The Secret of Denver: Binary Translation & Code Optimization

As we alluded to earlier, NVIDIA’s decision to forgo a traditional out-of-order design for Denver means that much of Denver’s potential is contained in its software rather than its hardware. The underlying chip itself, though by no means simple, is at its core a very large in-order processor. So it falls to the software stack to make Denver sing.

Accomplishing this task is NVIDIA’s dynamic code optimizer (DCO). The purpose of the DCO is to accomplish two tasks: to translate ARM code to Denver’s native format, and to optimize this code to make it run better on Denver. With no out-of-order hardware on Denver, it is the DCO’s task to find instruction level parallelism within a thread to fill Denver’s many execution units, and to reorder instructions around potential stalls, something that is no simple task.

Starting first with the binary translation aspects of DCO, the binary translator is not used for all code. All code goes through the ARM decoder units at least once before, and only after Denver realizes it has run the same code segments enough times does that code get kicked to the translator. Running code translation and optimization is itself a software task, and as a result this task requires a certain amount of real time, CPU time, and power. This means that it only makes sense to send code out for translation and optimization if it’s recurring, even if taking the ARM decoder path fails to exploit much in the way of Denver’s capabilities.

This sets up some very clear best and worst case scenarios for Denver. In the best case scenario Denver is entirely running code that has already been through the DCO, meaning it’s being fed the best code possible and isn’t having to run suboptimal code from the ARM decoder or spending resources invoking the optimizer. On the other hand then, the worst case scenario for Denver is whenever code doesn’t recur. Non-recurring code means that the optimizer is never getting used because that code is never seen again, and invoking the DCO would be pointless as the benefits of optimizing the code are outweighed by the costs of that optimization.

Assuming that a code segment recurs enough to justify translation, it is then kicked over to the DCO to receive translation and optimization. Because this itself is a software process, the DCO is a critical component due to both the code it generates and the code it itself is built from. The DCO needs to be highly tuned so that Denver isn’t spending more resources than it needs to in order to run the DCO, and it needs to produce highly optimal code for Denver to ensure the chip achieves maximum performance. This becomes a very interesting balancing act for NVIDIA, as a longer examination of code segments could potentially produce even better code, but it would increase the costs of running the DCO.

In the optimization step NVIDIA undertakes a number of actions to improve code performance. This includes out-of-order optimizations such as instruction and load/store reordering, along register renaming. However the DCO also behaves as a traditional compiler would, undertaking actions such as unrolling loops and eliminating redundant/dead code that never gets executed. For NVIDIA this optimization step is the most critical aspect of Denver, as its performance will live and die by the DCO.


Denver's optimization cache: optimized code can call other optimized code for even better performance

Once code leaves the DCO, it is then stored for future use in an area NVIDIA calls the optimization cache. The cache is a 128MB segment of main memory reserved to hold these translated and optimized code segments for future reuse, with Denver banking on its ability to reuse code to achieve its peak performance. The presence of the optimization cache does mean that Denver suffers a slight memory capacity penalty compared to other SoCs, which in the case of the N9 means that 1/16th (6%) of the N9’s memory is reserved for the cache. Meanwhile, also resident here is the DCO code itself, which is shipped and stored as already-optimized code so that it can achieve its full performance right off the bat.

Overall the DCO ends up being interesting for a number of reasons, not the least of which are the tradeoffs are made by its inclusion. The DCO instruction window is larger than any comparable OoOE engine, meaning NVIDIA can look at larger code blocks than hardware OoOE reorder engines and potentially extract even better ILP and other optimizations from the code. On the other hand the DCO can only work on code in advance, denying it the ability to see and work on code in real-time as it’s executing like a hardware out-of-order implementation. In such cases, even with a smaller window to work with a hardware OoOE implementation could produce better results, particularly in avoiding memory stalls.

As Denver lives and dies by its optimizer, it puts NVIDIA in an interesting position once again owing to their GPU heritage. Much of the above is true for GPUs as well as it is Denver, and while it’s by no means a perfect overlap it does mean that NVIDIA comes into this with a great deal of experience in optimizing code for an in-order processor. NVIDIA faces a major uphill battle here – hardware OoOE has proven itself reliable time and time again, especially compared to projects banking on superior compilers – so having that compiler background is incredibly important for NVIDIA.

In the meantime because NVIDIA relies on a software optimizer, Denver’s code optimization routine itself has one last advantage over hardware: upgradability. NVIDIA retains the ability to upgrade the DCO itself, potentially deploying new versions of the DCO farther down the line if improvements are made. In principle a DCO upgrade not a feature you want to find yourself needing to use – ideally Denver’s optimizer would be perfect from the start – but it’s none the less a good feature to have for the imperfect real world.

Case in point, we have encountered a floating point bug in Denver that has been traced back to the DCO, which under exceptional workloads causes Denver to overflow an internal register and trigger an SoC reset. Though this bug doesn’t lead to reliability problems in real world usage, it’s exactly the kind of issue that makes DCO updates valuable for NVIDIA as it gives them an opportunity to fix the bug. However at the same time NVIDIA has yet to take advantage of this opportunity, and as of the latest version of Android for the Nexus 9 it seems that this issue still occurs. So it remains to be seen if BSP updates will include DCO updates to improve performance and remove such bugs.

Designing Denver SPECing Denver's Performance
Comments Locked

169 Comments

View All Comments

  • MonkeyPaw - Wednesday, February 4, 2015 - link

    Maybe I missed it, but can you comment on browser performance in terms of tabs staying in memory? I had a Note 10.1 2014 for a brief time, and I found that tabs had to reload/refresh constantly, despite the 3GB of RAM. Has this gotten any better with the Nexus and Lollipop? Through research, I got the impression it was a design choice in Chrome, but I wondered if you could figure out any better. Say what you want about Windows RT, but my old Surface 2 did a good job of holding more tabs in RAM on IE.
  • blzd - Friday, February 6, 2015 - link

    Lollipop has memory management issues right now, as was mentioned in this very article. Apps are cleared from memory frequently after certain amount of up time and reboots are required.
  • mukiex - Wednesday, February 4, 2015 - link

    Hey Josh,

    Awesome review. As I'm sure others have noted, the Denver part alone was awesome to read about! =D
  • gixxer - Thursday, February 5, 2015 - link

    There have been reports of a hardware refresh to address the buttons, light bleed, and flexing of the back cover. There was no mention of this in this article. What was the build date on the model that was used for this review? Is there any truth to the hardware refresh?

    Also, Lollipop is supposed to be getting a big update to 5.1 very soon. Will this article be updated with the new Lollipop build results? Will FDE have the option to be turned off in 5.1?
  • blzd - Friday, February 6, 2015 - link

    Rumors. Unfounded rumors with zero evidence besides a Reddit post comparing an RMA device. All that proved was RMA worked as intended.

    The likelyhood of a hardware revision after 1 month on the market is basically 0%. The same goes for the N5 "revision" after 1 month which was widely reported and 100% proven to be false.
  • konondrum - Thursday, February 5, 2015 - link

    My take from this article is that the Shield Tablet is probably the best value in the tablet market at the moment. I was really shocked to see the battery life go down significantly with Lollipop in your benchmarks, because in my experience battery life has been noticeably better than it was at launch.
  • OrphanageExplosion - Thursday, February 5, 2015 - link

    I seem to post this on every major mobile review you do, but can you please get it right with regards to 3DMark Physics? It's a pure CPU test (so maybe it should be in the CPU benches) and these custom dual-core efforts, whether it's Denver or Cyclone, always seem to perform poorly.

    There is a reason for that, and it's not about core count. Futuremark has even gone into depth in explaining it. In short, there's a particular type of CPU workload test where these architectures *don't* perform well - and it's worth exploring it because it could affect gaming applications.

    http://www.futuremark.com/pressreleases/understand...

    When I couldn't understand the results I was getting from my iPad Air, I mailed Futuremark for an explanation and I got one. Maybe you could do the same rather than just write off a poor result?
  • hlovatt - Thursday, February 5, 2015 - link

    Really liked the Denver deep dive and we got a bonus in-depth tablet review. Thanks for a great article.
  • behrangsa - Thursday, February 5, 2015 - link

    Wow! Even iPad 4 is faster than K1? I remember nVidia displaying some benchmarks putting Tegra K1 far ahead of Apple's A8X.
  • behrangsa - Thursday, February 5, 2015 - link

    Anyway to edit comments? Looks like the K1 benchmark was against the predecessor to A8X, the A7.

Log in

Don't have an account? Sign up now