Appreciate the context for that comparison, but imagine that CPU for probably 50% or less the cost and with a DDR4 controller. Sounds like a valuable product to me.
Maybe somebody could give us an estimated performance per dollar based on this some previous AMD CPU releases..
The 5960x already is DDR4. Also, Intel's supposed to be releasing the Broadwell-E 6950X in a few weeks, which will have 10 cores. That being said, the 6950X is also supposed to be $1500. The real fun will be the bang/buck comparison against the 6900K, which is going to have 8 cores and be $1000. Regarding bang/buck, it's hard to say, but I'd probably say that AMD places their part at the $800 price point, which would definitely give it a decent bang/buck.
Another wrench in comparing things is that AMD's promising some interesting peripheral support, like 64 lanes of pcie, usb 3.1 and direct nvme support, which will further complicate direct comparisons until release date. Intel's hanging on with the x99 chipset, which is stuck with usb 2.0 and pcie 2 for the support lanes.
On an overclocked 2600k now, with the ageing Sandy Bridge, I'm in the market for a >4 core CPU myself.
It's a shame the AMD is nearly half a year away, I need the invoice dated June 30 at the latest for tax purposes. So if the AUD isn't too horrible I'll probably just get a 6850k, water-cooled and overclocked, with the Samsung 950 M2 drive and 32gb DDR4 ram with room for 64. For my work in Visual Studio/NCrunch/Resharper this will be a killer combo, especially for the I/O offered by 4 PCIE lanes to the SSD.
Can't wait to see Intel get some competition again though.
> Appreciate the context for that comparison, but imagine that CPU for probably 50% or less the cost and with a DDR4 controller.
Not only that, the differences between Haswell (2014) and Broadwell (2016) are fairly modest. And nobody buys the $999 version anyway -- if they can make something that competes with the i5 at the $300 price point they're in business.
The i5-6685R is $288. Several of the mobile versions are more.
Regardless, you get the point. It matters much more that they have something competitive with the Intel processors that constitute the bulk of market volume than that they can beat the crazy expensive one that doesn't.
Sure, but Intel's chips still won't be the best "value for the money" if Zen is as good as AMD claims it is.
Intel will never go toe-to-toe with AMD on pricing, at least not in the long term (1+ years), and unless it's approaching bankruptcy, which I assume isn't the case for now.
Intel's chips are the best value for the money right now because AMD totally messed up its last microarchitecture. They bet on something (high parallelization) that didn't pan out. And they suffered greatly for it. Kind of how Sony went with the Cell processor in the PS3 (although Sony did make the best of it in the end).
But Zen completely changes that equation. Zen is basically a direct and very close competitor to Intel's current microarchitecture and they should have rather similar performance.
However, It doesn't mean AMD's chips will necessarily beat Intel's chips in "peak performance". Because for one, Intel still has a better process technology, and second, AMD still had to compete with a moving target (albeit a very slowly moving target - Skylake is only like 20-25% faster in single-thread performance compared to Sandy Bridge).
But I imagine Zen is going to be within 10% or so of Intel's equivalent chips in single-thread performance, but potentially much better at multi-threaded performance if AMD actually sells twice as many Zen cores in its similar priced chips (8-core Zen vs quad-core Core i6, quad-core Zen vs dual-core Core i5, etc). On top of having more cores, AMD may also offer lower pricing points that outweighs the 10% difference in single-thread performance, but that remains to be seen (I probably wouldn't do that if I were them, as double the cores at a similar pricing point should be more than enough value for money).
Intel has the fastest processors but it doesn't have the best value for the money, in the entry level and mid level consumer market (sub-$300 range).
Concrete example: I recently built a Linux video transcoding machine to reencode H.264 BluRay movies to smaller H.264 files [1]. A $110 AMD FX-8300 is able to reencode an average BluRay movie around 47-50 fps, while you would need to spend more than twice that amount of money to match this performance with Intel (the performance of the FX-8300 at this tasks falls between a $200 Intel i5-4590 and a $250 Xeon E3-1231 v3).
There's also the power cost, depending on how much you're running that... went from an 8350 to an i7-4790K early last year, the power cost difference is pretty significant.
Even if the AMD chip runs 100 watts hotter and you're encoding constantly all year that's still only $100, small compared to the chip price difference. In datacenters where you have to think about density, cooling, etc then power efficiency gets more important but not so much for consumers except in terms of fan noise.
It has this bad reputation but in my experience it gives a very decent quality. I may not have stretched the compression ratio too. But at the age of 8TB consumer hard drives...
It gives a completely and quantifiably awful quality for storage, archival, and typical media center purposes. Transcoding at high efficiency is not what Quick Sync is for. It is for streaming, where it is quite fine. This is known.
I'm not repeating cargo cult science; I benchmarked it because I didn't believe it either (which is good! be suspicious!) and it would have been quite handy for a startup I was developing. But alas. Looks like dogshit or costs too many bits.
Be careful not to confuse "looks fine to me" with quantifiable.
and I honestly can't see the difference between the QuickSync and regular H264 samples. I am sure there are some tiny subtleties that I miss and perhaps my screen is not good enough to render the difference, but if we are talking about that level of degradation, and unless you are really a purist, the 10x performance gain is really worth it.
I think that (given a supported motherboard) being able to access ECC ram is a bigger thing... AMD on ASUS can do ECC at a price point much better than Xeon. Power usage is another issue though.
A big reason to consider AMD is that they tend to always just enable pretty much all their features for every chip rather than cherry pick what they enable for each market segment like Intel does.
yes it's about time, for example, that we had ECC ram capability in consumer-grade CPUs instead of having to pay double/triple for a XEON that doesn't even outperform (much).
I've got ECC (unregistered but still) in my "consumer grade" desktop[1] today, and have for a while; while many motherboards don't say whether they support ECC or not, some do and most AMD CPUs do as well.
Same reasons as ECC on a server, or for that matter, being able to do a lower cost home or small office server. The same rules apply on a desktop too though, just less risk of impact from corrupt data in memory for a consumer PC vs server. That said, most offices are using consumer grade CPUs and hardware too.
>Because for one, Intel still has a better process technology
One of the reasons I've had some hope for Zen, personally, is because AMD has put manufacturing and tech sharing agreements in place with Samsung ahead of the Zen launch. Intel still arguably has a process lead, but Samsung is one of the only companies in the world that has the resources to compete.
A couple of years ago, I thought Intel's process lead might be a semi-permanent advantage. Now I'm seeing some hope that may finally be coming to an end.
> Intel will never go toe-to-toe with AMD on pricing, at least not in the long term (1+ years), and unless it's approaching bankruptcy, which I assume isn't the case for now.
That's why Intel still keep their own fabs and AMD don't have fab anymore.
This is usually not true for very high end CPUs once they get older, they might even get more expensive. Here in Germany, the 5960X is still >1000 EUR for example.
Maybe somebody could give us an estimated performance per dollar based on this some previous AMD CPU releases..