The Bulldozer Review: AMD FX-8150 Testedby Anand Lal Shimpi on October 12, 2011 1:27 AM EST
The Pursuit of Clock Speed
Thus far I have pointed out that a number of resources in Bulldozer have gone down in number compared to their abundance in AMD's Phenom II architecture. Many of these tradeoffs were made in order to keep die size in check while adding new features (e.g. wider front end, larger queues/data structures, new instruction support). Everywhere from the Bulldozer front-end through the execution clusters, AMD's opportunity to increase performance depends on both efficiency and clock speed. Bulldozer has to make better use of its resources than Phenom II as well as run at higher frequencies to outperform its predecessor. As a result, a major target for Bulldozer was to be able to scale to higher clock speeds.
AMD's architects called this pursuit a low gate count per pipeline stage design. By reducing the number of gates per pipeline stage, you reduce the time spent in each stage and can increase the overall frequency of the processor. If this sounds familiar, it's because Intel used similar logic in the creation of the Pentium 4.
Where Bulldozer is different is AMD insists the design didn't aggressively pursue frequency like the P4, but rather aggressively pursued gate count reduction per stage. According to AMD, the former results in power problems while the latter is more manageable.
AMD's target for Bulldozer was a 30% higher frequency than the previous generation architecture. Unfortunately that's a fairly vague statement and I couldn't get AMD to commit to anything more pronounced, but if we look at the top-end Phenom II X6 at 3.3GHz a 30% increase in frequency would put Bulldozer at 4.3GHz.
Unfortunately 4.3GHz isn't what the top-end AMD FX CPU ships at. The best we'll get at launch is 3.6GHz, a meager 9% increase over the outgoing architecture. Turbo Core does get AMD close to those initial frequency targets, however the turbo frequencies are only typically seen for very short periods of time.
As you may remember from the Pentium 4 days, a significantly deeper pipeline can bring with it significant penalties. We have two prior examples of architectures that increased pipeline length over their predecessors: Willamette and Prescott.
Willamette doubled the pipeline length of the P6 and it was due to make up for it by the corresponding increase in clock frequency. If you do less per clock cycle, you need to throw more clock cycles at the problem to have a neutral impact on performance. Although Willamette ran at higher clock speeds than the outgoing P6 architecture, the increase in frequency was gated by process technology. It wasn't until Northwood arrived that Intel could hit the clock speeds required to truly put distance between its newest and older architectures.
Prescott lengthened the pipeline once more, this time quite significantly. Much to our surprise however, thanks to a lot of clever work on the architecture side Intel was able to keep average instructions executed per clock constant while increasing the length of the pipe. This enabled Prescott to hit higher frequencies and deliver more performance at the same time, without starting at an inherent disadvantage. Where Prescott did fall short however was in the power consumption department. Running at extremely high frequencies required very high voltages and as a result, power consumption skyrocketed.
AMD's goal with Bulldozer was to have IPC remain constant compared to its predecessor, while increasing frequency, similar to Prescott. If IPC can remain constant, any frequency increases will translate into performance advantages. AMD attempted to do this through a wider front end, larger data structures within the chip and a wider execution path through each core. In many senses it succeeded, however single threaded performance still took a hit compared to Phenom II:
At the same clock speed, Phenom II is almost 7% faster per core than Bulldozer according to our Cinebench results. This takes into account all of the aforementioned IPC improvements. Despite AMD's efforts, IPC went down.
A slight reduction in IPC however is easily made up for by an increase in operating frequency. Unfortunately, it doesn't appear that AMD was able to hit the clock targets it needed for Bulldozer this time around.
We've recently reported on Global Foundries' issues with 32nm yields. I can't help but wonder if the same type of issues that are impacting Llano today are also holding Bulldozer back.
Post Your CommentPlease log in or sign up to comment.
View All Comments
TiGr1982 - Monday, October 17, 2011 - linkIndeed, much much better performance was expected from BD. I was an AMD focused PC buyer since 2005, at AMD "golden age", when I purchased AMD Turion-based laptop. That CPU was actually better than the corresponding Intel competitor at the moment - Pentium M Dothan, as probaly some people remember.
We know the rest of the story since then till now...
But the released BD-based product in its current state seems to be barely concurrent at all on the desktop market. Presumably, its popularity will be much lower, than in case of previous Phenom II lineup...
TiGr1982 - Monday, October 17, 2011 - linkBy "concurrent", I actually meant "competitive".
psiboy - Tuesday, October 18, 2011 - linkWhy are there no benchmarks with it overclocked... especially gaming? Would be relevant as these processors are shipping unlocked as standard.. all I'm asking for is a reasonable overclock on air to be included...
eldemoledor25 - Tuesday, October 18, 2011 - linkI think they rushed all wanting to position their review as the first, if you read the other post of the network goes bullozer better positioned than the i7 2600K in many things over which a pricipio dicen.el problem was in the bios the asus and gigabyte motherboards, released immature bios fact overclock would hold more, as you may ASRock and MSI makes a bulldozer to 4.6 ghz be better than the i7 and i5 5.2GHz oc do not believe me check this and read well.
1.-http: / / www.madboxpc.com/foro/topic/161318-la-verdad-sobre-el-amd-fxo-bulldozer/page__st__20
Greetings to all!!!
Martin281 - Wednesday, October 19, 2011 - linkWell, the situation among AMDs CPU is still the same...good ideas, great expectitions and manufacturing delays resulting in inappropriate results compared to Intel. Bulldozer would have been a way competitive 2 years ago, not these days. At this point AMD desperately needs way higher clock speeds and core optimizations to be competitive..the predicted 10-15% performance per watt increasing each year is really funny when compared to planned intel´s cpu roadmap (just known information that 1Q/2012 to-be-introduce ivy bridge´s TDP in top performance class is to drop from 95W to 77W, that is almost 20% only in power consumption - not to mention performance boost caused also by 22nm manufacturing process). I am worried, that the performance gap between intel and AMD cpus is going to broaden in the near future without "any light in the darkness bringing the true competition in the CPU field".
siniranji - Wednesday, October 19, 2011 - linkwaiting for the BD to come, but now, what a disappointment, but AMD should continue to compete with intel, otherwise, there wont be any battle to watch. I love to see a good pricing from AMD.
loa567 - Wednesday, October 19, 2011 - linkI think you are wrong on one point, about the FPU. You claim that one bulldozer module has the same FP capacity as earlier AMD processors. However, in reality it has twice the (theoretical) capacity Whereas each K8/K10 core had one 128-bit FP unit, each bulldozer module has 2 x 128 bit FP units. They can work together as one 256-bit, when used with the new instructions (AVX and others). See for example this page for details: http://blogs.amd.com/work/2010/10/25/the-new-flex-...
However, it is strange that this does not show in performance. Could anyone explain this to me?
Pipperox - Thursday, October 20, 2011 - linkIt does show on performance.
In SiSoft Sandra, 4 Bulldozer modules easily beat 6 Thuban cores.
Same goes for floating point intensive rendering tasks such as Cinebench and 3dsMax.
beck2050 - Friday, October 21, 2011 - link"in single threaded apps a good 40-50% advantage the i5 2500K enjoys over the FX-8150."
These are the apps most people use, duh.
core for core Bulldozer is epic fail. This is not going to be a popular desktop chip at all. As for servers, AMD's share has dropped from 20% to 5,5% in the last few years. I doubt this chip will be the savior.
richaron - Friday, October 21, 2011 - linkThey have lost ground in the server market, so a radical new design wont make a difference...? I admire your logic.
For the record I specifically look for programs/games which are multithreading, it often shows good programming on the whole. Unless of course there are other factors limiting the system (like net speed, or gpu). Perhaps I'm just ahead of the curve compared to you're average troll, duh.