If all you care about is the IGP and don't care about power consumption and don't mind using closed-source drivers if you need real 3D performance, then the A10 is nice.

If, like me, you care about CPU, then the very-high end Haswell parts are about twice the price but deliver more than twice the performance in a lower power envelope. You can also buy Haswell parts for lower prices that are still comfortably ahead of the A10 at any CPU-bound load and have even lower power envelopes.

Oh and if the GPU is really that vital, there's are these things called "under $100 discrete cards" that are easily ahead of any IGP made by any company and are not insanely expensive.

When I say they are ahead, let me put it in perspective: Next year's highest-end Kaveri parts will -- in a theoretically perfect world -- approach the performance of an HD 7750 that's by no means high-end. Oh, and they are guaranteed to have a higher power envelope and substantially lower CPU performance than the systems that I have already been running for several months. Oh, and don't expect them to be given away for free, and yes, you have to buy a new motherboard just to use one.

You're doing it wrong.
Signed, Arch Linux user who has multiple Haswell machines running perfectly... oh and with open source drivers too.

Yeah.. as a raspberry pi early adopter, lemme tell you something: You're full of it. It's irrelevant if some flavor of Linux has a particular version targeted at particular ARM platforms because AMD's ARM platform will have its own requirements that will require quite a bit of software work to ensure that a full, production-quality software stack can get running on it from day zero.

Trust me, if everything ARM + Linux was completely perfect, then Torvalds wouldn't be on his high-horse about ARM SoCs. ARM's customization is wonderful... for vendors that want to lock you in to a particular platform even if most of the software is open source. ARM's non-standardization and proprietary nature is a giant PITA in a whole bunch of other areas though.

Actually, Atom has had extremely competitive idle power draw going back several generations... the trick was that the idle power draw was best on the embedded platforms that were not widely released prior to Clovertrail and *not* on the desktop platforms where the separate chipset alone used more power than the CPU.

That's not my point, however. At best I'd expect those A57 parts to have performance parity with Avaton under load... and Intel has already solved the idle power draw issues, especially when it comes to SoCs.

Hey, if your servers sit around doing absolutely nothing all day, then I have a better idea: Turn them off and use zero power.

ARM fanboy quote from 2013: "So what if the [ARM server part] power envelope is larger if they spend more time in a lower power state?"

Intel fanboy quote from 2008: "So what if the [Intel Atom parts] power envelope is larger if they spend more time in a lower power state?"

Watch the wheel o' time turn & turn.

The Kaveri-based APUs in servers are certainly not going to be great for every workload, but for servers that can take advantage of GPU compute, they give AMD a unique advantage in a competitive server environment.

Those ARM parts on the other hand have proven one thing: Just because ARM (and more importantly, Qualcomm) make good chips for smartphones doesn't mean that ARM is magic and can avoid physics.

  The 8 core Cortex A57 parts on AMD's roadmap for late 2014 have a 50% higher power envelope than the high-end 8-core Avoton parts that Intel has on sale *this year* (30 watts vs. 20 watts). By the time they launch, Intel will either have launched or be on the verge of launching 14nm microserver parts. These things are a nice prototype, and AMD is easily the best vendor for ARM servers since it has experience in the server world, but ARM ain't about to take over the server room at this pace.

I'm seeing lots of hyperbolic statements from people like you who:
1. Are convinced that the NSA runs Intel because... uh... conspiracy?
2. Are convinced that Via (to name one example) isn't on the NSA list.. which is why this is just a thinly disguised 2 minutes of hate on Intel with a veneer of "it's for the people!"
3. While you are 100% convinced that every number from RdRand is an NSA conspiracy, you flat out refuse to believe in real security holes that actually happened and that are actually documented.

If you have that level of paranoia, then you ought to just stop using any instruction of any kind on any big-bad Intel CPU (and ARM/ARM/etc. for that matter).

Basically, you are saying that even though it is easy to verify that the output of RdRand is random (and it is) that RdRand is "unauditable". Well guess what? If one instruction on a big-bad Intel CPU is "unaditable" then they ALL ARE. Add instructions? Can't be trusted. Mul instructions? NSA backdoors. Comparison ops? Obviously doctored to let NSA code sneak through?

You know what an "audited" software PRNG from a bunch of NSA-hating hackers is? A bunch of instructions that are executed on a CPU, that's what it is. But... we've just shown that all instructions on the CPU are really NSA backdoors, so even your magic "open source" PRNG is just another NSA backdoor! Have fun!

Ooh look.. an AC... do you work for the NSA? Are you here intentionally trying to plant a false flag about RdRand to push people into using crappy RNGs that are easier for you guys at Club Meade to break?

If I were the NSA, I'd be doing everything in my power to get the paranoid types to *NOT* trust good security solutions because of phantom magical "backdoors" that don't exist, while I would simultaneously exploit *REAL* security holes to spy on those exact same people. Just sayin...

You know.. I've seen plenty of real security research that says that the RdRand RNG is actually very good and produces very high quality output.

Here's just one set of results showing that the output is truly random, so-called NSA "backdoors" or not:
http://smackerelofopinion.blogspot.com/2012/10/intel-rdrand-instruction-revisited.html

You know what *ISN'T* truly random? When guys just like you who are all paranoid about the NSA went and broke OpenSSL in Debian for over 2 years in the name of "fixing" code: http://research.swtch.com/openssl

Oh, and are you and the petitioners going to be intellectually honest and demand the complete removal of Via Padlock support from Linux, or is this only an anti-Intel fanboy rant thinly disguised as "sticking it to the man?"

Guess what the NSA loves: When lemmings throw away real security solutions because they think the NSA engineers every transistor in every piece of hardware. Go ahead and try to put together your own crypto solutions, the NSA *wants* you to do that, because they are a hell of a lot smarter than you are.

Uh... deliberately weakened from what exactly? The magical multi-gigabit random number generator that doesn't exist in earlier chips?

It's pretty easy to go look at randomness and test it you know.... and Intel's RNG has stood up to testing and scrutiny by a whole bunch of real security researchers, not just paranoid basement dwellers who see the NSA around every corner.

If this petition is actually about people who think that an RNG is some evil NSA plot then I have news for you: THE NSA IS PROBABLY #1 IN LINE SUPPORTING THIS PETITION BECAUSE THE RDRAND GENERATOR PROBABLY MAKES THEIR LIVES A WHOLE LOT HARDER COMPARED TO CRAPPY DIY RNGS THAT THE PARANOID BASEMENT DWELLERS "INVENT" TO AVOID BIG-BAD INTEL!!!

Shouldn't we be welcoming RdRand with open arms? It's a mathematically proven high-quality random number generator that lets chips like Ivy Bridge & Haswell produce large amounts of true random data (not a simple PRNG data) at multi-gigabit speeds.

There are some excellent slides describing RdRand here: http://software.intel.com/en-us/tags/20757

I would strongly recommend using it wherever feasible as it is a great boon to security in Linux.

So is some AMD/ARM fanboy saying that it's not fair that AMD/ARM haven't bothered to implement RdRand yet so therefore nobody should be allowed to use it? How about we extend that logic to other pieces of hardware? Say, when AMD comes out with an improved GPU, let's say that Linux shouldn't support it because Intel doesn't have the same hardware.. fair is fair right?

Now that Dear Leader Obama is the president and has decided that we all need to give war a chance, NPR has taken to calling anyone who doesn't want his war to be an "isolationist."

  You'll note that this term was never used against people who disagreed with wars in Afghanistan or Iraq... instead those people were "anti-war" or "pro-peace". We basically need another Republican as president so that the press can go back to attacking the president instead of being his trained lapdog.

Yes you are right... it is unrealistically favorable to AMD that is since if you had bothered to look at the charts you'd note that the benchmark was a CPU-only test that gave AMD the advantage of being able to run the GPU at very low power since it isn't being stressed and redirect the power consumption to the CPU...

Oh and they also tested with discrete GPUs that completely relieve the APU of having to expend any energy on the IGP at all.

These chips are slightly faster (given equal core counts) than their predecessors but not in any interesting way.

  However, you have to remember that these are really server chips that are repurposed for high-end desktop use. The one vital metric where these chips shine is in their power consumption (or lack thereof): Techreport did a test where the 6-core 4960X running full-bore is using about the same amount of power as a desktop A10-6800K part ( http://techreport.com/review/25293/intel-core-i7-4960x-processor-reviewed/9 )

That level of power efficiency will do wonders in the server world and these chips (and their 12-core bigger brothers) should do quite well in servers.