AMD Ontario APU pictured,die size ~77mm^2

[-Boris-]

Savantu, you know damn well that Intel specifies TDP to the decimal while AMD only has three classes. If a chip is at 10W it gets the 18W TDP range. If it is at 5.5W it is put in the 9W TDP range. It seems to me that the T40N will have higher performance than any dual core atom SOC, at the same or lower power.

[duploxxx]

one for the intel fanboys who are arguing the fact of the impossible 1W bobcat powerdraw.... an official amd slide

http://www.xbitlabs.com/news/cpu/dis...rformance.html

and aimed for following market:

“[Ontario] is a dual-core system-on-chip implementation and APU of the upcoming “Bobcat” core for ultrathin notebooks, netbooks and <20W new market products. Ontario is designed to offer a performance PC experience in a low-power design,”

so pls stop all speculation and compares, all are based on assumptions

[Hornet331]

Savantu, you know damn well that Intel specifies TDP to the decimal while AMD only has three classes. If a chip is at 10W it gets the 18W TDP range. If it is at 5.5W it is put in the 9W TDP range. It seems to me that the T40N will have higher performance than any dual core atom SOC, at the same or lower power.

And you know damn well that intel does the same...

Jfyi, intel also overspecs TDP on atoms, the 330 is speced with 8W but actually needs somewhere ~4W-5W range.

[-Boris-]

And you know damn well that intel does the same...

Jfyi, intel also overspecs TDp on atoms, the 330 is peced with 8W but actually needs somewhere ~4W.

Well I hope both overspec a bit. But if they just have ranges just like AMD they have at least 15 of them, some of them only 0.1W - 0.2W wide in the lowest power segment. And only 1W difference in the mid segment. It's nothing you can compare with AMDs 9W jump.

Therefore, if an Atom processor has a TDP of 6.5W you know it can't be to far off. That's why you can't compare them that way with AMDs Ontario where it could be 8W off.

[savantu]

Anyone want to bet that Savantu owns Intel stock?
..

i'll bet $1 that I don't. Accepted ?

Yep, clearly Ontario is already Epic Fail. It's not sub 1 watt and thats just a disgrace. Performance won't matter because we know its AMD and that is automatic fail, especially anything remotely related to BD arch. (as we know from its proven horribly lower IPC).
If I'm not buying it (because it fails) nobody else will.

Yeah, the Atom killer ended squarely in the 9-20w range, just where single and dual core SB will be.

He is financial analysist, most probably focused on Intel so yes, it's more than clear that he does. Thats what makes this all so worthless. People aren't arguing just for the sake of the truth, but the sake of money and profession. It gets bloody when money is involved, as seen with the overclocking scene. Same happens here. There was one fanboy who even admited that he had Intel stock which fed his passion.

It's waste of time to argue with someone who refuses to take a neutral stance on the subject. Which is understandable though; Intel's failures don't bring him the bread and the butter, Intel's success does.

As always, when don't have anything intelligent to say attack the person. Easier.

^^ Add on that that llano is c*ap too,being indefinitely slower than SB on CPU side,having GPU that is "obviously" memory BW starved and being delayed a year+ .
This concludes the joke section for today

The joke is on you. I could dig some older posts of yours on Ontario and Llano and show how certain you were of their features and schedules.
It would be more than embarrassing, but hey, most here have shorter memories than a squirrel and fail to connect what was discussed weeks and months ago to what is happening now.
One even asked "where is the Llano delay, AMD said it was planned for 2011 ?"...
No wonder why the human specie is in danger. Most have their gray cells in permanent C6 state.

Why are you predicting Bobcat's performance? How can you be sure that a 1.5Ghz Atom will compare to a 1Ghz Bobcat? My guess is that the Bobcat at 1Ghz will be faster than the 1.5Ghz Atom, so in your Case #1, AMD would have superior performance + superior GPU for the same power. Clear win I would say?

Because I can. And my prediction is up to 50% faster clock per clock. What's yours so we can revisit this in 4 months and see who was right ?

Also, its apparent that the disabled cores and GPU's are sucking power, or they are just worse binned with higher voltages. See the single core Bobcat: 1.5Ghz 18W. Dual core without GPU: 1.4Ghz 18W. Makes no sense when there is a 1.6ghz Bobcat + GPU for 18W.

Since parts are power gated, no they shouldn't be using any power.
Did you happen to have heard by any chance of the concept of having different CPU bins ?
You see, not all parts have the same power characteristics. You have parts which are leakier and more power hungry. Instead of tossing them away, you put them in another TDP range. That's why some single cores will burn a lot of power, but instead of being discarded they are sold at a discount in a higher TDP range.

It seems fairly obvious that the ontario lineup posted and atom lineup extend into different market segments. They overlap in the middle where high-end Atom and low-end Ontario meet. But mostly it looks like AMD is targeting these first chips to fill the gulf between Atom and CULV that Intel has left open. It should be a nice boost to all those disgustingly slow netbooks and tablets.

If AMD is working on a sub 1W design, it clearly isn't in this lineup. It might not be ready, need a new process, or whatever. But if I was AMD I wouldn't be to eager to enter that sector of the market anyway because ARM is going to eat x86 alive.

This forum would definitely need at least a dozen posters like ^^. Otherwise, it is just crawling in the primeval slime from an intelectual POV.

a few things,
first a bobcat cpu with no gpu does not sound like something we would see in consumer products like netbooks, it has to get a gpu from somewhere else, so why get a single core, power sucking chip, in a devise that runs on a battery. sounds like they have a special purpose that we wouldnt see in conventional forms.

second, a single core with gpu is 9W, a duel core with gpu is also 9W, the difference is a 200mhz drop to add that second cpu. sounds like its pretty low power to me. the fact everything is 5 or 9 or 18 watts makes it really tough to infer any exact power consumption per core.

Did you happen, by any chance, to hear about wafer defects ? You see, wafers have a thing called defect density. When you have a defect, instead of throwing away that die, you could disable, let's say, the GPU if the defect was located there.

no. paul is an experienced engineer who knows his stuff although he is biased.

savantu is an experienced shill who cant help it when he takes things out of context.

Yeah, I'm taking things out of context. Probably because of me Bobcat can't touch handhelds and low power stuff even with a 10 ft pole. Then again, Llano being delayed to the summer by AMD's own admission is another thing taken out of context. Probably they feel no need to hurry up and save what can be saved from corporate refresh cycle.

Shills are also apologists and deniers of reality. A category where many would find a perfect fit.

one for the intel fanboys who are arguing the fact of the impossible 1W bobcat powerdraw.... an official amd slide

http://www.xbitlabs.com/news/cpu/dis...rformance.html

and aimed for following market:

“[Ontario] is a dual-core system-on-chip implementation and APU of the upcoming “Bobcat” core for ultrathin notebooks, netbooks and <20W new market products. Ontario is designed to offer a performance PC experience in a low-power design,”

so pls stop all speculation and compares, all are based on assumptions

So, we have the slide which says "sub 1w capable". Too bad they forgot to show that slide to the design team. Maybe, then, they would have made it in-order.

[Hornet331]

It seems fairly obvious that the ontario lineup posted and atom lineup extend into different market segments. They overlap in the middle where high-end Atom and low-end Ontario meet. But mostly it looks like AMD is targeting these first chips to fill the gulf between Atom and CULV that Intel has left open. It should be a nice boost to all those disgustingly slow netbooks and tablets.

If AMD is working on a sub 1W design, it clearly isn't in this lineup. It might not be ready, need a new process, or whatever. But if I was AMD I wouldn't be to eager to enter that sector of the market anyway because ARM is going to eat x86 alive.

Indeed, the current roadmaps suggest this.

[Manicdan]

we should have a rule. for every negative thing you say about a product, you have to say something positive too. see how many people can show their true colors by not being able to follow that rule

[Hornet331]

we should have a rule. for every negative thing you say about a product, you have to say something positive too. see how many people can show their true colors by not being able to follow that rule

Lol stupid rule.. if something is bad, its for a reason.

[Solus Corvus]

Yeah, I'm taking things out of context. Probably because of me Bobcat can't touch handhelds and low power stuff even with a 10 ft pole. Then again, Llano being delayed to the summer by AMD's own admission is another thing taken out of context. Probably they feel no need to hurry up and save what can be saved from corporate refresh cycle.

Shills are also apologists and deniers of reality. A category where many would find a perfect fit.

I'd hardly call chumbucket a shill. He has been quite evenhanded in AMD vs Intel, IMO.

I'd indeed suggest that you are taking things out of context. It should be obvious that these TDPs aren't aimed at things like smartphones. With some decent x86 cores and a decent GPU it would make more sense in something like a netbook or tablet. You could watch video and do 3d and still get decent battery life. It might be a very good product for those sectors.

Your argument seems to be "it's pathetic because it doesn't fit in a market segment I picked that it clearly wasn't designed to fit in". Well we agree, nobody is going to put this ontario in a smartphone.

[Manicdan]

Lol stupid rule.. if something is bad, its for a reason.

if there is a debate, then its both good and bad. and if certain people cannot properly handle a debate by understanding others perspectives, then i think they should go back to playing in the sandbox with other kids. if everyone agrees its a bad product, then no need to have a forum discussion about it

[RaV[666]]

Yeah, the Atom killer ended squarely in the 9-20w range, just where single and dual core SB will be.

At a different die size and different price and with most probably better IGP (for sure on the software side)

One even asked "where is the Llano delay, AMD said it was planned for 2011 ?"...
No wonder why the human specie is in danger. Most have their gray cells in permanent C6 state.

It was, it was meant to be available in Q1 2011.Shipping end of Q4 2010(and thats what we know is pushed back)

Since parts are power gated, no they shouldn't be using any power.
Did you happen to have heard by any chance of the concept of having different CPU bins ?
You see, not all parts have the same power characteristics. You have parts which are leakier and more power hungry. Instead of tossing them away, you put them in another TDP range. That's why some single cores will burn a lot of power, but instead of being discarded they are sold at a discount in a higher TDP range.

The eOntario roadmap isnt official one, it could very well be that there will be scrap parts destined for VERY low niche markets (better to sell broken chip for low $ than to scrap it).It IS NOT however official ontario roadmap.

Atom isnt in the too many (any?) handhelds too, bobcat from a consumer standpoint is targeted at nettops/netbooks/tablets, which it fits perfectly, with better performance than atom(most probably MUCH better)

So, we have the slide which says "sub 1w capable". Too bad they forgot to show that slide to the design team. Maybe, then, they would have made it in-order.

Thats exactly kind of comment that goes in the troll bin.This slide is about one bobcat core without GPU.We dont have such a product in the most imminent lineup, it doesnt mean it isnt possible tho.Amd does not have money for many different dies.What was the point of such comment ?

[Chumbucket843]

Yeah, I'm taking things out of context. Probably because of me Bobcat can't touch handhelds and low power stuff even with a 10 ft pole. Then again, Llano being delayed to the summer by AMD's own admission is another thing taken out of context. Probably they feel no need to hurry up and save what can be saved from corporate refresh cycle.

Shills are also apologists and deniers of reality. A category where many would find a perfect fit.

correct me if i am wrong here but the slide said "sub 1watt capable". this does not mean that it has to consume =<1 watt but if they want to bring a 1 watt chip to the market they can, and will probably have no trouble beating intel, who really dont give a rats ass about the low power market. your an analyst. right? well, havent you noticed that AMD is taking low risks? they are fireselling magny cours, they are going for small die gpu's even when they dont need to, reusing lots of IP, etc. why would they launch a chip for a market that they would have to create when there are existing ones left untapped?

and you do tend to blow things out of proportion, especially with the subject of validation for example. do you really think that they would tape out a chip that could not even function? validation is more behavioral testing and fixes can be done in microcode thus not cause a 6 month delay.

[savantu]

...


Thats exactly kind of comment that goes in the troll bin.This slide is about one bobcat core without GPU.We dont have such a product in the most imminent lineup, it doesnt mean it isnt possible tho.Amd does not have money for many different dies.What was the point of such comment ?

Read again what he said .

one for the intel fanboys who are arguing the fact of the impossible 1W bobcat powerdraw.... an official amd slide

He's countering the released bin specification ( 800MHz, single core, no GPU being 5w ) with a powerpoint slide. This isn't hilarious, this is sad, the level of denial.
To any rational mind, Ontario will perform very well in the 8-18w range. But it is far from touching the really low power stuff, smartphones and handhelds. The AMD slide , not to call it vaporware, must be referring to future iterations of the product.

[ajaidev]

IMO compare Ontario to CULV and Arrandale ULV chips, after all its performance is 90% main stream.

EDIT: Forgot to write a reason since 18 W TDP Celeron and Pentium processors are rumored to come soon, i reckon they will be competition to Ontario. The slowest Arrandale we have now a days is the i3-330UM @ 1.2ghz but in most cases i dont think a 1.6Ghz Ontario will beat it CPU wise but i can be wrong.

[flippin_waffles]

Bonic does not measure memory & multithreading performance.
But you can get some impression about the perf diff. between Atom D510 and Athlon x2 3250e.
http://www.anandtech.com/bench/Product/110?vs=116
As you can see, the average perf. diff. in the real world is much less then 2x.

What's real world got to do with anything? AMD has been promoting real world testing and usage scenarios for years now, yet has fallen on deaf ears for the most part.

-journalists/reviewers still use 800x600 gaming resolutions to test CPU performance.
-journalists/reviewers still use SPi to test CPU performance.
-journalists/reviewers still use $1000 CPU's to to test $100 GPU's
-journalists/reviewers still compare $1000 CPU's to $100 CPU's.
-journalists/reviewers still test notebook battery life at the lowest and highest extremes. Nothing in between.

And the list goes on and on...

Boinc is a lot more real world than anything listed above.

[RaV[666]]

Read again what he said .

You mean chumbucket ? I read it just now as he posted after me, yup. Hes right and youre wrong :P
If it didnt get to you yet i will kind of repeat.
AMD isnt targeting smartphone market with current ontario chips, while bobcat core has the potential ability to enter that market if needed.
Not the ontario dual core chip with gpu fused in .

[Hans de Vries]

Certainly. Bobcat isn't living at all to the hype you and others have created. Where is the sub 1w CPU that informal was bragging about ?

Where it is? on the die of course!

The sub 1W capable core is on the 10 times larger SOC die of course.
Do you really think AMD is lying in all these presentations?

Do you realize that driving the memory bus to a SODIMM or other
external memory may already take several Watt? That's why smart
phones have the DRAM in the package.

Have you done your homework and read all the fine print of how
some ultra low TDP's are measured and defined. Do you understand
that you are comparing with an SKU which is NOT binned for ultra
low power consumption at all but is just the cheapest bin for small
embedded applications. The 5W part is defined with (cheaper)
1.5V memory I/O devices and not even low power 1.35V LPDDR or
a more appropriate mobile POP setup.

You found this on an investor board as always didn't you? From
"somebody" who claims that an Atom core is an order of magnitude
more power efficient as a Bobcat core, yes? and you do believe this.....


Regards, Hans

[Hornet331]

What's real world got to do with anything? AMD has been promoting real world testing and usage scenarios for years now, yet has fallen on deaf ears for the most part.

-journalists/reviewers still use 800x600 gaming resolutions to test CPU performance.
-journalists/reviewers still use SPi to test CPU performance.
-journalists/reviewers still use $1000 CPU's to to test $100 GPU's
-journalists/reviewers still compare $1000 CPU's to $100 CPU's.
-journalists/reviewers still test notebook battery life at the lowest and highest extremes. Nothing in between.

And the list goes on and on...

Boinc is a lot more real world than anything listed above.

-valid (do you want to test gpu performance on a cpu review?, also tehy test more realworld resolution)
-hardly anymore, show me a recent review who does.
-valid, since with flexible multis you get the exact same performance a lower cpu would offer
-valid, since you want to see what performance you get for what money
-valid, since they cover both extremas so you know you can expect a runtime inbetween those two points. Also your implying they don't test normal usage scenarios.

Boinc cpu test falls in the same categorie as sandra or spi, pure synth test. If they would have provided actual WU/day figures it would be far more usefull.

[LightSpeed]

Because I can. And my prediction is up to 50% faster clock per clock. What's yours so we can revisit this in 4 months and see who was right ?

You want mine? 70%. Im not here to see who was right or wrong in 4 months, might be you or me or neither of us, we'll know that in time. However, let me paste the case you pointed out again:

Dual cores
Bobcat: T40N - 9W, 1.0 GHz, dual core, GPU, LVDDR3
vs.
Atom: N550 - 8.5W, 1.50 GHz, dual core, GPU

Evenly matched performance wise and power wise, GPU favours Ontario

So over here, the TDP is the same, performance is the same (by your predictions), and Bobcat has a much faster GPU. Clean win I would say, atleast in this case?
So is AMD superior to Intel in this occasion ? Or are you having a hard time admitting it? Or are you gonna say we cant predict the performance of Bobcat's GPU, the Atom's might be superior? You also completely ignored by next paragraph, let me paste it again for you:

Also, its apparent that the disabled cores and GPU's are sucking power, or they are just worse binned with higher voltages. See the single core Bobcat: 1.5Ghz 18W. Dual core without GPU: 1.4Ghz 18W. Makes no sense when there is a 1.6ghz Bobcat + GPU for 18W.

Hence, while its perfectly OK to compare those SOC's mentioned above as well since they are a part of the lineup, the dualies with GPU seem to have much superior perf/watt and lets see what they consume at 800Mhz. Should be lower than 9W

[flippin_waffles]

-valid (do you want to test gpu performance on a cpu review?, also tehy test more realworld resolution)
-hardly anymore, show me a recent review who does.
-valid, since with flexible multis you get the exact same performance a lower cpu would offer
-valid, since you want to see what performance you get for what money
-valid, since they cover both extremas so you know you can expect a runtime inbetween those two points. Also your implying they don't test normal usage scenarios.

Boinc cpu test falls in the same categorie as sandra or spi, pure synth test. If they would have provided actual WU/day figures it would be far more usefull.

-Not valid. The premise is real world scenarios. How does gaming at 800x600 reflect CPU performance in the real world? How about CPU usage scenarios that reflect what customers do with their computer.
-Granted SPi is used in limited cases today. But is real world different today than it was 3 years ago when it was pervasive across almost every review? I suppose I should have said sandra and other synthetic tests.
-Not valid, not real world, which means something a customer uses in the real world.
-Valid. Doesn't belong in the context of a real world discussion. My point here is journalist/reviewers comparing $100 to $1000 CPU's in pure performance alone, rather than a performance/price comparison which happens a lot.
-Not valid. The runtime vary so much between fully idle and fully loaded, that it tells buyers exactly zero. How about some real world testing with something customers actually do with their notebooks. Have you ever bought one to let it sit in the corner and idle? Or buy one to fold until the battery runs out, recharge it and start all over again?

"Boinc falls in the same category as sandra and spi, pure synthetic". One is as valid as the other.

The Phoronix Test Suite is probably the best software a reviewer could use to indicate real world performance.

http://www.phoronix-test-suite.com/

And it runs on Win7.

[Chumbucket843]

superpi is not synthetic. people just use that word to cover up bad performance.

if you knew a thing or two about the algorithms used you might not consider them synthetic. if it is then essentially all that work fourier did was useless for science and computation.

[Manicdan]

-Not valid. The runtime vary so much between fully idle and fully loaded, that it tells buyers exactly zero. How about some real world testing with something customers actually do with their notebooks. Have you ever bought one to let it sit in the corner and idle? Or buy one to fold until the battery runs out, recharge it and start all over again?

this is a very important thing that needs to be tested carefully. if your using your laptop to watch a movie, how many can you watch before it dies. it factors in cpu efficiency and screen brightness (which is important that they be set to a similar standard if you test multiple, or atleast mention the brightness used and how it felt). or if you work with excel all day, it might be idle 90%, but when it is loaded does it use half power or full power.

i think some macros need to be made to mimic light use for facebooking and word processing and see how many hours can be obtained. and again, the screen can play a big part and needs to be mentioned but is probably often forgotten about.

when testing a system, all factors are affecting the result, and a good reviewer knows know to isolate certain parts and play fairly

[flippin_waffles]

superpi is not synthetic. people just use that word to cover up bad performance.

if you knew a thing or two about the algorithms used you might not consider them synthetic. if it is then essentially all that work fourier did was useless for science and computation.

Are there a lot of real world modern consumer software that still use x87 code? I can't think of any off the top of my head, but maybe i've missed a niche that you could pigeon hole my argument into.

[Hornet331]

Your to fixated about realworld.. since most of the reviews only condisder this as one aspect.

[Hornet331]

Are there a lot of real world modern consumer software that still use x87 code? I can't think of any off the top of my head, but maybe i've missed a niche that you could pigeon hole my argument into.

PhysX.