[SemiAccurate] Contra-revenue comes back in a big way at Intel

[piesquared]

AMD wasnt attacking the low-end tablet market, they where targeting the 10-11.5" Windows tablets and 2-in-1.
Mullins was clearly the best x86 product, but when your competitor giving you free the design tablet and the APU then no matter how good your product is nobody going to even bother with you. Not to mention that AMD at the time was in its worst possible financial position to even try to compete in such a environment.

This is the correct history, and the sentiment of everyone at the time. The passage of time allows people to attempt to change that history, it's important to correct those attempts when they rear their ugly head. Mullins was far and away the better option for that category, and the CEO of AMD stated that the contra revenue caused significant headwind against the sale of their processors.

 

[frozentundra123456]

AMD wasnt attacking the low-end tablet market, they where targeting the 10-11.5" Windows tablets and 2-in-1.
Mullins was clearly the best x86 product, but when your competitor giving you free the design tablet and the APU then no matter how good your product is nobody going to even bother with you. Not to mention that AMD at the time was in its worst possible financial position to even try to compete in such a environment.

Well you certainly werent going to attack the high end price wise with Atom or Beema/Mullins. Anybody who would pay more than 250.00 for a device with any of those chips is crazy.

 

[flash-gordon]

From the other thread it seems Cannon Lake is a total failure. So, the contra-revenue is to sell those cpus?

 

[oak8292]

Intel's contra revenue was not really focused on AMD. Another major expenditure for Intel was x86 compatibility for Android. This was part of the urgency of selling 40 million x86 tablets. Without a base of machines the App developers would not test against Android/x86. Intel was counting Windows tablets when they fell short of the target but I have a feeling that was not their intent.

AMD as far as I know did not have a version of Android working on their x86 tablet and was dependent on Microsoft for their tablet effort. Windows netbooks and tablets on Atom and ARM proved to be less than compelling. The iPad and low powered ARM tablets had already killed the Windows netbook market.

 

[Roland00Address]

Didn't AMD fail to gain traction in the tablet market because their offering was a power hog?

AMD chips could be power hogs, but the best of the best chips from the AMD wafer were competitive with Intel from a power standpoint. Remember the best of the best chips could be run at lower voltages, lower voltages has a quadratic effect on power consumption where just by lowering your voltage by a little bit can make a massive decrease in power use both on idle cpu / gpu cores but also loaded fully active cpu / gpu cores. This is one of the reasons why die shrinks are so important for the silicon industry for die shrinks besides putting more transistors per wafer often allow you to have lower silicon voltages which in turn allow you to have better power efficiency.

• So my point with the previous paragraph is that it is clear if you cherry picked AMD tablet cpus they could have been competitive with Intel tablet cpu with power consumption while also having similar or maybe even slightly better cpu and gpu performance. The problem is you are comparing the best AMD tablet chips to the average Intel chip. It is not clear AMD could do the same thing with the rest of the cpus from the wafer. Furthermore even if AMD could make their average tablet cpu compeitive on power doing it on price and wanting to compete in this very cheap market is a whole different kettle of fish. Aka people are having half a different discussions at the same time and change one detail and everything shifts.

But no AMD chips and no Intel chips were competitive with ARM from power per watt standpoint. There were times they came close but key word was close.