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
Gasaraki88 - Friday, October 21, 2011 - linkI think the main competitor for Intel in the future is going to be the ARM processor makers. As Intel goes in to that space with the x86 and the ARM chips getting faster and faster and Windows 8 supporting ARM, you get a mix and soon ARM chip will invade the desktop/laptop market.
AMD is done.
ppro - Monday, October 24, 2011 - linkI decided to get this new cpu from www.amd.com
navair2 - Monday, October 24, 2011 - linkI decided to try AMD when I "inherited" my brother's older socket 939 hardware some years ago, then built my own using a Phenom II X4 940 BE.
At the time it was released, the 940 wasn't too far behind the i7 920 in many respects, plus it was about $70 cheaper...I was very satisfied with my decision. However, after 3 years of advancement by both companies and watching Intel ONCE AGAIN come up with something that gives excellent performance with ever-increasing power reduction, I was on the fence about Bulldozer even before the reviews came out.
Once I saw the majority of the reviews, I knew what side of the fence to be on for obvious reasons..."Bulldozer" just didn't hit the expectations I thought it should, especially when it comes to load power consumption. Perhaps in a couple years when it matures, but I didn't feel like waiting for AMD to iron out all the wrinkles.
My next build is already done and sorry to say, it's NOT AMD. For what I do the i5-2500K is just too good to pass up at combo prices that result in a $200 processor ( less than what I payed for my X4 940 when IT was new).
Best wishes AMD, I hope you can make "Bulldozer" work, but for now "BD" stands for "Big Disappointment". I'll check back with you in a year or so to see how things are doing.
johnsmith9875 - Tuesday, October 25, 2011 - linkIntel's first "dual core" was actually 2 processors on one chip.
They could have saved a lot of engineering time by merely shoehorning two X6 Thuban processors together at 32nm and sell it as a 12 core. Now that would have rocked!
Poxenium - Wednesday, October 26, 2011 - linkDoes anybody remember the first Intel processors with the entirely new architecture called Core Duo, Conroe-L or something ? They were pretty lousy at first, with slightly higher performance than the previous generation, but constantly overheating. later the Core 2 Duo was a complete success, not to mention the first generation iCore processors and of course Sandy Bridge.
Considering the fact that these Bulldozer processors are AMD's first attempt at a completely new architecture, I say that both performance and power consumption are at reasonable levels. Upcoming models will surely do a lot better.
shbdf - Friday, November 4, 2011 - link
== ( http://www.amtata.com )====
== ( http://www.amtata.com )====
Handbags(Coach l v f e n d i d&g) $35
Tshirts (Polo ,ed hardy,lacoste) $15
Wolfpup - Friday, November 4, 2011 - linkI'd love to see how well the SMP client runs on an "8 core" Bulldozer part compared with a quad Sandy Bridge, and for that matter a 6 core Phenom 2 and 6 core Nehalam.
It SEEMS like it should do really well, right? Or not? Because basically an 8 core Bulldozer is a quad core when it comes to floating point, right? And Folding uses a lot of floating point? Or...?
Also, if it really has double the transistor count of Sandy Bridge...where is the performance? It seems like even on heavily threaded stuff it's just kind of about equal with Sandy Bridge, which doesn't seem right....
JumpingJack - Sunday, November 6, 2011 - linkExcept for the fact even with that 'professional' software the competitor is just as fast or faster consuming 30% less power.
it is unfortunate for AMD and their fan base bit BD is definitely a dud.
JumpingJack - Sunday, November 6, 2011 - linkConsidering the power consumption and the reported problems with many games, e.g. Dues Ex, Portal 2, Shogun ... I would see this more appealing around the $180 mark. The 1100T is a better buy if you must do AMD.
JumpingJack - Sunday, November 6, 2011 - linkThe statement is partially true. There are quite a few apps that the Thuban outnguns BD and many cases where it out performs on energy effiency as well.