[FW]Llano / Bulldozer ES ???

[Hornet331]

You mean to say intels double cost, double Wattage and a mere 3 times the die size cpu is faster? Jesus who'd have believed that?

That PoS will get annihilated by the lowest end Llano at the same TDP.

The only thing that counts for the consumer in the end is price, performance and consumption (in that order). Die size is pretty much irrelevant, it can be 10 times the size if every thing else is ok.

I was quite suprised to see that the 2100 consumes less in idle then the E350. You pretty much get that what you pay for, ION2+Atom is in the same price range as the E350. For somewhat simmilar pefromance.

If Llano will offer an overall better performance then SB (or IB later this year) it will command a higher price, you don't get anything for free.

[sergiojr]

I was quite suprised to see that the 2100 consumes less in idle then the E350.

It depends on motherboard design. There is a remark in that article saying that MSI E350IS-E45 consume only 7,3 W in idle (2,5 W less then system with 2100T).

[Hornet331]

It depends on motherboard design. There is a remark in that article saying that MSI E350IS-E45 consume only 7,3 W in idle (2,5 W less then system with 2100T).

Its still impressive considering its double the TDP and that the MSI E350IS has no WLan and no USB 3 compared to the Gigabyte E350N or the Zotac H67-ITX WiFi. Youd probably can push it even further down with a board that also lacks this featuers like the ASRock H67M-ITX.

I hope that llano can reach simmilar lowerpower stats as SB.

[sergiojr]

Its still impressive considering its double the TDP and that the MSI E350IS has no WLan and no USB 3 compared to the Gigabyte E350N or the Zotac H67-ITX WiFi..

There is USB3 on MSI board and I suppose WiFi was disabled in power consumption tests.
Actually you can see power consumption figures of the MSI board yourself (it's the same author so testing methods are the same and numbers are totally comparable).
http://www.xbitlabs.com/articles/cpu..._12.html#sect0
It seems that Gigabyte design is a shame (consumes 9W more at full load than MSI board).

[Particle]

The only thing that counts for the consumer in the end is price, performance and consumption (in that order). Die size is pretty much irrelevant, it can be 10 times the size if every thing else is ok.

I believe his point about die size is price. Die size to a large degree dictates the price floor of what you have to sell a product for since a die that is 2 units in size costs more than double to manufacture than what a die 1 unit in size would.

[Hornet331]

There is USB3 on MSI board and I suppose WiFi was disabled in power consumption tests.
Actually you can see power consumption figures of the MSI board yourself (it's the same author so testing methods are the same and numbers are totally comparable).
http://www.xbitlabs.com/articles/cpu..._12.html#sect0
It seems that Gigabyte design is a shame (consumes 9W more at full load than MSI board).

No they are different boards, there is the MSI 350IS and the MSI 350IA. Just looked it up, the IS doesn't have USB 3 and no Wlan and the IA only has USB 3 but also no Wlan, so yeah if you cut Wlan you get some nice engergy saving, cause depending on what wlanchip they use they can consumes quite much.

[sergiojr]

No they are different boards, there is the MSI 350IS and the MSI 350IA. Just looked it up, the IS doesn't have USB 3 and no Wlan and the IA only has USB 3 but also no Wlan, so yeah if you cut Wlan you get some nice engergy saving, cause depending on what wlanchip they use they can consumes quite much.

The reviewed MSI board definitely has 2 USB3 ports at back panel. Just look at photos in the review. Model number is really strange. Probably it was a prepoduction sample or some special OEM version.
And disabled wi-fi doesn't consume much.
As for 2100T 2x power consumption at full load and 2,5W more consumption at idle is already a good result. Why are you expecting that it can be even better?

[JF-AMD]

Wait... we know die sizes of Llano and Bulldozer? Since when?

There have been plenty of guesses about die size. Nobody has ever gotten it right.

I have an E350N gigabyte board at home. Copying all of my data from my server to this box (i.e. maxing out the NIC and drive controller) I am averaging ~35W at the wall. 3.5" HD, no optical drive, no I/O cards.

[MrMojoZ]

Why are you expecting that it can be even better?

Brand is probably the lead factor here, but thats just based off things said in the past.

[kl0012]

AMD Expects Fusion "Llano" to Offer Three Times Higher Graphics Performance than Core i "Sandy Bridge".

According to AMD internal documents seen by X-bit labs, which describe performance of AMD's new products, AMD A8-3550 APU (quad-core, 4MB cache, 400 stream processors, Radeon HD 6550 graphics core at 594MHz) delivers three times higher performance in 3DMark Vantage test compared to Intel Core i5-2300 (quad-core, 2.80GHz/3.10GHz clock-speed, 6MB cache, Intel HD graphics core at 850MHz/1100MHz): P3335 versus P1007 marks.

[ajaidev]

This may mean that AMD A8-3550 will be priced near i5-2300. I am not sure how much a HD5650 scores but i recall around P3500 pls correct me if i am wrong, in that aspect the performance seems really good. I am still waiting for a proper review after all Intel engg. again and again say there are bottle necks all over.

[Manicdan]

so how fast is the gpu on the A8-3560? or the two P models with an extra 35 watts of TDP

these are looking great for something like an imac where you cannot change out your gpu, and hopefully will take full use of opencl in the future.

[Nintendork]

The "P" moniker means 35w of TDP for the cpu, so probably a high clocked quad core vs non "p" versions.

A8 3550 vs A8 3550P


Same specs except TDP and cpu speed.

[flexcore]

The "P" moniker means 35w of TDP for the cpu, so probably a high clocked quad core vs non "p" versions.

A8 3550 vs A8 3550P


Same specs except TDP and cpu speed.

Where you find the chart?

[jimbo75]

AMD Expects Fusion "Llano" to Offer Three Times Higher Graphics Performance than Core i "Sandy Bridge".

According to AMD internal documents seen by X-bit labs, which describe performance of AMD's new products, AMD A8-3550 APU (quad-core, 4MB cache, 400 stream processors, Radeon HD 6550 graphics core at 594MHz) delivers three times higher performance in 3DMark Vantage test compared to Intel Core i5-2300 (quad-core, 2.80GHz/3.10GHz clock-speed, 6MB cache, Intel HD graphics core at 850MHz/1100MHz): P3335 versus P1007 marks.

An i5 2500K scores about P1600, i7 2600K about P2k I guess.

Those are all 95W desktop cpu's. This A8-3550 is 65W, possibly for high end notebooks.

Where you find the chart?

I'd guess xbit http://www.xbitlabs.com/news/cpu/dis...Documents.html

[Nintendork]

That's desktop. AMD mobile ones tops at 45w(current PII X4 mobile).

Llano mobile chip:
A8-3510MX (who knows if it's the high end one)

The one of the Llano vs SB video
A8 3510MX vs i7 2630QM 2Ghz

[jimbo75]

That's desktop. AMD mobile ones tops at 45w(current PII X4 mobile).

They need graphics too.

Llano mobile chip:
A8-3510MX (who knows if it's the high end one)

The one of the Llano vs SB video
A8 3510MX vs i7 2630QM 2Ghz

That was the 1.8ghz llano. You are probably right about it but I see no reason why 65W for an APU can't be doable on mobile.

Also I'm thinking the "6" in the number scheme is reserved for desktops.

[Manicdan]

Same specs except TDP and cpu speed.

those specs specifically say that gpu clocks can be different by mentioning TBD

if they were the same then it would have been listed as the same

[kl0012]

An i5 2500K scores about P1600, i7 2600K about P2k I guess.

Those are all 95W desktop cpu's. This A8-3550 is 65W, possibly for high end notebooks.



I'd guess xbit http://www.xbitlabs.com/news/cpu/dis...Documents.html

Around P2000 on i5-2520M (which is 35W). Any way, the number from xbitlabs is pretty low (1k less then for HD 5570 which was rumored to be equal to llano).

[jimbo75]

Around P2000 on i5-2520M (which is 35W). Any way, the number from xbitlabs is pretty low (1k less then for HD 5570 which was rumored to be equal to llano).

I have a few doubts about that tbh.

The i7 2820QM scores P2k in Vantage.



It may well be the case that the i5 5250M scores the same, If so it's clearly not indicative of actual gaming performance.





As you can see the 2520M isn't even close.

[Olivon]

AMD to Refresh Vision Brand Outlook Along with Launch of New Chips - XBit Labs

[Musho]

Dual graphics! Nice! Looks like Hybrid Crossfire might really take of to provide low power usage on low graphical workloads by using just the IGP, while giving great boosts in high workloads by using crossfire! Might make for some very interesting new laptops!

[Mats]

Thin, Lost-logo-style fonts seems to be the new black. WP7, Nokia Pure, and now AMD. Not that they're identical, tho.

[zerazax]

Not gonna lie, I was looking at Zacate for my HTPC rebuild, but with Llano apparently coming out a lot sooner than expected, I might well go for Llano if they come in Mini ITX flavors

That GPU looks awfully nice for casual gaming/indie games on the HTPC and will do all the HTPC functions nicely

[Macadamia]

Thin, Lost-logo-style fonts seems to be the new black. WP7, Nokia Pure, and now AMD. Not that they're identical, tho.

umm what?
Segoe 7/Nokia Pure are way more humanist sans-serifs than what AMD's using. They're meant for interfaces (and in Pure's case, big sizes for print marketing too) so they have to look pristine and have that human touch of uncertainty, aimed for readability.

What AMD's using here is either Avenir or Futura (or Twentieth Century, no idea since it's 5 in the morning), something very geometric (the O looks like a circle). It really has no other purpose than looking classy and recognizable (legibility), like the London Transport signage.

The Lost wordmark? meh, trash.