AMD New Horizon

[Vode]

5 minutes ago, Swatson said:

It's not skewed if intel's cpu got the speed boost too, which it did... Both companies cpu's performed the same, which means the 6900k was faster than expected too. DO YOU THINK AMD WANTS TO MAKE THE 6900k LOOK FASTER THAN IT IS?

Sorry for the caps but I think people aren't getting it.

 

Of course 3rd party benchmarks trump all, but it's really stupid to just assume amd skewed something when the results appear to be quite equitable.

That's true but it's not a scientific, repeatable or verifiable result.

 

People want transparency and repeatable, replicable results. You don't tend to trust stuff if it doesn't match those criteria.

[S w a t s o n]

3 minutes ago, Vode said:

That's true but it's not a scientific, repeatable or verifiable result.

 

People want transparency and repeatable, replicable results. You don't tend to trust stuff if it doesn't match those criteria.

Absolutely, like I said 3rd party benchmarks trump all, but it's completely unfair to assume AMD skewed anything. They could have quite possibly but from the results we saw there is no reason to assume they did.

 

Waiting for 3rd party benchmarks and implying a company skewed results are two different things.

[Carclis]

13 minutes ago, Swatson said:

It's not skewed if intel's cpu got the speed boost too, which it did... Both companies cpu's performed the same, which means the 6900k was faster than expected too. DO YOU THINK AMD WANTS TO MAKE THE 6900k LOOK FASTER THAN IT IS?

Sorry for the caps but I think people aren't getting it.

 

Of course 3rd party benchmarks trump all, but it's really stupid to just assume amd skewed something when the results appear to be quite equitable.

I understand what you're saying but it makes absolutely no sense. Why would you change the results of your own benchmarks to make them impossible to replicate if it doesn't even change the result? They even supplied the blender files as proof and so that people could match the results and nobody can.

[Vode]

1 hour ago, Swatson said:

Absolutely, like I said 3rd party benchmarks trump all, but it's completely unfair to assume AMD skewed anything. They could have quite possibly but from the results we saw there is no reason to assume they did.

 

Waiting for 3rd party benchmarks and implying a company skewed results are two different things.

People are just pissed because AMD didn't disclose full details and are getting emotional and overreacting. 

 

While I don't agree that the results are skewed it's very expected behaviour for people to assume they are.

[S w a t s o n]

7 minutes ago, Carclis said:

I understand what you're saying but it makes absolutely no sense. Why would you change the results of your own benchmarks to make them impossible to replicate if it doesn't even change the result? They even supplied the blender files as proof and so that people could match the results and nobody can.

They supplied the files but not the settings used, as far as I know.
Again, why would they allow the stock 6900k perform better than it should if they were skewing results? It's not impossible to replicate, you just dont have all the details yet.

 

I should mention that to truly have a scientific method for replication you need to know every component in the 6900k system and probably the windows settings/state. Were they connected to a network, how were they captured? Even the little shit like that takes cpu cycles.

[A Damn Crab!]

From what i saw the Ryzen chip will be price competitive to the 6700K not X99.

[anotherriddle]

It seems like AMD RYzen could be more efficient than Intel (because of lower TDP). Are there any numbers on idle power consumption? I am planing a build, but am waiting for what AMD pulls out of their hat. If they can match (or God forbid beat) idle power consumption of Intel I am sold.

[Citadelen]

8 minutes ago, anotherriddle said:

-snip-

AMD will be integrating the technologies they use in their APUs into Zen, so at Idle I believe you could see power draw of even less than a watt.

[anotherriddle]

9 minutes ago, Citadelen said:

AMD will be integrating the technologies they use in their APUs into Zen, so at Idle I believe you could see power draw of even less than a watt.

Thanks, I just droped my coffee.

Can you back that up somehow, do you have a link? Honestly, with that core count, below 3 watt idle would be impressive. I need to look up on AMD APUs now.

[Space Reptile]

23 minutes ago, anotherriddle said:

Thanks, I just droped my coffee.

Can you back that up somehow, do you have a link? Honestly, with that core count, below 3 watt idle would be impressive. I need to look up on AMD APUs now.

my 2600 has under 5w in idle according to coretemp and cpuz*

 

 

*me thinks it was cpu z

[Dabombinable]

24 minutes ago, anotherriddle said:

Thanks, I just droped my coffee.

Can you back that up somehow, do you have a link? Honestly, with that core count, below 3 watt idle would be impressive. I need to look up on AMD APUs now.

Either way, what ever the configuration of the Zen APU ends up like, it needs to be far better than the POS A8 4555M that I was using as an "upgrade" from my Phenom II N970+Mobility Radeon HD5650 (the old laptop was more powerful in every single way, except file compression).

All I can say is, uad core my ass.

[anotherriddle]

13 minutes ago, Space Reptile said:

my 2600 has under 5w in idle according to coretemp and cpuz*

 

 

*me thinks it was cpu z

yes, but that is half the cores and I don't really trust software on estimating power consumption. From my experience the real value is often very different from what software says.

[Prysin]

3 hours ago, Space Reptile said:

but the demos look nice  

iv'e seen 3DCG porn that looks nice. Doesnt make it any more real

[Prysin]

50 minutes ago, Dabombinable said:

Either way, what ever the configuration of the Zen APU ends up like, it needs to be far better than the POS A8 4555M that I was using as an "upgrade" from my Phenom II N970+Mobility Radeon HD5650 (the old laptop was more powerful in every single way, except file compression).

All I can say is, uad core my ass.

Bristol Ridge alone is way faster then your puny N970. If ZEN is +40% ish ontop of Bristol, then you're looking at the 450-550 point range for a "pure" quad core, and i'd guess around 350-430 for a 2C/4T "i3" SKU

 

Dual channel FX-8800P, which is Carrizo, is already up at 277 points in Cinebench R15 Multi Thread, according to Anandtech. Single channel configs however is much less potent.

http://www.anandtech.com/show/10000/who-controls-user-experience-amd-carrizo-thoroughly-tested/13

 

This means that Bristol Ridge which is even slightly higher clocked should hit somewhere near 300 points... Meaning ZEN hould hit in the 450ish range for a laptop SKU

[Dabombinable]

6 minutes ago, Prysin said:

Bristol Ridge alone is way faster then your puny N970. If ZEN is +40% ish ontop of Bristol, then you're looking at the 450-550 point range for a "pure" quad core, and i'd guess around 350-430 for a 2C/4T "i3" SKU

 

Dual channel FX-8800P is already up at 277 points in Cinebench R15 Multi Thread, according to Anandtech. Single channel configs however is much less potent.

http://www.anandtech.com/show/10000/who-controls-user-experience-amd-carrizo-thoroughly-tested/13

I suppose that it at least the IPC is similar to that of Hapertown (2007). If AMD manages to catch up after a decade of releasing CPU with inferior architectures, its going to be one hell of a feat.

[Prysin]

2 hours ago, Carclis said:

Well it appears to be skewed and is reasonable to assume so until AMD's results can be replicated.

we knew already a year ago that if everything lined up with best case scenarios, ZEN would hit just shy of broadwell level performance. We could extrapolate that from AMDs claims of 40% IPC over Carrizo (Excavator), which was already 15-20% faster then Steamroller. And by calculating that, the best case would be right smack inbetween Broadwell and Haswell. 

Last night, AMD claimed to have surpassed their 40% goal, this means Broadwell performance IS within reach. And thus, these tests DO make sense. 

 

HOWEVER, i still think they going to drop the ball by pricing their products way too high. Either AMDs own products, of the motherboard manufacturers will

[Prysin]

2 minutes ago, Dabombinable said:

I suppose that it at least the IPC is similar to that of Hapertown (2007). If AMD manages to catch up after a decade of releasing CPU with inferior architectures, its going to be one hell of a feat.

merely 1.5 decades "behind" at worst. 10-15 years ish since Athlon 64.

[Paragon_X]

The SenseMI stuff is gonna be interesting if the CPU can cooperate with the chipset on the mobo to monitor power delivery mosfets temps and all those systems. please AMD dont pull a stupid move everything seem really good atm.

[Dabombinable]

1 minute ago, Prysin said:

merely 1.5 decades "behind" at worst. 10-15 years ish since Athlon 64.

I still wish that the days of the Slot A Athlons would come back. Intel was shitting bricks back then and rushed things a lot (which lead to my Abit VP6 supporting 2x Pentium III that got withdrawn due to severe stability issues-if I only had 2x Pentium III 1333EB-1.8V would have had them stable, with OC headroom at 1.95V)

[Dabombinable]

1 minute ago, Paragon_X said:

The SenseMI stuff is gonna be interesting if the CPU can cooperate with the chipset on the mobo to monitor power delivery mosfets temps and all those systems. please AMD dont pull a stupid move everything seem really good atm.

A stupid move like not having a proper temperature sensor like they did with the Phenom II?

[Phentos]

9 minutes ago, Prysin said:

HOWEVER, i still think they going to drop the ball by pricing their products way too high. Either AMDs own products, of the motherboard manufacturers will

If AMD does this, they will deserve what would be coming for them. Ryzen is their last shot and they know it. They absolutely have to be smart and price the chip well, and that means undercutting Intel substantially. 

 

If the motherboard manufacturers overprice the motherboards, well...

[Misanthrope]

2 hours ago, Swatson said:

Absolutely, like I said 3rd party benchmarks trump all, but it's completely unfair to assume AMD skewed anything. They could have quite possibly but from the results we saw there is no reason to assume they did.

 

Waiting for 3rd party benchmarks and implying a company skewed results are two different things.

Not completely: they've done so in the past.

[Prysin]

10 minutes ago, Misanthrope said:

Not completely: they've done so in the past.

And intel havent?

 

Pfft. Believing in any companies in-house benchmarks is like religion. You can believe all you want, but one day someone will come along and prove how wrong you are

[Misanthrope]

Just now, Prysin said:

And intel havent?

No new products announced by intel we can talk about them when we're discussing them. This "everyone does it" mentality doesn't fucking helps.

[Red Hardware]

1 hour ago, Phentos said:

If AMD does this, they will deserve what would be coming for them. Ryzen is their last shot and they know it. They absolutely have to be smart and price the chip well, and that means undercutting Intel substantially. 

 

If the motherboard manufacturers overprice the motherboards, well...

Hi

Motherboard manufactures won't do that , they want to sell their boards so lower is better and there is no reason for them to do that , so i think we'll see a good pricing on chips and boards .