A Comparison: SSE4, AVX & VMX

Is this actually what AVX will have or just a guess at this point. 1Tflop SP from an 8 core x86 would be pretty impressive!
Yes, this is what Haswell's implementation of AVX2 and FMA will be capable of. It hasn't been officially confirmed yet, but it's easy to deduce as the only logical answer.

We know for a fact Haswell will support FMA, and we also know Sandy Bridge has a separate ADD and MUL execution unit. They can't go for a single FMA unit with Haswell, since that would dramatically cripple legacy performance. They also can't go for an ADD+FMA or MUL+FMA combination, because then the same port is needed by MUL and FMA or ADD and FMA respectively, and with typical Instruction mix frequencies this actually results in lower performance due to port contention!

So under the safe assumption that they want the extra transistors to pay off, the only sane option is dual FMA units. This also simplifies scheduling. And note that Bulldozer already has dual FMA (even though it's 128-bit each, note that it's on 32 nm).

This also isn't all that incredible compared to what we've come to expect from GPUs. And Intel clearly is putting a lot of Larrabee's technology into AVX2.
Ive no doubt Haswell will be capable of hitting 4 Ghz but I doubt Intel will clock it that high given the lack of competition. I'm fairly sure intel could have been releasing stock 4ghz CPU's since Sandybridge if they'd have felt the need.
AMD is clearly aiming to hit 4 GHz sooner rather than later. Regardless of superior IPC, the market will demand Intel to follow suit (or steal their thunder). Also for what it's worth 3.9 GHz would actually suffice for 500 GFLOPS out of a quad-core, and we're at 3.8 GHz Turbo Boost frequencies already.
 
Last edited by a moderator:
Cheers Nick, its a pretty exciting prospect. I kinda wish we could see Haswell in a console just so we can see what such a monster would be capable of if fully utilised for games. Good point about 4ghz too, I guess if AMD hit it then Intel will have little choice but to match for marketing reasons.
 
AMD is clearly aiming to hit 4 GHz sooner rather than later. Regardless of superior IPC, the market will demand Intel to follow suit (or steal their thunder). Also for what it's worth 3.9 GHz would actually suffice for 500 GFLOPS out of a quad-core, and we're at 3.8 GHz Turbo Boost frequencies already.
A quick note, AMD has already hit 4 GHz base clock with a Bulldozer SKU. No thunder was to be had.
 
A quick note, AMD has already hit 4 GHz base clock with a Bulldozer SKU. No thunder was to be had.
Thanks for the heads up, I totally missed that. Unfortunately the FX-4170 is only a dual-module chip, not quad-module. It sacrifices cores for clock speed, and consumes a whopping 125 Watt. That's not much of a victory over Intel. It seems to me that they're merely putting out a feeler to get a sense of the market's response before we truly enter the 4 GHz era.

A quad-core Haswell chip would offer four times the peak FP throughput, and definitely consume less. Fortunately Piledriver looks like an improvement on the power consumption front, but AMD has to put AVX2 on the roadmap sooner rather than later to keep up.
 
Thanks for the heads up, I totally missed that. Unfortunately the FX-4170 is only a dual-module chip, not quad-module. It sacrifices cores for clock speed, and consumes a whopping 125 Watt. That's not much of a victory over Intel. It seems to me that they're merely putting out a feeler to get a sense of the market's response before we truly enter the 4 GHz era.
It's not so much a feeler as much as it's a case of an architecture that's supposed to run in that range.
If they had their way, there would have been an 8 core running several hundred MHz above where the 8150 is at.
The architecture is not able to overcome its many tradeoffs in per-clock performance until it does.

A quad-core Haswell chip would offer four times the peak FP throughput, and definitely consume less. Fortunately Piledriver looks like an improvement on the power consumption front, but AMD has to put AVX2 on the roadmap sooner rather than later to keep up.
It still needs to be able to handle native AVX, much less AVX2.
I'm in a relatively sour mood with regards to AMD today, so I was going to snark that the "keep up" part was out of the question several years ago.
 
Back
Top