AMD's Barcelona to "stink" in ORACLE - Page 3

Last message on previous page:

Indeed.

We have not seen any details about the microarchitecture yet, so we can't conclude.


How much more details do you know about Core 2?

Let me just ask you, does it use score boarding or reservation stations? How many entries of TLB does it have, with what replacement algorithm? Does it tag caches with virtual or physical addresses, and in what format does it store instructions in its L1 cache?

All you think you know about Core 2 are those marketing-oriented high-level specs - not much different from what we know about Barcelona.

How? Have you seen it? Do you know how it performs?

No, I haven't seen it, much like you haven't seen the shared L2 smart cache of Core 2. Oh yeah, do you think the latter a good design? Don't you?

BTW, Barcelona's L3 is a large shared victim cache. If you still don't know what I mean, then you're not yet qualified for the answer; sorry. :-p

cc-NUMA has nothing with the scalability of Barcelona! It has with the scalability of its platform.


And Barcelona works and supports such a platform, unlike any Intel x86 chips.

I don't believe to their undefined bold claims, until they support them with benchmarks.

Good, I already said you don't need to. Go get your candy now.

They are estimating too much, but they haven't provided any evidence or explanations about what they are estimating and how. For example I think that it is absolutely unreasonable to expect Barcelona to outperform Clovertown by 40%.

We will see whether you think is right or wrong when it's out. You better eat your own post when there are some apps that Barcelona outperforms Clovertown by 40%.

1. lower performance/clock
2. lower frequencies


1. With lower price and much lower power consumption, a 2% less performance/clock is nothing.
2. Be sure it will pick up shortly, and go higher than Windsor.

They promised Brisbane to be available in 2006, but they didn't kept their promise.

It is available, to the OEMs and the channel, but not to the end users. Did AMD call you and say it'll be available to you and your neighbors?

Yes, QuadFX(not 4x4) is a platform, C2Q is a CPU.

So you know, don't you?

Nope, both are for same purposes desktop/workstation PC's.

Still you claim the platform and the processor are for the same purpose? Do you know the difference between a platform and a processor?

So?

So what?

Yes.show us.

Yes.wow, something for you to brag about..

We haven't seen a review with 4 GPUs yet. Have you? What's the big deal about the 4 graphics cards, when only two can be uses in SLI ?
What's not the big deal to have 4 different displays from one machine?
Ah... I know, you really love to buy two or more C2D for the same purpose.

If you mean graphics and video editing, encoding and rendering, then yes. If you mean two 3D games running on two monitors, then no because no one can play two 3D games at same time.

Who plays 2 games at the same time?

What do you mean by hacking around for SLI?

I don't know. I usually find the word "hack" associated with C2D+SLI on the Internet.

I think that you haven't checked and you don't know what is Quad-SLI. Quad-SLI was introduced on Intel platform using Pentium EE CPU, the Dell XPS 600. And yes it is possible on a Core2 platform with 680i/650i chipset. So, you may go check. Here is a reputable link: http://www.nvidia.com/object/IO_28569.html

And that link refers to a Core 2 system? Are you sober?

Are you sure?

Do I have to be?

Oh, please enlight us with your definitions for mega-tasking. We just came from Mars and we just saw a PC for the first time.

How unfortunately, because if so you wouldn't understand. (Read: stop replying like a nutcase.)

Damn! You have no idea what are you talking about.....
1. QFX does not outperform Core2 Quad.

You are wrong. It depends on the application.

Are you sure the structure of 680i is stronger? Because if so, then you are again wrong.

I guess if that makes QFX (the platform) a failure compared to C2Q (the processor), then the whole Intel's Pentium-4 line is a failure, at its initial release, compared to both Athlon XP and Pentium-3.

I think that you are consuming too much AMD marketing. That has nothing with the real life.

I guess your understanding of real life is just too limited.

Because you have no idea how is NUMA working.

No, because I know what is NUMA properly supported, whereas you apparently don't.

Yes I can. But you can also learn by your self. Here is a nice paper:http://www.research.rutgers.edu/~w [...] ulz.pdf.gz.


Why don't you just admit that you can't answer? That paper does not compare memory access latency in multiprocessor NUMA and UMA at all.

If NUMA is properly supported, then remote memory access should be very rare, and mostly from local cache. The latency will be lower than a highly congested FSB.

Is the number ~70% pulled out of your ass?
100% *(1333 / 1067) - 100% = 24.93% != ~70%

You are just ignorant. In the first order, power consumption is proportional to square of frequency. 1.3*1.3 = 1.69.

We will see whether you think is right or wrong when it's out. You better eat your own post when there are some apps that Barcelona outperforms Clovertown by 40%.

Sciencemark and Cinebench


I guess if that makes QFX (the platform) a failure compared to C2Q (the processor), then the whole Intel's Pentium-4 line is a failure, at its initial release, compared to both Athlon XP and Pentium-3.

Pretty much everyone on these forums knows that.

They promised Brisbane to be available in 2006, but they didn't kept their promise.

Last I checked the calendar said 2007 today. And let's keep in mind:

From DailyTech

AMD senior vice president of technology development Douglas Grose claimed "We'll be producing early products probably in Q2 of 2008, with full production in the second half."

18 months. Minimum. That's a helluva long time in the CPU industry!

Hey, when was the last time you gave me a nice big kiss on my hairy a$$ and slipped me the tongue?

I don't think I ever did... maybe you should ask your dog or your neighbor's cat? I mean if you usually get that drunk... you could ask some others, too.

EDIT: about the "pretty much everyone" part, that sentence alone is cut out of its context. I was actually saying if everyone thinks Pentium-4 is a failure (compared to AXP and P3), then everyone is ignorant.

In the end, P4 proves to be a better performing processor than both AXP and P3. It does not scale, but some of its tricks and tweaks are useful in Core 2 and will be used in future Intel processors, too. Of course, those who can only comprehend simple arguments will always have a simple view of the world: Netburst sucks.

That's not what you wrote.

Yes, P4 was a complete failure at it's release. It was getting beat up by P3 Tualatins, and a 1.4GHz Athlon was giving the 2.0GHz P4 fits. Fully understood. Nothing out of context.

http://www.intel.com/design/proces [...] 248966.pdf

32 entry RS.

16 entries for inner level large/4kB pages, 256 entries for outer level large & 32 entries for outer level 4kB pages.

pseudo-LRU

physical

SFENCE

BS! We were talking about Barcelona's architecture, and your questions about Core2 have nothing to do with it!
Now when I answered your questions about Core2, answer me the same questions about Barcelona!

I wasn't talking stories about how Core2 shared L2 is a brilliant design, but you were about Barcelona's L3. So, if you know something about Barcelona's L3 shared cache, that have leaded you to your "brilliant design" conclusions, please share it with us.

Although Barcelona & Core2 shared caches are differently implemented, I would need to know some exact numbers about Barcelona's shared L3, before we can talk about which one is better. For example the latencies are important:

Large for K8. We'll have to see how large the 2MB are for Barcelona.

Oh, Mr. Smart@$$, I know what you mean. The victim cache....it will be shared between the cores as exclusive, proportionally to the needs of the cores.

BS! Honestly, I would recommend you to read more, but post less.

Damn, I thought that you are more intelligent.

Yeap, we'll see the same about you too.

You are missing the point of the forum.

So, "various workloads" is "some apps"?
So, what about the "some apps" in which Clovertwon will outperform Barcelona?

It is. Instead of improving the performance/clock, AMD degraded it.

I don't think so. On their roadmap, the highes clocked parts for this year are the 90nm. Unless you count on unexpected paper launching.

BS, no OEM was offering Brisbane in 2006.

Look after the interviews with Henri Richard and tell me if I have interpreted something wrong.

Like you, obviously.

Nope, QuadFX and Core2 Quad are for the same purpose: PERSONAL COMPUTING, x86 desktop/workstation. Maybe some people will by QuadFX as a space heater, but it's purpose is really to be used in a PC.
http://www.intel.com/products/proc [...] /index.htm
http://www.amd.com/us-en/Corporate [...] 83,00.html

Do you know the difference between a stupid question and a stupid conclusion?

Are you blind or stupid?

I thought that you were the one bragging about something...

Oh, Mr.4Heads. For what purposes are you going to use 4 different displays on a PC(you know, personal computer)?

I don't have such needs, and quadFX is not required for such purpose. You can have 4 displays on a single graphics card with a Semrpon. Have you heart about Matrox?

Maybe someone with two heads and 4 arms.

So?

No, you don't have to be sure about anything. You are the smartest and most educated man on the earth, since you are born. Also you can read our thoughts and you can predict what we are going to think.

Ok, it depends on the application. In most real-life applications Core2 Quad paltform outperforms QuadFX. Am I right now?

I am so sure.

http://www.hexus.net/content/item. [...] 505&page=3
Please provide some evidence to debunk me!

No, it makes the QFX platform a failure alone.

Oh yeah, AMD positive spin, Intel negative spin. Although Northwood was not that much of a failure it can't change the fact the Netburst was a disaster. So, how the failure of P4 fits in our discussion and how is it an argument that QFX is not a failure?

You guess stupid. PM me when you will get rid of the green iron blocks welded on your eyes.

ROFL! Teach us about properly supported NUMA!

OMG, why do you think that I can't answer?

Damn you, read the f*king paper!

Tell us more please

Are you talking about NUMA in platforms with FSB or HTT?

Nope. I think the same for you, but I didn't told you so far. I also think that you are a noob.

Nope! You are wrong.
Ptotal = Pstatic + Pdynamic
Pdynamic = C * V * V * F, where C is cumulative capacitance, V is voltage and F is frequency.

Well, at least others are not going to consume his BS seriously.

Well, at least others are not going to consume his BS seriously.


Dang - 75% of what you were talking about went way over my head. 8O

Mind if I PM you when I put together specs for a new rig?

Frankly, I don't see how you can compare a processor that doesn't exist to one that does with any kind of seriousness.

That's true. I haven't seen Barcelona yet, only the specs. The specs seem right, the promise seem reachable. Only a few refuse to accept them - well, just plainly say you don't want to believe in AMD, and stop using any half-baked technical arguments to diss Barcelona (and DCA).

Clovertown is the Code name for quad core Xeons which are derived from Kentsfield, Brisbane is the name for the latest generation of 65nm Athlon64 and its subfamilies, and Barcelona is the name of the new unreleased AMD platform, also know as K8L or more recently K8.
So the socket breakdown is like this:
Clovertown - LGA771
Brisbane - AM2 (940)
Barcelona - LGA1207

Can you show us a Quad FX system running Alan Wake? Intel obviously can't run 4 heavy tasks across 4 displays smoothly, because no one has the resources or interest to demonstrate so And why are you even touting the ridiculousness of 4 displays as an actual advantage?

Its bad enough we have the horde and intelliots to deal with, now the Horde has to bring in reinforcements from another site?

Baron???? That you???

Nope, could be Charlie but certianly not 9-inch. However, he likes to ignore data and deride the the conversation.... pretty much the staple for an AMD fanatic.

So much for people on this forum, exactly like what Baron has described. You deride the conversation, then start personal markings (if not attacks) and blame the others, when you were found plain wrong.

And to gOJDO:

a) No, I'm not the Chinese. I'm not "native" American, but I didn't come to US to Los Angeles.
b) I guess you got that image by searching "abinstein"? FYI, here is more information about this guy http://360.yahoo.com/abinstein
c) You failed to get your knowledge of NUMA (among others) right, so you try to switch attention to me personally, and even fail that, too. You are such a loser. You truly reveal your own nature.

Just for the record, the above statements, especially the "respected member of FUDZone", is total BS. First, IMO, this is more a FUDrumZ than any zone. Second, if you were referring to AMDZone, then I have to say I'm a relatively new member there, and there are many members more respected than me.

Also, at AMDZone, people (including me) actually question pro-AMD opinions as much as pro-Intel ones. Discussions there tend to be heated, but rarely get as low as what you (gOJDO and JJ) have made it here.

Just for the record (again), I've never insulted your intelligence. IMO, or maybe IMHO, anyone who thinks that "P4 is a failure compared to AXP and P3" is ignorant. Does that insult you? If so, my apology.

Or, if you have examples where I insulted you, please kindly specify, and I will apologize to you if apology is due. Otherwise, feel free to lock this thread or ban the ID. I really don't care.

Baron???? That you???

Nope, could be Charlie but certianly not 9-inch. However, he likes to ignore data and deride the the conversation.... pretty much the staple for an AMD fanatic.

So much for people on this forum, exactly like what Baron has described. You deride the conversation, then start personal markings (if not attacks) and blame the others, when you were found plain wrong.

And to gOJDO:

a) No, I'm not the Chinese. I'm not "native" American, but I didn't come to US to Los Angeles.
b) I guess you got that image by searching "abinstein"? FYI, here is more information about this guy http://360.yahoo.com/abinstein
c) You failed to get your knowledge of NUMA (among others) right, so you try to switch attention to me personally, and even fail that, too. You are such a loser. You truly reveal your own nature.


Well lets see. The actual "personal markings (if not attacks)" I saw in your quote were made by you

followed by

Are you one of those individuals who feels you have the right to post attacks in a public forum, but no one else has the right to respond in kind?
If so, perhaps you should limit yourself to AMDzone

This is so FUD ; you cut the "if" part and the "then" from the first quote of my statement above.

Can you show us a Quad FX system running Alan Wake? Intel obviously can't run 4 heavy tasks across 4 displays smoothly, because no one has the resources or interest to demonstrate so And why are you even touting the ridiculousness of 4 displays as an actual advantage?

To say having 4 displays ridiculous, you have to first make sure there is no one wanting it. In fact, I usually have two different displays in front of me when I'm working, sometimes three, and thus I really like the QFX concept. I agree, however, that presently it is too noisy, too power hungry (partly due to the MB and GPUs), and too few choice of MB & setup options, which is why I'm not getting one until later. I guess and hope these will improve over time.

Just for yet another record, I do not fully agree with AMD's 4x4 decisions, such as only the more expensive (FX) CPUs could be used, and no registered/ECC memory at the moment. (Well, I guess its their business tactics to separate markets, but it's not to our interests.) But that doesn't mean QFX by itself is a worthless concept - IMO, it could be very useful if implemented right, and it leverages K8's NUMA nature well. I guess this is the main difference between me and those dissing QFX for the sake of online reviews.

Eggxaktly. I for one learn alot even if the truth resides mostly on one side, I can also learn from the other side as well. Usually a correction from fud leads to more knowledge for me anyways

If two processors are working on two different processes with few shared data between them (as per definition of megatasking), then snooping as MEOSI won't induce much if any overhead.

A cache miss requiring a call to memory still has to snoop to ensure it fetches the freshes copy of the data.

That is true. But his original claim that "every memory transaction requires the CPU to check the other CPU first" is wrong. As you said, only when there is cache miss would the local processor consult a remote cache.

For cHT, the remote cache access can happen in parallel with the main memory read (for the local cache miss), and extra latency would be small. The major overhead is bandwidth, whose characteristic is heavily app-dependent, but well under cHT capability for 4P and less.

QFX is not a flop. All of you idiots were actually thinking AMD was trying to compete with C2Q. It wasn't. It was a show of innovation and a way to differentiate FX and X4. The only previous difference was an unlocked multiplier.

How could they expect 4 cores to overpower 4 cores that are clock for clock faster? It still may end up my next upgrade if AMD raises the RAM to 8GB.

Barcelona is a 90% new core. The specs have impressed everyone that saw them. Even more info has come out since the first analysis last year. Who are you?

Barcelona is a BEAST. PERIOD! The only way to beat K8 is to have more available sockets. Itanium goes up to 64 and Power goes to 64. They are the leaders of TPC-C.

I think Barcelona at 16 sockets (64 cores) will actually crack at least the Top 20, maybe even the Top 10. It depends on how good their L3 logic is as that and unganging should make that happen. At the minimum L3 will allow eight 1 hop NUMA nodes rather than the 4 possible now.

lol fanboy

QFX is not a flop. All of you idiots were actually thinking AMD was trying to compete with C2Q. It wasn't. It was a show of innovation and a way to differentiate FX and X4. The only previous difference was an unlocked multiplier.

How could they expect 4 cores to overpower 4 cores that are clock for clock faster? It still may end up my next upgrade if AMD raises the RAM to 8GB.

Barcelona is a 90% new core. The specs have impressed everyone that saw them. Even more info has come out since the first analysis last year. Who are you?

Barcelona is a BEAST. PERIOD! The only way to beat K8 is to have more available sockets. Itanium goes up to 64 and Power goes to 64. They are the leaders of TPC-C.

I think Barcelona at 16 sockets (64 cores) will actually crack at least the Top 20, maybe even the Top 10. It depends on how good their L3 logic is as that and unganging should make that happen. At the minimum L3 will allow eight 1 hop NUMA nodes rather than the 4 possible now.

lol fanboy

dont mock him, just give him a hug and hope he goes away