2016 Macbook Pro 15" CPU is Slower than 2015 [Everything Known So Far and Questions]

[Kris.K]

I am creating this so we can have one place that summarizes all the information we know about why a MacBook Pro running the 2016 2.9Ghz i7-6920HQ would be slower than the 2015 2.5Ghz i7-4920HQ. I will keep this thread updated with information as it comes in.

Theories So Far:
We need to debunk or prove these. Your help is appreciated

1. 2015 CPU Turbos upto 4Ghz where as 2016 only 3.8Ghz. - littlepud
2. The Haswell chip in the 2015 Macbook Pro is a Crystalwell with Iris Pro, which effectively has an additional 128 MB L4 cache over the 2016 Macbook Pro. - littlepud
   
3. On Ivy bridge even though official Intel documentation says the turbo max speed is only for one core, terraphantm experienced that all 4 of his boosted. Was this the same for Haswell/Broadwell and has it changed for Skylake? - terraphantm
4. A combination of things:
   1- Apple wanted to preserve battery life so they made some compromises at the expense of performance. They have upped the Mhz on the RAM even though Intel says it only officially supports upto 1866Mhz on DDR3, Apple has gone 2133mhz but increased the latency and subsequently slowed down the memory for battery life which has effectively created a bottleneck for the machine.
   
   Reason to think this:
   2015 memory latency GeekBenchmark: 12.5 Moperations/sec
   2016 memory latency GeekBenchmark: 7.64 Moperations/sec
   
   2- The 2015 CPU having a higher turbo speed (4Ghz vs 3.8Ghz + other differences - Intel and littlepud)
   3- Using DDR3 instead of DDR4 for the 2016 CPU. They chose DDR3 because LPDDR4 is not supported by Skylake (Reddit user) (Skylake works faster on DDR4 than DDR3 - Corsair)
   - Me and My friend/coworker Peter (CTO at ServersAustralia)
5. Being held back because Mac OS is far behind by still using OpenCL 1.2 and OpenGL 4.2 when there has been many developments in the API - SoyCapitanSoyCapitan
   
6. Please suggest or link others in the comments below.

Questions that need Answering:

1. What latency is the 2015 memory running? [Need CPUz screenshot]
2. What latency is the 2016 memory running? [Need CPUz screenshot]
3. Is there any kind of setting or way to allow better performance [timing?] when plugged in? [If the theory holds, since the reasoning for LP was battery power, so why limit when connected to power?]
4. Suggest other questions and i'll add up here as long as its reasonable.

Additional Benchmarks we need run: [For i7-4920HQ 2015 and i7-6920HQ 2016]

1. Memory, Memtest? - Leman
2. Compubench- Leman
3. Luxmark- Leman
4. 3dmark- Leman
5. unigine benchmarks - Leman
   
6. What are the CPU speeds on the 2016 when being benchmarked? [Don't know how to do this]
7. Suggest any other benchmarks that could be helpful in figuring this out.

Interesting Links:

Thank you NAYo2002 for the find.

Good Review of Macbook Pro 15" by Verge

Interesting Benchmarks:
Please suggest or post good examples and whichever is the best will get included.

Additional information:

Geekbench 4 Benchmarks

Single-CPU Scores
MacBook Pro (15-inch Retina Mid 2014) - 4538
MacBook Pro (15-inch Retina Mid 2015) - 4487
MacBook Pro (15-inch Late 2016) - 4508

Multi-CPU Scores
MacBook Pro (15-inch Retina Mid 2014) - 14978
MacBook Pro (15-inch Retina Mid 2015) - 14744
MacBook Pro (15-inch Late 2016) - 14058

Links:
2014 https://browser.primatelabs.com/v4/cpu/1042571
2015 https://browser.primatelabs.com/v4/cpu/942360
2016 https://browser.primatelabs.com/v4/cpu/1047696

[doublepost=1479387619][/doublepost]Reserving this post for my own personal opinion so the first post stays neutral.

Hmm.. @Moderators are you able to unmerge my second post?