Tuesday, February 24th 2015

AMD's Excavator Core is Leaner, Faster, Greener

AMD gave us a technical preview of its next-generation "Carrizo" APU, which is perhaps the company's biggest design leap since "Trinity." Built on the 28 nm silicon fab process, this chip offers big energy-efficiency gains over the current-generation "Kaveri" silicon, thanks to some major under-the-hood changes.

The biggest of these is the "Excavator" CPU module. 23 percent smaller in area than "Steamroller," (same 28 nm process), Excavator features a new high-density library design, which reduces die-area of the module. Most components are compacted. The floating-point scheduler is 38% smaller, fused multiply-accumulate (FMAC) units compacted by 35%, and instruction-cache controller compacted by another 35%. The "Carrizo" silicon itself uses GPU-optimized high-density metal stack, which helps with the compaction. Each "Excavator" module features two x86-64 CPU cores, which are structured much in the same way as AMD's previous three CPU core generations.
The compaction in components doesn't necessarily translate into lower transistor-counts over "Kaveri." In fact, Carrizo features 3.1 billion transistors (Haswell-D has 1.4 billion). Other bare-metal energy optimizations include an 18% leakage reduction over previous generation, using faster RVT components, which enables 10% higher clock speeds at the same power-draw (as "Kaveri"). A new adaptive-voltage algorithm reduces CPU power draw by 19%, and iGPU power draw by 10%. AMD introduced a few new low-power states optimized for mobile devices such as >9-inch tablets and ultra-compact notebooks, which reduces overall package power draw to less than 1.5W when idling, and a little over 10W when active. In all, AMD is promising a conservative 5% IPC uplift for Excavator over Steamroller, but at a staggering 40% less power, and 23% less die-area.
The integrated-GPU is newer than the one featured on "Kaveri." It features 8 compute units (512 stream processors) based on Graphics CoreNext 1.3 architecture, with Mantle and DirectX 12 API support, and H.265 hardware-acceleration, with more than 3.5 times the video transcoding performance increase over "Kaveri." For notebook and tablet users, AMD is promising "double-digit percentage" improvements in battery life.

Now, if only AMD can put six of these leaner, faster, and greener "Excavator" modules onto an AM3+ chip.
Add your own comment

85 Comments on AMD's Excavator Core is Leaner, Faster, Greener

#1
Sony Xperia S
"Now, if only AMD can put six of these leaner, faster, and greener "Excavator" modules onto an AM3+ chip."

I hope also so because if they do it, I would immediately go a buy the plaform with it.
Posted on Reply
#2
TRWOV
AMD themselves stated that Piledriver was the end of the line for AM3+

I think the next big CPU from AMD will feature a new socket and new architecture. Maybe Zen?
Posted on Reply
#3
Sony Xperia S
TRWOV
AMD themselves stated that Piledriver was the end of the line for AM3+

I think the next big CPU from AMD will feature a new socket and new architecture. Maybe Zen?
Yes, unfortunately this Excavator core is very mobile systems centric and kind of "useless" for a desktop CPU. :(

Which is a shame because I do NOT care about anything mobile now. /#egoismmodeswitchedon
Posted on Reply
#4
Bjorn_Of_Iceland
Hopefully it would deliver.. else this excavator would be digging them a deeper grave.
Posted on Reply
#5
btarunr
Editor & Senior Moderator
6x Excavator AM3+ could be this generation's "Thuban" before Zen.
Posted on Reply
#6
BiggieShady
So IPC is going up 5% and power consumption is going down 30%. If they use this architecture improvements in their desktop CPUs, they are better off with upping the clocks than adding more cores IMO.
Posted on Reply
#7
Sony Xperia S
BiggieShady
So IPC is going up 5% and power consumption is going down 30%.
This might be true only in a limited area of the frequency / power curve....
Posted on Reply
#8
arbiter
TRWOV
AMD themselves stated that Piledriver was the end of the line for AM3+

I think the next big CPU from AMD will feature a new socket and new architecture. Maybe Zen?
AM3+ being killed off was well over due. It was old, gotta replace socket with something once and a while else cpu's get held back by old and slow parts. I mean that as not a bash but it was time that, 3 and half years old (october 2011).
Posted on Reply
#9
Sony Xperia S
arbiter
AM3+ being killed off was well over due. It was old, gotta replace socket with something once and a while else cpu's get held back by old and slow parts. I mean that as not a bash but it was time that, 3 and half years old (october 2011).
Only if you consider the socket itself as a bottleneck spot which is not serious.
Posted on Reply
#10
Exceededgoku
Genuine question, how does AMD expect to compete with Intel when they're process node size is double that of the Skylake CPU's?
Posted on Reply
#11
comperius
Did you saw any Skylake CPU's on the market? They are delayed in 2016.

So realistically it has no sense to compare who has what.

As long as the CPU doesn't use too much power, I don't care if it's made in 100nm techology :)
Posted on Reply
#12
Assimilator
comperius
Did you saw any Skylake CPU's on the market? They are delayed in 2016.
End of August 2015 is 2016 now? Interesting.
Posted on Reply
#13
Mussels
Moderprator
these could make for some very, very nice 9-10" windows tablets
Posted on Reply
#14
RejZoR
Looks quite decent...
Posted on Reply
#15
Sony Xperia S
Exceededgoku
Genuine question, how does AMD expect to compete with Intel when they're process node size is double that of the Skylake CPU's?
AMD is expecting to move to a new process with Zen sometime in 2016.

At that time, the process gap will be largely gone.
Posted on Reply
#16
Aquinus
Resident Wat-man
Exceededgoku
Genuine question, how does AMD expect to compete with Intel when they're process node size is double that of the Skylake CPU's?
Intel and AMD measure node sizes differently. They're actually not as dissimilar as you think they are, but that isn't to say that Intel doesn't have a smaller process than AMD, it's just over exaggerated because of how it's measured. I wish I could find the article, but IIRC it describes how Intel measures more of the width of any given circuit where AMD excludes the very outside edge where Intel includes it, or something along those lines. So in reality, something like 16nm for Intel would be something like 20nm by how AMD measures it. It's important to keep that in mind and if I end up finding the article, I'll add it.
Posted on Reply
#17
micropage7
power consumption looks ok then we need some benchmark
Posted on Reply
#18
Sony Xperia S
Aquinus
Intel and AMD measure node sizes differently.
Somewhat.

Aquinus
They're actually not as dissimilar as you think they are, but that isn't to say that Intel doesn't have a smaller process than AMD, it's just over exaggerated because of how it's measured.
Posted on Reply
#19
comperius
Assimilator
End of August 2015 is 2016 now? Interesting.
No, I read Q3 2015 is out of the picture and the CPU's wil arrive in late Q4 2015 or Q1 2016. That's why I wrote this info. You can check this info online via google search.
Posted on Reply
#20
Aquinus
Resident Wat-man
Sony Xperia S
Somewhat.




Out of curiosity, do you recall where you got that picture from?
Posted on Reply
#22
TheinsanegamerN
arbiter
AM3+ being killed off was well over due. It was old, gotta replace socket with something once and a while else cpu's get held back by old and slow parts. I mean that as not a bash but it was time that, 3 and half years old (october 2011).
Yeah, you have to replace the socket (and chipset) eventually, but AMD didn't do that. They killed AM3+ without offering a newer socket, or offering 6-8 core chips for FM2+. This is after they tried to pump up everyone on the MOAR CORES system. If I have to buy a quad core, the i5 smears the a10 across the walls. I'm still waiting for an upgrade from the fx6300.....
Posted on Reply
#23
GhostRyder
Interesting to say the least, would really be nice to see the power reduction numbers in action on a laptop (Or other mobile device) to see how much better the battery life is on the chip compared with the steam roller chips. Would also be nice if the GPU received a nice step up in platform because that would make the mobile media house laptops more appealing with that chip inside.

But we have to wait for further details/proof before we decide anything.
Posted on Reply
#24
alwayssts
I preferr this now often-referenced article:

https://www.semiwiki.com/forum/content/3884-who-will-lead-10nm.html

Up to 14nm Intel/20nm Fab Club they seem fairly similar enough to state one notch larger node from AMD is similar to one size smaller from Intel.

When you start getting to 14nm from the CPA, things start to get switched around. My best guess (and the numbers more-or-less follow the logic) to this is Intel will be targeting the low-power/lower-end of the voltage curve in favor of density, where-as Samsung/GF may target higher performance in effort to be more TSMC-like (notice how the CPA/TSMC essentially end in the same place). I haven't gone deep where they got their 14/16nm numbers from in relation to version of process, but it meshes pretty cleanly with TSMC shooting for roughly 10% smaller chips with 16ff+ vs earlier 16nm (which would bring them in line with the CPA 14nm process for density), and the CPA having 'early' vs 'plus' versions of the 14nm process, which the later is supposed to improve speed/power consumption (read: be more like TSMC).
Posted on Reply
#25
Jorge
Sony Xperia S
"Now, if only AMD can put six of these leaner, faster, and greener "Excavator" modules onto an AM3+ chip."

I hope also so because if they do it, I would immediately go a buy the plaform with it.
Unfortunately AMD dropped the ball and should have released an Excavator based AM3+, but they are NOT going to do this. Zen will be the upgrade for AM3+ users but it will require a new socket/mobo. While Zen will be a big performance jump from the current FX Vishera series, it's taken YEARS too long for AMD to deliver this performance upgrade. AMD should have definitely offered an Excavator cored AM3+ late in 2014 so AM3+ users had a viable upgrade path.

Exceededgoku
Genuine question, how does AMD expect to compete with Intel when they're process node size is double that of the Skylake CPU's?
Many people do not understand that node size below ~32Nm does not buy much in actual CPU performance. What is offers is lower power consumption and higher transistor density with lower unit costs. It's pointless to rush to lower die size based on cost for each new iteration. Anyone who has paid attention knows that Intel has not achieved any significant gains in CPU performance with their last three node drops. That should be no surprise if you understand the process.

Technically informed people are not buying Intel chips based on the node size. In fact Intel just delayed their 14Nm Skylake due to slow uptake on their current products and development issues with the die shrink.

Sony Xperia S
AMD is expecting to move to a new process with Zen sometime in 2016.

At that time, the process gap will be largely gone.
The process diff has little impact on actual CPU/APU performance other than primarily lower power consumption. Carrizo uses advanced power management to make significant reductions in power consumption.
Posted on Reply
Add your own comment