AMD Unveils Barcelona Quad-Core Details 206
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."
Memory Controllers (Score:5, Funny)
256 terabytes should be enough for anybody.
Re: (Score:3, Informative)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
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
wha? (Score:3, Funny)
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.
On snap! (Score:5, Funny)
BUUUUUUUUUURNED
Next week: Intel responds by telling us how fat AMD's mother is.
Re: (Score:2, Funny)
Mine goes to eleven (Score:2, Funny)
Wake me up when they have a processor that goes to eleven.
Re: (Score:2)
Re: (Score:2)
Re: (Score:3, Funny)
Intel's response? (Score:2)
How obscure is this? (Score:2)
It's the processor that runs like a dog with no nose!
Once again... (Score:5, Insightful)
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: (Score:2)
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 need
Re:Once again... (Score:4, Informative)
Re: (Score:2)
... 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... (Score:4, Informative)
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: (Score:3, Insightful)
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 th
Re: (Score:2)
You are wrong about the
Re: (Score:2)
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
Re: (Score:2)
"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.
Note to AMD: We don't care (Score:5, Insightful)
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 (Score:5, Informative)
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: (Score:2)
Re: (Score:2, Informative)
Re: (Score:3, Insightful)
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: (Score:2)
I think you chose a bad example. Everyone knows red ones go faster!
Re: (Score:2)
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: (Score:2)
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: (Score:2)
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 wou
Re:Note to AMD: We don't care (Score:5, Insightful)
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: (Score:2)
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
Re: (Score:2)
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: (Score:2, Informative)
Re: (Score:2)
Re: (Score:2)
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 (Score:4, Interesting)
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: (Score:2)
Re: (Score:2)
"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,
Amazing analysis (Score:5, Funny)
Intel: 4=2x2
Where do they hire these guys?
-Nano.
Re: (Score:2)
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. (Score:3, Informative)
Socket consideration (Score:2, Insightful)
O. Wyss
Re:Socket consideration (Score:4, Insightful)
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
Re: (Score:2)
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
Re: (Score:2)
And as for a spyware monitor, I don't even believe anyone should have to run one of those if they have
Re: (Score:2)
Also, if
Re:Socket consideration (Score:5, Informative)
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: (Score:2)
Re: (Score:2)
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 sever
Re: (Score:2)
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 proce
Re: (Score:2)
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 [slashdot.org].
O. Wyss
Re: (Score:2)
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.
Re: (Score:2)
Re: (Score:2)
Point is, you can still buy 939-pin [shoprbc.com] 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-p
Re: (Score:2)
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 performa
Stable hardware platform (Score:2)
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 soc
The MARKET demands? Your joking, right? (Score:2)
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: (Score:2)
I guess you won't buy Intel either... (Score:5, Informative)
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.
Re: (Score:2)
Quad-core vs. dual-dual-core? (Score:3, Interesting)
Re: (Score:2)
Of course, I'm just guessing.
Re:Quad-core vs. dual-dual-core? (Score:5, Informative)
Re: (Score:2)
Re: (Score:2)
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 t
Re:Quad-core vs. dual-dual-core? (Score:5, Informative)
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: (Score:2)
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 proportion
Re: (Score:2)
Suppose you're right and there is no advantage to using a dual-die approach. Why doesn't
Re: (Score:2)
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 keepi
Re: (Score:2)
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 th
Re: (Score:2)
"Otherwise, why would Intel EVER move to true dual-core
Re: (Score:2)
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
Re: (Score:2)
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, t
Re: (Score:2)
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: (Score:2)
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 G
Re: (Score:2)
As for your DMA comment it's actually better now. You can read
True QC versus MCM: (Score:5, Informative)
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: (Score:2, Funny)
Nobody wants her, and you get to reap the rewards of extra mammary performance(or is that memory?)
Well, Yeah... (Score: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 (Score:2)
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? (Score:2)
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 themsel
Re: (Score:2)
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: (Score:2, Offtopic)
Re:Intel's Ben Sandler? (Score:4, Funny)
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 [slashdot.org] by entering your nickname, uid, or email address.
Re: (Score:2, Insightful)
Re: (Score:2, Funny)
Re: (Score:2, Funny)
There's a nice image to drink your coffee to...
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
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
Re: (Score:2, Funny)
(Manuel with thick Spanish accent:) Mr. Fawlty! I'm from Barcelona, I know *notheeeng*!
Re: (Score:2)
Re: (Score:2)
1975 & 1979 actually. God I feel old.
Re: (Score:3, Interesting)
Re: (Score:2)
And in some cases the Pentium D performs quite well [clustermonkey.net]
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: (Score:2)
Actually what he said was
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 again