[wccftech] AMD Kaveri APU A10-7850k Engineering Sample Spotted and Benchmarked

[MangoX]

http://wccftech.com/amd-kaveri-apu-...spotted-benchmarked-cpuz-gpuz-screens-leaked/

Seems fake, IMO.

 

[inf64]

I was the one to find the results via google and of course wccftech rips off everything. Originally I posted it on my blog, SA forums and here.
And no, it's not "fake".

 

[Ventanni]

Looks promising. Won't be long before the real reviews make headline

 

[ShadowVVL]

$173-$189 is rather close to the i5 4650k or non-k and the igpu will still be held back by ddr3.
The i5 is just a better buy IMO.

 

[ShintaiDK]

$173-$189 is rather close to the i5 4650k or non-k and the igpu will still be held back by ddr3.
The i5 is just a better buy IMO.

I dont think it will cost more than 139$ at release.

 

[inf64]

Yeah, Richland will go down ~10% in price and Kaveri will sit on top of it (~15% more, my guess). There will be some initial price gouging like always.

 

[Homeles]

I was the one to find the results via google and of course wccftech rips off everything. Originally I posted it on my blog, SA forums and here.
And no, it's not "fake".

It's their job to post news and distribute it to the masses, bro.

 

[inf64]

They can post credits for the finds, bro . This way it seems like they found it which is not true (like in 99% of their "news articles").

 

[ShintaiDK]

Wccftech is a terrible site. Stealing, making up stuff etc.

 

[JDG1980]

The MP ratio of 3.53 is a tipoff that that screenshot is fake. Are we supposed to believe that implementing separate decode units for each core had no positive effects on multi-thread scaling? Even when that was one of the primary goals of Steamroller? 3.53 is essentially the same scaling as currently seen on existing Piledriver/Richland parts.

Presumably they took "screenshots" by pointing a camera at the screen so it would be harder to spot the Photoshop artifacts.

 

[inf64]

The MP ratio of 3.53 is a tipoff that that screenshot is fake. Are we supposed to believe that implementing separate decode units for each core had no positive effects on multi-thread scaling? Even when that was one of the primary goals of Steamroller? 3.53 is essentially the same scaling as currently seen on existing Piledriver/Richland parts.

Presumably they took "screenshots" by pointing a camera at the screen so it would be harder to spot the Photoshop artifacts.

No it's not... The ST score was gotten at 7850K's stock ST turbo of 4Ghz, while in MT portion it runs at base clock of 3.7Ghz. The scaling corrected for this difference is 3.82x which is 7% higher than PD. This is perfectly in line with geekbench scores for Kaveri ES (which is shows 7-8% improvement in MT scaling Vs Piledriver).

 

[Homeles]

Wccftech is a terrible site. Stealing, making up stuff etc.

Not to mention some authors that could seriously use an eyebrow trim.

 

[PPB]

So lets see, it actually has a higher MP scaling because of the absent turbo in MT tests.

Anyone's got a 6800K turbo OFF run on CB 15 to see MT scaling? Should be in the Area of 3.2x

 

[OCGuy]

I was the one to find the results via google and of course wccftech rips off everything. Originally I posted it on my blog, SA forums and here.
And no, it's not "fake".

So are you breaking NDA or are you stealing the results "via Google"?

You seem very confident in the numbers for something that you didn't do yourself.

 

[inf64]

So are you breaking NDA or are you stealing the results "via Google"?

You seem very confident in the numbers for something that you didn't do yourself.

I didn't post the results so I don't brake any NDA( nor I have any NDA signed). I just found them 1st via google, after the source website posted them online. This must be very hard concept to grasp I guess.
And yes, I know the result is correct for 7850K.

 

[OCGuy]

I didn't post the results so I don't brake any NDA( nor I have any NDA signed). I just found them 1st via google, after the source website posted them online. This must be very hard concept to grasp I guess.
And yes, I know the result is correct for 7850K.

I don't think you understood what I was trying to say.

You criticized wccftech for just re-posting unoriginal material, while at the same time admitting that is all you did as well.


Although I agree that wccftech is complete garbage, I thought I would point out the irony.

 

[inf64]

I credited the source. Also I wouldn't post the link if I knew it was "fake" (which is not, hence I posted it). The Japanese website obviously bought the chip or borrowed it from some shop. These people (wccftech) just stole the news piece from others and made it appear like they have found it in order to generate traffic on their "website". They also posted all sorts of fake "news" they picked up in the past, last of which was "new Phenom III" garbage "news". This just tells you they don't care about posting genuine information for the readers, they just want the clicks to their website.

 

[BallaTheFeared]

Not to mention some authors that could seriously use an eyebrow trim.

Comedy Gold! :thumbsup:

 

[Shivansps]

That is just FAKE, thats a screen of the Windows 8 "Microsoft Basic Render Driver".

 

[EightySix Four]

That is just FAKE, thats a screen of the Windows 8 "Microsoft Basic Render Driver".

Probably because they don't have the right driver for the graphics due to it being unreleased...