Fuad: Barcelona up to 50% faster than Kentsfield

D

defiant007

2[H]4U
Joined
Feb 27, 2006
Messages
3,497
http://www.fudzilla.com/index.php?option=com_content&task=view&id=543&Itemid=1

The K10 part codenamed Barcelona is up and running. It's not ready for launch, but AMD has already got parts ready and working.

We don't have many numbers but at least in specfp_rate2000 Barcelona ends up a bit less than 50 percent faster than Intel quad core codenamed Kentsfield.


We are quite sure that this is the best score but it definitely sounds impressive. AMD chaps are very confident that K10 marchitecture and native Quad core is the way to go and Intel will probably have to pay the price for duct taping its chips. AMD will come to collect in second half of 2007.
Click to expand...

Damn you Fuad....now I have to wait even longer to upgrade to see whether your bullshit has any truth to it
 
I don't believe it in the least bit.

Perhaps if they started sending out some sample chips to reviewers it would be easier to swallow. I also don't see that they can quickly whip up a processor and have it ready in a matter of a few months... I would say very late 2007 or even early 2008.
 
InorganicMatter said:
It's Fuad, he has less credibility than George Bush.
Click to expand...


Wow thats pretty bad lol

I also don't beleive this either. I don't see how AMD can come from so far behind in such a short period of time, and to surpass Intel by 50% in performance.
 
I also don't beleive this either. I don't see how AMD can come from so far behind in such a short period of time, and to surpass Intel by 50% in performance.
Click to expand...


AMD didn't just put together Barcelona as a response to core 2, it has been on their roadmap for a long time.
It is part of the original plan for A64, hence the on die memory controller and hypertransport....:rolleyes:
 
Lou Cypher said:
AMD didn't just put together Barcelona as a response to core 2, it has been on their roadmap for a long time.
It is part of the original plan for A64, hence the on die memory controller and hypertransport....:rolleyes:
Click to expand...
Not to mention that this is only one benchmark, specfp, so it's entirely possible that the chip isn't close to 50% faster across the board; it's more possible that either it's very competitive with C2D or slightly faster.
 
Fuad said:
We don't have many numbers but at least in every single Barcelona ends up a bit less than 82.4 percent faster than the weaksauce communist Intel quad-core codenamed "Kentsfield". Honestly, Barcelona sounds soo much cooler.

We are quite sure that this is the best score and it definitely sounds impressive. AMD chaps are very confident that K10 marchitecture and native Quad core is the way to go and Intel will probably have to pay the price for duct taping its chips and slaying goats and sacrificing children for their pact with the devil! AMD will come to collect in second half of 2007 and defeat the communistic competitor Intel.

Ohh, I'm sorry. I took a break from snorting coke off $5 hookers at the local brothel to come up with that gem of info made up of words I've heard bandied around my local Best Buy. -Fuad
Click to expand...

Eh, I'm bored at work, what can I say? And the original quoted article was poorly, poorly written.
 
Maybe it will, maybe it won't. Most likely, AMD gets the performance crown back for a while, and then Intel will eventually take it back again, and so forth.

Rinse, repeat.

In the meantime, all this Fuad character has managed to do is to get some traffic to his site.
 
I have some really good feelings about Barcelona being a Conroe Killer....It is going to have some long legs as well if my information is correct. 2008 is going to be a good year for AMD once Barcelona sales start moving forward back into the server market and parts beyond R600 are realized.
 
Unabomber said:
Maybe it will, maybe it won't. Most likely, AMD gets the performance crown back for a while, and then Intel will eventually take it back again, and so forth.

Rinse, repeat.

In the meantime, all this Fuad character has managed to do is to get some traffic to his site.
Click to expand...

Yes, it's called "Hit Phishing". Say something controversial, wrong clueless and even negative traffic mean hits. That means more ad dollars. Look at the Sha-Koo-Koo dewd's traffic for proof? It doesn't matter if those hits are from folks saying he doesn't know what he's talking about or etc....:rolleyes:
 
Can't spell fraud without fuad. I hate these rumors. I Don't believe squat until i see a reparable site bench. It's just a typed statement.

"The 8600gt will be faster than the 8800gtx by 22.345% in 3dmark06!!!!"~ stoney_titan There thats my random statement for the day! ;)

Edit: Well if Elmer Fuad get's paid doing it why don't I?
 
AtomicMoose said:
And this is what it's all about...

It doesn't matter if you like this Fuad guy or not. The information does not appear to be bogus. :)
Click to expand...


Well I will say this. If you make up enough bullshit and pass along rumors from enough people you HAVE TO get something right sooner or later. Fuad is not considered a reliable source by myself. Hence I figured I would throw my 2 cents in on this subject.
 
Exactly....since you decided to weigh in, I figured this did have some truth behind it. ;)
 
I want this to be true. Come on AMD if you go then we have to deal with INTEL only.
 
We'll see, tho I'm not surprized because AMD all along has been saying this if everybody wasn't blind to that fact. AMD clams a 40% over all performance advantage in clock speed. And FPU performance by 3.6x. By now do you really think AMD would be lieing when its running real parts? Give me a break. I think Intel has brain washed too many ppl of what AMD is really capable of. This is just another netburst like in 2003 and when the opterons come, they send them to the cleaners just like 2003.

Don't be like oh you where right all along and I told you so... when you see how much of a advantage AMD will have over the conroe core. :rolleyes: Its really not that hard to beleave. Its only the advancment of progress. Do you really beleave conroe will stay compeditive forever? No. Its a boomarang battle. One min its on AMD's side then Intels side then AMD's and all until the end of time. Wolfdine and Yorkfield will make things intresting again but no current conroe can match a stock K10 opteron quad core unless you OC it around 4ghz from conroe. And what about how devistating a OC'ed opteron K10 would be? OCed to 3ghz even would give a 20% more advantage out of it. And these are not K8's so they would have a unknown OCing limit over current AMD tech because it has totally different capabilities. We have no idea if it will still get only a 3 to 3.5ghz OC or if it can really do conroe clocks. If it can this will seriously give conroe a run for its life. Specially since AMD plans much faster K10 parts in the future around 2008 when they make 45nm chips will most likely be when this happens tho no cpu has been ever limited at launch so we could expect clock raises if needed anyways. Tho OCers will already know what the max is when they launch and can actoully see for themselfs.

Until then we only have AMD to go by. And frankly I don't see why they would lie on performance when they are getting 50% over intel on the benchmarks they can't make public yet until launch. Its just how AMD is. They are quite, they don't spread hype like Intel does. AMD has their ES team in side the labs, Intel gives ES's to the ppl for wide spread testing. This is just not how AMD does things so live with the fact. We had to deal with this before when the K8's 1st came out. They had their time, now its K10's turn to do the same. AMD didn't spread bull before, why do you think they would again? This Pro stuff is getting really childish. But ahh beleave what you want. Its none of my consurn. Just don't be like oh that guy was right, now what do we tell him. I beleave in AMD is all that means. Means nothing else.
 
I also believe the AMD will deliver, they have no reason to foolishly lie about a product that is ready for mass production.
 
Man, it's been 5 months since AMD first gave a specfp_rate2000 score and they still can't offer anything new.
 
Tetrahedron said:
why because he cant spell? big deal.. alot of intelligent people lack spelling skill, engineers, scienitist, etc..

I also believe the AMD will deliver, they have no reason to foolishly lie about a product that is ready for mass production.
Click to expand...

engineers especially, our spelling and grammar is so bad, the school of engineering at my university was thinking of adding english classes to the core classes to help improve our spelling.........bad spelling and grammar is a real problem....... I have to triple, check everything I write........ I get razzed for it all the time on msn, to the point I use the spell checker plugin for amsn, and used the spelling extension on firefox before it became a main feature.
 
specfp_rate2000? What about specfp_rate2006? What about all that crying about Intel using "outdated" benches?
 
Serge84 said:
We'll see, tho I'm not surprized because AMD all along has been saying this if everybody wasn't blind to that fact. AMD clams a 40% over all performance advantage in clock speed. And FPU performance by 3.6x.
Click to expand...

Serge84 said:
AMD = Keep quite, show NO benchmarks until release date has been met. Make NO HYPE.
Click to expand...

So which is it Serge? How about you make up your mind before you start talking?
 
I wouldn't be suprised if AMD managed to pull off that kind of performance boost with Barcelona. IIRC, there are a bunch of ex-DEC (read Alpha) engineers at AMD... draw your own conclusions :)
 
w1retap said:
The point was anybody can make retarded political bashings. It is meaningless, especially in this thread.
Click to expand...
Exactly, so why bring it up by nitpicking me? ;)
airbatica said:
I wouldn't be suprised if AMD managed to pull off that kind of performance boost with Barcelona. IIRC, there are a bunch of ex-DEC (read Alpha) engineers at AMD... draw your own conclusions :)
Click to expand...
I will be very surprised if they pull it off. Like all things ATI/AMD, I believe they will come to the ball game 6 months late with a 5-10% increase.
 
InorganicMatter said:
Exactly, so why bring it up by nitpicking me? ;)

I will be very surprised if they pull it off. Like all things ATI/AMD, I believe they will come to the ball game 6 months late with a 5-10% increase.
Click to expand...

Like they did with K8? Or like they did with K7? Or maybe like they did with K6-2?

Or maybe you were taliking about ATI? Like they did with r300? Or maybe like they did with r520?


Sorry but I cant think of a single product from either company that was bad... They both create kick ass hardware that had in its hayday no competition...
 
duby229 said:
Sorry but I cant think of a single product from either company that was bad... They both create kick ass hardware that had in its hayday no competition...
Click to expand...

O come on... you know the GeForce 3 was better than the Radeon 8500
and the Pentium 2 was at least as good as the K6-2/3
and the P4 *did* beat the K7 after Northwood came out


now, K8 kicked Intel's ass so hard, and R300 made Nvidia cry....

but still, not everything AMD or ATI has done has been golden
 
duby229 said:
Like they did with K8? Or like they did with K7? Or maybe like they did with K6-2?

Or maybe you were taliking about ATI? Like they did with r300? Or maybe like they did with r520?


Sorry but I cant think of a single product from either company that was bad... They both create kick ass hardware that had in its hayday no competition...
Click to expand...

Not bad, just mediocre. To date, K8 is AMD's only home run, and R300 is ATI's only huge trump. All the other parts have fluctuated between better or worse than the competition. K7, K6, and K6-2 all were beaten by equivalent Intels, and ATI's stuff, while always slightly better than NVIDIAs, is consistently 6 months late and consumes twice as much power.
 
Can't wait to pop in a new processor. Who wants dibs on my 5000+? :p
 
You guys do know AMD, while running at 2.4GHz, is already 40% faster than Intel's X5355 at 2.66GHz, right?

All you have to do is go to SPEC.org.

Here are the two systems...

1. AMD 8216 2.4GHz (4 socket, 8 cores)

2. Intel X5355 2.66GHz (2 socket, 8 cores)

While AMD is claiming the 40% advantage, that is not what is important considering they already have it, and should not be important to people waiting to upgrade, what is important are two things...

1. Looking at Integer performance, they should be able to pull slightly ahead of Intel, who has had about 30% advantage as of now.

2. PERFORMANCE/PRICE/POWER... You will now have the performance of a 4 socket machine in a lower priced 2 socket system while using the same power you would have used in the 2 socket system with dual cores.

These two companies will be alot closer than most people on this forum think, the thing to watch is when Intel releases the new chipset with dual 1600 FSB's.

An example of this is to look at the results for a slower clocked Intel system in Spec_fp2000, Intel E5320 1.86GHz (2 socket, 8 cores), compared to the previous link, it's within 25% of the performance while giving up 45% in clockspeed, this shows how big a negative impact Intel's FSB has on high bandwidth applications/benchmarks.

If you add FSB it should help Intel alot, and should make it somewhat competitive with Barcelona in high bandwidth applications/benchmarks.

This is just my two cents, but hopefully you will look at the links and see that the 40% claim is and has been attained.
 
Some of you seem to be forgetting how badly the P4 was whooping on the Athlon XP's when the northwood core was introduced. It was a far worse beating than AMD is receiving today, not to mention, AMD did not have nearly the market share, or resources they have today. Despite their position back then, they delivered the Athlon64 and claimed the performance lead for 2 years! If they did it then, there is absolutely NO reason they can't do it now. Not to mention the K8 architecture is far more scalable than the K7 ever was.
 
RamonGTP said:
Some of you seem to be forgetting how badly the P4 was whooping on the Athlon XP's when the northwood core was introduced. It was a far worse beating than AMD is receiving today, not to mention, AMD did not have nearly the market share, or resources they have today. Despite their position back then, they delivered the Athlon64 and claimed the performance lead for 2 years! If they did it then, there is absolutely NO reason they can't do it now. Not to mention the K8 architecture is far more scalable than the K7 ever was.
Click to expand...

Err.. the Northwood was at best 10% faster than an Athlon XP. It is far worse now for AMD, as the fastest C2D is ~20% faster than the fastest X2, plus they don't have an answer for quad core yet (QuadFX doesn't really count IMO).

It's funny you mention K8 being 'far more scalable' than K7 because I still happen to run an old XP-M 2500+ in my spare PC, and it does 2.6GHz on air easily. I could probably push it to 2.7 - 2.8GHz with extra volts and an extra loud fan, but what's the point? In it's day it was awesome, even now it'll match up to an A64 3200+, pretty good for a 3 year old chip. And remember, I achieved 2.6GHz on AMD's 130nm process. How far are 65nm Brisbane overclocking nowadays? 2.8 - 3.0GHz. Point taken?
 
Yeah, maybe 10% in the things the AXP did well at, there were a lot of tasks that the Northwood was WAAAY better at... Just to refresh your memory...

http://www.anandtech.com/showdoc.aspx?i=1690&p=4

Also, scalable and overclockable are two different things. I thought by now people would be over the MHz myth. Take your 3GHz XP-M against a single core A64 running at the same clock rate and compare the two. C2D's are also OCing to arond 3GHz, are you saying your XP-M is nearly as scalable as C2D? Point taken? I think not.

They do have an answer to quad core, they just haven't released it yet.
 
RamonGTP said:
Yeah, maybe 10% in the things the AXP did well at, there were a lot of tasks that the Northwood was WAAAY better at... Just to refresh your memory...

http://www.anandtech.com/showdoc.aspx?i=1690&p=4

Also, scalable and overclockable are two different things. I thought by now people would be over the MHz myth. Take your 3GHz XP-M against a single core A64 running at the same clock rate and compare the two. C2D's are also OCing to arond 3GHz, are you saying your XP-M is nearly as scalable as C2D? Point taken? I think not.

They do have an answer to quad core, they just haven't released it yet.
Click to expand...

Wow, Sysmark2002, it's funny, back in 02 AMD fans were saying how 'biased' the benchmark was in favor of Intel, and how, 5 years later, here you are using it as 'proof' that a NW was streets ahead of the Athlon XP when it clearly isn't. Take it from me, I also used to own a P4-C @ 3.3GHz before it died on me, and the XP-M @ 2.6GHz is easily a match for it.

So what exactly is your interpretation of 'scalable'? If you mean in performance, things are all relative. K7 was able to 'scale' to within 10% of Northwood at it's peak, K8 is stretching it to get within 20% of Core2.

AMD also wasn't in a load of debt back in 02, so while your attempts to paint a rosy picture are admirable, it doesn't take away from the truth - AMD is struggling financially, and really needs to hit a home run with K10, otherwise all the hard work and momentum/marketshare gained with K8 would be wasted.
 
harpoon said:
AMD also wasn't in a load of debt back in 02, so while your attempts to paint a rosy picture are admirable, it doesn't take away from the truth - AMD is struggling financially, and really needs to hit a home run with K10, otherwise all the hard work and momentum/marketshare gained with K8 would be wasted.
Click to expand...

I agree with this statement 100%, the K10 definitely needs to flex its muscle, and I think it will.

As far as scalable, what I mean is that the core architecture responds well to simple core improvements like die shrinks, clock speed increases, pre-fetching optimizations, cache optimizations, memory controller improvements, etc etc.

The K10 is not drastically different from the K8, it basically incorporates all these optimizations and then some. From an engineering standpoint, going from K8 to K10 is far easier than the transition from K7 to K8. If AMD was able to pull off K7>K8 with success, they'll be able to do the same for K8>K10
 
RamonGTP said:
I agree with this statement 100%, the K10 definitely needs to flex its muscle, and I think it will.

As far as scalable, what I mean is that the core architecture responds well to simple core improvements like die shrinks, clock speed increases, pre-fetching optimizations, cache optimizations, memory controller improvements, etc etc.

The K10 is not drastically different from the K8, it basically incorporates all these optimizations and then some. From an engineering standpoint, going from K8 to K10 is far easier than the transition from K7 to K8. If AMD was able to pull off K7>K8 with success, they'll be able to do the same for K8>K10
Click to expand...

I know this is getting off topic, but wasn't K8 just K7 with enhancements and an IMC? The performance jumps from from K8 to K10 is (anywhere from 40% to 80% per clock, at least according to AMD) is far greater K8 achieved over K7.
 
You must log in or register to reply here.
Back
Top