Slashdot Mirror
AMD Says It's 'Ambidextrous,' Hints It May Offer ARM Chips
J. Dzhugashvili writes "Today at its Financial Analyst Day, AMD made statements that strongly suggest it plans to offer ARM-based chips alongside its x86 CPUs and APUs. According to coverage of the event, top executives including CEO Rory Read talked up an 'ambidextrous' approach to instruction-set architectures. One executive went even further: 'She said AMD will not be "religious" about architectures and touted AMD's "flexibility" as one of its key strategic advantages for the future.' The roadmaps the execs showed focused on x86 offerings, but it seems AMD is overtly setting the stage for a collaboration with ARM."
A PC(or laptop) running Windows 8(or any OS which supports both x86 and ARM) powered by a processor having full x86-64 support, and a low power ARM with a GPU capable of basic stuff like handling browsing and media playback
So, when you switch to a high requirement program (Gaming,encoding,VS,etc) the x86 cores turn on like a coprocessor and the work is handed to them
The ARM handles the UI and other stuff
It's a tough question. The Intel Atom has an edge on ARM, but it's not a big one, and while a high-performance ARM chip costs below $20, the Atom is significantly more. On the other hand, right now there are no ARM implementations that are really competitive on the PC front, and probably won't be until ARMv8 (64-bit) chips, or at least until Cortex-A15. A15 chips will probably come out in late 2012 and be a bit faster than the Atom, but a long way from Sandy Bridge and the other current Intel designs.
The most powerful general-purpose processor in the world (Power7) is a huge seller for IBM, and is a PowerPC implementation. PPC is also big in telecom applications, and Freescale does a number of fairly high-performance designs for that market.
The PPC used in the AmigaOne X1000 is a PA Semi PA6T - not very fast, designed as a low-power chip, and long-dead. Apple bought the company a few years ago, and I'm pretty sure new PA6T's are not being made. I suppose that speaks volumes about how many X1000's they reasonably expect to sell...
Wondering if a big state-of-the-art chip-fab like AMD getting into ARM processors might make sub-45nm ARM processors a possibility? AFAIK, only X86 chips are made like this just now. Could lead to fantastic performance-per-Watt chips coming off the line.
"I bless every day that I continue to live, for every day is pure profit."
Not technically true. Power7 belongs to the same family of architectures as PowerPC, but it's not really appropriate to say that Power7 is a PowerPC implementation. You might say that PowerPC is an uncle of Power7.
Its also worth noting that ARM has never been about performance until the semi-recent smart phone (mobile computing) surge. And even today, performance takes a backseat to power consumption. And it is here where ARM has always led the way. ARM vs Intel, ARM provides better price, better consumption, and very competative performance, albeit second place. But given the market to whch ARM is primarily focused on, ARM easily scores the win; in spite of Intels best efforts.
For those doing more traditional embedded development, Intel's offers are likely front runners. For those participating in the mobile computer segment, ARM, by far, is the very clear winner.
Does that mean it's using two ARMs at once?
(duck)
File under 'M' for 'Manic ranting'
... would it be possible (or I guess more importantly) worthwhile to put x86 cores WITH ARM cores on a single chip?
In addition to offering dual boot capabilities, it might be useful to run "Virtual" (or sort of virtual) machines at full speed. I've often thought it would be nice to run some of the thousands(!) of cellphone Apps that I have on my laptop. Although it might be tricky to implement multi-touch correctly, still I'd think there might be some utility.
Or maybe all CPUs today are very generalized RISCy architectures with everything taken care of in microcode (or maybe nowadays it's nanocode)? That would make it (comparatively) really easy to do, right?
The price tag is directly comparable, because ARM doesn't make processors, they sell licenses to designs. The only relevant metric is really performance at a given power point.
The closest competitor is Intel's Atom chips. At comparable power points, the current ARM chips seem to substantially outperform Atom chips, and the ARM chips scale far lower than Intel's do. It becomes a bit murkier at higher power levels, since until recently nobody was really making ARM chips that high, but we'll see a lot more competition in this field in the future with the ARM Cortex A15, which is intended to be a lot more scalable. The current design is planned to go from 1.0GHz single-core, up to 2.5GHz eight-core, depending on what the integrator wants. On top of that, they've got the new Cortex A7 that they've designed as an ultra-lower performance chip, which is intended to be a much simpler architecture that's still ISA-compatible with the A15. The intention is actually to put an A7 and A15 in the same SoC, so that the SoC can entirely turn off the A15 cores when only low performance is needed (like playing audio or video, since that's done almost entirely on a DSP). This is similar to what nVidia did with the Tegra 3, just taken even farther.
Much of this is a change of focus... Instead of beefy desktop CPUs running bloated OS, the focus is becoming more on portable devices.
Basically, this is "We're hanging in there in the desktop/laptop market, but rather than hang on to our piece of a shrinking pie, we want to get in on the pie that's getting bigger".
ARM is superior in low-power applications. It's highest-end CPUs maybe match Intel Atom, but often have far more peripherals (such as a fairly decent GPU and 1080p multi-format video decoding all on a tiny chip about the size of your thumbnail. Seriously - I can almost completely cover an OMAP4 with my thumb.)
retrorocket.o not found, launch anyway?
http://www.arm.com/files/downloads/ARMv8_Architecture.pdf
Power7 is fully compatible with the PPC 2.06 spec. How is it not a PPC ISA implementation?
AIUI ARM do HDL design of processor cores, then they pass that HDL on to other companies who make complete chip designs based on it. Those companies in turn pass the designs onto fabs (which may be in-house or external) for manufacture. IIRC some vendors also do their own HDL work and only license the basic architectural design from ARM.
note: i'm known as plugwash most places but i screwd up registering that here somehow in the past and now can't register
Since they have no products using that other architecture I think the word they were looking for is "Bicurious".
moi
You're confusing AMD and Intel. StrongARM was bought by Intel not AMD.
The real question is: Will the left hand know what the right hand is doing?
Modern architectures usually don't do that. There is a solution to this problem, but it's kind of MESI.
Ezekiel 23:20
The Tegra is basically an ARM SoC with an nVidia video system. Maybe they're looking at doing an ARM SoC with the ATI video core...
Seems some are working on bringing ARM into the server rack, and we can see the reason when we read about the kinds of power and cooling issues there are around some of the larger server farms.
comment first, facts later. http://chem.tufts.edu/AnswersInScience/RelativityofWrong.htm
Licensing fees.
AMD bleeds money to Intel for the x86 instruction set. At one point this was manditory since all the programs out there that were able to be run by a comparatively inexperienced computer users were written for the computers they could find at Radio Shack et al. Now that Microsoft and Google are popular and platform agnostic (Linux/Android vs win8) AMD has a window of opportunity to start from scratch and just offer a kernel patch to have your apps run on their chips. This new direction is going to be interesting to see execute. Intel was and is the gatekeeper of the consumer PC space now that is not so stable. Android can and has already been ported to x86,MIPS and a whole slew of variant ARM archetectures. To top it off millions of people use and enjoy Android, distributors like that they can make the products cheaply and must stay with the platform to keep their purchases/investments, lastly carriers love it cause they can lock you in for 3 years at premium rates.
It looks like this time Intel might have to tighten it's belt for a change.
A loop, by its nature, continues. If that didn't make sense, start reading this sentence again.
I never understood why file servers didn't use low power processors. Recently we've seen more and more ARM NAS devices, but I figured FTP servers and such would use these "lower end" processors simply because they only need to perform minimal computation to validate users and serve files.
Every time I start to have faith in humanity, I ruin it by driving to work between 7 and 8 am.
AMD builds a hybrid chip. It uses the ARM core for everyday tasks and then the x86 core when power is necessary. Kind of what Samsung does with their 5 core processor. Add in an AMD graphics core and that would bring some power.
Its also worth noting that ARM has never been about performance until the semi-recent smart phone (mobile computing) surge. And even today, performance takes a backseat to power consumption.
It was a long time ago, but not "never", when ARM was about performance and running circles around the 80286 and 68000 CPUs.
The problem with ARM is there are literally millions of x86 programs that have become an integral part of peoples lives, this is also why even though Linux has been getting better each year it fails to find any real gains. Everything from that camera that came with the photo software your Aunt Sue loves to Corel and Photoshop, from that bain of Linux geeks MS Office to Quickbooks/Quicken which is God in small business and rightly so.
The reason ARM is able to gain so much in mobile is because frankly geeks have never understood how normal users think, as someone who has to understand their needs or go out of business i think i can shed some light. you see to a geek that Droid or iPhone is a general computing device, to a normal user it doesn't even have an OS, its just "A screen with buttons i can google and play games on that I'll chunk when the contract is up" and that's it. they have been conditioned that nothing is compatible so assume when they chunk the phone the only thing they'll keep is the SIM card and that's that. Creates a lot of waste but is great for the carrier. Tablets to the consumer is the same, its a large mostly disposable flatscreen TV that can let them Google. There is no real attachment there, no real desire by the majority to develop long term rapport with programs. this is why ARM netbooks went nowhere because to them a netbook is NOT just a general computing device, its a "baby laptop that should do everything my big laptop does only slower, because babies are smaller than grownups" see how that works?
I think where AMD is on the right track and has a real shot is Fusion. Not 3 years ago i could walk into the local Walmart or staples and i'd be lucky if there was a single AMD machine, usually the cheapest machine in the house. Now I see AMD Fusion netbooks, laptops, all in ones, and even desktops, some going up to nearly $1000 in price and talking to some of the guys that i know working there they are brisk sellers. More and more the PC is not only the office machine, its also an entertainment center With the AMD Fusion chips not only do you get great battery life/lower electric bills, like my EEE E350 that gets 6 hours playing 720p and lets me HDMI into any 1080p set and watch videos, but you also get to have all your programs that you know and are familiar with and which frankly there is often no FOSS equivalent and probably never will be. There is no FOSS software that matches the features of Quickbooks or photoshop, and certainly nothing like the little quilting app I installed the other day for a customer on her new Acer AMD C60 netbook. while FOSS users would probably think its stupid and not waste time for her its a "must have" because it helps her to work up the patterns she is gonna use on her next quilt and to visualize what it will look like.
So I think the future is bright IF, and that's a BIG IF, AMD continues to play it smart. the new Vector based GPUs will lower the power footprint even lower while letting the APU use the GPU cores like a super fast floating point which will give any program using floating point a nice kick in the ass, and considering they've had to lower desktop output to keep up with all the orders for the Bobcat chips shows the OEMs think its the right path too. you can now get those chips in every form factor you can name, from HTPC to iMac style to netbooks and laptops. While i'm sure AMD never considered it a desktop chip the OEMs found that its more than good enough for the average user and its selling quite briskly so they made a good call there.
Finally there is one place where AMD has already fucked up, and that's the recent killing of the entire AM3 line. While consolidating to a few chips would have been smart IMHO killing the AM3 Stars chips when Bulldozer has neither the yields nor performance to take its place was just stupid. if you have an AM3 board I'd suggest you pop over to tigerdirect where they are selling Thu
ACs don't waste your time replying, your posts are never seen by me.
*BUT*, there comes a massive performance penalty which is that the clock rate now has to be twice as fast as a RISC processor in order to achieve the same results.
That's just complete bollocks.
A modern x86 processor (meaning... since the Pentium Pro in the mid 90s) is, internally, a RISC-like core with full OoO execution and so on and so forth.
Variable instruction decode is a pain in the ass and does add latency in the front end. This isn't great, but it is nowhere near a 50% reduction in IPC. Try more like 1-2% (measured via correlated cycle-accurate performance simulator), depending on how clever you get and in any case easily made up for by a clever widget or two.
Basically predictions of RISC eating x86 for breakfast were made over 15 years ago and never came to pass. Mostly by x86 morphing so that the difference was essentially irrelevant.
Your talk about northbridges sounds woefully out of date, too. This has nothing to do with ISA, and both major x86 vendors now have integrated northbridges.
You're closer to reality when talking about power. Regardless of the small IPC penalty, those decoders burn up a lot of power. There are ways to get around this, too, and for moderate perf moderate low power x86 does just fine. At the very low end of power, though, going to something like ARM makes sense.
The enemies of Democracy are
"The problem with ARM is there are literally millions of x86 programs that have become an integral part of peoples lives"
Not really. There are many ARM programs that have become and integral part of people lives. Android and IOS are two big example not to mention the apps that run on them.
Software is not as locked to an ISA as it once was. Microsoft and Apple have shown that with the move of Windows to ARM and the move of OS/X to x86.
Applications are not written in assembly anymore they are written in C++ or another high level language. Take your example of Photoshop? Moving Photoshop from Windows to Windows on ARM is probably a much simpler project a Windows and OS/X version. The same is true of Office.
I do think that AMDs Fusion is interesting but your reasoning on why people will keep use the x86 is not valid. They will only keep using x86 for as long as that is the best solution. IMHO x86 is endanger of being the next PDP-11 or VAX unless it can scale down to mobile and fast.
See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
AMD is clearly talking about using both x86 and GPU for compute work vs. focusing on x86 only... the ARM thing is just a wild speculation, or wishful thinking.
Never underestimate the bandwidth of a 747 filled with CD-ROMs.
Published benchmarks disagree with your assessment of ARM.
RSC(POWER1) is the most popular CPU architecture on Mars, and possibly in the solar system outside of Earth.
“Common sense is not so common.” — Voltaire
Well there's always encryption, but they could probably integrate an on-chip co-proc for those functions.
Laissez lire, et laissez danser; ces deux amusements ne feront jamais de mal au monde. - Voltaire
First off, Intel was selling ARM chips up until a few years ago. They snagged the famous "StrongArm" series off of DEC and rebranded it "XScale".
Second, ARM only recently established itself as THE x86 competitor. Go look up all the RISC architectures out there which were competing for dominance. If you needed high performance embedded, PowerPC has long been the way to go. SPARC has been competing in the embedded space. Hitachi made a go of it with their SuperH chips (eg. SH3).
Last but not least, MIPS is the old man of the bunch, and the one with the most fight left in it... It was always faster than ARM, even powering high-end (SGI) workstations and supercomputers in the past, while still competitive on the low-end. With China throwing it's weight behind MIPS as the basis of their domestic CPU development effort (Loongsoon/Dragon Chip), it's both advancing nicely, and being produced extremely inexpensively, which gives us things like the infamous $100 ICS Android tablet running on MIPS.
But more than that, ARM doesn't fit Intel's model... ARM just licenses the IP/Cores, and let's others fab them. Intel wants the whole pie. Even if they bought ARM, they couldn't stop existing licensees from continuing as before, and if they didn't keep ARM producing what customers wanted, switching over to MIPS wouldn't be that hard.
Slashdot gets worse every day... Pipedot: News for nerds, without the corporate slant
I agree w/ this. ARM is an overcrowded market already, and why would someone prefer AMD to an established vendor who's been making it for years, such as TI, Qualcomm, nVidia, Freescale - just about every other big name in the semiconductor industry? Why would anyone prefer AMD to those guys? AMD did a good thing when it first went to the x64, and they can make that a more RISCy CPU over time when memory is never less than 4GB, there are 64-bit versions of most apps and then they can start dropping 32-bit instructions from future implementations.
Otherwise, there is no reason why a particular instruction set by itself will make a CPU consume less power (beyond the traditional RISC vs CISC argument). So AMD needs to play to its strengths, which is the x64. Any Android vendor can already buy chips from the vendors I listed above, so why go to AMD? But if AMD makes an x86/x64 chip that's low power enough, it can potentially help save Microsoft's Windows 8 tablets from becoming a debacle, which again I agree w/ the other observation above that it will.
I can't say I've seen the driver issues you talk about.
Things tend (for me) to either work in linux, because the driver is supplied as a kernel module, or there's just no driver. In fact, for me, it's now considerably easier than windows. You don't even have to think about installing or rolling back drivers, because they're either just there already, or not available.
That's just me though, and what you're used to is a large part of it. I certainly do fall into the 1% here though.
Was 2004 the last time you tried linux?
I didn't really start using it seriously until about '06.
I told you already, Debian is the best at updates even across multiple versions. Also, OpenSuse and Fedora are where experiments happen and I doubt an "experienced Linux Admin" would be using them for something really important -unless she has the requisite depth to deal with a little breakage during an upgrade.
And posting again and again and again the same rant about an ABI isn't going to change the FACT that a stable ABI for drivers would make things worse for everyone. You keep trying to make comparisons to Windows in areas where it makes no sense at all to compare the two.
What's it going to take for you to realize that nobody cares if Linux gains market share on the desktop? You're making arguments (for years now) based the assumption that market share is a goal when in fact no one has such a goal.
Your perspective on this is so wrong and you've been corrected so many times by so many people I can't help but wonder if you've got some kind of learning disability or OCD or something. Seriously, not trying to be insulting at all. You've apparently started down this road in 2004 with some wrong assumptions that you just can't let go of, and that's really why you're so frustrated now.
Samsung took back my unlocked bootloader because Google wants me to rent movies. They're both evil.