AMD Enthusiasts - NOW is your chance

[MadDias]

@Maddias
Ive used a thuban and a deneb in my old 790FX mobo just fine, tested in my new am3+ and didnt noticed too big of a difference.It was motherboard maker problem with some and not with amd.

what is a cpu without a motherboard?
did u test deneb on am2+ and then on am3 (both use 790fx chipset)? just a ridicolous update to have ddr3 support, but the cpu definetly runs better.
i even had to contact msi tech support in order to get a working bios for the thuban on my msi gd70. the jump in architecture from deneb to thuban wasn?t that big, but it was still too much for msi to come up with a properly working bios.
that?s all i say about "drop in updates"

[AMDforME]

it would have to be....

1. faster than vishera
2. OC better than vishera
3. have much lower power consumption than vishera
4. priced very very competitively
5. if all of above are met then i would buy one
6. if non of the above are met then ill be buying a 4770k.


and

7. AMDforME works for AMD

If you read the criteria stated in the first post and follow up comments you'd know that the proposed AM3+ Steamroller core desktop CPU would essentially meet the above criteria that you have listed.

I believe we would need a thousand votes to make a difference. So start voting people!

I don't know if there are even close to 1,000 active AMD enthusiasts in this forum but the more who vote the better. It would take surveys in literally every PC enthusiast forum to get a sense of the real potential volume but the percentages tell the story. When 90+ % of current and non-owner AM3+ PC enthusiasts voting are willing to purchase a Steamroller based AM3+ CPU with the criteria stated which is primarily ~15% performance increase over Vishera, 125w TDP and similar pricing to Vishera, it's a no brainer that a Steamroller desktop CPU would be a viable business success. That is why it's important for people to keep voting because AMD management does take note of these types of surveys.

[Johan45]

I would love to see another release on the AM3+ platform but I'm not holding my breath. As AMDforMe just put it , very difficult to get a real number on something like this.

[AMDforME]

I would love to see another release on the AM3+ platform but I'm not holding my breath. As AMDforMe just put it , very difficult to get a real number on something like this.

The goal here is not to get a real number on the quantity of people who would purchase a drop-in socket AM3+ Steamroller core desktop CPU, but to see the percentages of current AM3+ owners and also non-owners that would buy one. If you posted this type of survey in every PC performance forum you could think of you might reach 20-30% of AMD enthusiasts so it's difficult to get an accurate sales volume number but based on Vishera sales, if the majority of current AM3+ users would buy a Steamroller CPU, then sales volumes should certainly be profitable and GloFo appears to have the capacity to make this happen.

With "yes" votes in the 90% range for a Steamroller drop-in, it's hard to argue that this wouldn't be a smart move by AMD.

[csimon]

I think that AMD is far too quiet on the future of AM3+. I think it's time for me to invest my monies elsewhere.

[demonkevy666]

Ekhm, on what basis are you making this assumptions ? You either know something we dont, or you are just pipedreaming.
I and many others would buy komodo too, however as stilt said they were canned (that could be a REAL centurion).

@STILT
Are there any functioning komodo ES`s for am3+ ?

here is a question I asked him in the "AMD FX.. Centurion" thread.

question what if it's 5ghz Turbo core ?

here is the reply.

It would not make much of a difference.
The CPU power plane is shared between the compute units so the effective Core VDD would still be at the level requested by 5.0GHz Turbo PState. The lower clock frequency on some of the CUs would lower the power consumption just by few watts.

In the current form the Turbo Core is retarded anyway IMO.
It could be much more efficient if it would not just rely on precalibrated values.

Now here is a fact the FX 9590 Has turbo core of 5.0ghz.

basically it's all speculation till it's out.

[Johan45]

The goal here is not to get a real number on the quantity of people who would purchase a drop-in socket AM3+ Steamroller core desktop CPU, but to see the percentages of current AM3+ owners and also non-owners that would buy one. If you posted this type of survey in every PC performance forum you could think of you might reach 20-30% of AMD enthusiasts so it's difficult to get an accurate sales volume number but based on Vishera sales, if the majority of current AM3+ users would buy a Steamroller CPU, then sales volumes should certainly be profitable and GloFo appears to have the capacity to make this happen.

With "yes" votes in the 90% range for a Steamroller drop-in, it's hard to argue that this wouldn't be a smart move by AMD.

Agreed now that I see your logic. This may hold true for many Enthusiast forums, the +90% but I'm starting to believe that AMD isn't too concerned with the "enthusiast market" . They're doing well in other areas and the last thing they need is to release another "bulldozer" if you know what I mean. They're future is quite delicate ATM.
If they could come up with something that would actually challenge/pas the performance of their Intel counterparts, that could be a saving grace and a revisit to the "Good Ole Days" but are they willing to take the chance. Appearances are everything these days, can make or break a company.

[haylui]

Agreed now that I see your logic. This may hold true for many Enthusiast forums, the +90% but I'm starting to believe that AMD isn't too concerned with the "enthusiast market" . They're doing well in other areas and the last thing they need is to release another "bulldozer" if you know what I mean. They're future is quite delicate ATM.
If they could come up with something that would actually challenge/pas the performance of their Intel counterparts, that could be a saving grace and a revisit to the "Good Ole Days" but are they willing to take the chance. Appearances are everything these days, can make or break a company.

they know best about their partner foundry...

[Johan45]

I'd love to see something

[zir_blazer]

here is a question I asked him in the "AMD FX.. Centurion" thread.

here is the reply.

Now here is a fact the FX 9590 Has turbo core of 5.0ghz.

basically it's all speculation till it's out.

The Stilt seems to have access to many AMD docs under NDA, but he may not necessarily have access to all. With the FX-9000, I think he got catched off guard because the mammoth undertaking in infrastructure (Motherboard and Cooling) required to make an uncompetitive 5 GHz Piledriver possible looked rather stupid for what it was worth. I also thinked that the idea was unrealistic and quite stupid, specifically after seeing that AMD actually DID it.

AMD usually upgrades some white papers or code with information about die Revisions, so you can easily identify actually physically different dies. I did that once back at the time. Chances are that Komodo Revision appeared on some NDA document or AGESA source code then got removed from it after the idea was scrapped.

[vario]

what is a cpu without a motherboard?
did u test deneb on am2+ and then on am3 (both use 790fx chipset)? just a ridicolous update to have ddr3 support, but the cpu definetly runs better.
i even had to contact msi tech support in order to get a working bios for the thuban on my msi gd70. the jump in architecture from deneb to thuban wasn?t that big, but it was still too much for msi to come up with a properly working bios.
that?s all i say about "drop in updates"

Heh, well i had thuban so no idea about deneb, however thuban is a deneb with two more cores so it is comparable, i switched from 790FX to 970,CPU in itself was performing and clocking near kidentical,only diff was in mem performance (duh).Whole problem is that most mb manufacturers wanted to sell you a new mobo so they werent exactly proffesional about their bios upgrades...
To be honest i was even more satisfied with mobo performance on am2+ than am3+,UEFI digi vrm apm turbo and all this variables make it harder to get top performance from a cpu.

That is why it's important for people to keep voting because AMD management does take note of these types of surveys.

I really mean no disrespect and i voted, but, how do you know that ? Are you an inside AMD employee ?
AMD made a lot of weird decisions lately ,if that would be all that we need, wouldnt there be roadmaps for AMD server segment with SR,wouldnt there be komodos in server and high end desktop ?
I can tell ypu right now, that there are MANY people out there that were just so dissatisfied with BD/PD that theyre still on their AM3+ and phenom cores,they dont want the hassle of changing mobos, they dont want intels cut down cpus, there are many people who need high MT power which could be achieved with komodos etc etc.But Amd decides to scrap komodo, scrap server markets and mid end desktop just when they finally are on the verge of releasing a good core.Insteda they release a higher clocked PD with insane wattage and price .
As for sockets.
People ,we have insane DDr3 speeds right now which pretty much over 2ghz are useless, 4channel would give us nothing,it barely gives anything on intels 2011.
DDR4 is also useless besides some power saving.
PCIE 3.0 IS usefull however only if you have SLId titans or at least crossfired 7970 ghz, however we know PD is too weak to drive high end SLI and crossfire seriously, so thats not a problem too.
AM3+ can be in market for another 2 years without too much of a hassle.
If you work in AMD, and you have any input about products, i can assure you, there is market for:
A new PD stepping
Komodo (10core PD`s)
SR core
all on am3+, because there are MANY people with am3+ boards and deneb/thuban cpus, and a lot of dissatisfied BD (and some PD) owners.
However when AMD does the switch to Fm2+ with at most 2module cpus, whole gaming /enthusiast market goes to crappy intel,because when people finally will make the change, theyre not gonna do downgrades.
However if SR for am3+ desktop would be even on the table, they would do first servers, because margins there are a lot bigger.And theyre not doing even that.

[mcnbns]

I had an XP-M and a 939 A64 and loved both dearly. I want to buy something awesome from AMD and I don't care what socket it has to go in.

[AMDforME]

The squeaky wheel generally gets the grease so the saying goes. It seems to me that if enough people really want a Steamroller AM3+ CPU with ~15% performance boost over Vishera, then they should say so. Right now AMD has no upgrade path for desktop users unless they want an APU and many folks do not. In addition the AM3+ mobo is still quite sufficient for most PC enthusiasts so to me it's a no brainer. We'll see how much noise AM3+ owners can make via these surveys.

[m_jones_]

The squeaky wheel generally gets the grease so the saying goes. It seems to me that if enough people really want a Steamroller AM3+ CPU with ~15% performance boost over Vishera, then they should say so. Right now AMD has no upgrade path for desktop users unless they want an APU and many folks do not. In addition the AM3+ mobo is still quite sufficient for most PC enthusiasts so to me it's a no brainer. We'll see how much noise AM3+ owners can make via these surveys.

We won't make a difference, I bet even if 10,000 people still voted they still wouldn't take notice. You seem very optimistic that AMD is willing to dump more and more money into a dying platform when they could be improving their APU lines which are far more popular, they just can't compete due to the amount of money they have and have a inferior fab compared to Intel.

[i found nemo]

Heh, well i had thuban so no idea about deneb, however thuban is a deneb with two more cores so it is comparable,

no the imc's were diff in thuban and deneb

We won't make a difference, I bet even if 10,000 people still voted they still wouldn't take notice. You seem very optimistic that AMD is willing to dump more and more money into a dying platform when they could be improving their APU lines which are far more popular, they just can't compete due to the amount of money they have and have a inferior fab compared to Intel.

so we should not try to tell them what we want? yes... i understand. what kind of consumer speaks out none? suuure.

[BeepBeep2]

lol. My post got deleted! Talk about censorship at it's worst.

15 guys in a forum aren't going to sway any opinions. Sales and marketing charts and graphs do that.

[m_jones_]

no the imc's were diff in thuban and deneb



so we should not try to tell them what we want? yes... i understand. what kind of consumer speaks out none? suuure.

We can tell them all we want but it won't make a difference for such a small niche forum when the money could be better spent in other areas. What about if they know they can't keep up with what Intel is currently out putting is there any point in them trying to compete?

[vario]

We can tell them all we want but it won't make a difference for such a small niche forum when the money could be better spent in other areas. What about if they know they can't keep up with what Intel is currently out putting is there any point in them trying to compete?

Well, for now Intel seems to be content with not competing even with itself.So a surprise attack from AMD ,IF they have the goods (ie the SR is good,optimized and not nuclear powered) would be successful.
And the "niche" isnt that small, there is really a big audience for this, cheap AMDers who dont want to change mobo in an intel swap, disappointed AMDers who bought BD or even PD and were underwhelmed that want something with claws to finally put in their mobos, people who need MT power without $ to buy intel 2011 setup.I have 5 friends with AM3+ mobos, one bought PD and was underwhelmed ,the rest is still on different phenoms because i could not with a pure heart recommend PD to them, and theyre gonna use this mobos for 2-3 years more.There are a lot of people who would shell out reasonable amount of money for a reasonable upgrade ,six core SR probably would be spot on for all these people, and 8 cores for the more power hungry like me.
To be honest i dont know how much money developing another die costs, but if theyre already developing SR for FM2+ , it cant be that much money.

Maybe we should spam whole AMD with email attack on this :P

[AMDforME]

We won't make a difference, I bet even if 10,000 people still voted they still wouldn't take notice. You seem very optimistic that AMD is willing to dump more and more money into a dying platform when they could be improving their APU lines which are far more popular, they just can't compete due to the amount of money they have and have a inferior fab compared to Intel.

Many folks don't seem to understand the situation...

Despite the myths and speculation perpetuated by the media and those with a crystal ball, AMD never said that they were abandoning the AM3+ platform. What they said was that they were:

1. Not going to allow Intel's Biz model to dictate AMD's Biz model

2. That AMD was going to pursue other opportunities - some of which Intel could not pursue and has not pursued

In addition AMD has never said that they would NOT deliver an AM3+ Steamroller core desktop CPU. It was removed from the road maps at the end of 2012 and some folks speculated that was the end of it, but AMD never confirmed this. For all we know an AM3+ Steamroller desktop CPU might show in H1 of '14, shortly after Steamroller based Kaveri is fully launched.

The point of this survey is to let AMD management know that the take rate of respondents which includes current AM3+ mobo and non-current AM3+ mobo owners is ~ 90% or better. Any company would be missing an excellent business opportunity with a take rate of 90+ %. The absolute numbers are not in question as AMD already knows the millions of Bulldozer and Vishera CPUs they have sold. Many of those folks would gladly upgrade to a ~15% better performing Steamoller compared to Vishera. According to this and other surveys the take rate looks like 90+ %.

As far as cost to make a Steamroller AM3+ socket CPU, it's a drop in the bucket as all the development work was done. All AMD needs to do is Fab the chip, which is simple in comparison to an entirely new CPU design as Steamroller is just a core evolution.

http://www.theregister.co.uk/2012/08...micro_opteron/

[BeepBeep2]

Many folks don't seem to understand the situation...

You sure that's not you?

Despite the myths and speculation perpetuated by the media and those with a crystal ball, AMD never said that they were abandoning the AM3+ platform. What they said was that they were:

Actually, AMD updated their 2014 roadmap which didn't seem to include new AM3+ products, and declined to comment on any unannounced products.

1. Not going to allow Intel's Biz model to dictate AMD's Biz model

2. That AMD was going to pursue other opportunities - some of which Intel could not pursue and has not pursued

Source? Creative interpretation? Intel is unable to persue something that AMD can? Intel is roughly 1000% bigger than AMD.

In addition AMD has never said that they would NOT deliver an AM3+ Steamroller core desktop CPU. It was removed from the road maps at the end of 2012 and some folks speculated that was the end of it, but AMD never confirmed this. For all we know an AM3+ Steamroller desktop CPU might show in H1 of '14, shortly after Steamroller based Kaveri is fully launched.

Usually removing a CPU from a roadmap means the end, I'm pretty sure that banking on a surprise release of CPU from AMD is a pipe dream. This industry does not work that way, partners want to know exactly what is going on so that they can plan for a release or begin marketing strategy, etc.

The point of this survey is to let AMD management know that the take rate of respondents which includes current AM3+ mobo and non-current AM3+ mobo owners is ~ 90% or better. Any company would be missing an excellent business opportunity with a take rate of 90+ %. The absolute numbers are not in question as AMD already knows the millions of Bulldozer and Vishera CPUs they have sold. Many of those folks would gladly upgrade to a ~15% better performing Steamoller compared to Vishera. According to this and other surveys the take rate looks like 90+ %.

Okay, so let's say the listen to you (or us, or whatever) and start right away, they might be able to have a product to retail on 32nm by December 2014...
Unfortunately, they can't do that, they have sales figures for current enthusiast products, they aren't pretty, and AMD has very limited resources between $$$$ and brains at this point in time.

See here:

Imagine you are the head of the desktop division and the board comes to you and says "here's $5M, go bring us another product for the roadmap."

Do you put that into something that brings in $500K-$1M in revenue or do you put it into something that brings in $10-20M in revenue.

With a full portifolio and all of your R&D dollars accounted for, adding something like this means you have to either A.) hire more people, spend more capital or B.) cancel an existing project.

So, generally speaking, when you ask to add anything to the product roadmap, you are essentially having to make a tradeoff. Nobody is going to ask for more more money in order to lose money nor is anyone going to ask to cancel a profitable program to replace it with an unprofitable program. It just doesn't happen.

Based on this conversation I have a feeling nobody here is in the semiconductor business, so let me try to explain it to you.

Engineering samples are designed to validate the design and for partners to validate their systems...
(cut for brevity)
There are multiple steppings of samples, comparing one stepping to another tells you nothing...
(cut for brevity)
When we make engineering samples, we target the best yield, not the highest performance. Then we have more chips to work with. Typically you get a couple of the highest clock speed anyway, those typically stay in our labs, so it is unlikely that you see those out in the wild.

We don't even do any real performance analysis until we get the final silicon. Prior to that, we are totally focused on design validation.
...
...
...
...
Imagine making a cake. You start with a pile ingredients. You put it all in a mixing bowl. You stir it up. You put it in the oven. You let it cook. You take it out. You put frosting on it.

You are all looking at bowlful of batter and arguing about whether or not it is appetizing.

As far as cost to make a Steamroller AM3+ socket CPU, it's a drop in the bucket as all the development work was done. All AMD needs to do is Fab the chip, which is simple in comparison to an entirely new CPU design.

http://www.theregister.co.uk/2012/08...micro_opteron/

You have to be kidding. Desktop and Server products both have to go through extensive R&D. There are slight differences at design level between Desktop and Server products such as there were between Lisbon/Thuban/Instabul, and Server parts are tuned for efficiency, while any Desktop released FX processor will be tuned toward performance.

[zir_blazer]

Good JF-AMD quotes. I recall back at the time many people asked him for a Dual Processor Motherboard with support for overclocking and he said there was no market for such a thing and how enthusiasts was a niche segment. That guy got severely discredited due to Bulldozer performance predictions and the rage against him that followed, but most of his other comments regarding business decisions stay true.

AM3+ and FM2 got two main differences. AM3+ does not have support for PCIe lanes coming out of it, and the connection to the Chipset is made using old HyperTransport. On FM2, besides that, Hyper Transport got replaced for 4 PCIe lanes to connect to the Chipset if I recall correctly. So basically, for Steamroller to be on AM3+, besides Kaveri they need a separate die that could fit on the AM3+ infrastructure. And no, they will NOT invest resources in R&D to do it for a niche market.

[imamage]

Well to be honest , as a current owner of Socket FM2 and AM3+ PC
As AM3+ is a "dead-end" platform , and AMD won't release any new chipset for it, why bother ?
If they launch a AM3+ SR , I will buy one for sure , but I rather AMD focus on the "new platform"

Maybe release a "6Core" SR on FM2+ (not likely as too large die-size and require new design , but hey , you can do that with Thuban , why not make a 3Module SR for FM2+ ??? )

Plus the FCH chipset provide more modern feature set support (USB 3.0 and SATA3 , especially)
is easier to sell both OEM and DIY market.

But I hightly doubt it would be possible as FM2+ is maxed out up to 95W TDP

[AMDforME]

Folks this thread is NOT a debate over if AM3+ is dead or not. This thread is a vote on a Steamroller AM3+ socket.

If you want to discuss AMD's Biz model or your ideology start a thread of your own.

[BeepBeep2]

Then don't start a discussion about AMD's business model?