rishiswaz :
"The Nexus 9 is the first Android tablet to have an Nvidia Denver CPU that has very high single-threaded performance. However, it doesn't have a very high multi-threaded performance due to the fact that it's only a dual-core CPU."
What are you basing these claims on? If it's your own speculation then you know pretty damn little about the Denver K1, or in SoC's in general...
You do understand that the Denver K1 employs a 7 way superscalar, higher core frequency, and an L1 cache that blows the Snapdragon 805's clear out of the water. Early benchmarks are showing this chip beating out the Note 4, iPhone 6 +, S5 , pretty much everything on the market to date.
Honestly why make baseless assertions like that?
Your support is rather baseless as well unless you have benchmarks to prove otherwise. He is going off of what is known, it is a dual core 64 bit SoC with a Kepler GPU. We really don't know how Denver is going to be until there are hard numbers behind it. Unless you have proof besides Nvidia's spec sheet you have no argument either.
The iPhone 6 outperforms the Galaxy S 5 in a lot of benchmarks, a dual core 1.3 Ghz A8 vs a quad core 2.4GHz Qualcomm Snapdragon 801. Denver may well blow everything out of the water but it is still a big question mark.
You just gave an example (iphone6) of WHY the author should have never said what he said. What we know is quad doesn't mean much (yet?) on mobile as Apple usually proves. Also as Anthony said there are benchmarks out already for Denver. Not to mention as extremetech says:
http://www.extremetech.com/computing/187862-nvidia-details-64-bit-denver-tegra-k1-claims-haswell-class-performance-for-first-64-bit-android-chip
"in reality a dual-core CPU should be more than capable of handling most workloads (most apps and games are still bottlenecked by single-threaded performance)."
Considering many games don't efficiently use a quad on even PC's yet (better perf at times, but not scoring 2x a dual regularly), I don't think dual is a problem on mobile yet...LOL.
http://hothardware.com/News/Leaked-Nexus-9-Benchmark-Results-Show-DenverBased-64Bit-Tegra-K1-Performing-Very-Well/
3166 for Nexus 9 multicore, 1903 single core.
iphone 6: geekbench single core:1624 muti core:2911
3-core iPad Air 2: Geekbench single core - 1812, Multi - 4477
http://www.redmondpie.com/ipad-air-2-features-3-cores-2gb-ram-55-faster-than-iphone-6-6-plus-68-than-ipad-air/
One thing not to forget here though, is this is 28nm Denver vs. 20nm Tricore apple for ipad2air. I can't wait for NV's shrink scores
The A8 is ~89mm^2 (a8x obviously bigger with 3rd core etc), K1 is 123mm^2 IIRC, so a shrink will probably come in under 89 unless they up gpu or cpu (meaning change gpu out for maxwell at 20nm, no idea if K1 will be shrunk or not). I can't remember if there is a difference in size of K1 32vs.64, I don't think Denver version has been sized yet, so don't quote me. Either way K1 is impressive here toppling A8 20nm in iphone6. Granted you'd have to cut some speed to match phone's size, but it's still going to probably match apple and this is PRE-SHRINK of K1.
http://browser.primatelabs.com/android-benchmarks
Galaxy Note 4 Single core 1161 (LOL, note this is better than K1 A15 at 1079, but also note S805 loses to this), multicore for note4 8 core is 3914. But I'd say while this is a winner here, how many apps effectively use 8 cores?
Also note the Note4 with S805 scores 3169 (K1 32 scores 3211 here). Again, I suspect there won't be many places a Denver will be hindered for being dual in a usage scenario.
http://www.engadget.com/2014/11/03/google-nexus-9-review/
Some odd scores in there, so I'm thinking Lollipop or something here needs a bit more polish maybe. CF/Quadrant/Sunspider here sucked vs. Shield, yet we know geekbench is virtual tie in multicore. We'll find out more as we get more reviews. Also I thought Denver was supposed to be 2.5Ghz, but this runs 2.3. Maybe they had to take a cut to ensure enough chips for an early/large release? We'll see as more Denver based devices come out I guess.