• What
    is this?

    You've landed on the AMD Portal on AnandTech. This section is sponsored by AMD. It features a collection of all of our independent AMD content, as well as Tweets & News from AMD directly. AMD will also be running a couple of huge giveaways here so check back for those.


Cache Is Not the Only, Or Even the Main, Culprit

Most people pointed to high latency caches as a reason for subpar Bulldozer performance, but the real explanation of why Bulldozer's performance was underwhelming is a lot more complex. First of all, in most applications, an OOO processor can easily hide the 4-cycle latency of an L1 cache. Intel introduced a 4-cycle latency cache three years ago with their Nehalem architecture, and Intel's engineers claim that simulations show that a 3-cycle L1 would only boost performance by 2-3% (at the same clock), which is peanuts compared to the performance boost that is the result of the higher clock speed headroom.

Secondly, a dedicated 4-way 16KB cache, although relatively small, is hardly worse than Intel's 8-way 32KB data cache that is shared by two threads. The cache is also predicted lowering the power to search, so the Bulldozer data cache organisation does have its advantages.

Considering that SAP and Libquantum tell us that Bulldozer's prefetching works quite well, the 20-cycle L2 cache latency might not be a showstopper after all in server and HPC applications. We noticed also that the large 2MB cache offers (much) higher hit rates than the 512KB L2 cache of the older Istanbul/Magny-Cours cores. So while the L2 cache latency is not an advantage, we definitely have doubts that it is a major factor.

We do agree that it is a serious problem for desktop applications as most of our profiling shows that games and other consumer applications are much more sensitive to L2 cache latency. It was after all one of the reasons why Nehalem was not much faster than the older Penryn based CPUs. Lowly threaded desktop applications run best in a large, low latency L2 cache. But for server applications, we found worse problems than the L2 cache.

The Real Shortcomings: Branch Misprediction Penalty and Instruction Cache Hit Rate

Bulldozer is a deeply pipelined CPU, just like Sandy Bridge, but the latter has a µop cache that can cut the fetching and decoding cycles out of the branch misprediction penalty. The lower than expected performance in SAP and SQL Server, plus the fact that the worst performing subbenches in SPEC CPU2006 int are the ones with hard to predict branches, all points to there being a serious problem with branch misprediction.

Our Code Analyst profiling shows that AMD engineers did a good job on the branch prediction unit: the BPU definitely predicts better than the previous AMD designs. The problem is that Bulldozer cannot hide its long misprediction penalty, which Intel does manage with Sandy Bridge. That also explains why AMD states that branch prediction improvements in "Piledriver" ("Trinity") are only modest (1% performance improvements). As branch predictors get more advanced, a few tweaks here and there cannot do much.

It will be interesting to see if AMD will adopt a µop cache in the near future, as it would lower the branch prediction penalty, save power, and lower the pressure on the decoding part. It looks like a perfect match for this architecture.

Another significant problem is that the L1 instruction cache does not seem to cope well with 2-threads. We have measured significantly higher miss rates once we run two threads on the 2-way 64KB L1 instruction cache. It looks like the associativity of that cache is simply too low. There is a reason why Intel has an 8-way associative cache to run two threads.

Desktop Performance Was Not the Priority

No matter how rough the current implementation of Bulldozer is, if you look a bit deeper, this is not the architecture that is made for high-IPC, branch intensive, lightly-threaded applications. Higher clock speeds and Turbo Core should have made Zambezi a decent chip for enthusiasts. The CPU was supposed to offer 20 to 30% higher clock speeds at roughly the same power consumption, but in the end it could only offer a 10% boost at slightly higher power consumption.

Server Workloads: There Is Hope

If there is one thing this article should have made clear, it's that server applications have completely different demands than SPEC CPU or workstation software. They are much more limited by MLP, come with lower IPC, and are more scalable. They also come with a much larger memory footprint and punish small, low latency caches with high miss rates. Therefore a higher latency but larger L2 cache assisted by good prefetchers can perform adequately.

We strongly believe the concepts behind Bulldozer are sound ones for the professional IT world. The trade-offs are well made for these workloads, but there seem to be four show stoppers. So far we found out that the instruction cache, the branch misprediction penalty, and the lack of clock speed are the main reasons why Bulldozer underperforms in the server world.

The lack of clock speed seems to be addressed in Piledriver with the use of hard edge flops and the resonant clock edge, which is especially useful for clock speeds beyond 3GHz. That means "Abu Dhabi" might be a pleasant surprise. AMD has done it before: in 2007, "Barcelona" (K10 architecture) started at a very dissapointing 2GHz and with worse single-threaded performance than expected. At the end of 2008, a slightly improved version of this architecture (Shanghai) was running at 2.7GHz and had a cache that was three times larger with slightly lower latency. So let's hope that "Abu Dhabi" can repeat the "Shanghai stunt".

But what about the fourth show stopper? That is probably one of the most interesting ones because it seems to show up (in a lesser degree) in Sandy Bridge too. However, we're not quite ready with our final investigations into this area, so you'll have to wait a bit longer. To be continued....

Branch Prediction Analysis


View All Comments

  • Aone - Monday, June 04, 2012 - link

    Bulldozer's conception was wrong from the scratch.
    I told it a few time, let's me explain it here again.

    I'm sure everyone of you do remember AMD's own words "one BD module has 80% of throughput of two independent cores".
    What does this mean in figures?
    Let's take the performance of one core as 1.0 point. Therefore two BD modules would have 3.2 points or in other words less than 10% than 3.0 (performance of three independent cores).
    Should I remind that with development of independent cores AMD wouldn't had wasted resources (engineering, transistors, money and time) on design and debugging the shared logic. The chip could have been much smaller due to the fact that the chip would have had only 1MB L2 and 2MB L3 per each core and no shared logic. And all of those released resources could have been allocated for development of a more advanced core.

    You see that packing two cores inside a one module was wrong even on the conceptional level. I'm very curious who was the main supporter and decision maker of this approach in AMD.

    AMD must through away BD conception and return to standard practice. The only question remains: Does AMD have long enough TTL to do it?

    BTW, I recommend to look through Spec results again. The comparison of 12c Opteron 62xx w/ 12c Opteron 61xx is of special interest. And let's not forget that Opteron 62xx submissions have higher freq, faster memory and as well as more advanced compiler version and extended instruction set.
  • TC2 - Monday, June 18, 2012 - link

    I'm agree in 100%!!!
    The BD uA is "unsuccessful" port from graphics uA. There is many and major drawbacks! Note for example one - to write an optimal software you must adopt an application at algorithmic level (in sense of thread specialization)! This is because the both BD-cores are not the same! Also they shares L1 IC, the number of elements is high, ... and many others uA weaknesses.

Log in

Don't have an account? Sign up now