Our Zen CPU will outperform Intel Broadwell-E: AMD

Makes me wonder why they fell so far behind though.

Because people are sheep and bought Intel anyway because of the brand name. R and D is expensive and Intel had deeper pockets.
 
Massive USA company vs little Canada company? But we do need to be thankful for AMD - without the competition we would most probably be way behind on current released tech.

There's also intels dirty business practices resulting in unfair competition & antitrust related law suites in the us & eu which they lost or settled.
 
They have a decent enough share of the GPU market, so it's not like they're broke.
 
IIRC, Intel really dropped the ball with their Tracecache architecture and deep instruction pipeline on their P4, and AMD hit gold with their much larger L1 cache and larger L2 cache on the Athlon. The situation reversed later when Intel removed the trace cache and deep pipeline (closer to P3 design) with their core 2 architecture, which is also the first desktop CPU to use a real 128-bit ALU, which made apps that were heavily using SSE instructions run nearly 2x faster without any code change.
 
What AMD DIDN'T say, however, is that it'll only comlete in select bemchmarks under certain circumstances which have no real world bearing.

Has everyone forgotten every prior AMD release? They always make these promises. According to AMD's slides, Bulldozer was going to be up to twice the speed of Sandy Bridge.

Pull up any of their prerelease slides - Istanbul, Bulldozer, Piledriver, Lagos, etc. They ALL promised to absolutely annihilate Intel.
 
Wasn't that always the case or what am I missing here?
Was originally supposed to be Q2/Q3 2016 but it has been known for at least a year now it won't be ready till 2017.

[XC] Oj101;18180353 said:
What AMD DIDN'T say, however, is that it'll only comlete in select bemchmarks under certain circumstances which have no real world bearing.

Has everyone forgotten every prior AMD release? They always make these promises. According to AMD's slides, Bulldozer was going to be up to twice the speed of Sandy Bridge.

Pull up any of their prerelease slides - Istanbul, Bulldozer, Piledriver, Lagos, etc. They ALL promised to absolutely annihilate Intel.
Where were these promises? AMD never made the claims that were circulating in the media and even warned against people making unrealistic claims until samples could be benchmarked.
 
My AMD FX-8150 8 core been running fine for some time. Don't think I will change it soon, even if the Zen is affordable.
 
Was originally supposed to be Q2/Q3 2016 but it has been known for at least a year now it won't be ready till 2017.


Where were these promises? AMD never made the claims that were circulating in the media and even warned against people making unrealistic claims until samples could be benchmarked.

They didn't? These slides from AMD disagree with you.

fx-launch-deck-07-r3-wm.jpg

amdfxpressdeck_15a_dh_fx57.jpg

amdfxpressdeck_22a_dh_fx57.jpg

Until tested by a neutral third party, I don't believe ANY performance claims from AMD.

The wPrime part is my favourite. A stock 3.4 GHz Core i7-2600K scores around 3m40, while an FX-8150 needs around 4.5 GHz to equal it.

Or is my favourite part "up to 56x faster"? :D

AMD has had these slides for Bulldozer, Piledriver, Interlagos, Magny Cours, Istanbul, etc. They've ALWAYS shown AMD to be faster.

Obfuscating results is something AMD does best, they even did it with the RX 480 performance where "two RX 480s edge ahead of a single GTX 1080 in a CPU bound benchmark."

How about showing us what ONE can do in a non CPU bound benchmark using a game or benchmark that has some real world application?
 
Last edited:
Our Zen CPU will outperform Intel Broadwell-E: AMD*

* In obscure benchmarks
* No promises
* high temperatures might apply

I have a R9 290x. This winter it kept me nice and warm. This summer I will be building a vent system just for my pc and drill a 200mm hole through my wall to ensure proper cooling for my card.**
 
Our Zen CPU will outperform Intel Broadwell-E: AMD*

* In obscure benchmarks
* No promises
* high temperatures might apply

I have a R9 290x. This winter it kept me nice and warm. This summer I will be building a vent system just for my pc and drill a 200mm hole through my wall to ensure proper cooling for my card.**

Try 2 x R9 280x's. I'm quiet amused that there is still paint on the wall where the PC stands. Awesome cards though.
 
Our Zen CPU will outperform Intel Broadwell-E: AMD*

* In obscure benchmarks
* No promises
* high temperatures might apply

I have a R9 290x. This winter it kept me nice and warm. This summer I will be building a vent system just for my pc and drill a 200mm hole through my wall to ensure proper cooling for my card.**

Which PC case (and cooling) do you use?
 
[XC] Oj101;18180835 said:
They didn't? These slides from AMD disagree with you.

[SNIP][/SNIP]

Until tested by a neutral third party, I don't believe ANY performance claims from AMD.

The wPrime part is my favourite. A stock 3.4 GHz Core i7-2600K scores around 3m40, while an FX-8150 needs around 4.5 GHz to equal it.

Or is my favourite part "up to 56x faster"? :D

AMD has had these slides for Bulldozer, Piledriver, Interlagos, Magny Cours, Istanbul, etc. They've ALWAYS shown AMD to be faster.

Obfuscating results is something AMD does best, they even did it with the RX 480 performance where "two RX 480s edge ahead of a single GTX 1080 in a CPU bound benchmark."

How about showing us what ONE can do in a non CPU bound benchmark using a game or benchmark that has some real world application?
So? Everyone uses custom benchmarks. That is a pretty accurate representation of the real performance. It doesn't show the 30-50% gains that 3rd parties claimed it will have before launch. AMD even warned against such claims being realistic so I would much rather trust them than 3rd party sources.

Our Zen CPU will outperform Intel Broadwell-E: AMD*

* In obscure benchmarks
* No promises
* high temperatures might apply

I have a R9 290x. This winter it kept me nice and warm. This summer I will be building a vent system just for my pc and drill a 200mm hole through my wall to ensure proper cooling for my card.**
AMD's tests show about a 40% performance increase over excavator. It was designed that way. This is not fiction.
 
So? Everyone uses custom benchmarks. That is a pretty accurate representation of the real performance. It doesn't show the 30-50% gains that 3rd parties claimed it will have before launch. AMD even warned against such claims being realistic so I would much rather trust them than 3rd party sources.


AMD's tests show about a 40% performance increase over excavator. It was designed that way. This is not fiction.

I don't know if you need to gel up your hair to reach the point I'm making. AMD themselves claimed the FX-8150 to be , and I quote, "up to 56x faster [than 2600K]." Look at the slides published by AMD - the FX-8150 is supposed to be faster than the i7-980X in eight different games as well as 3DMark11. Guess what? It isn't. The FX-8150 is supposed to be significantly faster than the 2600K in media encoding and WinRar decompression. Guess what? It isn't. AMD's benchmarks have historically not only been obscure cherry-picked results but also straight up BS.
 
[XC] Oj101;18189049 said:
I don't know if you need to gel up your hair to reach the point I'm making. AMD themselves claimed the FX-8150 to be , and I quote, "up to 56x faster [than 2600K]." Look at the slides published by AMD - the FX-8150 is supposed to be faster than the i7-980X in eight different games as well as 3DMark11. Guess what? It isn't. The FX-8150 is supposed to be significantly faster than the 2600K in media encoding and WinRar decompression. Guess what? It isn't. AMD's benchmarks have historically not only been obscure cherry-picked results but also straight up BS.
We're clearly not looking at the same things. The "up to 56x faster" does not state against what. The slides show nothing of the general hype that was made up against Bulldozer. I have seen cases at the time where it was faster with encoding and compression. There are instances where AMD is indeed significantly faster than Intel, unfortunately those are not in general use, so the benchmarks are not inaccurate. All in all my point still stands that AMD never made the claims that were circulating in the media and even warned against making them.
 
[XC] Oj101;18189049 said:
I don't know if you need to gel up your hair to reach the point I'm making. AMD themselves claimed the FX-8150 to be , and I quote, "up to 56x faster [than 2600K]." Look at the slides published by AMD - the FX-8150 is supposed to be faster than the i7-980X in eight different games as well as 3DMark11. Guess what? It isn't. The FX-8150 is supposed to be significantly faster than the 2600K in media encoding and WinRar decompression. Guess what? It isn't. AMD's benchmarks have historically not only been obscure cherry-picked results but also straight up BS.

Up to 56x times faster in certain floating point calculations, not in everything. (they'll be but one little piece in a benchmark)
And benchmarks comparing: http://www.anandtech.com/bench/product/434?vs=287
Stuff like blender, the AMD processor was faster. In lots of games the AMD processor lost due to terrible single thread performance.

Sadly those games listed are not on the anandtech website: http://www.anandtech.com/bench/product/434?vs=142
 
Top
Sign up to the MyBroadband newsletter
X