AMD AM4 CPU, APU And Motherboard Spotted On Zauba –Spring 2016 Launch Imminent

[MrTeal]

I think there is a big difference in trying to shoot down posters who are in anyway enthusiast about a possible new true alternative from AMD and jumping for excitement.

That is why the extreme effort in disparaging an enthusiastic and possibly hopeful pro AMD poster is very childish and quite immature.

Who benefits?

I would say there's a difference between being enthusiastic about AMD's new products, and being intentionally obtuse. It's possible that Ian took MSI's statement out of context or it was poorly worded, but suggesting that one of the Big 4 motherboard vendors is simply not interested in releasing products for AMD's highly hyped new products is just silly. It's not some random wccftech "according to an anonymous poster on a web forum" source, the VP of Motherboard Sales for a $3B company that has more than a couple dozen AM3+ and FM2+ SKUs that will be replaced by AM4. If he says they're looking forward to AM4 and will work closely with AMD, but they'll start work in the spring and they don't think AMD has the details all finalized, that's about as confirmed as you're going to get. Demanding a source (presumably from AMD) that has come out and said all the details on the AM4 platform aren't finalized yet is just ridiculous.

BTW - I'm pretty sure wccftech is wrong on the Zaruba motherboard. FOC might mean full operational capability in some (military) circles, but in importing and exporting an FOC item is usually free of charge or free of cost. It says nothing about its functionality.

 

[Dresdenboy]

BTW - I'm pretty sure wccftech is wrong on the Zaruba motherboard. FOC might mean full operational capability in some (military) circles, but in importing and exporting an FOC item is usually free of charge or free of cost. It says nothing about its functionality.

Thanks for pointing that out.

 

[Kuiva maa]

Well truth be told , MSI was left out in the initial badge of Fury X cards in favor of Asus and Sapphire . Even today they offer no plain Fury, just a reference Fury X ,and it has been half a year, I don't remember another big vendor ignoring a big chip before. I wouldn't be surprised if they would lag behind with boards too.

 

[Arachnotronic]

Well truth be told , MSI was left out in the initial badge of Fury X cards in favor of Asus and Sapphire . Even today they offer no plain Fury, just a reference Fury X ,and it has been half a year, I don't remember another big vendor ignoring a big chip before. I wouldn't be surprised if they would lag behind with boards too.

They ignored Fury X probably because they had enough of an understanding of the market to realize that it wouldn't do well against the GTX 980 Ti. MSI has put out some stellar 980 Tis.

 

[Vesku]

MSI was also the most vocal about peak FX 8000 series power draw being too high even at stock, heavily suggesting retail FX 8000 chips did not stay within the parameters AMD gave motherboard makers.

Wouldn't surprise me if MSI was taking a wait and see approach with AM4.

 

[Kuiva maa]

They ignored Fury X probably because they had enough of an understanding of the market to realize that it wouldn't do well against the GTX 980 Ti. MSI has put out some stellar 980 Tis.

They haven't ignored Fury X and now they offer nano too. They ignore the plain Fury which makes no sense since it is the high volume card of the three. The other two are reference models so the best explanation is they Can't be bothered to design a custom pcb for Fiji. Sapphire, Asus, Gigabyte have their own designs, MSI is the only big vendor without. They don't seem to have a very close relationship with AMD.

 

[dark zero]

They ignored Fury X probably because they had enough of an understanding of the market to realize that it wouldn't do well against the GTX 980 Ti. MSI has put out some stellar 980 Tis.

However the best of the GTX 980 Ti's were Asus or EVGA ones.

 

[kawi6rr]

Stating the truth makes me a fanboy?

Two of the "truths" you posted were not the truth and yes you are a fanboy it's clear with your posts. I bolded what you are insinuating are truths but are not unless you have a source.

 

[Phynaz]

Two of the "truths" you posted were not the truth and yes you are a fanboy it's clear with your posts. I bolded what you are insinuating are truths but are not unless you have a source.

I already posted my source. I see no need to post it again, since you will still not be convinced as long as you are wearing your AMD glasses.

 

[Vortex6700]

AMD has to be late to the party and inferior. They make more money when intel doesn't pay the oems not to buy from them.

 

[The Stilt]

MSI was also the most vocal about peak FX 8000 series power draw being too high even at stock, heavily suggesting retail FX 8000 chips did not stay within the parameters AMD gave motherboard makers.

That´s because MSI used and still uses POS quality components, combined with poor engineering. Any board built to meet AMD infrastructure specifications will easily be sufficient at stock.

Bulldozer released with peak TDP of 125W and the design requirements have not changed since. All boards which support 125W TDP rated 15h CPUs needs to be able to provide 110A sustained current (TDC) and 145A peak current (EDC) without overheating or blowing up, in temperatures up to 62°C. To put that into perspective, the fastest mainstream FX (FX-8370) CPU is rated for 94.6A.

So any board which fails or throttles at stock clocks does not meet AMD specifications for the given TDP envelope. There are quite few boards (usually advertized as "Gamer" or "Extreme" targeted) which do just that D:

 

[LTC8K6]

That´s because MSI used and still uses POS quality components, combined with poor engineering. Any board built to meet AMD infrastructure specifications will easily be sufficient at stock.

Bulldozer released with peak TDP of 125W and the design requirements have not changed since. All boards which support 125W TDP rated 15h CPUs needs to be able to provide 110A sustained current (TDC) and 145A peak current (EDC) without overheating or blowing up, in temperatures up to 62°C. To put that into perspective, the fastest mainstream FX (FX-8370) CPU is rated for 94.6A.

So any board which fails or throttles at stock clocks does not meet AMD specifications for the given TDP envelope. There are quite few boards (usually advertized as "Gamer" or "Extreme" targeted) which do just that D:

8370 is Piledriver, not Bulldozer, though.

What about the FX-9XXX chips?

 

[The Stilt]

8370 is Piledriver, not Bulldozer, though.

What about the FX-9XXX chips?

Doesn´t matter since the infrastructure is the same for both. The fastest Bulldozer is rated for 95.0A (FX-8150), so basically no difference in that aspect either.

FX-9K parts belong to AM3r2 infrastructure group F, whereas 125W TDP models belong to group A. The group F motherboards need to be able to provide 140A (TDC) sustained current and the same 145A peak current (EDC) in order to meet the specifications.

AFAIK there are no boards that were designed to meet "group F" requirements specifically, since the infrastructure group was created shortly before FX-9K series was released. However boards like ASUS Crosshair V Formula / -Z and Sabertooth 990FX are able meet the requirements thanks to their heavy duty VRM and cooling.

 

[LTC8K6]

Why would AMD tolerate board mfgs gimping their CPUs?

 

[The Stilt]

Why would AMD tolerate board mfgs gimping their CPUs?

What could they do about it? Tell the ODMs that they won´t be getting any chipsets to make the motherboard with, unless they make designs which meet their specs?

Unfortunately I think the case is that AMD is in no position to make any kind demands or give ultimatums...

 

[DrMrLordX]

No. But even then its irrelevant with the Zauba information posted. Because it doesn't have anything to do with the matter. Just think on when Fiji was spotted for example. Or Carrizo.

Um, the Zauba information has everything to do with the matter. The linked article specifically compared Carrizo's launch vs. the first signs of Carrizo board + APU samples being spotted in Zauba printouts. According to that comparison, the appearance of a Zauba printout showing shipments of Myrtle samples indicates a March 2016 launch for Myrtle.

Why would AMD tolerate board mfgs gimping their CPUs?

Who knows? As The Stilt indicated, what can they really do to stop it, other than frustrate ODMs/OEMs by making it impossible to create too-cheap-to-be-useful configurations? Take a look at what the OEMs did to Carrizo . . .

 

[LTC8K6]

What could they do about it? Tell the ODMs that they won´t be getting any chipsets to make the motherboard with, unless they make designs which meet their specs?

Unfortunately I think the case is that AMD is in no position to make any kind demands or give ultimatums...

AMD fans would not buy a board if AMD came out and said it was gimping their chips and not meeting AMD specs.

I wouldn't buy a board that AMD did not approve.

If I knew that a board would not let my FX-8350 run wide open, I would certainly not buy it.

If I found out on my own after buying the board, I would return it if I could, or not buy in the future.

 

[DrMrLordX]

Well, several OEMs came out with entire laptops that did not let the FX8800P run "wide open". In fact, nearly all of them did. Hell they gimped the 8700P too.

If AMD buyers refused to touch those laptops, AMD would have sold no Carrizo APUs at all . . .

 

[Phynaz]

Not when the mocking posts has all the characteristics of being posted by a gang of bigots.

That simple word, truth, or to be more precise your truth. The fact that there are different values here means no universal truth, else there would be very short arguments.

Just relax and let people enjoy themselves.

Just let the ADF spread lies and misinformation.

Not gonna happen bub.

 

[LTC8K6]

Well, several OEMs came out with entire laptops that did not let the FX8800P run "wide open". In fact, nearly all of them did. Hell they gimped the 8700P too.

If AMD buyers refused to touch those laptops, AMD would have sold no Carrizo APUs at all . . .

That could be because the chip cannot reasonably run wide open in a laptop...?

I suspect most laptop buyers have no idea what is inside their laptop.

It's enthusiasts and gamers who would notice and complain / not buy.

 

[frozentundra123456]

Well, looks like we are back to SOP, blaming everyone but AMD for anything that goes wrong with AMD's products. Must be intel payola or manufacturers deliberately gimping (why???) the very products that they are trying to sell. Now that makes a lot of sense.

 

[cbn]

Well, several OEMs came out with entire laptops that did not let the FX8800P run "wide open". In fact, nearly all of them did. Hell they gimped the 8700P too.

If AMD buyers refused to touch those laptops, AMD would have sold no Carrizo APUs at all . . .

That could be because the chip cannot reasonably run wide open in a laptop...?

I suspect most laptop buyers have no idea what is inside their laptop.

It's enthusiasts and gamers who would notice and complain / not buy.

Even us an enthusiasts have a hard time figuring it out:

http://forums.anandtech.com/showpost.php?p=37919768&postcount=2665

I do think OEMs that used Carrizo in 35W configuration (was it just Lenovo with the Y700?) should have labeled it so on the outside.

Otherwise how is a person to know whether or not they had the souped up 35W version or the 15W version?

If I was someone looking at a FX-8800P laptop and didn't know it was 35W, I would be inclined to pass on it if I saw another FX-8800P next to it that was thinner, lighter and cheaper.

 

[Dresdenboy]

Unknown as to what will be the clockspeed "wall". It's using the same process as Carrizo, which is GF28A (forget which process it really is, but it isn't 28shp ala Kaveri). Apparently the voltage scaling is improved thanks to process refinements similar to what GV-A1 brought to the table for Kaveri earlier this year.

Sorry for my late answer.

There are many walls, but no hard ones, as timing, area, power are targets in a multi objective optimization. So there is no 3 or 5 GHz wall for given processes.

 

[DrMrLordX]

That could be because the chip cannot reasonably run wide open in a laptop...?

Depends, how many other laptop CPUs have had TDPs of 35W-45W? Apparently it's possible in the y700 until the VRMs overheat.

I suspect most laptop buyers have no idea what is inside their laptop.

ditto. Though they can notice difference in performance . . .

Well, looks like we are back to SOP, blaming everyone but AMD for anything that goes wrong with AMD's products.

Okay, what did AMD do wrong? It's a 35W chip being sold in 15W TDP configurations.

Must be intel payola or manufacturers deliberately gimping (why???) the very products that they are trying to sell. Now that makes a lot of sense.

If it helps them to cut costs, maybe, though in some cases it defies explanation, such as when they include dGPUs at extra cost and then compensate by selling Carrizo chips in boards built to handle Carrizo-L with Carrizo-L cooling solutions. But hey there're entire threads of people griping over the treatment given to Carrizo. At least for us enthusiasts, Bristol Ridge will be available with whatever AM4 board we care to use as host.

Sorry for my late answer.

There are many walls, but no hard ones, as timing, area, power are targets in a multi objective optimization. So there is no 3 or 5 GHz wall for given processes.

Right, but it's pretty well-known that Kaveri (for example) has a wall on air/water of around 4.7 GHz. LN2 barely gets it past 5.2 GHz. If Bristol Ridge can at least reach the same clockspeeds, it'll be interesting.

 

[The Stilt]

It's a 35W chip being sold in 15W TDP configurations.

It´s actually 15W chip, which can be configured up to 35W Currently there are no > 15W TDP models available, but it will change soon. All currently available Carrizos should allow cTDP configuration 12-35W, or at least I´ve not seen any which wouldn´t have.

I must admit that I don´t like the way AMD has been marketing Carrizo (business as usual...). For instance there is a huge confusion about the TDP:

Carrizo (8K series or 6th gen) defaults to:

- 15W TDP
- 18W BBB (Boost TDP limit on battery)
- 25W STAPM PPT (Boost TDP limit on AC, availability depends on timer and temperature)
- 12-35W cTDP (ODM configurable, if not set then 15W)

Yeah...