• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

HUGE article on Barcelona core

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
I agree 100%.

Intel will have the manufacturing advantage. Even if AMD is say ~5% more efficient, Intel has a very good architecture on it's hands, and can just ramp clock speeds to compensate.

Sucks, but this is the truth.

That would only lead them into the same hole they got themselves in with the P4 architecture.
 
I'm looking forward to the new AMD cpu's, but I'm only interested in the dual core version. I have no use for quad core.

*edit: also, I hope they get the 65nm process down, so the chips will OC well.
 
Originally posted by: firewolfsm
I agree 100%.

Intel will have the manufacturing advantage. Even if AMD is say ~5% more efficient, Intel has a very good architecture on it's hands, and can just ramp clock speeds to compensate.

Sucks, but this is the truth.

That would only lead them into the same hole they got themselves in with the P4 architecture.
It would be a a bridge until their next generation uarch (Nehalem) comes out in 2008 where IPC is suppose to get a significant boost.
 
Originally posted by: munky
I'm looking forward to the new AMD cpu's, but I'm only interested in the dual core version. I have no use for quad core.

*edit: also, I hope they get the 65nm process down, so the chips will OC well.


If the price and performance are right I would love a quad core. Just today I was compressing 3 movies, watching TV in a window, browsing the internet, printing some large document, and editing some photos in photoshop.

My dual can handle about half of that load after that it starts to slow down. A quad is great when I've gotta get a lot done and I don't have time to wait.

Then again I was thinking my old P4 3.06 would have taken about 5 times longer to accomplish what I did today in 2 hours!

More speed is always better. Especially if it comes at a good price point.

It would be so great if AMD came out guns blazing and put Intel on the defense again. A good old pricewar the likes of which we haven't ever seen in the computer industry would be a lot of fun for us.

I can just imagine a 3GHz single chip quad core solution for $200!

 
I think the AMD fans are getting a little too hopeful although I really hope AMD can pull a rabbit out of their hat.

They have the fan's chanting "native quad versus dual core" when AMD recently debutted the AMD Quad FX which was a "quad" solution spread over two chips. :roll: And this is all noise on a chip that is still fairly far in the future.

I truly hope AMD offers a competitive product because competition is what keeps the prices down but AMD needs to do more than just keeping talking... they need to just offer up a chip that does the talking. And forget about telling us that their cpu's are the most energy efficient when they are sitting idle... hey, turning them off when nobody uses them saves even more energy, lol.
 
I think 20% real world performance gain over Core 2, as Core 2's 25% in most benchmarks sometimes more and rarely less over X2. If it can lower prices by 50% like Core 2 did it's a winner and I will buy one or three.
 
Originally posted by: Conky
They have the fan's chanting "native quad versus dual core" when AMD recently debutted the AMD Quad FX which was a "quad" solution spread over two chips. :roll:
True, QuadFX is a joke; I completely agree. Hopefully more enthusiast boards (lower-priced ones) will be released, besides the Asus Quadfather. Having a dual-socket, eight-core solution with regular DDR2 is a VERY appealing proposition for enthusiasts.
And this is all noise on a chip that is still fairly far in the future.
It's March and production starts in June/July.
And forget about telling us that their cpu's are the most energy efficient when they are sitting idle... hey, turning them off when nobody uses them saves even more energy, lol.
Hundreds of trillions of idle clock cycles occur every day across the globe. An overwhelming majority of CPU cycles worldwide occur at idle.

And that's only concerning desktops. Please explain how a company turns off computers at a data center or server/render farm? Are they going to shutdown and startup hundreds of nodes every time a task arises, or peak usage occurs? Lowering idle power consumption is a BIG deal for businesses.
 
QuadFX would have been much more attractive than what it is now if there is an alternative board with only 1 NB. Semantics aside (quad-GPU, yeah right), having two of those burning nForce chips (sitting next to each other) take the whole thermal/power issue to absurdity. Then again, that'd decisively hurt the Opteron market so probably AMD itself didn't want anyone to produce one-chip board with 2 CPU sockets for QuadFX CPUs. In the end, it's obvious that even AMD knew this platform is a low-volume, show-off stuff. (hence the stop-gap)

I thought (well, more likely wished) the QuadFX platform would show a tidbit of the Torrenzza initiative, but apparently that wasn't the case. But Barcellona is really interesting. One thing I worry about is its pricing at launch. Hope it's not too expensive to be prohibitive for us enthusiasts.
 
Originally posted by: Acanthus
Most of the information thats out there cant be extrapolated into real world gains. It's too vague to predict where or even if we will see large gains.

We will see at launch.

I also find the clockspeeds disappointing unless there is an enormous gain in IPC, im still pointing at problems with AMDs 65nm.

Remember that these are quad-core. the 2.3GHz parts, anyway. 2.3GHz at 95W is pretty decent (45W per "pair"), especially if it outperforms the 130W part that Intel already has out there (possible in multi-socket systems but I'll believe that when I see it). We'll be seeing a few speed bumps in Q3, too, and these are just Opteron parts. AMD usually has higher clocks on the desktop side.
 
2.3GHz at 95W is pretty decent (45W per "pair")

So does that mean that the high porformance dual cores, Kuma, will now be 45W?
Sweet! This is likely what I will be getting by the end of the year if they are available.
 
love this one! best technical article on Barcelona architecture I've came across so far! way to go Anand!!
 
I'm psyched for Barcelona, and very earger to see the peformance numbers.

Unfortunately, like much of ATI/AMD lately, it's a very late response to an already solid product from a competitor.

I'm still happily running my dual core Opteron 165 @ 2.5 GHz (no big reason for me to upgrade yet, my system is still fast), but I'd love something new to try out, and if I would be to upgrade today, my only real choice is Intel.
 
Originally posted by: Viditor
Originally posted by: zsdersw
Originally posted by: Viditor
EETimes
"Advanced Micro Devices upcoming Barcelona processor will sport floating-point performance 42 percent higher than Intel's current top-of-the line CPU, the Xeon X5355 also known as Clovertown"
"Mario Rivas, general manager of AMD's microprocessor group, said Barcelona will provide a double-digit leap in integer performance over the quad-core Xeon, though he declined to be more specific. Henri Richard, chief of sales and marketing at AMD, said Barcelona will have a significant integer performance lead over Intel's quad-core chips"

42% improvement in FP and double digit improvement in integer over quad core Xeon.

No skepticism on your part for any of that like you had with Conroe just before it's release?

No need for me to be skeptical Z, you have enough skepticism for 10 of us...I'm just trying to keep the information flowing and the thoughts balanced. 🙂



Actually though I like you, I think your idea of balanced is a bit skewed....When has there been a time when you haven't only posted rosy pictures of this release and AMD's future offerings? You dont.....You dont post rosy rumors of intel offerings yet only put STOCK in marketing fluff from the company itself...What the heck do you think the company is going to say? Conroe had some big claims...remember HT claims....All overestimated a bit...the truth lies somewhere in the middle. 45% gains? More like half that most likely...So that would be more like 20-25%. Which would get it back to and just ahead of the clock performance of the C2D....

Ofcourse I avoid these threads because until we see hard benchmark numbers outside of the chronic liars known as marketers we dont really know....Just a bunch of ASSumptions. I can wait until we at least some previews....

I only step in when I see this one sided opinion....
 
Originally posted by: Duvie
Originally posted by: Viditor
Originally posted by: zsdersw
Originally posted by: Viditor
EETimes
"Advanced Micro Devices upcoming Barcelona processor will sport floating-point performance 42 percent higher than Intel's current top-of-the line CPU, the Xeon X5355 also known as Clovertown"
"Mario Rivas, general manager of AMD's microprocessor group, said Barcelona will provide a double-digit leap in integer performance over the quad-core Xeon, though he declined to be more specific. Henri Richard, chief of sales and marketing at AMD, said Barcelona will have a significant integer performance lead over Intel's quad-core chips"

42% improvement in FP and double digit improvement in integer over quad core Xeon.

No skepticism on your part for any of that like you had with Conroe just before it's release?

No need for me to be skeptical Z, you have enough skepticism for 10 of us...I'm just trying to keep the information flowing and the thoughts balanced. 🙂



Actually though I like you, I think your idea of balanced is a bit skewed....When has there been a time when you haven't only posted rosy pictures of this release and AMD's future offerings? You dont.....You dont post rosy rumors of intel offerings yet only put STOCK in marketing fluff from the company itself...What the heck do you think the company is going to say? Conroe had some big claims...remember HT claims....All overestimated a bit...the truth lies somewhere in the middle. 45% gains? More like half that most likely...So that would be more like 20-25%. Which would get it back to and just ahead of the clock performance of the C2D....

Ofcourse I avoid these threads because until we see hard benchmark numbers outside of the chronic liars known as marketers we dont really know....Just a bunch of ASSumptions. I can wait until we at least some previews....

I only step in when I see this one sided opinion....

I can agree with this, we will have to see how Barcelona compares the products it will be going against namely Clovertown and Tigerton, when both products from the respective companies are available.

I can't really put much stock in what AMD says until I see some hard figures, on actual machines.
 
Originally posted by: Grabo
'Excuses' and 'excuses'; they can either keep pace with Intel or they can't, and in the latter case the consumers, who would rather purchase the temporarily best number cruncher than attempt to keep the balance between two struggling giants, will be the losers.
I disagree. We didn't get $120 dual-cores until they weren't keeping up with Intel. As long as they can compete at some level, we win.
 
Another advantage on Intel is its 45nm is socket compatible with current mobo. Although AMD quad is also socket compatible, I think they need new mobo to fully exploit its potentials like the PLL per core etc etc. Its an interesting battle come 2H. We know Penryn core's only survival is if it can push more clock speed.
 
Originally posted by: ahock
Another advantage on Intel is its 45nm is socket compatible with current mobo. Although AMD quad is also socket compatible, I think they need new mobo to fully exploit its potentials like the PLL per core etc etc. Its an interesting battle come 2H. We know Penryn core's only survival is if it can push more clock speed.
Intel has said nothing about Penryn being compatible with current motherboards, and everything I've read says the opposite (at least in the case of desktop boards). Intel makes a LOT of money from chipset sales; I wouldn't be surprised if we see a double launch (CPU and compatible-chipset) in Q407/Q108.

Intel has maintained socket compatibility since the P4 E (Prescott) era, but they've changed chipset requirements with every new product launch (Pentium D and Core 2 Duo).
 
Originally posted by: Duvie

Actually though I like you, I think your idea of balanced is a bit skewed....When has there been a time when you haven't only posted rosy pictures of this release and AMD's future offerings? You dont.....You dont post rosy rumors of intel offerings yet only put STOCK in marketing fluff from the company itself...What the heck do you think the company is going to say? Conroe had some big claims...remember HT claims....All overestimated a bit...the truth lies somewhere in the middle. 45% gains? More like half that most likely...So that would be more like 20-25%. Which would get it back to and just ahead of the clock performance of the C2D....

Ofcourse I avoid these threads because until we see hard benchmark numbers outside of the chronic liars known as marketers we dont really know....Just a bunch of ASSumptions. I can wait until we at least some previews....

I only step in when I see this one sided opinion....

I guess I wasn't being clear...I don't mean that my own opinion was balanced (though I have stepped in on several occasions when an AMD zealot has espoused nonsense about Intel and I've made some very pro-Intel comments), but that a balancing of the especially pro-Intel rantings was required.
In this case, I'm not very skeptical about the numbers because it's pretty much what I and most of the engineers I work with have been expecting.
You must keep in mind several things...

1. The statement covers a specific CPU (Cloverton vs Barcelona).
2. FP - I don't think anyone would be surprised at a >40% lead by Barcelona, considering that even the old K8 is faster at FP than Woodcrest.
3. Int - A greater than 10% advantage isn't surprising here either as in the case of this specific matchup, 8 cores of the old K8 is already faster than 8 cores of Cloverton.

Where I see some of the problems occuring here are when the details aren't read. It should be obvious (even before the reviews) that for 2P quad-core, AMD will easily be the leader...even at the lower clockspeeds.
Even with the older architecture, they are already faster at >4cores
Anandtech review

"Two 2.4GHz Opteron 880 processors are as fast as one Xeon 5345, but four Opterons outperform the dual quad core Xeon by 16%. In other words, the quad Opteron system scales 31% better than the Xeon system"

The problem is that 99.9% of the people reading here have no use or experience with that kind of setup (you may be one of the few exceptions), but they translate that to meaning that desktop processers will have the same degree of advantage...of course this is not true.
I then receive the usual comments about "healthy skepticism", etc...
I guess what I should say is to go back and reread the thread, but I can only sigh so often. 🙂

The bottom line is that I can say with a fair certainty that the Barcelona will be quite superior to Cloverton, but I have no idea what will happen with the Stars processers until the reviews start to come out (unless Barcelona gives us more insights).
 
Originally posted by: Viditor
though I have stepped in on several occasions when an AMD zealot has espoused nonsense about Intel and I've made some very pro-Intel comments

Must've been very difficult for you to do that.

 
Originally posted by: zsdersw
Originally posted by: Viditor
though I have stepped in on several occasions when an AMD zealot has espoused nonsense about Intel and I've made some very pro-Intel comments

Must've been very difficult for you to do that.

And as Z has courageously demonstrated here, many posters will judge what you're saying without actually reading all of it...
 
Originally posted by: Viditor
And as Z has courageously demonstrated here, many posters will judge what you're saying without actually reading all of it...

I read everything in your comment. I quote only what I'm responding to.
 
Originally posted by: zsdersw
Originally posted by: Viditor
And as Z has courageously demonstrated here, many posters will judge what you're saying without actually reading all of it...

I read everything in your comment. I quote only what I'm responding to.
Well, at least we now know for sure that you're as pro-Intel as Viditor is pro-AMD. If we need any definitive performance numbers about Penryn before it's released, we'll look you up.😉
 
Back
Top