Low & Mid end Kaveri getting late - AMD mistake that will cost them?

[ElFenix]

i was about to come in complaining about no new amd mobile parts for years, and i come to find out that though richland is 6000 series on the desktop, it's a 5000 series part for mobile. i thought the laptops i was looking at had trinity.

being halfway informed is worse than being uninformed. but that's a really bad naming convention, because even uninformed people will say "hey, i can buy a 7000 series part for desktop, but only 5000 for mobile?"

 

[USER8000]

I envision a different problem. Customers will see a $350 15" A6(beema) laptop and a $400 15" A6(kaveri) laptop. They wont understand why the price difference exists and will choose the cheaper beema A6, further eroding AMDs big cores.

I know people who bought Core i5 and Core i7 laptops thinking there were quad cores,but in fact they were dual cores,and they wanted a quad core chip.

It does not get any better with the Celerons,were the BT and Haswell based ones are not clearly differentiated in many UK stores including the largest ones.

AMD and Intel both have weird naming schemes.

 

[erunion]

I know people who bought Core i5 and Core i7 laptops thinking there were quad cores,but in fact they were dual cores,and they wanted a quad core chip.

It does not get any better with the Celerons,were the BT and Haswell based ones are not clearly differentiated in many UK stores including the largest ones.

AMD and Intel both have weird naming schemes.

Agreed on all parts. But Intel's scheme doesn't water down their brands, I think AMD's will.
(The exception being celeron/pentium, but I think Intel plans to transition those brands to majority atom skus anyways).

 

[frozentundra123456]

The average consumer doesnt know each SKU name and specs, hell even I dont remember every AMD and Intel SKU naming and specs and i have to find them online.



Maybe, but that happens the last 30 years or more. It is the job of the retailer to inform the customer if he/she is uniformed. Or the Consumer have to make a research about the products available in the price point he/she is interested in. That is how its done all those years, why the sudden complaining and bizarre commends about OEM inability to understand each SKUs market place etc ??

Retailer informing the customer?? Are you serious? The retailer trying to sell me something would be the last person I would trust to give me unbiased advice or even expect to be knowledgable. Maybe in the old days of small shops that stayed around and tried to develop a good reputation that could happen. But in this day of big box stores, minimum wage short term employees and commission sales, it is hardly going to happen. Not to mention a big portion of the sales now are from on-line where there is not salesperson to even ask a question. Actually though, I think the new Beema/Mullins nomenclature is not as bad as Kabini was or the new Celeron/Pentium atoms.

 

[AtenRa]

One more thing, I havent seen if it can be done in any other motherboard as of now but ASUS A88XM-Plus can change the TDP of the A10-7700K all the way down to 45W from within the BIOS. You just dial the TDP number you want and you have an A8-7600 at 65W or 45W TDP.
I will upload photos tomorrow

Default value is at Auto,
Lowest value is 45, you can raise by 1 point each time, highest value is 65.

I will test various values and report the power consumption and Frequencies when ill get more time.

 

[AtenRa]

Retailer informing the customer?? Are you serious? The retailer trying to sell me something would be the last person I would trust to give me unbiased advice or even expect to be knowledgable. Maybe in the old days of small shops that stayed around and tried to develop a good reputation that could happen. But in this day of big box stores, minimum wage short term employees and commission sales, it is hardly going to happen. Not to mention a big portion of the sales now are from on-line where there is not salesperson to even ask a question. Actually though, I think the new Beema/Mullins nomenclature is not as bad as Kabini was or the new Celeron/Pentium atoms.

Then you only left with having done your own research to find out what each SKU brings to the table. If you do your research you should be able to know that Beema/Mullins A-series is not the same as Richland A-series.

 

[Erenhardt]

snip

Damn, I would love to play with it but I don't think I can justify another system

 

[raghu78]

I envision a different problem. Customers will see a $350 15" A6(beema) laptop and a $400 15" A6(kaveri) laptop. They wont understand why the price difference exists and will choose the cheaper beema A6, further eroding AMDs big cores.

Mobile Kaveri models are in the A8 and A10 range. You could have die harvested A6 Kaveri but those will be severly gutted in CPU core count (1M/2C) and graphics cores (256 or 192) cores.

 

[pw257008]

Mobile Kaveri models are in the A8 and A10 range. You could have die harvested A6 Kaveri but those will be severly gutted in CPU core count (1M/2C) and graphics cores (256 or 192) cores.

yeah i see no point in anything other than a very low end 2 core part from AMD

edit: poor grammar. what i meant was: i see no point in a 2 core part from AMD except at the very low end.

 

[shady28]

Don't be silly. If its confusing for customers it is bad for OEMs. OEMs will reject bad marketing. Look at WinRT, OEM were very vocal about it causing confusion and they refused to support it


Kaveri and Beema will be in similar form factor budget laptops.

Your argument is true but spans all of the major chipmakers for the past 10 years.

Do you really think a typical consumer looking at a PC knows the difference between an Pentium G620 and a Pentium G3220? They may infer that the G3220 is faster due to the larger number - but that's an inference that won't always be true. For example, an i5-4440S is slower than an i5-4430.

I agree with your sentiment that the chip naming convention is retarded - but it's been retarded for a decade. If anything, AMDs is easier to figure out than Intels.


Consumers basically stopped understanding the chip "names" when they stopped actually naming things like "Pentium", "Pentium MMX", "Pentium II", "Pentium M(obile)" etc with a Mhz rating.

As soon as they started doing all this "E6xxx" "E8XXX" "P8XXX" "Q8XXX" etc garbage, most people stopped trying to keep up. To this day, off the top of your head, can you tell me which is faster : E6850 or E8300? Most people would look at the Mhz on the 8300 and assume its slower, but of course we know better than that. Most of the time. Well it depends on what you're doing. Why would a consumer want to deal with that?

I have no doubt that factored into things like the rise of Apple. It just works. Who cares what's in it. Do you know what kind of CPU your microwave has? Why should we care?

 

[erunion]

Mobile Kaveri models are in the A8 and A10 range. You could have die harvested A6 Kaveri but those will be severly gutted in CPU core count (1M/2C) and graphics cores (256 or 192) cores.

As far as I know only DT chips have been announced. Richland mobile offered A4s through A10s. I presume Kaveri will offer A6s. But it is possible they will move Kaveri towards the high end with beema filling the gap.

 

[pw257008]

Your argument is true but spans all of the major chipmakers for the past 10 years.

Do you really think a typical consumer looking at a PC knows the difference between an Pentium G620 and a Pentium G3220? They may infer that the G3220 is faster due to the larger number - but that's an inference that won't always be true. For example, an i5-4440S is slower than an i5-4430.

I agree with your sentiment that the chip naming convention is retarded - but it's been retarded for a decade. If anything, AMDs is easier to figure out than Intels.


Consumers basically stopped understanding the chip "names" when they stopped actually naming things like "Pentium", "Pentium MMX", "Pentium II", "Pentium M(obile)" etc with a Mhz rating.

As soon as they started doing all this "E6xxx" "E8XXX" "P8XXX" "Q8XXX" etc garbage, most people stopped trying to keep up. To this day, off the top of your head, can you tell me which is faster : E6850 or E8300? Most people would look at the Mhz on the 8300 and assume its slower, but of course we know better than that. Most of the time. Well it depends on what you're doing. Why would a consumer want to deal with that?

I have no doubt that factored into things like the rise of Apple. It just works. Who cares what's in it. Do you know what kind of CPU your microwave has? Why should we care?

:thumbsup: spot on. Another example: Most DT i3s will be faster than most mobile i5s. The i3s lack turbo, but don't need it, as their base clock is higher than a mobile i5 turbo anyway. This holds true for any of the U level i7s as well, and in single or lightly threaded applications (Intel's bread and butter vs. AMD), even some of the mobile quad core i7s will be slower than a desktop i3. See this for a couple examples: http://ark.intel.com/compare/75117,75119,77771

 

[raghu78]

As far as I know only DT chips have been announced. Richland mobile offered A4s through A10s. I presume Kaveri will offer A6s. But it is possible they will move Kaveri towards the high end with beema filling the gap.

Beema is for E1,E2,A4,A6. Kaveri is most likely for A4,A6,A8,A10. Fully enabled Beema is 2 tiers below fully enabled Kaveri. So there is no confusion here. Even Intel has Haswell Celerons and Baytrail Celerons. The difference is Baytrail Celerons are fully enabled and Haswell Celerons are die harvested.

 

[frozentundra123456]

Beema is for E1,E2,A4,A6. Kaveri is most likely for A4,A6,A8,A10. Fully enabled Beema is 2 tiers below fully enabled Kaveri. So there is no confusion here. Even Intel has Haswell Celerons and Baytrail Celerons. The difference is Baytrail Celerons are fully enabled and Haswell Celerons are die harvested.

Not sure what you mean by the two teirs below Kaveri. Are you saying a Beema A6 is the same performance as a Kaveri A6? Still seems to me to be confusing. You have the same model class for two different architectures.

It would have been clearer to just give all beema models the "Exxxx" nomenclature and all Kaveri models the "Axxxx" nomenclature, just like there is no overlap in naming between Kaveri and FX (at least not yet).

I do agree though that Intel's naming scheme is at least as bad or worse.

 

[pw257008]

Not sure what you mean by the two teirs below Kaveri. Are you saying a Beema A6 is the same performance as a Kaveri A6? Still seems to me to be confusing. You have the same model class for two different architectures.

It would have been clearer to just give all beema models the "Exxxx" nomenclature and all Kaveri models the "Axxxx" nomenclature, just like there is no overlap in naming between Kaveri and FX (at least not yet).

I do agree though that Intel's naming scheme is at least as bad or worse.

Depending on the SKU and task, Kabini and mobile Richland A6s will probably trade performance blows. Kabini A6 has double the cache and double the cores (2 MB instead of 1 MB) and comparable IPC to Richland, though lower clock speeds. I'd assume the case will be similar with Kaveri v Beema comparatively, since Kaveri boosts IPC while Beema boosts clock speeds, especially in cases where it would have fallen furthest behind Richland (single threaded workloads). Now, this is a CPU side discussion. GPU-wise, Kaveri will probably have a larger advantage.

 

[AtenRa]

A few notes about APU naming,

Mobile Richland/Kaveri have the letter (M) at the end of the naming. So, an A6-xxxxM is Richland/Kaveri. Also, Richland/Kaveri iGPU also have the leter (G), 8xxxG

Kabini iGPU is HD series
Beema/Mullins iGPU is R series

DT Richland A6-xxxx, iGPU = HD8xxxD
Mobile Richland A6-xxxM, iGPU = 8xxxG
Kabini A6-5xxx, HD8xxx
Beema A6-6xxx, iGPU R series
Mullins A10-Micro, iGPU R series

Easier now ??

Edit: DT Richland naming starts from 64xx, Beema have up to 63xx.
So you may have Richland A6-6400K but you have Beema A6-6310

 

[shady28]

:thumbsup: spot on. Another example: Most DT i3s will be faster than most mobile i5s. The i3s lack turbo, but don't need it, as their base clock is higher than a mobile i5 turbo anyway. This holds true for any of the U level i7s as well, and in single or lightly threaded applications (Intel's bread and butter vs. AMD), even some of the mobile quad core i7s will be slower than a desktop i3. See this for a couple examples: http://ark.intel.com/compare/75117,75119,77771

You're example is better than mine

Yeah, I have often wondered how many people went to buy a desktop, saw a $800 i5 desktop rig and then saw a $500 laptop i5, and figured "hey I can get the same processor in the laptop so I'll just buy that."

Of course *we* know that a laptop i5 would have to be huffing it to keep up with a desktop i3.

I think this, among many other factors, has helped speed along the decline of the desktop and the PC market in general.

What would make a lot more sense is their code names.

Haswell Dual
Haswell Quad
Haswell Quad Extreme
Haswell Dual M
Haswell Quad M

Can you guess what chips I'm referring to above?
Haswell D = i3 desktop
Haswell Q = i5 Desktop
Haswell Q Extreme = i7 desktop
Haswell DM = i3/i5 mobile
Haswell QM = i7 mobile

What they are doing now is like selling you a car and telling you it's an S197.

You know what an S197 is? It's a freaking Mustang.

 

[shady28]

A few notes about APU naming,

Mobile Richland/Kaveri have the letter (M) at the end of the naming. So, an A6-xxxxM is Richland/Kaveri. Also, Richland/Kaveri iGPU also have the leter (G), 8xxxG

Kabini iGPU is HD series
Beema/Mullins iGPU is R series

DT Richland A6-xxxx, iGPU = HD8xxxD
Mobile Richland A6-xxxM, iGPU = 8xxxG
Kabini A6-5xxx, HD8xxx
Beema A6-6xxx, iGPU R series
Mullins A10-Micro, iGPU R series

Easier now ??

Edit: DT Richland naming starts from 64xx, Beema have up to 63xx.
So you may have Richland A6-6400K but you have Beema A6-6310

It still really makes no sense to use such obtuse naming conventions. It's like you need a Captain Crunch decoder ring to figure out current CPUs.

It's like instead of calling a car a Ford Fusion, or Chevrolet Tahoe, they started advertising the part of the VIN number that identifies the model.

Look at this awesome new 1xHHD124dH!

Really?

 

[pw257008]

You're example is better than mine

Yeah, I have often wondered how many people went to buy a desktop, saw a $800 i5 desktop rig and then saw a $500 laptop i5, and figured "hey I can get the same processor in the laptop so I'll just buy that."

Of course *we* know that a laptop i5 would have to be huffing it to keep up with a desktop i3.

I think this, among many other factors, has helped speed along the decline of the desktop and the PC market in general.

What would make a lot more sense is their code names.

Haswell Dual
Haswell Quad
Haswell Quad Extreme
Haswell Dual M
Haswell Quad M

Can you guess what chips I'm referring to above?
Haswell D = i3 desktop
Haswell Q = i5 Desktop
Haswell Q Extreme = i7 desktop
Haswell DM = i3/i5 mobile
Haswell QM = i7 mobile

What they are doing now is like selling you a car and telling you it's an S197.

You know what an S197 is? It's a freaking Mustang.

:thumbsup:

 

[blackened23]

Remember back when we could just use clockspeed as the naming convention. Before the MHZ/GHZ race ended anyway.

My take on it is, especially with mobile devices, consumers don't pay attention as much as they used to. But OEMs do care. Apple will not put a sub-par media tek SOC in their ipad, for instance. They want high end products for their high end devices, within reason of course - of course Apple can't put the 4930MX in their macbook air, but they put the best CPU possible in for the best combination of performance per watt and battery life with the Haswell i5's. So the OEMs care more with the mobile products - the premium products won't use sub-par CPUs or components, generally speaking. But it still has to fit within the proper performance per watt metrics, idle standby, battery life and all that sort of thing. But..yeah at face value some of the naming conventions are strange. If you want to be informed on that sort of thing, and power users generally are, you can be. The average joe probably just buys their favorite brand that gets great reviews at websites like theverge or engadget, without necessarily having that knowledge.

 

[jpiniero]

Haswell Dual
Haswell Quad
Haswell Quad Extreme
Haswell Dual M
Haswell Quad M

Can you guess what chips I'm referring to above?
Haswell D = i3 desktop
Haswell Q = i5 Desktop
Haswell Q Extreme = i7 desktop
Haswell DM = i3/i5 mobile
Haswell QM = i7 mobile

What they are doing now is like selling you a car and telling you it's an S197.

You know what an S197 is? It's a freaking Mustang.

You do realize that's what Intel basically used for the original Core 2? I think Intel is overdue for a brand overhaul, but they are clearly not going back to something like that.

 

[shady28]

Remember back when we could just use clockspeed as the naming convention. Before the MHZ/GHZ race ended anyway.

My take on it is, especially with mobile devices, consumers don't pay attention as much as they used to. But OEMs do care. Apple will not put a sub-par media tek SOC in their ipad, for instance. They want high end products for their high end devices, within reason of course - of course Apple can't put the 4930MX in their macbook air, but they put the best CPU possible in for the best combination of performance per watt and battery life with the Haswell i5's. So the OEMs care more with the mobile products - the premium products won't use sub-par CPUs or components, generally speaking. But it still has to fit within the proper performance per watt metrics, idle standby, battery life and all that sort of thing. But..yeah at face value some of the naming conventions are strange. If you want to be informed on that sort of thing, and power users generally are, you can be. The average joe probably just buys their favorite brand that gets great reviews at websites like theverge or engadget, without necessarily having that knowledge.

Exactly - people don't look at the benchmark sites much, if they look at all they're actually more likely to be reading reviews on bestbuy.com or amazon.com than even engadget and theverge.

With that said, one of the biggest things on any high end mobile device SoC is its GPU. There's a reason for that - the buggers are higher rez now than most people's desktop monitors, and smooth scrolling UI along with decent 3D is high on the list of user experience needs when you're using a touch screen.

AMD may now be better positioned in that regard, but I don't think either Intel or AMD are doing anything outstanding. They get knocked around pretty hard by the A7, high end Qualcomm chips, and even the older A6X in the iPad 4.

So to your point about high end parts - Bay Trail isn't it, and I doubt Kaveri is either. Look at where the iPad 4 (A6X), iPad Air (A7), and SnapDragon 800 sit in relation to the Z3470 Asus T100 \ BayTrail FFRD :

i.e. :

 

[blackened23]

Yes, customers care less about SKU naming but OEMs care very much about performance per watt, idle standby, and overall balanced performance. So my point is, even though the end buyer may not be the know it all of every SKU intel offers, rest assured that the OEM will get a premium part if they're positioning it in a premium device. If they're designing a low cost budget device (AMD's area in mobile devices) then maybe not.

AMD doesn't really have anything meaningful to compete with the U and Y core Haswell CPUs for macbook type devices. Do end consumers know this? Not all of them. But Apple does know this. And that's why AMD is not in the macbook air/pro and intel is.

That's my point: even if the consumer is none the wiser, if it's a premium product, the OEM will put the premium parts in. Period. Hence Apple chose the best CPU possible for the macbook air and the macbook pro. Even if their consumers don't know, it ultimately doesn't matter. They (Apple) know what's best in terms of the best performance per watt for the performance area that they're targeting.. and that is the ULV Haswell CPUs by a mile.

If the OEM wants to make a cheap budget device, then they will use cheap budget parts. The customer may be a complete idiot but the OEM isn't. The OEM will act accordingly to make their device the best possible. Therefore they won't put a crap CPU in just because the consumer is none the wiser - the consumer will find out anyway once the unit gets reviewed.

 

[shady28]

Yes, customers care less about SKU naming but OEMs care very much about performance per watt, idle standby, and overall balanced performance. So my point is, even though the end buyer may not be the know it all of every SKU intel offers, rest assured that the OEM will get a premium part if they're positioning it in a premium device. If they're designing a low cost budget device (AMD's area in mobile devices) then maybe not.

AMD doesn't really have anything meaningful to compete with the U and Y core Haswell CPUs for macbook type devices. Do end consumers know this? Not all of them. But Apple does know this. And that's why AMD is not in the macbook air/pro and intel is.

That's my point: even if the consumer is none the wiser, if it's a premium product, the OEM will put the premium parts in. Period. Hence Apple chose the best CPU possible for the macbook air and the macbook pro. Even if their consumers don't know, it ultimately doesn't matter. They (Apple) know what's best in terms of the best performance per watt for the performance area that they're targeting.. and that is the ULV Haswell CPUs by a mile.

If the OEM wants to make a cheap budget device, then they will use cheap budget parts. The customer may be a complete idiot but the OEM isn't. The OEM will act accordingly to make their device the best possible. Therefore they won't put a crap CPU in just because the consumer is none the wiser - the consumer will find out anyway once the unit gets reviewed.

You seem to be confusing Apple with (all other) OEMS, and revising Apple's history a bit.

My first Mac was an iBook G3, 2nd was an eMac G4. The first Apple devices with Intel chips were the Core Duo line. Core Duo was not better than AMD's Athlon X2 line on the desktop, though Intel mobile was arguably better at the time. That didn't happen until Core 2.

I think Apple went with Intel because it was a safe bet. Apple needed a mobile chip though. The G5 at the time was fine for high end desktops, but laptops were booming and they needed a good mobile G4 which wasn't forthcoming. So they stopped fighting fate and went x86. I think they had their share of taking risks with CPU makers.

Most OEM tablets are cheap crap. Samsung puts MediaTek processors in the Galaxy Tab 3 7.0 They use a 2 year old Exynos chip in the Tab 3 8.

The Tab 3 10.1 used a Cedar Trail Atom - and got nailed on review after review for sluggish performance. You can pick one of those up for ~$200 now. The Tab 4 10.1 upgraded to a Snapdragon 400 - bottom of the Qualcomm Snapdragon line...

Their premier Galaxy Tab Pro products use Snapdragon 800 and Exynos Octa.

Again, I think neither AMD nor Intel really has a premier high end SoC for tablets yet.

Intel should be worried though. The importance of the iGPU is underrated by a PC industry that is used to desktops. an industry that for a decade derided integrated GPUs as garbage due to the abysmal performance of Intel iGPUs. Intel has a direct competitor in AMD in this segment, but even if AMD fails it does not mean that Intel will win. It may just mean that neither of them can hack it in mobile SoCs.

By all rights Intel should win easily. They have the best process tech and loads of cash.

They are doing something wrong to get one-upped by AMD on x86 SoC for tablets (even if only for a little while), and have their last gen mobile part in what was probably their most significant non-Windows 8 OEM design win (Tab 3 10.1) replaced by a low end Qualcomm part.

 

[Shivansps]

I was unable to figure out the intentions of the APU naming until Beema and Mullins, expecially MUllins.

Lest take a close look of AMD APU lineup from the start.

Ontario and Zacate had diferent names, and IGP name is in the same scale, with correct names acording to performance.

Kabini and Temash, Kabini has OK, but Temash jumped to a diferent scale, it seems to be intended to create confusion, intended to hide what the A4-1250(and 1200) was, and those where actually used in several 14" and 15" notebooks.
Still if we look carefully, the IGP is the same on all models and they share the correct numbering, so there is not really a reason to make a fuss about.

OK, what is this? Mullins now have both A and E models, with numbers that are over the beemas numbers? its hard to explain to someone that a A4-6210 R3 and E2-6110 R2 are better than a A10 Micro-6700T R6, they where not doing that on temash and Kabini, and why is jumping all the way to A10 and R6? why they skip the A6 and R4??? there is no way to explain that.
And the IGP naming is ridiculus, how on earth we get here? How the exact same IGP can be named R2 on Beema and R6 on Mullin? what is that? there is no way to defend that. They never did that before. Hell, even on E Beema is called R2 no matter if its a a full 150mhz diff on there.

There the MICRO name on then, no one will know what that even means, people gona be taken away by the name, A10? R6? woooow, same way that happens with Intel, I3, I5, I7, people does not figure out the difference on mainstream to ultrabook ones. But not even Intel screws that much up with the igp names.

I can guarantee you, a lot of people broght the A4-1250 and 1200 not knowing what they where getting, we gona see those Mullins mainstream notebooks and its gona be worse.
What people get is: higher number = better, no one figures out what the U means on Intel, and no one will figure out what the micro and T means.

If i translate that to Intel models, its like the Z3770 where called "I7 Micro-4900T with Intel HD5 Graphics", i guess more people whould be complaining if so, im sick of people defending this sort of things just because its AMD. Like when Intel called Celeron and Pentiums to BT-M and D, i did not like it and a lot of people complained, when AMD did the same with desktop Kabini no one said anything, im very tired of this.