PDA

View Full Version : Liano/Zambezi CPUs price leaked!


Pages : [1] 2

AthlonXP1800
05-21-11, 06:27 AM
We are now close to the launch of Zambezi CPUs, the price of the CPUs leaked.

4 core FX-4110 $190/117
6 core FX-6110 $240/147
8 core FX-8110 $290/178
8 core FX-8130P $320/196

http://translate.google.com/translate?hl=en&sl=auto&tl=en&u=http%3A%2F%2Fwww.donanimhaber.com%2Fislemci%2Fha berleri%2FAMDnin-Bulldozer-ve-Fusion-Liano-islemcileri-icin-ilk-fiyat-bilgileri.htm

AMD cutted CPUs price a while ago, you can find Phenom II X4 955 for nearly 80 as AMD are in processing to retire Athlon II and Phenom II CPUs line for the arrival of Zambezi and Fusion Liano CPUs.

It great to see Zambezi will use the same price range to ceased fears that it would be more expensive so actually it will be cheaper than current mainstream Sandy Bridge CPUs. High end Sandy Bridge E will expect to be bloody expensive than mainstream Sandy Bridge CPUs. :)

Viral
05-21-11, 09:42 AM
Nothing is confirmed here, these prices are rumours.

AthlonXP1800
05-21-11, 09:56 AM
Nothing is confirmed here, these prices are rumours.

Same thing with leaked FX series CPU retail box photos and banners, nothing is conformed just rumours.

Bearclaw
05-21-11, 10:12 AM
When are these coming out?

AthlonXP1800
05-21-11, 10:19 AM
When are these coming out?

It will be launch on 7 June 2011 at Computex accorded to leaked Gigabyte slide.

http://lensfire.blogspot.com/2011/04/bulldozer-processor-release-date-new.html

That almost 2 weeks away. :)

Edit: Either Gigabyte slide made a typo with Computex 2011 date as Computex open on 31 May and close on 4 June 2011 or Computex decided to change the date.

AMD announced they will launch FX series CPUs at E3 on 7 June instead.

http://www.tomshardware.com/news/AMD-Llano-AMD-Sabine-AMD-Bulldozer-AMD-Fusion,12688.html

Intel17
05-21-11, 10:53 AM
AMD needs to freaking confirm something already!

Redeemed
05-21-11, 11:04 AM
Such low pricing might not be a good sign...

That's kinda' what I'm thinkin' as well. :o :(

Bman212121
05-21-11, 11:41 PM
I wouldn't really call them low... $320 USD is roughly the same as an i7 2600k.

LydianKnight
05-22-11, 03:38 AM
Nah, something being more expensive doesn't always equate to it being better... that's just a stupid assumption we tend to make (although this depends on a given individual's opinion) and something keeps us (sometimes, not always) from having some interesting products in favor of the 'usual' products (more expensive and with more 'reputation').

So would you feel more comfortable if that CPU costed $550? That's just nonsense... this is not a I-throw-myself-out-for-a-couple-of-dollars appreciation scale but come on... $300 for the (supposed) top-of-the-line AMD offering is awesome (and I even think it's a bit more expensive than their last CPU line was, which makes me even more assured my purchase of an i5 2500 will be even more 'justified' haha).

:p

LydianKnight
05-22-11, 04:50 AM
It's not about "more expensive = better" from a consumer perspective.

I have a hard time believing that AMD likes to sell a killer product that low. They didn't do that the last time when they had a killer CPU (AMD X2 4800+ and FX series).
Of course Intel put a lot of pressure on the competition with their nicely performing new low end and mid range line up this year. But this is a "8 Core" CPU with high Turbo modes. It should blow away a normal Sandy Bridge. Why not sell it at a higher price?
It seems strange...

Not that strange... it's well known AMD is at a second place, so maybe it's part of their strategy... for this generation, they will offer us a nice product at a (relatively) low price and if they have something up to their sleeve for a refresh, drop it at a slightly higher price.

Still, I highly doubt they will release another CPU with the same price levels as in the past. Close or equal to Intel in the low/mid/mid-high end? Sure, but not in the high end anymore, and that's good for consumers, so I think it's a nice move on their side :)

Intel17
05-22-11, 07:33 AM
Look -- AMD is a company and it's out to make money. They price their items according to how well they perform.

If the BD 8-core is going to go head-to-head with Intel's Gulftown, then it will be priced in its league. If it's got no chance in hell, then it will be priced lower. But AMD isn't going to deliberately kill their margins just to claim some market share -- they're going to do this thing called product segmentation!

I'm not totally convinced that these prices are correct. You're telling me that I can pay less-than-double for an 8 core over a 4 core? Please.

Redeemed
05-22-11, 10:25 AM
It's not about "more expensive = better" from a consumer perspective.

I have a hard time believing that AMD likes to sell a killer product that low. They didn't do that the last time when they had a killer CPU (AMD X2 4800+ and FX series).
Of course Intel put a lot of pressure on the competition with their nicely performing new low end and mid range line up this year. But this is a "8 Core" CPU with high Turbo modes. It should blow away a normal Sandy Bridge. Why not sell it at a higher price?
It seems strange...

Blow SB away? Where'd you get this impression? Last I heard it should be about on par with SB, hence the pricing.

Redeemed
05-22-11, 01:22 PM
"Should" is the keyword. AMD's top of the line product, especially with those specs, should be way faster than a normal SNB.

But first that slide on which it's just put up against a normal SNB and now this way too low pricing. I'm not sure if those are good signs. AMD can't play the integrated GPU card with Bulldozer.

If you were expecting something that'd topple your 980x from AMD then you had your expectations set to high.

Me, I'm just glad they have an architecture that's apparently clock-for-clock competitive with Intel. This gives them leg room to grow and improve. :)

Roadhog
05-22-11, 01:32 PM
pretty disappointing if it is going to be equal to or slower than sandy bridge.

Intel17
05-22-11, 02:12 PM
http://news.mydrivers.com/1/194/194271.htm

ViN86
05-22-11, 04:11 PM
Nah, something being more expensive doesn't always equate to it being better... that's just a stupid assumption we tend to make (although this depends on a given individual's opinion) and something keeps us (sometimes, not always) from having some interesting products in favor of the 'usual' products (more expensive and with more 'reputation').

So would you feel more comfortable if that CPU costed $550? That's just nonsense... this is not a I-throw-myself-out-for-a-couple-of-dollars appreciation scale but come on... $300 for the (supposed) top-of-the-line AMD offering is awesome (and I even think it's a bit more expensive than their last CPU line was, which makes me even more assured my purchase of an i5 2500 will be even more 'justified' haha).

:p

Last time they had a better performing product they priced it to match. Hence why the FX chips carried a $1k pricetag. Companies don't sell products as cheap as they can unless the performance isn't there.

Ninja Prime
05-22-11, 07:20 PM
I expected the 8 Cores to be priced against Intel's 6 Cores (not Extreme Editions). This doesn't mean that the FX performance will be underwhelming but it sure seems a bit strange.

You don't understand AMDs design then. Its more akin to a 4 core with hyperthreading than a 8 core. Hence why it is priced at the same as the i2600, I would expect it to beat the i2600 at most things though. Originally they were not even going to call them "8 cores" they were going to call them 4 modules. I suppose the marketing of saying "zomg 8 cores, thats more than 6!" won out over the logical naming scheme.

Make no mistake, these are meant to compete with the current sandy bridges. I don't know that they really intend to compete with future 6 core sandy bridges at all, they have 16 core (8 module) server products for server markets and the 6 core high end market is so tiny and 6 core does almost nothing for the average consumer over 4 cores anyway, I think they won't even bother, that is their marketing currently, with nothing to compete with intels highest end, why change?

Viral
05-22-11, 07:40 PM
You don't understand AMDs design then. Its more akin to a 4 core with hyperthreading than a 8 core. Hence why it is priced at the same as the i2600, I would expect it to beat the i2600 at most things though. Originally they were not even going to call them "8 cores" they were going to call them 4 modules. I suppose the marketing of saying "zomg 8 cores, thats more than 6!" won out over the logical naming scheme.

Make no mistake, these are meant to compete with the current sandy bridges. I don't know that they really intend to compete with future 6 core sandy bridges at all, they have 16 core (8 module) server products for server markets and the 6 core high end market is so tiny and 6 core does almost nothing for the average consumer over 4 cores anyway, I think they won't even bother, that is their marketing currently, with nothing to compete with intels highest end, why change?

I'm not sure if you understand AMDs design. It's nothing like 4 cores with HT, it's just 8 integer cores in lots of two sharing a single FP unit (a beefy one that's even more capable than SB's). That's an 8 core processor, it's recognized by every OS out there as having 8 cores. You don't call every processor before the Pentium Pro a half core, do you?

For clarification AMD never stated they were going to call this a "4 module" processor, that was just speculation by the public.

WRT these prices, I wouldn't take anything to the bank just yet. They're on par with SB, which is a good sign, but doesn't mean they might not still be faster. I think there's no question that they will be faster at most properly threaded tasks. Games, they will probably be on par for now. That's just my guess anyhow.

As for server, I would think that is far less of a concern. Everything on servers is massively threaded and with 16 real cores, I think AMD should be pretty comfortable in the server realm... especially with their competitive pricing and low power use (if following on the trend from Magnycours).

Ninja Prime
05-23-11, 03:20 AM
That's an 8 core processor, it's recognized by every OS out there as having 8 cores.

So are 4 core sandy bridges, so I guess they are 8 cores too right? They are not 8 full cores. They are 4 cores, with parts of another core in there for a second thread via hyperthreading. At best you might call them 1.5 cores.

seastar
05-23-11, 04:18 AM
Well, remember the Radeon HD 4870? AMD priced it way lower than what competing products from Nvidia was and caused Nvidia to lower their prices.

Viral
05-23-11, 07:15 AM
So are 4 core sandy bridges, so I guess they are 8 cores too right? They are not 8 full cores. They are 4 cores, with parts of another core in there for a second thread via hyperthreading. At best you might call them 1.5 cores.

Really? "parts of another core"? It's a complete integer core, nothing to do with hyperthreading.

Hyperthreading aims to constantly saturate the schedulers, making sure they are always active and thus ready to feed the pipeline. That's it, there's no more magic going on or extra "parts" of a core.

Bulldozer is two integer cores. In the same way as HT, AMD plan to keep the schedulers busy by lowering the ratio of schedulers per cores (3 for 2int and 1fp vs. 2 for 1int and 1fp in K7-10). This sounds stupid and that it would decrease performance, but a) the schedulers aren't the same as K10 ones, and b) the aim is efficiency, both in power, decreasing pipeline stalls and obtaining higher clocks. We'll just have to wait and see how it pans out. My guess is that it will give better performance in some applications, and worse in others.

The way I see it, this is 2011. No one should be buying single cores for a desktop PC or laptop anymore. It makes sense to design an efficient dual core with shared resources, reduced power and die size as the very lowest end chip you can produce. This is what Bulldozer is and it makes sense to me. How it actually performs will be the telling factor though.

Intel17
05-23-11, 09:39 AM
(a beefy one that's even more capable than SB's).

Yeah? Proof?

Redeemed
05-23-11, 09:43 AM
Really? "parts of another core"? It's a complete integer core, nothing to do with hyperthreading.

Hyperthreading aims to constantly saturate the schedulers, making sure they are always active and thus ready to feed the pipeline. That's it, there's no more magic going on or extra "parts" of a core.

Bulldozer is two integer cores. In the same way as HT, AMD plan to keep the schedulers busy by lowering the ratio of schedulers per cores (3 for 2int and 1fp vs. 2 for 1int and 1fp in K7-10). This sounds stupid and that it would decrease performance, but a) the schedulers aren't the same as K10 ones, and b) the aim is efficiency, both in power, decreasing pipeline stalls and obtaining higher clocks. We'll just have to wait and see how it pans out. My guess is that it will give better performance in some applications, and worse in others.

The way I see it, this is 2011. No one should be buying single cores for a desktop PC or laptop anymore. It makes sense to design an efficient dual core with shared resources, reduced power and die size as the very lowest end chip you can produce. This is what Bulldozer is and it makes sense to me. How it actually performs will be the telling factor though.

The more I read your posts, the more I want to head over to AMDZone and start picking their brains about AMD's design of Bulldozer. :p :lol:

Ninja Prime
05-23-11, 05:01 PM
Really? "parts of another core"? It's a complete integer core, nothing to do with hyperthreading.

Hyperthreading aims to constantly saturate the schedulers, making sure they are always active and thus ready to feed the pipeline. That's it, there's no more magic going on or extra "parts" of a core.

Bulldozer is two integer cores. In the same way as HT, AMD plan to keep the schedulers busy by lowering the ratio of schedulers per cores (3 for 2int and 1fp vs. 2 for 1int and 1fp in K7-10). This sounds stupid and that it would decrease performance, but a) the schedulers aren't the same as K10 ones, and b) the aim is efficiency, both in power, decreasing pipeline stalls and obtaining higher clocks. We'll just have to wait and see how it pans out. My guess is that it will give better performance in some applications, and worse in others.

The way I see it, this is 2011. No one should be buying single cores for a desktop PC or laptop anymore. It makes sense to design an efficient dual core with shared resources, reduced power and die size as the very lowest end chip you can produce. This is what Bulldozer is and it makes sense to me. How it actually performs will be the telling factor though.

There are two integer cores, but they share a single front end and back end. That makes them one core IMO. I like AMDs idea better than Intels, but it remains to be seen how much better or worse it will do in the real world.

Viral
05-23-11, 06:59 PM
There are two integer cores, but they share a single front end and back end. That makes them one core IMO. I like AMDs idea better than Intels, but it remains to be seen how much better or worse it will do in the real world.

Yes they share a front end, one that's 50% more capable than K10's. Same story with the FP unit, it can be seen as one big shared 256-bit unit, or two 128-bit units similar (not the same) to K10's. You do have a point about sharing the front end though and we will just have to wait and see how that goes. AMD state that this module approach gives 90% of the performance of a theoretical non-modular BD with dedicated hardware in each core.

There are clearly going to be downsides, but also many upsides to performance, not just power and die size savings; like shared L2 cache between 2 cores. Single threaded app's will have access to a whole 2MB of L2 cache.