock-speed, cache, core integration, architecture etc.

Yes, but saying "Oooh, our chip is true quad core and yours isn't" doesn't on its own say anything about final cost or performance. They might as well have said "But ours are red".

It's a clue that it might perform better with all other factors being equal, which they're not, so it's still a useless statement. Until they benchmarks are out, it's all just speculation.

Re:Note to AMD: We don't care

[aminorex]

"...what matters is benchmarks..."

Lies, damned lies, and statistics. I both agree and disagree. Throughput on applications is what matters to end users. Synthetic benchmarks are useful (and so matter) in as much as they identify specific architectural performance characteristics for a given implementation. They are less than useful (and do not matter) when they do not correspond in a predictable way to throughput results.

"...vs total power usage..."

For your application, perhaps. Most home and office users don't care about the power dissipation of their CPU, as long as the cooling rig is zero-maintenance. GPUs completely overwhelm small variations in CPU for gamers these days. For high-throughput computing systems, there is a major shared/distributed memory split. For shared memory systems (i.e. capable of scaling throughput on multithreaded applications by increasing CPU counts), interconnect scalability matters more than any thing else, and AMD wins handily. For distributed memory systems, blade farms, etc, scalability and rank density will be determined by power dissipation, and there, finally, I can agree with your comment, and Intel may have a (very small) lead. It's a rather small slice of a diverse market, however.

Re:Note to AMD: We don't care

[name*censored*]

Well maybe YOU don't care about how psychic circus midgets consume power, but uh....

Wait what was the question?

Re:Note to AMD: We don't care

[Visaris]

Note to AMD: We don't care about the implementation details. We care about performance, cost, and power consumption; the clock speed, cache sizes, and how cores talk to each other is irrelevant.

AMD it taking the route that will give better performance. I hear you saying that soldering some copper pipes with rubber-bands would be fine as long as it would perform. The point is that it will work... just not very well.

If you don't think I'm right, look at Intel's own product roadmap. They plan to release a new version of Kentsfield that has all four cores on one peice of Si, with a shared cache, just like AMD is about to do... only later in 2007 after AMD's version comes out. When the two major chip companies move in the same direction, usually that means it is the right one. The only difference is that AMD is going to get there sooner because they didn't bother to play around with this MCM (Multi-Chip-Module) junk. Intel just wants to get to market first; they don't seem to put quality first.

Re:Note to AMD: We don't care

[andr0meda]

AMD it taking the route that will give better performance. I hear you saying that soldering some copper pipes with rubber-bands would be fine as long as it would perform. The point is that it will work... just not very well.

If you don't think I'm right, look at Intel's own product roadmap. They plan to release a new version of Kentsfield that has all four cores on one peice of Si, with a shared cache, just like AMD is about to do... only later in 2007 after AMD's version comes out. When the two major chip companies move in the same direction, usually that means it is the right one. The only difference is that AMD is going to get there sooner because they didn't bother to play around with this MCM (Multi-Chip-Module) junk. Intel just wants to get to market first; they don't seem to put quality first.


This sounds very XP. I like XP. Yet, someone please explain why all managers seem to believe that he copper pipes soldered together will hold and throw them fortunes? Is it the attractivity of the gamble in and on itself?

Re:Note to AMD: We don't care

[Visaris]

Yet, someone please explain why all managers seem to believe that he copper pipes soldered together will hold and throw them fortunes? Is it the attractivity of the gamble in and on itself?

Intel will be first to market with QC chips. At that time, Intel's product will be better... because AMD won't have one yet. That is why Intel is taking the route it is. Good for business? Maybe... it just puts Intel that much furhter behind in the long run, only for some short term gains... I don't have a business degree, but I don't think it's worth it.

Re:Note to AMD: We don't care

[Kijori]

They might as well have said "But ours are red".

I think you chose a bad example. Everyone knows red ones go faster!

Re:Note to AMD: We don't care

[afidel]

Intel will be first to market with QC chips. At that time, Intel's product will be better...

No, it won't. Because their quad core will be so resource starved that for 90+% of real world application it won't perform any better than dual core and will consume significantly more power.

Re:Note to AMD: We don't care

[mhectorgato]

I guess it all depends on what real-world applications your talking about. If you're referring to Word/Excel/Web/etc then an AMD QC won't be much quicker than a AMD DC as well. If your real-world apps refer to multi-threaded activity then: Acorrding DailyTech's benchmarks - comparing to a similarly clocked Core2Duo to Kentsfield (2+2), it's about 90% faster in 3D Studio Max 8, Cinebench 9.5 and TMPG Encoder; about 70% faster in Windows Media encoding. According to HardwareSecret compared it to a Core2Extreme (10% faster clock speed) it's 80% faster in POV-Ray, 50% faster in Sony Vegas 7.0a. XBitLabs compared it to a Core2Extreme (10% faster clock speed) and it's 54% faster in the 3D Mark 06 CPU tests. AnandTech estimated it was 51.4% faster than a similarly clocked Core2Duo in Divx 6.2.5 with XMPEG. The article concluded thusly: "With only a 266MHz difference in clock speed, the new Core 2 Extreme QX6700 isn't too hard of a choice to make. When Intel introduces a lower cost 2.40GHz Core 2 Quad version, things may get a little more complicated, but at the very high end we would rather have four slightly slower cores than two slightly faster cores. We expect that there will be some improvements in multitasking performance, especially if you have a decently fast I/O setup, and don't forget the performance boost you'll get in well threaded applications"

Re:Note to AMD: We don't care

[MrNaz]

You don't care about tech details? What on Earth are you doing reading Slashdot then? That's like reading BRW and then saying "Don't bore me with money stuff, I just want my liabilities to go away!"

Re:Note to AMD: We don't care

[Jeppe Salvesen]

Yeah. I meant "useful" benchmarks.

As for the total power usages, these quad-cores are meant for servers. And in servers, heat production matters.

It also kinda matters for the desktop users - I am currently being annoyed by the noisy fan on my coworker's desktop.

Re:Note to AMD: We don't care

[Chris Burke]

Yes, but saying "Oooh, our chip is true quad core and yours isn't" doesn't on its own say anything about final cost or performance.

If you know something about computer architecture it does. This is MPF, so it is reasonable to assume a good portion of the audience would understand the implications of that statement.

Just because that statement by itself does not produce a benchmark score for you doesn't mean that it is not a useful statement.

Re:Note to AMD: We don't care

[aka1nas]

Pretty much every benchmark around shows that the K8 doesn't benefit signficantly from extra cache. From 512KB to 1MB you get maybe 3% or 4% more performance if you are lucky. The IMC saps quite a bit of the gains that having more cache brings you as the penalty for a cache miss is reduced.

Re:Note to AMD: We don't care

[afidel]

I'm talking about databases or middleware logic servers or non-trivial parallel scientific workloads like weather forcasting. Also the fact that double the cores only got you 50% more speed in many of those benchmarks proves that it (or the apps running on it) doesn't scale well. I just fundamentally think that Intel is going market driven instead of technology driven again on this one and that as Netburst proves that's not the smart way to go long term. Intel really needs to put their intelligent and talented engineers to work on making a scalable design or else they are going to have trouble in the next couple years as the performance gains from the Core2 architecture fade due to inability to feed additional cores.

Re:Note to AMD: We don't care

[mhectorgato]

I agree that AMD solution is better for high-end processing. But what percentage of customers need that extra few % of performance that a true QC would derive vs 2xDC? What percentage of customers need the extra % of performance of Core2 2xDC vs Core2 Quad?

This is definately a marketing move, this will sure help regain market share. But it's also a technical move, due to the fact if I bought a Xeon5100 server, When this system starts to bog down, I can go from 4 cores to 8 cores relatively painlessly.

This is also only a stop gap solution, as Intel is coming out with a true QC later in '07. They are already working on a replacement technology for the current FSB, like AMD's approach - CSI (I believed it was named)

Intel is working on multiple fronts simultaneously - 2xDC; QC; multiple FSBs on server mobos; 45n; Geneseo, CSI, etc ...

Although I'm a self-pronounced Intel fan boy, I can clearly see the technical superiority of AMD's approach. Intel also is following AMD's lead, as well as the other way around.

Re:Note to AMD: We don't care

[dfghjk]

so if it turns out that AMD's design is faster then AMD is better? Is that what you are trying to say. For someone bent on refuting the parent's claim, you did a pretty poor job. The claim was that all that mattered was performance (and that is irrefutable).

Re:Note to AMD: We don't care

[dfghjk]

"Just because that statement by itself does not produce a benchmark score for you doesn't mean that it is not a useful statement."

It's not a useful statement nor even a factual one. A quad-core processor has 4 cores on one physical CPU. Those 4 cores are not required to be on a single die to qualify. Hopefully you understand enough about "computer architecture" to get it but most likely not. AMD's ongoing condemnation of Intel's use of MCM is nothing but FUD.

Re:Note to AMD: We don't care

[Chris Burke]

Well there isn't some tech language board like there is with the French language, but usually the usage of multi-core implies on a single die, not an MCM. Still Intel can legitimately call it that, AMD is just pointing out that there is a distinction between an MCM and a "true" multi-core, which there is.

And the condemnation isn't FUD -- there are implications of that choice that affect the performance of the resulting part. Does it make the Intel quadcore part bad? Not necessarily, but worse than it would be were it a single die.

Re:Note to AMD: We don't care

[Chris Burke]

Yes, implications like lower yield, which will result in higher prices and probably lower clock rates until the process improves.

That's one of them, yes. AMD will have to work hard to get their yields up.

That isn't going to mean lower clock rates as a result. I mean, it's a new process (65nm), so that will take time to ramp both the frequency and yields, but that is irrespective of the choice to have a single quad core die. The things causing the faults that make a processor inoperable are different than the things that cause variations in transistor speed, and while the former occur more or less at random and are thus more likely with larger error, the latter tend to be more widely spread across the wafer.

I would expect that the higher die area would result in a percentage loss across all speed bins, but not to significantly change the ratio of bin splits.

Speaking of clock frequencies, though, be aware that one of the reason why an MCM is worse is because it creates a greater signal integrity problem for the FSB. Intel's bus already has major issues, the reason why the MP parts have always had lower bus frequencies than the non-MP parts. This makes it worse.

If you doubt the superiority of single-die solutions, then watch Intel. They are going to move to a single-die quadcore part as soon as they finish developing one, just like they did with their dual-core parts which also started life as MCMs. The only reason they are making the MCM is for time-to-market reasons -- which should be obvious, because with their huge capacity and the best process in the world, they certainly aren't worried about the yield issues of a quad-core die.

Not that there's necessarily anything wrong with that (though I'd wait to buy the real deal instead of the MCM myself), it's Intel's way of being able to say "we did it first", and pointing out the difference between their and AMD's solution is AMD's way of saying "no you didn't".

Re:Note to AMD: We don't care

[dfghjk]

"but usually the usage of multi-core implies on a single die"

says who? It means on a single chip. The fact that most chip have only one die does not make what you say true.

"AMD is just pointing out that there is a distinction between an MCM and a "true" multi-core, which there is."

only when you've bought into the falsehood of your first claim. Single die is no more "true" multi-core than MCM is.

"And the condemnation isn't FUD"

It is in every way. The purpose of such claims is to cast doubt on the performance of the Intel product until AMD can bring theirs to market. It's classic FUD and nothing other than that. Intel's MCM will deliver 4 cores on a single processor. That's a true quad core part regardless of what AMD says and the fact that they say otherwise proves that AMD is practicing FUD.

"but worse than it would be were it a single die."

Certainly, but a single die solution can't be brought to market as quickly. A processor that I can buy today is infinitely faster than one I can't.

Re:Note to AMD: We don't care

[Chris Burke]

says who? It means on a single chip.

Says who? Like I said, it's used that way in industry a lot, but it varies. Multi-core can mean per chip or per die, usually depending on the context. E.g. as an architect I wouldn't call an MCM multi-core because from my perspective the other core is on the other side of the FSB.

Chip, by the way, is another one of those words that can vary depending on context to mean either a die or package, so you probably should have asked for clarification from whoever told you that it means "on a single chip".

A good example of another way of looking at this is IBM's Power 5, which is a dual-core chip, and when packaged onto a gigantic MCM with 3 other Power 5 the result is called "eight-way", using the terminology for an MP system.

only when you've bought into the falsehood of your first claim. Single die is no more "true" multi-core than MCM is.

No, there is still a distinction, because you can call an MCM multi-core, fine, but it isn't multi-core done right. You are focusing entirely on the "true multi-core" aspect, whereas I'm focusing on the "architercurally significant differences in design" aspect. You're taking it entirely as an issue of "is Intel allowed to call their product multi-core", a literalist view. I was taking "true" in a less existential form, one that means "multi-core done right".

It is in every way. The purpose of such claims is to cast doubt on the performance of the Intel product until AMD can bring theirs to market. It's classic FUD and nothing other than that. Intel's MCM will deliver 4 cores on a single processor. That's a true quad core part regardless of what AMD says and the fact that they say otherwise proves that AMD is practicing FUD.

You're right, it is of course a marketing ploy to cast doubt on the performance of the Intel product, and in that sense it is FUD. I got caught up in the new definition of "FUD" meaning "bullshit", which this isn't, because the fact is that you should doubt the performance of Intel's MCM as a direct result of it not being a single die. By the same token it's FUD when AMD casts aspersions on Intel's bus architecture and MP scalability, even though those aspersions are for the most part accurate.

At the same time, Intel's MCM itself is nothing but a marketing ploy. "Intel's MCM will deliver 4 cores on a single processor". Check off that bullet point on the marketing brochure. Is it a good implementation? Well, no, but the bullet point is still checked off, and besides they'll do it "right" in the near future anyway.

Certainly, but a single die solution can't be brought to market as quickly. A processor that I can buy today is infinitely faster than one I can't.

Being to market faster with an inferior product helps Intel, but not really the customer, who in most cases would be better served by waiting until Intel comes out with their four-cores-on-one-die product, which they absolutely will. The MCM is a stop-gap whose sole function is not to deliver a quality product to customers but to prevent AMD from being the first to have a quad-core part. It's innevitable short life, just like the short life of the dual-core MCM, shows what Intel really thinks of their "true" multi-core.

AMD has a legitimate point to make, but you've taken issue with their wording. That's fine, I'm not going to defend their marketing department, but there is a legitimate point to be made and that's the point I'm making.

Re:Note to AMD: We don't care

[dfghjk]

"Says who? Like I said, it's used that way in industry a lot, but it varies. Multi-core can mean per chip or per die, usually depending on the context. E.g. as an architect I wouldn't call an MCM multi-core because from my perspective the other core is on the other side of the FSB."

OK, you just stated that it's so a second time without anything to back up the claim. As an architect, I would call an MCM part multi-core because from my perspective there are multiple cores on the part. Your FSB argument is irrelevant.

"Chip, by the way, is another one of those words that can vary depending on context to mean either a die or package, so you probably should have asked for clarification from whoever told you that it means "on a single chip"."

Except that "chip" is well established as slang for a part that has pins coming out of it. No one confuses "chip" with "die", though they could, so I chose to not be anal about it. If I don't use "chip" then what do I use to refer to the part?

"A good example of another way of looking at this is IBM's Power 5, which is a dual-core chip, and when packaged onto a gigantic MCM with 3 other Power 5 the result is called "eight-way", using the terminology for an MP system."

I don't see how it's another way of looking at the issue at all. Multicore processors ARE MP systems. If IBM offers 8 cores on a single part then it's an 8 core processor. MCM predates multicore dies.

"No, there is still a distinction, because you can call an MCM multi-core, fine, but it isn't multi-core done right."

Sure, that's the AMD propaganda. "Done right" is a matter of achieving your goals the best way possible, so going onto a single die is only "done right" when it's better than MCM. Time to market is obviously an issue for Intel, so they've done multi-core "right" by using MCM instead. No doubt that using a single die has technical merits and Intel will switch to that when the time comes.

"...whereas I'm focusing on the "architercurally significant differences in design" aspect."

Yes, and your fallacy is declaring that your values are more "right" than mine or Intel's. (not saying my values are the same as Intel's, I'm simply defending them from FUD)

"You're taking it entirely as an issue of "is Intel allowed to call their product multi-core", a literalist view. I was taking "true" in a less existential form, one that means "multi-core done right"."

Yes, mine is correct in reality while yours is correct in some "less existential form". Intel will do a one-die quad-core part when it makes technical sense for them to do so. The MCM part, just like the previous dual one, is a early market product that AMD won't match. That doesn't mean it's not a true quad-core part.

"I got caught up in the new definition of "FUD" meaning "bullshit", which this isn't"

Unfortunately, /.'ers misuse the term and ruin it for those who might use it correctly.

"At the same time, Intel's MCM itself is nothing but a marketing ploy. "Intel's MCM will deliver 4 cores on a single processor". Check off that bullet point on the marketing brochure. Is it a good implementation? Well, no, but the bullet point is still checked off, and besides they'll do it "right" in the near future anyway."

This is just more crap. There will be nothing wrong with the Intel MCM part. Yes, it will likely be less good than future processors, but then what isn't? Yes, Intel is doing it in order to to offer a quad part early. How horrible of them to do that! The MCM part is not "a marketing ploy", it's a part you can buy.

"Being to market faster with an inferior product helps Intel, but not really the customer, who in most cases would be better served by waiting until Intel comes out with their four-cores-on-one-die product, which they absolutely will."

In most cases maybe, but not in all cases. If there were no customers then Intel wouldn't offer it and AMD wouldn't be publishing FUD to sell against it. If customers

Re:Barcelona

[Plammox]

If anybody at AMD had watched Fawlty Towers, maybe they would have opted for Madrid instead.

(Manuel with thick Spanish accent:) Mr. Fawlty! I'm from Barcelona, I know *notheeeng*!

Amazing analysis

[nanoakron]

AMD: 4=4
Intel: 4=2x2

Where do they hire these guys?

-Nano.

Re:Amazing analysis

[l0cust]

Oh I know I know ! Since:

AMD: 4=4 Intel: 4=2x2

=> A = D = I = '4', M = n = '=', t = l = '2' and e = 'x'
=> They hire these guys at Hideous Crytograhy Inc.

Now where is my cookie !

Re:Amazing analysis

[ozbird]

Intel: 4=2x2

4=2x2, or 4=2+2? And where does AMD's 4x4(=16?) fit in?
1st grade maths, yet I'm still confused by what it all means... ;-)

Hmmmm Wrong.

[Solokron]

Looks like someone RTFA a bit wrong. Ben Sander works for AMD. He is one of their media presenters. Here are a few of the events he has done: http://www.cpd.iit.edu/cpd/events.htm http://www.ewh.ieee.org/r4/chicago/foxvalley/meet. thru.mid2005.html http://www.instat.com/FallMPF/06/conf1.htm http://mtv.ece.ucsb.edu/MTV/index_files/program-mt v.txt

Re:Hmmmm Wrong.

[AcidPenguin9873]

Ben Sander is actually a lead for one of the Performance Modeling team at AMD, meaning he's a hardcore CPU architect, not a "media presenter". It just so happens that because he's the lead, he gets to present his team's findings at media and conference events.

Some of us do.

[the_hoser]

I care. I think this kind of tech is cool. If you read into it a bit, it gives you a decent enough idea of how much better it will perform. It's not always accurate, but it's usually correct.

And, accorting to the article, this was all coming from the mouth of an intel person, not AMD.

Re:Some of us do.

[J.R. Random]

"And, accorting [sic] to the article, this was all coming from the mouth of an intel person, not AMD."

Actually, according to the article, the person providing the information was Ben Sander, who works for AMD:

"The details of Barcelona discussed in this article were presented by Ben Sander, who led the performance modeling group for Barcelona."

Yes, the guy who submitted the Slashdot summary screwed up. But use a little common sense. Is an Intel spokesman going to give us the inside skinny on a new, unreleased, AMD chip?

Socket consideration

[wysiwia]

I won't buy any AMD processors anymore until AMD clears its socket plans and guaranties a minimum of 3 year availability for processors on a socket. See also http://hardware.slashdot.org/comments.pl?sid=19821 5&cid=16242757.

O. Wyss

Re:Socket consideration

[somersault]

Why - do you think todays processors won't still be useful in 3 years? Most games don't take advantage of current technology for a year or more I'd say, and your office applications/OS are going to run fine on any of today's decently specced systems (3000+, 3Ghz Pentium, doesn't even matter if they only have one core). The only people that can truly make use of multicore chips would be scientists and people who do any other kind of intensive parallel processing, like like graphics rendering. In 3 years you'll probably want a new mobo anyway to take advantage of whatever new-fangled technology has come out. I guess you could say I'm becoming less of a geek these days even though I'm an IT manager, but if my computer works, and plays the games I like sufficiently (say 1280x1024@60fps with details maxed out), I don't see the need for upgrading my processor (I'd upgrade my graphics card before anything else, since graphics cards come out more often and usually would have a larger effect on performance from one generation to the next).

Since most of the chipset is becoming integrated into the processor these days then your argument will make more sense over time, but if you were more patient and waited for things to come down in price, as they always do, and rather quicker than I expect sometimes, then you'd be able to buy a new mobo, ram and processor for the same price as the new processor would have cost 6 months previously (not meant to be a perfect example, I haven't been following the prices of stuff since I built my last system a couple of years ago, but the idea is sound :p )

Re:Socket consideration

[wolrahnaes]

As the person who responded to your last post explained, that's just not possible with the K8 architecture as it is. The memory controller is on-die and memory technology is evolving, therefore the interface between the processor (where the controller is) and motherboard (where the DIMMs are) must also change.

The closest to a solution we have would be going back to Pentium 2/3 style processor-on-a-card designs which would move the memory slots to an expansion card shared with the processor which would then have a HyperTransport interface to the motherboard.

This works, as some motherboard manufacturers (ASRock on the 939DUAL for one) have implemented something along these lines for AM2 expandability. The problem lies in laying out the circuitry for this new slot, not to mention the incompatibility with many of the large coolers we often use today. It also would become even more complex when faced with another one or two extra HyperTransport lanes as found on Opteron 2xx and 8xx chips, respectively.

AMD made a compromise when they designed K8. On the one hand, the on-die memory controller improves latency by a huge amount and scales much better by completely eliminating the memory and FSB bottlenecks that Intel chips get in a multiprocessor environment. On the other hand, new memory interface = new socket, no way around it.

From what I understand, the upcoming Socket F Opterons will have over 1200 pins in their socket so as to allow both a direct DDR2 interface and FB-DIMM. If I understand FB-DIMM technology correctly, it should end this issue by providing a standard interface to the DIMM which is then translated for whatever type of memory is in use. Logically this will trickle down to the consumers in another generation. For the time being however, AMD has stated that the upcoming "AM3" processors will still work in AM2 motherboards, as they will have both DDR2 and DDR3 controllers.

Re:Socket consideration

[beezly]

Indeed, Socket F has 1207 pins. There are some snippets of information and some more links available at http://en.wikipedia.org/wiki/Socket_F. We're delaying the upgrade of our cluster to wait for Socket F systems to become available (so we can compare them against Intel's latest offering at that point).

Re:Socket consideration

[wysiwia]

As the person who responded to your last post explained, that's just not possible with the K8 architecture as it is. The memory controller is on-die and memory technology is evolving, therefore the interface between the processor (where the controller is) and motherboard (where the DIMMs are) must also change.

Yet that doesn't matter more than the last time you responded. It's no problem to merge a new core (or multi cores) with a memory controller for the 939 socket. It's not even a big problem to put several memory controllers on the same die and just connect the fitting one with the socket so only a single die is sufficient. There are no technical obstacles for AMD to provide new cores with old interfaces, it's just marketing considerations.

Besides I'm sure AMD already thinks about how to fix the socket AM2 low acceptance. It's a fact, customers are annoyed with AMD because of AM2 and it's market share will shrink unless AMD comes up with a solution fast. IMO the 939 is probably the cheapest at least on short terms.

O. Wyss

Re:Socket consideration

[tomstdenis]

First off, that isn't true. Things like Vcc sources have to move around to acommodate new designs. You're also disregarding the move to DDR2 which has a different interface as well.

You've been able to get 939 and 940 pin boards for a LONG WHILE [even now given AM2 is out]. Sure 754-pin has disappeared but AMD doesn't even sell 754-pin desktop processors anymore [laptops being the exception].

You might as well bitch out Intel for not being able to get Super Socket 7 motherboards anymore for your P54C processor.

Tom

Re:Socket consideration

[C_Kode]

A Meat Socket has a longevity of 5 years... Maybe thats what you need. :)

Re:Socket consideration

[kimvette]

The only people that can truly make use of multicore chips would be scientists and people who do any other kind of intensive parallel processing, like like graphics rendering.

The way people use PCs is drastically changing. Now SMP benefits any gamer, anyone transcoding video (not everyone does it? Uh, Windows Media, digital camcorders, Windows Media Center|MythTV|Other PVR app, and in the case of Windows users, running various spyware in the background without totally dragging down the system ;), and other now-not-so-cutting-edge activities which are rapidly becoming commonplace. Add to that image processing (digital camera apps, Photoshop w/plugins, etc.) which nearly everyone does, printing large photos on an inkjet printer, and now you've argued that 90%+ of home users can now gain (knowingly or unknowingly) from SMP.

Re:Socket consideration

[somersault]

Personally out of that I only play games and do photo editing, but I accept that more and more people will be able to 'benefit' from multicore, though the examples you list can all be done fine on a single core processor. When you want to do them at the same time, then you have more of an argument, but again most single core processors can handle multitasking/threading and have done so for over a decade.

And as for a spyware monitor, I don't even believe anyone should have to run one of those if they have a decently secure OS, application, and browsing habits.. :p

Re:Socket consideration

[wysiwia]

First off, that isn't true. Things like Vcc sources have to move around to acommodate new designs. You're also disregarding the move to DDR2 which has a different interface as well.

You haven't done any chip design, haven't you? What's the speed inprovement of DDR2 against DDR?

But that's all not the point, people are simply annoyed with AMD's socket policy, rightfully or not. Just read http://hardware.slashdot.org/article.pl?sid=06/09/ 29/0542214.

O. Wyss

Re:Socket consideration

[tomstdenis]

sockets change because the technology changes. A different die design may need more Vcc inputs, different memory technology may need more I/O pins etc. Also newer opterons are adding HT links which definitely require more pins.

Also DDR2 is just double-data rate memory [hence the name]. The diff between DDR1 and DDR2 is the electrical spec is different, the process is different, there are different memory commands and the frequency is higher.

In theory a dual-channel DDR2-800 should top out at a max of 12.8GiB/sec compared to 6.4GiB with current technology. Of course you'll only max out 60-75% of this [due to timing and overhead] but 75% of 12.8GiB/sec is still more than 75% of 6.4GiB/sec.

The fact that you can still buy 939-pin gear today even though the standard is AM2 for desktops means you should shut your whiny little trap. AMD has plans to include backwards compatibility in future sockets as far as I understand.

Finally, why do you need to upgrade everyday? If you have a 939-pin AMDX2 4800+ today chances are it'll still be a good processor 5 years from now. So if you treat your box right [e.g. put it on a UPS, clean out the dust, etc] it should last a long time. Hell, I have friends still running Pentium II boxes...

So quit your uninformed bitchy little rant and get real.

Tom

Re:Socket consideration

[drinkypoo]

Psssst P54C installed to socket 7, as did P55C. It's K6 that went into SS7, which would simply also accomodate P5[45]C.

Re:Socket consideration

[tomstdenis]

It's hyperbole for a reason.

Point is, you can still buy 939-pin boards today. So even though AMD is going through new sockets you're not stuck if you need a replacement.

Also keep in mind AMD is the company bringing on-board memory controllers, HT links and the like. HT v3.0 is around the corner and it promises higher bandwidth, lower latency and more versatility.

Sure Intel is stuck on 775 today [with no less than a 4 or 5 diff incompatible chipsets] but they're also the company NOT bringing you point-to-point links, integrated low latency memory controllers or real HPC numbers. So I don't know how much we should be cheering them on [*]

Tom

[*] Note that I think the C2D is really awesome core. It runs ice cold, can overclock like mad, undervolt at the same time and gets an IPC that rivals the Athlon64. If Intel could just figure out the server side of things they'd be a real competitor...

Re:Socket consideration

[aka1nas]

Where are you getting that AM2 has low acceptance? Enthusiasts aren't happy with it as it's not faster than 939 and Conroe outperforms it. Dell and the other OEMs don't seem to care as it's still fast enough and cheap. AMD still has pretty firm control over the value segment as the K8s are preferrable over the netburst junk Intel is still trying to sell off.

Re:Socket consideration

[The_Dougster]

Where are you getting that AM2 has low acceptance? Enthusiasts aren't happy with it as it's not faster than 939 and Conroe outperforms it. Dell and the other OEMs don't seem to care as it's still fast enough and cheap. AMD still has pretty firm control over the value segment as the K8s are preferrable over the netburst junk Intel is still trying to sell off.

I went with the AM2 when I recently built up a cheapie X2 3800 box for home. Maybe its not right there at the bleeding edge of technology and performance, but it is a nice upgrade over my former XP-2400+ system which I passed over to the wife.

For the (roughly) $500 that it took to assemble it, I got a Athlon64 X2 3800+, 1G DDR2-800, 320G SATA-2 drive, GF 7600GT PCIE, DVD burner, Socket AM-2 motherboard, power supply, and case. That is hard to beat for the price IMO, and it should be upgradeable to one of the Barcelona quad cores when they become available.

Whatever, it seems to be running well at least :-)

Re:Socket consideration

[kimvette]

Considering that your average "Dude, you're getting a $299 Dell" special comes with Spyware preloaded, Windows' being on entry-level budget systems where they are priced to the point where the kindly-preinstalled spyware subsidizes the cost of the machine, I do not think that Linux is going to be much of an option for most Joe Sixpack-type consumers, and with the upcoming vista, unfortunately, it is likely that big-box-branded preinstalled spyware is going to be a whole lot more difficult to remove.

Also, if you do photo editing, if you have sprung for Photoshop and use "droplets" (basically, Photoshop macros tied to a desktop icon), SMP will cut down processing time considerably. If, for example, you went on a week-long trip with your family to Disney (I know, this is slashdot, so the likelihood of having a spouse is infinitely close to zero, hahaha, I had to toss in a /. cliche in here somehow) and want to downsample your 6MP photos to something web-friendly, a droplet makes the job much easier, faster, and SMP will make the job go twice as fast, or 3x faster if you have a four-way box combined with a decent HDD.

Re:Socket consideration

[Emetophobe]

The only reason AMD switched from Socket 939 to Socket AM2 was because they were moving from DDR1 to DDR2. You can't stick DDR2 in your DDR1 motherboard, so you would need to buy a new motherboard that supports DDR2 anyway. The new socket AM2 is still 939 pins, they just changed the layout of the pins so that people don't try to put an AM2 CPU in a S939 motherboard. That's the downside of AMD having the memory controller on the cpu.

I bought a socket 939 system this spring knowing AM2 was coming out in a couple months. I rather have older, more tested tech, then brand new bleeding edge tech that hasn't had the time to mature. In a couple years when Conroes have matured and the motherboard line up is better, I will probably switch back to Intel.

Re:Socket consideration

[somersault]

I use the GIMP actually ;) Oh you mean an icon where you can drag and drop to perform a function. I remember dragging and dropping onto apps on our Mac Classic, was very handy..

Reminicient of the early dual-cores?

[TheThiefMaster]

"but AMD claims that its quad core is true quad core, while Intel's is two dual-cores grafted together."
I'm sure that this was true about intels early dual-cores, if they've done it again then we can expect some truely awful performance from their quadcores...

Re:Reminicient of the early dual-cores?

[masklinn]

It was indeed, Intel didn't have integrated ("true") dual-core (AMD-style) before the Core architecture. Pentium-D's two cores, for example, had to use the FSB to communicate with one another, they didn't have a specific, fase, core-to-core bus. In the end, they were no better than regular dual-core, except that you only needed a single socket.

Re:Reminicient of the early dual-cores?

[deadline]

And in some cases the Pentium D performs quite well

We have been using dual processor systems for a while. No one really complained that they were not "dual core" and got quite a lot of work out of them. Gluing two cores together was quicker and easier than a true dual core, and if it is the same price as a single core, I'll take it.

Re:Reminicient of the early dual-cores?

[civilizedINTENSITY]

"I think you meant to say no better than regular dual processors as dual cores are better with the inter-processor communication."

Actually what he said was

Pentium-D's two cores, for example, had to use the FSB to communicate with one another, they didn't have a specific, fast, core-to-core bus.

The was no "inter-processor communication" except for the *same* communicatin path that two physically separate CPUs would have to use: the FSB. A strange way to design a system unless you are only competing against dual single core CPUs.

Quad-core vs. dual-dual-core?

[Kopretinka]

Can anyone please shed some light on the difference (for the user) between a true quad-core and a dual dual-core processor? I expect a quad-core can be cheaper because it is more integrated, but is that it?

Re:Quad-core vs. dual-dual-core?

[glwtta]

Probably has to do with memory/cache access and total available bandwidth between the cores. Memory architecture seems to be the one area where the Core still can't touch the Opteron.

Of course, I'm just guessing.

Re:Quad-core vs. dual-dual-core?

[Phleg]

A "true" quad-core means that all of them share the same L2 cache, AFAIK. Basically, performance benefits as they can all use the same high-speed memory cache for L1 misses. This is also extremely useful in the case of multiple processes which aren't bound to a CPU. If process A is scheduled on processor 1, then 2, then 3, then 4, there are going to be a lot of cache misses (since it's in no CPU's L1 cache). With two dual-cores bolted on to each other, processes switching from processors 1-2 to 3-4 are going to incur severe performance penalties as any relevent memory is fetched over the memory bus from RAM.

Re:Quad-core vs. dual-dual-core?

[Phleg]

As a silly analogy, imagine two cars strapped to each other versus a single car with dual engines but lots of shared components where it makes sense to do so. The one that actually had some engineering and design behind it will likely make better use of resources, rather than the ad-hoc, bolted-together solution.

Re:Quad-core vs. dual-dual-core?

[DohnJoe]

actually, there's three levels of cache in the opteron: L1 and L2 are CPU bound, L3 is shared.
They claim that this improves performance with virtualization


From the article:
Barcelona uses a three-stage cache architecture. The L1 cache is 64KB, the L2 cache is 512KB and the L3 cache is 2MB. The L1 and L2 caches are dedicated to a particular core, while the L3 cache is shared among all cores. Note that the L3 cache has been engineered to be variable in size, so that different products may offer different L3 cache sizes. The L1 and L2 caches are exclusive, as with current Opterons and Athlon 64s. This means that the L1 and L2 cache don't hold copies of the same data.

Re:Quad-core vs. dual-dual-core?

[tomstdenis]

As others pointed out, inter core communication has to hit the FSB. That makes things like owning/modifying/etc cache lines slower as you have to communicate that outside the chip.

There are also process challenges. Two dies take more space than 4 cores on one die since you have replicated some of the technology [e.g. FSB interface driver for instance]. Space == money therefore it's more costly.

If one dual-core takes 65W [current C2D rating] than two of them will take 130W at least [Intels ratings are not maximums]. AMD plans on fitting their quadcore within the 95W enveloppe. Given that this also includes the memory controller you're saving an additional 20W or so. In theory you could save ~55W going the AMD route.

Also currently, C2D processors have lame power savings, you can only step into one of two modes [at least on the E6300] and it's processor wide. The quad-core from AMD will allow PER-CORE frequency changes [and with more precision than before] meaning that when the thing isn't under full load you can save quite a bit. For instance, the Opteron 885 [dual core 2.6Ghz] is rated for about 32W at idle down from 95W at full load. I imagine the quad-core will have a similar idle rating.

Tom

Re:Quad-core vs. dual-dual-core?

[automattic]

Hey, if those cores were boobs, I'm almost positive you could pick up the chick with 4 teets (have over 2 and they're not considered boobies anymore, bub) a lot cheaper than scoring two racks of twins.

Nobody wants her, and you get to reap the rewards of extra mammary performance(or is that memory?)

Re:Quad-core vs. dual-dual-core?

[smithmc]

  As a silly analogy, imagine two cars strapped to each other versus a single car with dual engines but lots of shared components where it makes sense to do so. The one that actually had some engineering and design behind it will likely make better use of resources, rather than the ad-hoc, bolted-together solution.

Meanwhile, Intel got their two-cars-strapped-together out first, thus meeting the needs of some people who might've needed a solution like this, and will have a huge market lead on AMD by the time they get around to releasing Barcelona.

Re:Quad-core vs. dual-dual-core?

[dfghjk]

"As others pointed out, inter core communication has to hit the FSB."

Whereas AMD's cores all share a single on-die memory controller. Just because AMD has HT and a memory controller built in doesn't mean that it has a significant advantage. In a multi-CPU system it's a different story.

"Two dies take more space than 4 cores on one die"

The aggregate size of the dies is meaningless. Intel's design requires two dies but they are cheaper, perhaps even less than half the cost, of AMD's die. Cost is proportional to die size and large dies get very expensive.

"Space == money therefore it's more costly."

Absolutely incorrect.

"AMD plans on fitting their quadcore within the 95W enveloppe."

Just how exactly does a single die uniquely enable this?

As for the power claims, the question is whether a single quad die has inherent power benefits over two dual dies. Intuitively the answer is little to none. AMD may or may not have real power advantages over Intel but may things go into that, the least of which is how many dies on the chip. In order to add cores without adding to peak power consumption, design work has to be done.

Re:Quad-core vs. dual-dual-core?

[tomstdenis]

I don't know where you went to school but slapping two dies on a chip instead of putting one moderately larger [than a single] die is going to cost more. There is going to be more die surface area when you have two independent processors on the chip since you duplicate a lot of house keeping [e.g. FSB interface, clocks, etc]. And space is money since it limits the number of dies per wafer and ultimately their yields.

Suppose you're right and there is no advantage to using a dual-die approach. Why doesn't AMD just HT link two dies on one chip then? The truth is the quad-core layout and design is the more optimal approach over dual-die processors.

Also the cores can talk over the internal SRQ instead of polluting the HT links. Simply put, there is a lot of memory bandwidth in an AMD processor. You have the three HT links running at 4GB/sec each, then the memory interface at 6.4GB/sec.

Tom

Re:Quad-core vs. dual-dual-core?

[dfghjk]

"I don't know where you went to school..."

Apparently where you went to school they teach you to lead with an insult.

"...slapping two dies on a chip instead of putting one moderately larger [than a single] die is going to cost more."

Yes, but 2x larger is more than moderate, especially when a 2x larger die may cost 5x to make due to yields. May you should retake that class.

"There is going to be more die surface area when you have two independent processors on the chip since you duplicate a lot of house keeping..."

A modest amount of die space. Irrelevant anyway. What matters is the cost of two smaller dies versus one larger one.

"And space is money since it limits the number of dies per wafer and ultimately their yields."

You got that right. Too bad you don't understand that smaller dies improve yields.

"Suppose you're right and there is no advantage to using a dual-die approach"

You mean disadvantage. Perhaps AMD doesn't have access to MCM tech.

"The truth is the quad-core layout and design is the more optimal approach over dual-die processors."

Technically it is, but economically most likely not. Furthermore, there is a time to market advantage. Over time, Intel will have a single die solution as well.

"Simply put, there is a lot of memory bandwidth in an AMD processor. You have the three HT links running at 4GB/sec each"

How much more? HT links are not memory bandwidth.

Re:Quad-core vs. dual-dual-core?

[tomstdenis]

I think you are missing something...

Who is saying a quad-core is larger than two dual-core dies? What I am saying is a dual-core die takes $X mm^2, a quad-core takes $X+$Y mm^2 and that $Y $X.

Otherwise, why would Intel EVER move to true dual-core or quad-core? Why not just put 4 dies on the processor?

Also HT links ARE memory bandwidth. I can access memory from another NUMA node via HT while simultaneously performing an operation in my process local node. ...

Don't bother replying. I don't really care that much for what you have to say.

Tom

Re:Quad-core vs. dual-dual-core?

[dfghjk]

It isn't, but it is roughly twice the size of a single dual-core die. The cost of such a die would be 2x IFF the die size were reasonable, but at the sizes these processors use the actual cost will be much more. Designers of these processors have to balance die size with overall cost all the time and it's ludicrous to assume that the quad-core die will always be cheaper. Until process improvements support quad-core, the opposite will likely be true.

"Otherwise, why would Intel EVER move to true dual-core or quad-core?"

Because process shrinks enable Intel to pack more transistors on a die. In order to use those transistors they add more cores.

"Also HT links ARE memory bandwidth. I can access memory from another NUMA node via HT while simultaneously performing an operation in my process local node. ..."

No, they aren't. HT is a point-to-point link. I a UP quad-core AMD system there is only one memory controller. You can add all the HT links to it that you'd like but you don't increase memory bandwidth one bit.

"
Don't bother replying. I don't really care that much for what you have to say."

What's the matter, Tom, afraid you might learn something?

Re:Quad-core vs. dual-dual-core?

[tomstdenis]

At least for AMD they are going 65nm in their first quad-core so the die size will be comparable to the dual-core. HT links are used for PCI/PCIE devices too. Even in a single processor box you benefit from HT. That said in server setups HT links are gold because they are used for NUMA.

Anyways, I don't even know what we are discussing anymore. All I'm saying is dual-die == dumb, wait for a properly designed quad-core. That doesn't mean specifically AMD but hey if they hit it first all the power to them.

Intel's "me first" bullshit is not something professionals should be looking at. They're usually incomplete [e.g. reports that the power states are not fully supported in the dual-die are already coming out] and short lived. Their dual-core Core 2 Duo series are nice and worth a look if you are building a new box, but I'd stay away from the dual-die.

Tom

Re:Quad-core vs. dual-dual-core?

[dfghjk]

"At least for AMD they are going 65nm in their first quad-core so the die size will be comparable to the dual-core"

Intel will do the same with their next generation of course.

"HT links are used for PCI/PCIE devices too. Even in a single processor box you benefit from HT."

Yes, IO uses HT but that's not memory bandwidth either.

"That said in server setups HT links are gold because they are used for NUMA."

Yes, and in larger MP machines AMD has a definite advantage. This is a discussion of a single processor, though, and whether one die or multiple dies matter. There is only one memory controller in such a system, no NUMA. In that context HT is not a performance advantage. AMD's integrated memory controller (most likely) is but, again, that doesn't effect the one-die/two-die discussion.

It seems this argument for you is about Intel vs AMD, not one die vs MCM. Shame you can't separate the two.

"All I'm saying is dual-die == dumb, wait for a properly designed quad-core."

Yes, and you're wrong about that. MCM allows Intel to bring quad-core to market sooner. That's why it's used. "Properly designed" just shows your prejudice. A properly designed processor is one that meets its market objectives effectively and if using a single die makes you a year late to market then it's not "properly designed".

"They're usually incomplete [e.g. reports that the power states are not fully supported in the dual-die are already coming out] and short lived."

Blah blah. The part isn't out yet. Such reports are from people like you who are preconvinced that Intel's product can't be good. When it comes out it will be compared to what else is available at that time and there will be no AMD quad-core. Future products will always be better.

"...but I'd stay away from the dual-die."

and I'm sure you will. For the customer who may buy such a processor, he won't care whether it is dual-die or not. All that matters is how fast it is and how much it costs. The rest of the discussion is meaningless.

Re:Quad-core vs. dual-dual-core?

[tomstdenis]

Um, actually PCI/PCIE traffic ***IS*** memory bandwidth. Most devices on the market are memory mapped, and even those which art port based still use the same god damn bus. That all this is moved to an HT link and off the memory bus means you have more bandwidth.

Since you can't sort out even that little detail I'd like to just assume you're a clueless newb. Go hide now.

Tom

Re:Quad-core vs. dual-dual-core?

[dfghjk]

Again with the personal attacks. I guess it upsets you when you can't win an argument outright. Care you measure your technical penis with me, Tom? What are your qualifications?

Yes, I/O peripherals are frequently memory mapped, but even IO cycles are performed over the bus on Intel (vs HT on AMD) so it doesn't really matter what kind of cycles they are. What does matter is how much total bandwidth is used to program hardware. The answer, of course, is that it's a trivially small amount compared the the GB/sec numbers of overall bandwidth that the bus provides. It's a drop in the ocean compared the bandwidth differences between the memory controllers themselves. I'm sure you know that seein's how you're so technically elite. Care to continue your absurd defense of the value of HT in a single processor system?

The bulk of IO itself will occur using DMA to and from physical memory. The other stuff doesn't matter. In a typical PC, there will be 1 or 2 hard disk drives (100 MB/sec each), and ethernet port providing 10 or 100 MB/sec, and some low speed stuff. If the system optimistically can generate 200MB/sec of sustained DMA for IO, what fraction of that will be setup cycles for the hardware and how can that possibly make a blip on a 5GB/sec memory bus?

Of course, HT itself is and argument between Intel and AMD, not between single die and MCM. You've conceeded that you know nothing about the single die issue and are clinging to the idea that you can win the HT argument. HT is necessary for AMD because of the integral memory controllers and it's an architectural win in multiprocessor systems. It has no impact on the internal design of a single processor.

Re:Quad-core vs. dual-dual-core?

[tomstdenis]

If you can both fetch data/code for the processor to crunch and feed your graphics card with GL commands then you win. It's not always about bandwidth but about bus availability. For instance, most processors go far underutilized as it is. But in the "crunch" times latency is king and being able to return a result quickly is more important. So moving the I/O devices off the memory bus is a good thing in terms of bandwidth and also latency.

As for your DMA comment it's actually better now. You can read from memory and feed to the I/O at the same time.

As for the dual-die issue, there is going to be redudant logic in the 2nd die. That's unavoidable. It's another independent processor. Moving the cores, cache, APICs and etc into a single die will save on some of the front end. If there is an internal crossbar between the cores [e.g. not the FSB] then it's even better.

The only benefit for Intel of the dual-die is so they can beat AMD to the punch with an inferior product. I mean it doesn't even have proper power saving support yet. You can buy that crap if you want. But personally I'll wait for a properly worked out quad-core before I decide to take the leap.

Tom

Re:Quad-core vs. dual-dual-core?

[Emetophobe]

The common perception is that AMD's quad core approach is a true quad core chip designed from the ground up, where as Intels quad core is 2 dual core Conroes slapped together in one package. That's pretty much the jist of it anyway.

Re:Quad-core vs. dual-dual-core?

[dfghjk]

"So moving the I/O devices off the memory bus is a good thing in terms of bandwidth and also latency."

Only if there's something significant to move. Memory latency is a strong point with the AMD design but not because of HT.

"As for your DMA comment it's actually better now. You can read from memory and feed to the I/O at the same time."

If by that you mean that you can do memory cycles and do hardware programming on the bus concurrently then no you can't and no it doesn't matter. Both contend for the HT link and the IO programming is of no performance consequence. Regardless, the DMA in my example only accounts for 3 or 4 percent of total memory bandwidth.

"As for the dual-die issue, there is going to be redudant logic in the 2nd die. That's unavoidable. It's another independent processor. Moving the cores, cache, APICs and etc into a single die will save on some of the front end. If there is an internal crossbar between the cores [e.g. not the FSB] then it's even better."

Yes, and how much die size penalty is that? How much difference in cost does it make? You act like that matters so I'd like to see some numbers. The redundant logic you refer to is minor.

"The only benefit for Intel of the dual-die is so they can beat AMD to the punch with an inferior product."

Inferior to what? AMD will have nothing at all in that time frame. How's that for inferior!

"I mean it doesn't even have proper power saving support yet."

What is proper? Anything you've had before?

"You can buy that crap if you want. But personally I'll wait for a properly worked out quad-core before I decide to take the leap"

There you go again. "Proper" according to you is whatever AMD says it despite your inability to quantify just what makes MCM improper. I'm sure Intel will weep at your lost sale.

Re:Quad-core vs. dual-dual-core?

[h8macs]

For the user!?

Simple... their chain-letters can be forwarded MUCH faster. The average windows computer will then be able to much more quickly communicate with zombie masters. Your system will help to spread virii on a more vast scale in less time throughout p2p network sharing applications.

Users REALLY do not need more than a single core... and MOST do not need the speed they have. Hell I know developers that have MORE than they really need, or even really know how to use... Considering I still have my 200mhz in service... and I am a Systems Administrator by day... Gigolo by night... ;-)

Now who out there might need more cores than a gigolo!? hehehe rendering, servers and distributed computing is really what the dual/quad/octal cores are for. All else... well you may want to look at your software for the bottle-neck...

Stable hardware platform

[beezly]

You want the latest and greatest features, but you aren't willing to cope with changing your hardware to keep up?

CPU manufacturers don't change interface designs for fun. It costs them time and money to design a new interface. They do it because the market demands new technology.

Besides, looking at recent history, Socket A, 940 and 939 have had roughly 3 years. Socket 754 was a red herring that no one in their right mind should have bought if they were looking for platform longevity.

If you compare AMD's socket strategy to Intel's (http://www.tom.womack.net/x86FAQ/faq_time.html), AMD look pretty good at developing platforms with good "socket longevity"

Re:Barcelona

[Plammox]

In fact, I'm told that in the Catalan translation of Fawlty Towers Man(u)el says he's from Madrid.

But since this is an early eighties UK tv production, they surely didn't get all the Spanish vs. Catalan cultural differences right.

Re:Barcelona

[pubjames]

And the Spanish translation of Faulty towers had Manuel coming from Mexico I believe.

The MARKET demands? Your joking, right?

[PopeRatzo]

CPU manufacturers don't change interface designs for fun. It costs them time and money to design a new interface. They do it because the market demands new technology.

Show of hands: Who's been demanding new CPU technology? What percentage of the "market" has already gone to dual-core, and is clamoring for quad-core to run their apps?

You don't think maybe a manufacturer would push new technologies out the door to get new sales do you? "..the market demands.." my ass.

Re:Barcelona

[clickclickdrone]

>early eighties UK tv production
1975 & 1979 actually. God I feel old.

I guess you won't buy Intel either...

[Visaris]

I won't buy any AMD processors anymore until AMD clears its socket plans and guaranties a minimum of 3 year availability for processors on a socket.

I suppose that means you won't buy an Intel chip either. Look at what happened with Conroe. Core 2 Duo uses a socket with the same name as the P4 socket, the same number of pins too. But guess what? When Conroe came out there were less than a handful of reasonable boards out of the hundreds of models out, that would actually support it. The voltage requirements changed slightly, the BIOS requirements changed, and the end result was that upgrading to Conroe on a given board was hit or miss. I fail to see how Intel's MB upgrade situation is any better than AMD's. It sounds to me like you're falling for Intel's game: "We kept the socket name and number of pins the same, so that means we have better socket longevity." Sorry, but I'm not falling for it. I've read too many horror stories on the forums from Conroe upgraders that thought they could use their current P4 boards.

Don't get me started on Intel's TDP scam either (AMD's = max, Intel's = average). AMD may not always have the best tech, but I find them to be a much more straight-forward company, with fewer sneaky games designed to trick customers.

And why are we posting a story about AMD's tech said/written by an Intel employee? Sounds like it was biased before it even started to me.

True QC versus MCM:

[Visaris]

Intel's QC is really an MCM, or multi-chip-module. That means they have literally grabbed two Conroe (Core 2 Duo) chips off of the assembly line, and mounted them in a single package. From the outside it looks like a single chip, but inside, it has two, separate peices of Si, connected over the FSB. That is the problem: the two chips are connected to the same bus. A single chip presents one electrical load on the bus, and two chips present two loads. This means that the speed of the bus needs to be dropped. That is why kentsfield will have a slower bus speed than normal chips. If you think about it, this is the exact opposite of the situation you want. You have just added a core, so it would be nice to add more bus bandwidth. Instead, the Intel solution lowers the overall bus bandwidth, not to mention that it is a shared bus. The two cores fight each other over a very slow external bus, and this creates a performance bottleneck.

When all four cores are on a single peice of Si, all sharing a L3 cache, the chips don't need to fight over the external bus as much. The cores can share information between them internally, and do not need to touch the slow external bus to perform cache coherency and other synchronization. Also, true QC chip presents one load to the outside bus. This means that the bus speed does not need to drop because of electrical load.

There are many people who don't care how the cores are connected as long as the package works. The point is that the way the cores are connected have a direct impact on performance. We'll be talking about Intel vs. AMD cache hierarchy in 2007 when AMD uses dedicated L2 and shared L3 while Intel uses only shared L2. Expect cache thrashing on Intel's true QC chips with heavily threaded loads when it comes out. Next I'll hear people say that the cahce doesn't matter as long as it works. As long as it works for what? Single-threaded tiny-footprint benchmarks like SuperPi or Prime95? How about a fully threaded and loaded database or any other app that will actually stress more than the execution units?

Re:True QC versus MCM:

[dfghjk]

"What we are seeing with real applications under load and Intel's initial "quad core" pre-release hardware is an average 55% benefit over a shipping dual core when clocked 250mHz higher."

Who's "we" Mr. AC? Show me your test results.

"The Intel tech is being hampered by the MCM choice in their initial release."

Proof please.

"The fact is MCM as Intel has implemented it is hampering a chips potential the only question is how much."

The answer is close to 0. Prove me wrong.

"...you even conceed will get around to dropping MCM and going for a "true quad core chip" later."

No need to concede that. Intel and AMD have both made it abundantly clear that their focus will be more cores per die and per chip on future generations. It's not just AMD's idea and it's not to overcome any MCM problem. It's what is necessary to take advantage of new processes.

"AMD has NDAs out but some have said that even their initial AM2 based quad core chips like Barcelona get an 128% average benefit over dual core chips clocked 125mHz higher."

Of course they do, but that can't be verified because it's under NDA which you've apparently violated. In other words, it's bullshit.

"Replace "true" with "a supplier of well engineered products and technological leader..." ...who will bring their product to market about the same time as their competitor's second generation, single die product. If AMD did their own process engineering then they wouldn't be a generation behind Intel, would they?

Re:1+1+1+1==2+2==4

[automattic]

Think of it in terms of a hand of poker.

4 of a kind clearly beats 2 pair.............

I'm sure someone at AMD thought this up on poker night.

Intel obviously likes to keep things in pairs......like boobies, and AMD has decided to steer clear of grafting on another pair of mammary cores, er, glands.

Re:The MARKET demands? Your joking, right?

[saider]

You don't think maybe a manufacturer would push new technologies out the door to get new sales do you? "..the market demands.." my ass.

It is just an expression. It is a shorter way of saying "If a compelling product is not produced the market will preferentially buy a competitior and drive business down."

The market will buy dual-core if they think there is a benefit. Even if they have no need for the feature, if they think it will benefit them, they will buy it. Technically not a demand, but close enough from the point of view of the producer.

Re:The MARKET demands? Your joking, right?

[beezly]

Maybe I'm biased. I work in the High Performance Compute sector and we can never get enough CPU cycles!

Re:Question

[SirTalon42]

I believe all the Core 2 chips support Vanderpool already, and from AMD all the AM2 socket using chip support Pacifica IIRC. Though the IOMMU thats going to be in future AMD chips is what will REALLY help virtualization (i.e. so unmodified drivers can use DMA without any risk of breaking out of the VM).

Re:I'm sorry, but..

[MrNaz]

Eh? nVidia have a well supported chipset for AMD CPUs which has been in use for several years now.

Re:1+1+1+1==2+2==4

[MrNaz]

That's true, but only for extremely large values of 2.

One page version of article with no ads

[ubuwalker31]

http://www.extremetech.com/print_article2/0,1217,a =191008,00.asp

Well, Yeah...

[Nom du Keyboard]

Loyd Case considers that the shift isn't as major as Intel's move from NetBurst to Core 2,

Well, yeah. AMD was starting with a superior processor to NetBurst to begin with. If they haven't advanced as far over their previous designs as Intel has, perhaps it's because they didn't have as far to go to start with. Pretty stupid remark overall by Loyd IMHO.

48-bit addressing = 256TB Memory

[Nom du Keyboard]

the memory controllers now support full 48-bit hardware addressing, which theoretically allows for 256 terabytes of physical memory.

I always felt the IBM AS400 had a nice scheme with its revolutionary large address space at the time. Not only did every byte -- possibly even bit -- of main RAM memory have a unique address, but so did all the attached mass storage devices. With this type of addressing, one could bring that same type of architecture to the desktop.

Do people often open their cpus?

[OrangeTide]

Are you going to crack open your quad core Intel and suddenly be disappointed because they didn't put all the cores on one wafer of silicon?

I think two dual cores in one socket is as good as one quad core in one socket, everything else being equal.

I think we should complain to AMD that they didn't do eight cores, because intel already made four cores (true or not). AMD is just throwing around insults to cover up the taint of being a "me too!"

In the end I think we will just let sales numbers talk for themselves. And unless AMD's offering is significantly cheaper or better or something, the extra time Intel will have on the market will make them the clear winner in this competition.

Re:Do people often open their cpus?

[P3NIS_CLEAVER]

Consumers don't care but analysts that help set buy ratings on their stock care.. AMDs solution will cost less to produce, make less heat and perform better. Which would be a better bet in the stock market in the long run?

Re:Do people often open their cpus?

[OrangeTide]

Multiple chips in a single package has the advantage that you can manage yield problems much more effectively.

Share prices only indirectly impact a company (when they go for more funding). AMD doesn't make any money when thier share price shoots up, they make money when they sells lots of chips without spending too much money.

Re:Do people often open their cpus?

[tuxicle]

Multiple chips in a single package has the advantage that you can manage yield problems much more effectively.

How so? I thought you can't test most of today's chips until they're packaged. If there's two dies in the package and one is bad, both must be thrown away. Sounds like worse yield to me.

Also, if there's two chips in an MCM, shouldn't their power dissipation be equally matched, to prevent them from shearing apart due to uneven heating?

Re:Do people often open their cpus?

[OrangeTide]

you're right. you can't perform full diagnostics on the chip before it is packaged. tiny contact points that don't have wires yet and the chip would probably just go up in a puff of smoke if it wasn't in a package. you still can do optical validation, not that is worth that much.

I didn't consider the shearing from heating, how much you want to bet that there are patents for dealing with that.

cheaper prices are probably found in the volumes you do rather than any minor differences (like two wafers versus one big on in a package).

for either one if one to three cores have crapped you can still sell it as a single or dual core device. I would imagine. but likely nobody's yield would be so low for that to make sense. And you would probably would not want to mix real single/dual cores and quads that have been refurbed. (who knows, maybe Intel will try that)

I want an UltraSparc T2 soo bad. I wish they would make a little affordable devel workstation so us normal people could have one to devel on. Finally a machine powerful enough to run Java. (can execute 64 threads) ... even an UltraSparc T1 pulls 70 watts with eight cores. (32 threads)

"True" quad-core?

[Anonymous Freak]

AMD claims that its quad core is true quad core, while Intel's is two dual-cores grafted together.

Who cares? Really.

If Intel has quad-core released within the next two months, and AMD's quad-core solution is more than 6 months away, this just sounds like sour grapes. I'm not an Intel fan-boy or an AMD fan-boy. I follow processor because I like technology. AMD had the best processor before Core 2. Now Intel has the best. AMD is running behind Intel, and it looks like they will drop further behind in the next year. I have no doubts that AMD will recover, and in a few years, retake the performance crown. But they don't have it right now.

As I said, who really cares if it's four cores on one die, or two two-die cores? Only the most extreme zealots will give a damn. And they were already AMD fanboys, just looking for an excuse to decry Intel's offerings. Yes, eventually, AMD's quad-core offerings will most likely be superior to Intel's. But when they're 6 months (or more) behind, does it matter?

Does it matter that GM will be releasing a super-environmentally-friendly fuel cell car in five years when you can get a Prius now?

Bring some other things inside microprocessors

[v4vijayakumar]

Why adding more and more cores? Add other things too inside the microprocessors, like, memory, etc. Why not SOC?

Intel attempts emasulation while AMD Innovates!

[h8macs]

Rule of thumb I have used... stick with Abit motherboards I've had no issues either chipset (nVidia/VIA). For a second choive I go with Gigabyte, third MSI.... I typically do not look at others unless I need a dual-proc or more board.

I've had no major issues with linux on amd based chipsets since the old K6-2 days. The K6 series were less expensive and in that day with linux I could get equal to or better performance over their Win-Tel comparables. Unfortunately the bestbuy managers that were employed as middle managers in many of the "high tech" companies in the mid-late 90's were too sold on catchy names a swag...

Athlon was AMD's major leap forward and the gamers pounced, AMD64 was another layer and smart IT professionals valued the more affordable 64-bit procs... now once again AMD is poised to innovate... (though most manufacturers are still stuck on the OLD pentium buzz word some are throwing AMD procs in their boxes)...

Re:The MARKET demands? Your joking, right?

[PopeRatzo]

beezly, YOU deserve all the cycles you want. But you are hardly the "market" they're talking about. If they'd said "beezly demands..." or "The High Performace Compute sector demands..." I wouldn't mind so much.

I remember setting up a Beowulf cluster for some friends in Linguistics. They needed a sick amount of computing power and just couldn't afford a Kray.