• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

*** AMD "Zen" thread (inc AM4/APU discussion) ***

The bandwidth thing has been banded about for a long time. Lets stick to Zen progress rather than digging up old tittle tattle.

I dont think anyone realistically thinks anything groundbreaking is going to happen from AMD here. I am interested for various reasons but not expecting it to be trailblazing in any capacity or fantasising it can be competing with intels high end...
 
Th0nt said:
The bandwidth thing has been banded about for a long time. Lets stick to Zen progress rather than digging up old tittle tattle.

I dont think anyone realistically thinks anything groundbreaking is going to happen from AMD here. I am interested for various reasons but not expecting it to be trailblazing in any capacity or fantasising it can be competing with intels high end...
Click to expand...

I'm not sure what you're trying to get at here?

The bandwidth is absolutely an issue. They could do with at least doubling it if they want to make decent gains past ~512 shaders at ~900 MHz.

And this is indeed related to Zen. Zen could make very large gains in the CPU department, but if the GPU side doesn't improve much it'll be disappointing.

If AMD sort the bandwidth issue, they could very well create a 4 core, 1000 Shader, £150 APU that'll do 1080p60 gaming at medium/high settings consistently. That would be fantastic for a huge market.
 
http://wccftech.com/new-amd-zen-blender-benchmarks/

DebPzPp.png

pPLQYa3.png

The same blender benchmark back in Aug showed Summit Ridge had same performance as 2.8GHz 10C/20T Xeon E5 2680 V2 and 2.6GHz 8C/16T Xeon E5 2650 V2 behind it. I googled that Xeon CPUs, accorded to Intel ARK the CPUs was launched back in Q3 2012 and based on 22nm process so it is Ivy Bridge CPUs.

If 8C/16T Xeon E5 2650 V2 overclocked to 3GHz then it would matched 3GHz Summit Ridge performance so Zen with 40% improved IPC basically have same performance as Ivy Bridge.
 
AthlonXP1800 said:
http://wccftech.com/new-amd-zen-blender-benchmarks/

DebPzPp.png

pPLQYa3.png

The same blender benchmark back in Aug showed Summit Ridge had same performance as 2.8GHz 10C/20T Xeon E5 2680 V2 and 2.6GHz 8C/16T Xeon E5 2650 V2 behind it. I googled that Xeon CPUs, accorded to Intel ARK the CPUs was launched back in Q3 2012 and based on 22nm process so it is Ivy Bridge CPUs.

If 8C/16T Xeon E5 2650 V2 overclocked to 3GHz then it would matched 3GHz Summit Ridge performance so Zen with 40% improved IPC basically have same performance as Ivy Bridge.
Click to expand...

What about that render test they showed with 3 GHz 8-core Zen, vs the i7 6900k clocked at 3 GHz?

And Zen beat it by a tiny margin.
 
AllBodies said:
What about that render test they showed with 3 GHz 8-core Zen, vs the i7 6900k clocked at 3 GHz?

And Zen beat it by a tiny margin.
Click to expand...

While I'm very hopeful that was actually what happened, I do remember when AMD compared an FX8150 to an i7 980 and in Cinebench and the FX8150 won. And it turned out it wasn't an i7 980 at all, it was an i5 2500K.

But we'll see. At the moment performance estimates are all over the place :p
 
AthlonXP1800 said:
http://wccftech.com/new-amd-zen-blender-benchmarks/

DebPzPp.png

pPLQYa3.png

The same blender benchmark back in Aug showed Summit Ridge had same performance as 2.8GHz 10C/20T Xeon E5 2680 V2 and 2.6GHz 8C/16T Xeon E5 2650 V2 behind it. I googled that Xeon CPUs, accorded to Intel ARK the CPUs was launched back in Q3 2012 and based on 22nm process so it is Ivy Bridge CPUs.

If 8C/16T Xeon E5 2650 V2 overclocked to 3GHz then it would matched 3GHz Summit Ridge performance so Zen with 40% improved IPC basically have same performance as Ivy Bridge.
Click to expand...

Nope... the benchmark back in Aug you speak of was against a 3GHz i7 6900

Thats a slight margin faster than Broadwell, per thread per clock, Apples for Apples

The next faster CPU on that chart is a 10 core 20 thread Ivy at 2.8Ghz, vs the 8 core 16 thread Zen, so the faster Intel chip on that chart has 2 extra core / 4 extra threads, Apples for Oranges.

The 40% IPC referred to is vs Excavator, which is over 20% faster than Piledriver which is what you are referring to, Zen is actually over 60% faster than Piledriver.
 
Last edited:
AMD released benchmarks showing Bulldozer faster than Intel & also benchmarks showing Fiji faster than NV. Stop clinging to official shillmarks.

In the video they released of Zen "playing" Doom, it was a composite with pre-recorded footage spliced into the monitor frame. You can't get an accurate picture of a product by listening to the maker.
 
Last edited:
Martini1991 said:
Yes they did (with bulldozer).
They literally released a video showing their FX8150 besting a gulf town hex core in cinebench.
The video was later removed and re-added , the score for the Intel the same, the score for the AMD the same, but it was suddenly an i5
Click to expand...

Link to video?
 
Meh,AMD can delay as much as they can,but people will just not bother and buy an Intel system instead.

They keep doing it with their CPUs and their motherboards. There is zero excuse for any delay of BR by AMD. They are probably clinging to the hope,they can shift their few remaining Kaveri chips at RRP,whilst not realising the damage it is doing to them with an EOL socket,ancient CPUs and an old chipset. They seem to be at times doing more damage to their brand than their competitors! :(

AM4 should have been out by now with Bristol Ridge,yet if they cannot get that out in 2016,then how can anyone have any faith Zen won't be delayed again?

Its a modified version of a CPU launched LAST year on an ancient 28NM process.

They did it with BD - if it had launched when it was meant to(just before SB),it might have looked a tad better.

Llano launched with severely reduced clockspeeds and Kaveri launched late with less performance than expected.

The problem is that whilst they delay,Intel like clockwork releases their new chips.

So instead of the chips AMD says they are targetting,they end up sadly being compared to the newer Intel chips.

BR,might have some decent graphics improvements,but instead of Skylake they are going to be compared to Kaby Lake chips which have greater CPU and IGP performance.

Launching for OEMs first is all fine and dandy,but they should have actually had reviews of the chips out already,as these would have placed it in a better place against Skylake,not,Kaby Lake.

AMD need to understand,Intel is improving performance too - and even 10% more,is 10% AMD can't afford to give Intel in product comparisons.
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom