AMD Designing All-New CPU Cores For ARMv8, X86

crookedvulture (1866146) writes "AMD just revealed that it has two all-new CPU cores in the works. One will be compatible with the 64-bit ARMv8 instruction set, while the other is meant as an x86 replacement for the Bulldozer architecture and its descendants. Both cores have been designed from the ground up by a team led by Jim Keller, the lead architect behind AMD's K8 architecture. Keller worked at Apple on the A4 and A4 before returning to AMD in 2012. The first chips based on the new AMD cores are due in 2016."

Comments

Keller worked at Apple on the A4 and A4

[nitehawk214]

Probably worked on the A4 and A4 and the A4, as well.

Re:Keller worked at Apple on the A4 and A4

[flyingfsck]

No, no, that is obviously a typo. T'was the A4, Letter and Legal.

Re:Keller worked at Apple on the A4 and A4

[GodfatherofSoul]

You mean the A4 on the A4 on the A4?

Re:Keller worked at Apple on the A4 and A4

[nitehawk214]

Ahh, you figured it out, the multiple A4s were referencing different things.

So pick two of those.

Re:Been a long time since I cared

[werepants]

The last time I truly got excited about AMD was when the K6-2 came out.

What? During the P4 days AMD was ahead in almost every category in the benchmarks... did you miss that whole era? No denying the picture today is far less exciting, though.

Serious Question

[Anonymous Coward]

Is AMD just around so Intel doesn't get bogged down by anti-monopoly or antitrust penalties?

Re: Serious Question

[Anonymous Coward]

64 cores per U, 80% intel performance per core, at 12% intel price.

Re: Serious Question

[Junta]

Well, something of an oversimplification/exaggeration.

64 'cores' is 32 piledriver modules. That was a gamble that by and large did not pan out as hoped. For a lot of applications, you must consider those 32 cores. Intel is currently at 12 cores per package versus AMD's 8 per package. Intel is less frequently found with their EP line in a 4 socket configuration because the performance of dual socket can be much higher with Intel's QPI than 4 socket. AMD can't do that topology, so you might as well do 4 socket. Additionally, the memory architecture of Intel tends to cause more dimm slots to be put on a board. AMD's thermals are actually a bit worse than Intel's, so it's not that AMD can be reasonably crammed in but Intel cannot. The pricing disparity is something that Intel chooses at their discretion (their margin is obscene), so if Intel ever gets pressure, they could halve their margin and still be healthy margin-wise.

I'm hoping this lives up to the legacy of the K7 architecture. K7 architecture left Intel horribly embarrassed and took years to finally catch up with when they launched Nehalem. Bulldozer was a decent experiment and software tooling has improved utilization, but it's still rough. With Intel ahead in both microarchitecture and manufacturing process, AMD is currently left with 'budget' pricing out of desperation as their strategy. This is by no means something to dismiss, but it's certainly less exciting and perhaps not sustainable since their costs are in fact higher than Intel's cost (though Intel's R&D budget is gigantic to fuel that low-cost per-unit advantage, so the difference between gross margin between Intel and AMD is huge, but net margin isn't as drastic). If the bulldozer scheme had worked out well, it could have meant another era of AMD dominance, but it sadly didn't work as well in practice.

Re:Right, because that worked so well

[amorsen]

Transmeta was at the end of the era where decoding performance mattered. Keeping the translated code around was actually useful. These days decoding is approximately free on any CPU with half-decent performance -- the amount of extra die space for a complex decoder is not worth worrying about.

You can save a bit of power with a simpler decode stage, but you are unlikely to beat ARM Thumb-2 on power by software-translating x86 the way Transmeta did. Besides, most of the interesting code for low power applications is ARM or MIPS already, so what is the point?

Best of luck to them

[Dega704]

I was such an AMD fanboy ever since I built my first (new) computer with a K6-II. I have to admit I miss the days of the Athlon being called "The CPU that keeps Intel awake at night." After Bulldozer bombed so thoroughly I just gave up and haven't followed AMD's products since. I definitely wouldn't mind a comeback, if they can pull it off.

Re:Right, because that worked so well

[amorsen]

You cannot meaningfully do reordering and so on in software on a modern CPU. You do not know in advance which operands will be available from memory at which time. You have to redo that work every time you get to the code (unless it is in a tight loop, but modern x86's are REALLY good at tight loops) because circumstances will likely have changed -- and you cannot reorder in software every time, that is just too costly.

If you want to see an architecture which looks like it has a chance of breaking the limits on single-threaded performance, look at the Mill. In theory you could software-translate x86 to Mill code and gain performance, but it would be really tricky and no Mill implementations exist yet.