The site you link to (wccftech) seems to say something different
So either 16 cores + graphics or 32 cores with no graphics.
Anyone care to speculate on the cost of one of those processors?
I mean the 32Gb alone costs a lot.
The APU, dubbed an Exascale Heterogeneous Processor or EHP for short is the mother of all APUs with 32 Zen Cores, an absolutely huge Greenland graphics die and upto 32 GB of HBM2 memory all on a 2.5D interposer.
So either 16 cores + graphics or 32 cores with no graphics.Previously leaked information revealed the existence of a 16 Core Zen APU with HBM Graphics, and it now appears (via Fudzilla) that there is a monstrous 32 core processor in the works too minus the Greenland integrated graphics.
So either 16 cores + graphics or 32 cores with no graphics.
Anyone care to speculate on the cost of one of those processors?
I mean the 32Gb alone costs a lot.
Actually that's what AMD said in the beginning...Much more reasonable.
Actually that's what AMD said in the beginning...
16 Core + 32GB HBM targeted to HPC
32 Pure Core without any Graphics, targeted to servers
Both targeted to the very high end.
However as for desktop market, we are seeing this:
4 Core + 4 GB HBM to compete Crystalwell
8 Core without Graphics towards the Core i7 E
I don't think they are hyping any more than others in the business. For example just look at how Intel hailed Skylake as "its most important chip architecture in a decade". That is totally ridiculous, seeing how it performs in reality.
It's similar to nVIDIA now... since Intel iGPU is increasing dramatically their performance and nVIDIA GDDR5 reached their limits, their only hope is HBM2, similar to AMD. The problem? Supplys... AMD has the priority and nVIDIA is getting later. They will come late to the party.Apples and Oranges
When something Intel is overhyped, the disappointment is because the % increase over it's competition (which is itself these days) isn't as high as one may have hoped. When AMD is overhyped, the disappointment is because it's still not competitive.
With HBM iGPU prices (14nm) - From the A6, all are designed to OC, the rest can do, but with other waysSempron X2 - 2 Cores - 3.00 Ghz - 2Mb L3 - No HT - $50 - $55
Athlon X4 - 4 Cores - 2.70 Ghz - 2Mb L3 - No HT - $75 - $80
Phenom X4 - 4 Cores -3.00 Ghz - 4Mb L3 - Partial HT (2 cores with HT enabled, 2 cores with HT disabled) - $125 - $145
Phenom X6 - 6 Cores -3.25 Ghz - 4Mb L3 - Partial HT (2 cores with HT enabled, 4 cores with HT disabled) - $160 - $180
FX Hexa Core - 6 Cores - 3.25 Ghz - 6Mb L3 - With HT - $200 - $225
FX Octa Core - 8 Cores -3.50 Ghz - 8Mb L3 - With HT - $325 - $375
If you have others, launch that. Well... knowing Intel's pricing, AMD ones tends to be cheaper and will have lower performance.E1 - 1 Core - 2.75 Ghz - 2Mb L3 - 1GB "L4 HBM1" - With HT - $65 - $70
E2 - 2 Cores - 2.50 Ghz - 2Mb L3 - 1GB "L4 HBM1" - No HT - $80 - $90
A4 - 2 Cores - 2.80 Ghz - 4Mb L3 - 2GB "L4 HBM2" - Partial HT (1 core with HT enabled, 1 core with HT disabled) - $120 - $130
A6 - 2 Cores - 3.20 Ghz - 4Mb L3 - 2GB "L4 HBM2" - With HT - $140 - $160
A8 - 4 Cores - 3.0 Ghz - 6Mb L3 - 4GB "L4 HBM2" - Partial HT (2 cores with HT enabled, 2 cores with HT disabled) - $180 - $200
A10 - 4 Cores - 3.2 Ghz - 8Mb L3 - 4GB "L4 HBM2" - With HT- $250 - $275
Apples and Oranges
When something Intel is overhyped, the disappointment is because the % increase over it's competition (which is itself these days) isn't as high as one may have hoped. When AMD is overhyped, the disappointment is because it's still not competitive.
So the overhyping lies in making exaggerated claims compared to what is actually delivered. Note that this is regardless of what market position that item has. So Intel is clearly guilty of that with Skylake.Overhype: Make exaggerated claims about (a product, idea, or event); publicize or promote excessively
Not apples and oranges at all, according to Oxford English dictionary:
So the overhyping lies in making exaggerated claims compared to what is actually delivered. Note that this is regardless of what market position that item has. So Intel is clearly guilty of that with Skylake.
Quite obvious really. Otherwise the market leader in any product segment would be given carte blanch to make any ridiculous exaggerated claims about a product while failing to deliver accordingly, without it being called overhyping.
And sadly... the revolution was Sandy Bridge. Skylake seems to be a little revolution, like Ivy to Haswell... I miss the days of epic improvements on every aspect.The only ridiculous thing is your distorted interpretation. Do you have any proof/evidence that they were talking about pure performance when they called Skylake their most significant chip family of last decade?
It's not hype,you just have to understand what they are saying,it will be 40% more just like the i3 has 30something more ipc compared to a similar frequency pentium.Now their claim of 40% IPC improvement, OTOH, could well turn out to be hype.
IPC is neither ST nor MT speed. It is one of many factors.It's not hype,you just have to understand what they are saying,it will be 40% more just like the i3 has 30something more ipc compared to a similar frequency pentium.
It will be throughput/multithread speed and not single thread speed.
We need to wait until the reviews.That claim is so absurd. Saying it is the most important chip in a decade is in no way making any claim of performance. It could well be the "most important" simply because the PC market is declining and intel really needs more competitive products to combat ARM. Bulldozer was AMD's "most important" chip also, because it was the basic architecture they have depended on for several years. Pretty much sunk the company, but it still was the "most important" because there was no other alternative. Somehow I think Skylake will be more successful than that.
And if AMD were to claim Zen is the most important chip coming out for them, I would not call that hype in the least. It certainly is true, because it is their last chance to remain (become) a meaningful cpu company again. Now their claim of 40% IPC improvement, OTOH, could well turn out to be hype.
Not apples and oranges at all, according to Oxford English dictionary:
So the overhyping lies in making exaggerated claims compared to what is actually delivered. Note that this is regardless of what market position that item has. So Intel is clearly guilty of that with Skylake.
Quite obvious really. Otherwise the market leader in any product segment would be given carte blanch to make any ridiculous exaggerated claims about a product while failing to deliver accordingly, without it being called overhyping.
.. It looks to be a jack of trades chip, master of none.
.. Do you have any proof/evidence that they were talking about pure performance when they called Skylake their most significant chip family of last decade?
Yes I have, let me pull it out of my *** for your viewing pleasure. As a performance monster, skylake was certainly hyped as the next big thing..
Oh, I'm sure you can find a lot of stuff right there but Intel never made any official claim about Skylake performance as far as I can remmember.
Well, they did recently in some slides. Seems like the cpu was something like 10% in some obscure benchmark. Specint, or something?
Actually, if one wants to blame Intel for hype, I think Broadwell was more overhyped and disappointing, but performance and availability wise, than any recent release. Lets hope Skylake is not the same, especially as to availability.
But really, I dont blame AMD or Intel for the hype, as much as I blame the unblinded acceptance by some posters of claims by one company and total debunking of similar PR claims by the other. That and how they even take claims beyond any reason, and claim any test that shows their favorite chip not living up to the hype to be biased.
Apples and Oranges
When something Intel is overhyped, the disappointment is because the % increase over it's competition (which is itself these days) isn't as high as one may have hoped. When AMD is overhyped, the disappointment is because it's still not competitive.
The only ridiculous thing is your distorted interpretation. Do you have any proof/evidence that they were talking about pure performance when they called Skylake their most significant chip family of last decade?
If Intel says that Skylake is its "most important chip architecture in a decade", and it's nowhere near that by any metric in reality, Intel is still overhyping it no matter how you try to spin it.It's Apples and Oranges if you actually understood what I said. I'll give you a hint, I wasn't arguing the definition of "hype"
Considering that I actually acknowledged Intel products get over-hyped too when I said "when something Intel is over hyped" I find it quite astonishing that it still went over your head and you proceeded to waste your time looking up the meaning of the word "hype"
Actually now that I think about it it's rather funny.
When a new Intel CPU is on the way, Intel users complain about how it's likely going to be a incremental performance increase on the CPU side and worry about lack of OC potential
Huh? Who has said that they expect AMD to "leap frog multiple generations of Intel CPU's"?When a new AMD CPU is on the way, AMD users think it's going to leap frog multiple generations of Intel CPU's that have dominated AMD and suddenly be the performance king. When that doesn't happen, you wait for "software optimizations" for the next few years that never happen.
Who has said it's all about performance? At least I have not. Problem is that Skylake is not Intel's "most important chip architecture in a decade" by any metric.
Or would you really say it's more important than e.g. Conroe or Skylake? In that case in what way?
If Intel says that Skylake is its "most important chip architecture in a decade", and it's nowhere near that by any metric in reality...
15W Iris Pro SKUs and a brand new GT4e option for high-end notebooks/desktops (no AMD APU will touch it till 2017),