Here's one result I came up with Googling around for the register reuse cache: A compile-time managed multi-level register file hierarchy, by Nvidia's Bill Dally. It's only a citation, unfortunately, but it describes exactly the kind of configuration that the Maxas guy uncovered, and it was published in 2012, which should be around the time Maxwell was in its architecture definition stage.
The paper claims a reduced register file energy usage by up to 54%.
For no credible reason (you have yet to provide one anyway).
Indeed, it is not magic; it is quality engineering.
You have provided zero actual evidence that this is the case.
You have also provided zero evidence that Maxwell is less efficient for compute workloads.
Again, you have not yet provided any evidence that this is the case. Furmark power consumption is not evidence. If you don't understand why it is not evidence, I am not sure myself or anyone else here can help you. Indeed, the only actual compute benchmarks in this thread, that I am aware of, show Maxwell to be considerably more efficient than Kepler in compute workloads, sometimes over 100%+.extrajudicial said:Please explain why compute loads make maxwell only marginally (>10%) more efficient during compute
The only person here I see ignoring evidence is you.extrajudicial said:I'm appalled to see people putting their own personal favor and letting it trump evidence and common sense.
The Tegra K1 in the Nexus 9 is based on Kepler, not Maxwell.extrajudicial said:If maxwell is so power efficient why can't the nexus 9 manage 10 hours battery life?
And again you ignore my post completely.If it's not power gating as anandtech said then what? You criticize my explanation and provide none of your own besides saying "it's in the architecture"
Because Nexus 9 uses Kepler, not Maxwell?If maxwell is so power efficient why can't the nexus 9 manage 10 hours battery life? That's hardly impressive.
If maxwell is so power efficient why can't the nexus 9 manage 10 hours battery life? That's hardly impressive.
Thanks! Very interesting paper.The link you provided just times out.
Here is a link to Nvidia's paper on the subject:
https://research.nvidia.com/publication/compile-time-managed-multi-level-register-file-hierarchy
Yes. K1 = Kepler, 1 SMX. There is no Maxwell based mobile SOC yet.The Tegra K1 in the Nexus 9 is based on Kepler, not Maxwell.
Please explain why compute loads make maxwell only marginally (>10%) more efficient during compute. If it's not power gating as anandtech said then what?
It's more than twice faster than BT on Geekbench: http://browser.primatelabs.com/geekbench3/compare/1142302?baseline=620722How do you guys think this chip fares against Baytrail on the CPU front?
There are indeed too many benchmarks that are broken due to dedicated instructions. But having studied Geekbench code, it's not that bad as long as you don't forget it's a smallish benchmark (though not really a micro-benchmark such as dhrystone or coremark).I wonder if that thing is benchmarking the hardware encryption accelerator?
What is the alternative? SPEC has been mostly broken by compilers and autopar, and anyway it can't be run on most smaller devices (it requires 2GB). Javascript benchmarks are somewhat interesting but results vary a lot depending on the browser. So what non micro-benchmark do you propose that is available on many platforms?At that point in time I think I'm tired of micro-benchmarks, even if they "improved".
Year 2000 : dhrystones, raw MIPS and the like are dead, let's switch to application benchmarks instead!
Year 2014 : here's a multi-platform collection of ~40 synthetic benchmarks with autogenerated web pages.
It leaves me wondering if the bench is about what fits in L1, or L2.
There are now many Geekbench 3 AArch32 results for Nexus 9:
http://browser.primatelabs.com/geekbench3/search?q=htc+nexus+9
The single-core score is as high as ~ 2000 while the dual-core score is as high as ~ 3300-3500! That is incredibly good considering these are 32-bit results.