• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

Why Bulldozer's spotty performance is good news.

  • Thread starter Thread starter twilyth
  • Start date Start date
Steevo said:
All that being said, I'm sure it will be fun to play with, and yes, it is a upgrade from a X6 if you use software that will take advantage of it, but NO you CANNOT compare it to a 2600 for threaded performance.

$314 for a 2600

VS

$279.99 for a 8150

$30 less.

http://www.bit-tech.net/hardware/cpus/2011/10/12/amd-fx-8150-review/10

Load stock speed 82W more power. Means on average $17 more per year in power for the AMD system, not including cooling costs.
Click to expand...
Which is why we hope Piledriver takes it out completely. The AMD FX 8150 does fair well against the 2600 and the 2500, but not in all benchmark tests.
 
"AIDA64 Extreme Edition is a streamlined Windows diagnostic and benchmarking software for home users.

FX-8150 (Wins 8) vs. PII x6 1100T (Wins 4)
Despite Bulldozer winning most of those benchmarks over the older x6 1100T, it's still needs massive improvements in performance.

AIDA64 CPU AES (ALL HIgher is Better)
Core i5 2500K + z68 = 440,033
FX-8150 + 990FX = 346,134
PII x6 1100T + 990FX = 55,013

AIDA64 CPU PhotoWorxx
Core i5 2500K + z68 = 59,513
FX-8150 + 990FX = 48,200
PII x6 1100T + 990FX = 31,876

AIDA64 CPU Queen
Core i5 2500K + z68 = 35,087
FX-8150 + 990FX = 32,088
PII x6 1100T + 990FX = 32,256

AIDA64 CPU Zlib
Core i5 2500K + z68 = 217
FX-8150 + 990FX = 262
PII x6 1100T + 990FX = 245

AIDA64 CPU JULIA
Core i5 2500K + z68 = 11,689
FX-8150 + 990FX = 9,591
PII x6 1100T + 990FX = 12,659

AIDA64 CPU MANDEL
Core i5 2500K + z68 = 6,196
FX-8150 + 990FX = 4,793
PII x6 1100T + 990FX = 6,442

AIDA64 CPU SinJulia
Core i5 2500K + z68 = 3,272
FX-8150 + 990FX = 2,145
PII x6 1100T + 990FX = 3,215

AIDA64 CPU HASH
Core i5 2500K + z68 = 2,254
FX-8150 + 990FX = 3,672
PII x6 1100T + 990FX = 3,313

AIDA64 CPU VP8
Core i5 2500K + z68 = 3487
FX-8150 + 990FX = 3493
PII x6 1100T + 990FX = 3375

AIDA64 Memory Copy
Core i5 2500K + z68 = 20.361
FX-8150 + 990FX = 18.293
PII x6 1100T + 990FX = 10.93

AIDA64 Memory Read
Core i5 2500K + z68 = 18.892
FX-8150 + 990FX = 13.971
PII x6 1100T + 990FX = 8.773

AIDA64 Memory Write
Core i5 2500K + z68 = 18.853
FX-8150 + 990FX = 10.268
PII x6 1100T + 990FX = 7.08

AIDA64 Memory Latency (Lower Better)
Core i5 2500K + z68 = 46.65
FX-8150 + 990FX = 51.2
PII x6 1100T + 990FX = 51.2

http://www.bjorn3d.com/read.php?cID=2125&pageID=11106

Conclusion

It is extremely refreshing to see a new architecture from AMD after such a long time without one. Bulldozer is the world's first eight-core consumer CPU, and is marketed towards the consumer and enthusiast gaming crowd. The new chip does have many attractive features: unlocked multipliers across all models, a base frequency of 3.6GHz (unparalleled even by Intel), and a whopping eight cores per CPU.

Despite the nice specifications, there are some considerations we must take into account. Even though we say "eight cores", and AMD uses that term also, the Bulldozer architecture doesn't actually have eight true cores. The Bulldozer module's resource sharing of execution and front-end units means the FX chip will be weak in several ares of computing. Despite having the Turbo speed of 4.2GHz, the CPU falls very short in single-threaded (or even lightly-threaded) processes, even below the old Phenom II CPU's at lower clockspeed. We saw an example of this in some games, and in benchmarks like AIDA64, where the low floating point performance of the FX CPU resulted in performance lower than almost any other CPU.

The FX CPU also stands out in data encryption, though for the opposite reason. As new module architecture is designed not only for desktop system but also for server systems, we are glad to see that AMD has finally added AESNI instruction to its processor. In TrueCrypt, the FX CPU takes the top spot in every test. One of the advantages is the higher number of cores, and the addition of the AESNI instruction set extension. This allows the application to use the CPU for hardware acceleration. It is clear that for data encryption, the FX platform is the CPU of choice, so users who anticipate doing heavy encryption should definitely go for Bulldozer.

We hope Piledriver die shrink, which will be released in 2012, will retain Bulldozer's strength. We also hope that AMD will change some of the Bulldozer core structure to better suit lightly-threaded applications and games, as this is unfortunately an area where Bulldozer falls flat. This may be because AMD plans to implement Bulldozer on their Trinity APU's and perhaps move the bulk of floating point calculations to the graphics side of the APU, but we hope that the desktop CPU version of Bulldozer is also upgraded.

The FX CPU is by no means terrible. It does have a few shortcomings in its architecture that we hope AMD will improve upon. Nonetheless, it can perform solidly in multi-threaded programs and takes the cake in calculations that involve data encryption and integer calculations. It would be a good CPU for people who want the power of a full eight-core chip and do a lot of multi-threaded applications. Due to poorer single-threaded performance, it doesn't perform well in games. Overall, this CPU is better suited towards heavy-threaded productivity, and might be a better choice for those situations. We now await AMD's next iteration of Bulldozer to see how that performs.
Click to expand...
 
Last edited:
We also hope that AMD will change some of the Bulldozer core structure to better suit lightly-threaded applications and games, as this is unfortunately an area where Bulldozer falls flat.
Click to expand...

Exactly, and that's why im waiting.:)
 
no way, intels priceings on the up, what a surprise!, and a kick in the balls for all, and this thread
at this rate BD will only have to compete with the I5's based on cost by christmass anyway, and thats a fight it would win go intel lol not
 
CDdude55 said:
Exactly, and that's why im waiting.:)
Click to expand...
I fully agree 100%.
As for me I cannot wait, I need something to play with before I go nuts :laugh:
 
Ars Technica had a great overview of BD's advantages and failings. It's a little low-level in parts but is generally a very accessible article. If you really want to get under the hood, they reference 2 other articles that are much more demanding.

http://arstechnica.com/gadgets/news/2011/10/can-amd-survive-bulldozers-disappointing-debut.ars/1

One thing of note on page 2 is that the rumors about a registry fix while probably inaccurate if taken literally, do have some truth to them.

On top of all this, the instruction cache within each module has a flaw that can inflict about a 3% performance penalty whenever threads from different processes are scheduled to the same module. A fix for the issue has been incorporated into the Linux kernel, but the status of other operating systems is currently unclear.
Click to expand...
 
What's an improvement in single threaded performance?
How much?

Piledriver will obviously not catch up with Intel in that department, but how much is enough?
10% is not enough unless it's a low clocked PD that's benched first -- which may happen... or may not.
Then again, what is 'enough' depends on Ivy Bridge benchmarks.
But, in my opinion, single threaded performance needs to increase 25%, and that only brings it on par with the current i3-21xx cpu (the 2600K is 20% faster than that, single threaded), that is if we're talking fx-8150 to the PD equivalent.

10% is barely adequate, and downright dismal. Any Phenom II with a 3.4Ghz or higher clock speed is as fast or faster than the fx-8150 single-threaded. The only way 10% is OK, is if it's, lets say a 3ghz PD, and it's 10% faster (single threaded) than a 3.6 ghz fx-8150.

From the enthusiast perspective, it has to be more than 10% single threaded increase, clock for clock.

From the long term perspective, maybe it's OK, if they've got more improvements down the line.
It's quite obvious the sh!tty single thread performance delayed the FM2 socket 2nd gen APU... imagine... 2 bulldozer modules (4 integer cores) only being able to (at best) match the Athlon II cores in the FM1 APUs... so Piledriver modules are slotted in for the FM2 now.
They need to EOL the Athlons and Phenoms because of this... the 2nd gen APU will likely be the performance-equivalent of a current Phenom II + HD6850 in overall performance, just with a better memory controller.
The APUs are potentially interesting, from a certain point of view, if they're unlocked. Assuming the following:
1) the cpu power consumption problem is solved.
2) the 'power vs frequency vs performance' is fixed so that it doesn't top out at at ~4.3 Ghz, leading to dimishing returns and absolutely huge power consumption.
3) Third or Fourth generation APUs (Steamroller or whatchamacallit modules) + more than 800 stream processors or their next gen equivalent.

Personally, in the short term (6-12 months), I'm interested to see reviews of the new Ivy Bridge Q77 & Z77 boards and processors. In the long term, I'm interested to see the third and fourth gen APUs from AMD (especially unlocked versions)... unless Intel manages to somehow catch up in the GPU department.
 
Take a look at the article. This is the sort of thing they talk about. For example K10 has 3 128-bit FPU's per core. BD has 4 per module which amounts to a reduction in the number per thread - well, sort of. And this is what you see in the benchmarks.

The important thing for me was their explanation of how BD is an expression of AMD's philosophy. They think that FP operations should be offloaded to the GPU. That's why they had no problem forcing 2 cores to share FPU's. Plus you need 2 FPU for the new 256-bit AVX commands.

Also BD was supposed to be much faster. They made certain design sacrifices to achieve that but it didn't work.
 
A nice article, it neatly sums up everything. I don't dispute anything in the article.
But like I said, 10% is not enough, hence my interest in post-Piledriver APUs.
I can roll with a quad phenom II for a while, or a hexa core.
Or an Ivy Bridge.
I just don't see enough of a compelling reason to buy a Bulldozer cpu. And at this point, even if AMD delivers the promised 10%, I don't see a compelling reason to buy a Piledriver either. It's not looking to be a large enough performance boost in the 1 to 4 thread domain, which is essentially the core domain of typical PC usage.

I wait, so that I may see.
Time will tell.
 
I think that's about right for the average user. In fact, you're being pretty charitable considering the fact that on some benches it falls behind even a P2 X4 - not to mention X6's.

Had it come out at 4.4G, that would have helped, but you'd still have the same issues.

My main criteria is BOINC performance and from the scientific benches in the tech report article, I have to say that I'd be hard pressed to go with BD.

Of course I'll probably get one anyway - just like I still have my original Phenom 9600. And for most users, the main issue will be price. With decent performance and a catchy name, hopefully AMD can find a PR agency that can sell the shit out of those chips.
 
Inceptor said:
A nice article, it neatly sums up everything. I don't dispute anything in the article.
But like I said, 10% is not enough, hence my interest in post-Piledriver APUs.
I can roll with a quad phenom II for a while, or a hexa core.
Or an Ivy Bridge.
I just don't see enough of a compelling reason to buy a Bulldozer cpu. And at this point, even if AMD delivers the promised 10%, I don't see a compelling reason to buy a Piledriver either. It's not looking to be a large enough performance boost in the 1 to 4 thread domain, which is essentially the core domain of typical PC usage.

I wait, so that I may see.
Time will tell.
Click to expand...

Same here ill wait
Nothing i can't do right now with my PIIx4 for the next year or so , although i don't really need more ill still going to buy a 1090T before Christmas time & i am not going to spend any money on new board just to get the FX CPU... no thx!
 
They will sell to big agencies, no doubt about it.

twilyth said:
I think that's about right for the average user. In fact, you're being pretty charitable considering the fact that on some benches it falls behind even a P2 X4 - not to mention X6's.

Had it come out at 4.4G, that would have helped, but you'd still have the same issues.

My main criteria is BOINC performance and from the scientific benches in the tech report article, I have to say that I'd be hard pressed to go with BD.

Of course I'll probably get one anyway - just like I still have my original Phenom 9600. And for most users, the main issue will be price. With decent performance and a catchy name, hopefully AMD can find a PR agency that can sell the shit out of those chips.
Click to expand...
 
And if the FX 8150 was to hit a price tag of $200, it would further fly off the shelves. Many current AM3 and AM3+ owners feel Bulldozer is a great upgrade despite it's lack luster performance in many benchmarks.

For me Bulldozer makes for a great performance boost over my current setup, and I already have 16GB of DDR3-1866 along with the ASUS Crosshair V Formula and Corsair H100. They are sitting on my desk. :D
 
Damn! That i7 just got destroyed!
 
Super XP what about the BD at 4.7GHz?
 
Jizzler said:
Damn! That i7 just got destroyed!
Click to expand...

If getting destroyed equals achieving the same FPS while using way less power, cooler and more efficiently I don't know what destroyed means anymore.
 
Why are we looking at a single benchmark? When you have to try this hard to make something not look like a POS that should prod the logic center of your brain into giving up the ghost. Operative word there is "should".
 
I'm seriously missing something here...like what yall are talking about cause all i see is 4.7 for an intel chip and 4.7 for an AMD CHIP
 
jmcslob said:
I'm seriously missing something here...like what yall are talking about cause all i see is 4.7 for an intel chip and 4.7 for an AMD CHIP
Click to expand...

You missed the thread that was deleted. Chew* posted a Dirt 2 benchmark from both of them at 4.7 and they were pretty much equal with BD winning by a few frames.
 
so how do you determine if all cores were being utilized.
 
Damn_Smooth said:
He had 4 cores disabled.
Click to expand...
I believe he had 1 Core dispbled per Module, so 1 Core per Module was enabled.
Pay close attention to the specs.
16GB of Dual Channel DDR3-1866

AMD FX "Bulldozer" Review - (4) !exclusive! Excuse for 1-Threaded Perf.

What I'm about to deal with here is comparing 2CU/4C and 4CU/4C Bulldozers.
(CU stands for Compute Unit, or equivalently 'Module')
It can be an excuse for Bulldozer's initial poor single-thread performance:
benchmark tools are just under-optimized for that kind of architecture, but not virtually a 'poor' performance if some optimization is done.
Click to expand...
http://www.xtremesystems.org/forums/showthread.php?275873-AMD-FX-quot-Bulldozer-quot-Review-(4)-!exclusive!-Excuse-for-1-Threaded-Perf.



Dam who deleted my thread, I was in the middle of editing one of my posts.:D
 
Last edited:
Super XP said:
I believe he had 1 Core dispbled per Module, so 1 Core per Module was enabled.
Pay close attention to the specs.
16GB of Dual Channel DDR3-1866


http://www.xtremesystems.org/forums/showthread.php?275873-AMD-FX-quot-Bulldozer-quot-Review-(4)-!exclusive!-Excuse-for-1-Threaded-Perf.
Click to expand...

Yeah, I know. He still needs more than a single Dirt 2 benchmark to prove anything though.

Dam who deleted my thread, I was in the middle of editing one of my posts.:D
Click to expand...

I don't know for sure, but I have a good guess.
 
B3 Stepping Coming in Q1 2012

Have you seen this. It seems only the high end will use the B3 stepping and it's modifications and tweaks. If this CPU performs better than the 8150, I may pick it up.

AMD FX-8170
B3 Stepping :rockout:
3.90 GHz (Base Clock)
4.5 GHz (Turbo Core 3.0)
Q1 2012 (Release Date)

http://en.wikipedia.org/wiki/List_of_AMD_FX_microprocessors
 
You must log in or register to reply here.
Back
Top