Sunday, November 23rd 2014

AMD Mobile "Carrizo" Family of APUs Arrive in 2015

AMD (NYSE: AMD) today at its Future of Compute event announced the addition of its first high performance system-on-a-chip (SoC), codenamed "Carrizo", and a mainstream SoC codenamed "Carrizo-L" as part of the company's 2015 AMD Mobile APU family roadmap. In collaboration with hardware and software partners, these new 2015 AMD Mobile APUs are designed as complete solutions for gaming, productivity applications, and ultra high-definition 4K experiences. With support for Microsoft DirectX 12, OpenCL 2.0, AMD's Mantle API, AMD FreeSync and support for Microsoft's upcoming Windows 10 operating system, the 2015 AMD Mobile APU family enables the experiences consumers expect.

"We continue to innovate and build upon our existing IP to deliver great products for our customers," said John Byrne, senior vice president and general manager, Computing and Graphics business group, AMD. "AMD's commitment to graphics and compute performance, as expressed by our goal to improve APU energy efficiency 25x by 2020, combines with the latest industry standards and fresh innovation to drive the design of the 2015 AMD Mobile APU family. We are excited about the experiences these new APUs will bring and look forward to sharing more details in the first half of next year."

The flagship "Carrizo" processor will integrate the new x86 CPU core codenamed "Excavator" with next generation AMD Radeon graphics in the world's first Heterogeneous Systems Architecture (HSA) 1.0 compliant SoC. The "Carrizo-L" SoC integrates the CPU codenamed "Puma+" with AMD Radeon R-Series GCN GPUs and is intended for mainstream configurations. In addition, an AMD Secure Processor will be integrated into the "Carrizo" and "Carrizo-L" APUs, enabling ARM TrustZone across the entire family for the security commercial customers and consumers expect. Utilizing a single package infrastructure for "Carrizo" and "Carrizo-L," the 2015 AMD Mobile APU family simplifies partner designs across a broad range of commercial and consumer mobile systems.


"Carrizo" and "Carrizo-L," are scheduled to ship in 1H 2015, with laptop and All-in-One systems based on the 2015 AMD Mobile APU family expected in market by mid-year 2015.
Add your own comment

40 Comments on AMD Mobile "Carrizo" Family of APUs Arrive in 2015

#26
TRWOV
Damn, DDR3-2133 support? And I thought I was forward thinking when I got a 1866 Kit for my Athlon 5350
Posted on Reply
#27
vega22
RCoonAMD are focusing all their 20nm gear for consoles first and foremost to keep people (console makers) happy. Then console maker can sell millions more "slim" versions of everything. Not that AMD make any profit from it at all though. Next gen consoles will probably run intel, or even other chip makers...
that is their issue and solution in one go dude.

they need to get the consoles on 20nm asap so they can turn a profit on them otherwise they will be screwed. once they get that steady stream flowing we might then start to see some improvements for them.
Posted on Reply
#28
Dimsgr
Hello there :)

My 2 cents on the subject... I am currently on an nehalem i7 920, with promised low power consumption. Well, real life measures (on the wall) have shown almost double that, in contrast to various benchmarks over the internet (back then when it was realesed any way). My believe is, that at least at the past, Intel showcased bad company habits and ethics, with the consequence to directly be insulting and costly to customers. Why I should go to Intel again? The rush for Intel to show something new all the time, even if they are not ready to do so (as it has obviously happened to my case) points to the direction that, as someone else pointed, it is all about marketing; they are offering half baked products, regarding manufacturing abilities (or even with serious errata). Then, they are propably covering this by selling cherry picked hardware to the benchmarkin/testing sites. I wiil say no thank you, and I may prefer the approach of AMD, which is offering more mature products to the people
Posted on Reply
#29
HumanSmoke
DimsgrHello there :)
Well, hello to you new member!
DimsgrMy 2 cents on the subject... I am currently on an nehalem i7 920, with promised low power consumption. Well, real life measures (on the wall) have shown almost double that, in contrast to various benchmarks over the internet (back then when it was realesed any way). My believe is, that at least at the past, Intel showcased bad company habits and ethics, with the consequence to directly be insulting and costly to customers. Why I should go to Intel again? The rush for Intel to show something new all the time, even if they are not ready to do so (as it has obviously happened to my case) points to the direction that, as someone else pointed, it is all about marketing; they are offering half baked products, regarding manufacturing abilities (or even with serious errata). Then, they are propably covering this by selling cherry picked hardware to the benchmarkin/testing sites. I wiil say no thank you, and I may prefer the approach of AMD, which is offering more mature products to the people
Well done.
Mentioning the hardware concerned in the article: None....so your 2 cents is actually worth two cents less than that.

BTW: Who promised low power consumption for the i7 920? Intel only intro'ed per-core power management with Bloomfield. You'll also find that users and site reviews got what was expected (which would include me since I ran two i7 950's FWIW), namely higher power consumption for better overall performance.
FWIW, if you want to highlight Intel's "half baked" products, you might want to check AMD's track record.
Posted on Reply
#30
Prima.Vera
Crap DaddyThis. Sadly all bases are covered by Intel. Power and efficiency is all that matter, all devices need to be thinner, lighter, longer battery life.
Except for the price... ;)
Posted on Reply
#31
Crap Daddy
Prima.VeraExcept for the price... ;)
True. I think Core M 5Y70 goes for around $300 tray price, no wonder these ultrathin even fanless hybrids, 2 in 1 and whatever are so damn expensive.
Posted on Reply
#32
Dimsgr
HumanSmokeWell, hello to you new member!

Well done.
Mentioning the hardware concerned in the article: None....so your 2 cents is actually worth two cents less than that.

BTW: Who promised low power consumption for the i7 920? Intel only intro'ed per-core power management with Bloomfield. You'll also find that users and site reviews got what was expected (which would include me since I ran two i7 950's FWIW), namely higher power consumption for better overall performance.
FWIW, if you want to highlight Intel's "half baked" products, you might want to check AMD's track record.
Hello, nice to meet you here :)

Concerning the hardware mentioned, I remember measuring the i7 920 compared to the then reviews on the internet ans that is what I mean by saying "promised". Please, have in mind, I am not talking about some maybe 15% percentage, which would be acceptable, but to an almost double the consumption. And I am most likely not having a dude, as temperatures are just fine.

As for the track record of AMD, I have not checked about, but I have got, hitsorically, good CPUs out of them. Especially, Phenom ones where like cold mountains :) My conclusion is that I would be twice as sceptical to go Intel for 24/5 computing, where stability and consumption matters most; there is a reason Intel is not offering their Xeons on the latest manufacturing process at launch I might say. FWIW, as shrinking has become harder and harder, Intel itself might also be pulled to pay more attention to the processes it uses on manufacturing, so we may get better products as well.

Regards
Posted on Reply
#33
HumanSmoke
DimsgrConcerning the hardware mentioned, I remember measuring the i7 920 compared to the then reviews on the internet ans that is what I mean by saying "promised". Please, have in mind, I am not talking about some maybe 15% percentage, which would be acceptable, but to an almost double the consumption.
Never had the same problem...and my systems are all overclocked and under water. Power consumption was in line with voltage requirement. The 950 used more power than the Yorkfield-12M I was using prior, but power scaling was directly in line with both estimates and reviews that are tailored to people who OC.
DimsgrMy conclusion is that I would be twice as sceptical to go Intel for 24/5 computing, where stability and consumption matters most; there is a reason Intel is not offering their Xeons on the latest manufacturing process at launch I might say.
WTF are you talking about? Intel launched 22nm with desktop Haswell and the server Xeon range on the same day - June 2, 2013.
Posted on Reply
#34
EarthDog
I think his perception of "promised" less power consumption versus it's intent was skewed.

I also wonder how he tested power consumption in the first place... I'm betting it wasn't apples to apples...
Posted on Reply
#35
HumanSmoke
EarthDogI think his perception of "promised" less power consumption versus it's intent was skewed.
Probably mixing up a few bullet points at a guess. As Anandtech and a few of the other preview sites noted, Intel's only statement on Nehalem power consumption concerned their usual pessimistic "1% increase in performance for no more than a 1% increase in power consumption" company line (i.e. no gain in performance per watt). Intel more often than not are conservative in their estimations. Basic psychology - exceed public expectations and it adds to a favourable overall experience even if the product has limited impact. Better under-promise and over-deliver rather than the reverse.

Maybe our new member got a sad C0 overclocker rather than a sweet D0 SLBCH....assuming it wasn't just an excuse to put a knock on Intel (I'm 50/50 on that one)
EarthDogI also wonder how he tested power consumption in the first place... I'm betting it wasn't apples to apples...
I'm also guessing that the power consumption observation is anecdotal. Plenty of people came from a C2D and thought they could slap an Arctic Freezer 7 (or other budget/midrange tower cooler) and pump up the volume on a C0 i7 just as easily as they could on a E8400...and not realizing there's a marked difference between dialling in a 400+MHz FSB on a Penryn, and punching 1.4+V through a Bloomfield.
Posted on Reply
#36
Dimsgr
HumanSmokeNever had the same problem...and my systems are all overclocked and under water. Power consumption was in line with voltage requirement. The 950 used more power than the Yorkfield-12M I was using prior, but power scaling was directly in line with both estimates and reviews that are tailored to people who OC.

WTF are you talking about? Intel launched 22nm with desktop Haswell and the server Xeon range on the same day - June 2, 2013.
I am talking about manufacturing process, not microarchitecture, if that is the case here. Please, take a look at here

www.cpu-world.com/CPUs/Core_i7/Intel-Core i7-3770K.html or here, in the first 3 paragraphs

en.wikipedia.org/wiki/Ivy_Bridge_(microarchitecture)

It is obvious that Intel is wating for the manufacturing process to mature to begin production of server chips

As for my case... I may make clear that the measurement was in the wall with the proper instrument, many years ago. I do not have access to the meter right now but I would like to measure again.

Anyway, my experience, and maybe reality is that Intel (or x86 in general) has no luck in low power consumption, so instead of trying to make a pc mobile (or vice versa for ARM) I would prefer keep 'em designing nice table chips

best regards
Posted on Reply
#37
EarthDog
At the wall = total system power. In order to isolate the CPU power, you would need a clamping ampmeter on the 8 pin CPU power.
Posted on Reply
#38
HumanSmoke
DimsgrI am talking about manufacturing process, not microarchitecture, if that is the case here. Please, take a look at here

www.cpu-world.com/CPUs/Core_i7/Intel-Core i7-3770K.html or here, in the first 3 paragraphs

en.wikipedia.org/wiki/Ivy_Bridge_(microarchitecture)

It is obvious that Intel is wating for the manufacturing process to mature to begin production of server chips
I think you're grasping at straws.
Desktop Ivy Bridge launch (as per your link) April 29, 2012.
Server Ivy Bridge launch : May 14, 2012......13 days after the desktop variants.
This thirteen day delay in your mind proves Intel's "failure", and makes AMD's architecture/process offerings superior to Intel? Really? Lets, take a look at AMD's analogues:
Desktop Bulldozer launch: October 12, 2011
Server Bulldozer launch: November 14, 2011.......33 days after the desktop variants ( actually somewhat worse than that since AMD was supposed to launch Interlagos and Valencia before the desktop FX line). Even the difference between the following FX Piledriver and Abu Dhabi server parts was no shorter than the Ivy Bridge desktop and server parts.
Posted on Reply
#39
Dimsgr
EarthDogAt the wall = total system power. In order to isolate the CPU power, you would need a clamping ampmeter on the 8 pin CPU power.
I might say that this is what I should have done in the first place
HumanSmokeI think you're grasping at straws.
Desktop Ivy Bridge launch (as per your link) April 29, 2012.
Server Ivy Bridge launch : May 14, 2012......13 days after the desktop variants.
This thirteen day delay in your mind proves Intel's "failure", and makes AMD's architecture/process offerings superior to Intel? Really? Lets, take a look at AMD's analogues:
Desktop Bulldozer launch: October 12, 2011
Server Bulldozer launch: November 14, 2011.......33 days after the desktop variants ( actually somewhat worse than that since AMD was supposed to launch Interlagos and Valencia before the desktop FX line). Even the difference between the following FX Piledriver and Abu Dhabi server parts was no shorter than the Ivy Bridge desktop and server parts.
My link is actually a quick search on the site you provided; my actual link is that from Wikipedia, which states mid 2012 for the mainstreamm 22nm and 2013 for the server parts. Also, Cpu-world talks about "Introduction date", so perhaps could be a paper launch? All in all, I have nothing againts Intel (nor that they do care of course :) ), but I consider some points before making my decisions before purchasing some products e.g. a) A custom made PC out of i7 920 was a big difference , compared to the initial reviews out there b) A Dell laptop with a i7 720QM has enormous battery life (we all know Intel ships the best parts to special partners) c) a custom Phenom x3 720 (too many 920, 720, 720 :) ), granted with a 4650 from ATI, is icey cold at 60 watts idle.

best regards
Posted on Reply
#40
HumanSmoke
DimsgrMy link is actually a quick search on the site you provided; my actual link is that from Wikipedia, which states mid 2012 for the mainstreamm 22nm and 2013 for the server parts. Also, Cpu-world talks about "Introduction date", so perhaps could be a paper launch?
You are confusing "server parts" with processor families. The link you provided clearly states that 2013 was for Ivy Bridge-E/-EN/-EP/-EX, and that has very little to do with process issues. Rather it is because the HEDT/2P/4P/8P lines cadence is out of step with the mainstream parts (so ROI becomes the prime mover), and while they share an architecture with the Ivy Bridge (non-E) parts they aren't the same silicon - unlike AMD's Abu Dhabi Opteron for instance that is just an MCM package not unlike Intel's Kentsfield/Yorkfield "quads" - two Piledriver CPUs attached to the same substrate - and in many cases don't even use the same socket/pin count within the family.
Posted on Reply
Add your own comment
Nov 26th, 2024 11:32 EST change timezone

New Forum Posts

Popular Reviews

Controversial News Posts