Bulldozer Performance Issues - Motherboard Related?

[videogamer323]

Not sure if anyone has seen this yet. Thought I would share.
http://www.overclock.net/amd-cpus/11...board-may.html

On the other hand I found this as well, ex Engineer explaining what went wrong...
http://www.xbitlabs.com/news/cpu/dis...er_Fiasco.html

Be aware I am not grasping at straws or anything, just posted possibly relevant information that may explain the poor performance.

If this has already been posted or discussed before you can delete this, but I was unable to find anything.

[charged3800z24]

Very interseting indeed. I am curious to try out the new CHV bios 9911 with BD. I just loaded on the baord today with the 1100t. Not one issue yet. Need BD to see how it does.

[Lastviking]

I hope someone can test the BD with a Gigabyte board and make a compair with that board to.

[Hondacity]

im impressed with that first post, its so un-overclock.net-ish

hahahahha

[Oese]

There are always old employees that would tell something bad about a company if you look for it. How ever, parts may be true. But this in turn gives room for further improvement with newer steppings.

The board thing makes me wonder really, together with some bias in performance/watt which stems from maximum wattage is compared with performance that in most lightly threaded cases is hindered by an unoptimized scheduler.

In direct comparison, the perf/watt should be just better then with the old generation, but of course if you measure max wattage and compare it to average, maybe hindered performance you get some serious bias in comparison to cpu's that dont have problems with lightly threaded workloads...

[SBB]

This makes a lot of sense in gaming performance - we need some more game benches on bulldozer up here

[rintamarotta]

So gigabyte and asus boards are the reason why Bulldozer flops, AsRock and MSI give much better performance for Bulldozer.

@Lastviking
I have Gigabyte 990FXA-UD3 and BD is badly performing in it compared to reviews that use AsRock and MSI boards.

Anyway, why the hell AMD switched from hand crafting to inefficient Automated Design.

[videogamer323]

This makes a lot of sense in gaming performance - we need some more game benches on bulldozer up here

This is what I am waiting to see as well. I honestly do not see what all the ranting and raving is about that it is a complete failure. It is a CPU, it computes things, sure it didn't beat Intel's fastest and uses more power, but it introduced new tech that will be used in the future. Whether that is good or not, no one can know until they come out. So it is nothing more than a loss for the performance crown and honestly who believed it would kick the pants off Intel?

So gigabyte and asus boards are the reason why Bulldozer flops, AsRock and MSI give much better performance for Bulldozer.

@Lastviking
I have Gigabyte 990FXA-UD3 and BD is badly performing in it compared to reviews that use AsRock and MSI boards.

Anyway, why the hell AMD switched from hand crafting to inefficient Automated Design.

And I am betting the reason they switched to automated it so save money (something they desperately need), problem in doing that is they lost 20% performance and gained 20% power draw, which is probably what they would have needed to be neck and neck or even win some against Intel. In my eyes this is where AMD's leaders or whoever made the decision, failed. If they can straighten it out in future revisions then more power to them because they will be making more money or at the least using less to make the chips. Overall I definitely don't think it should have gotten the FX name in it's current state because it is anything but that. Problem is, how long ago was it decided that the FX name would be on these chips and how bad it would have looked if they changed it...

[Olivon]

Bullcrap$ ...

[informal]

How is 8150 performing under Linux in Phronix test suite? This is what Michael Larabel @ Phoronix posted yesterday (thanks to dresdenboy's blog):

Posted by Michael Larabel on October 14, 2011
Here's the first Linux benchmarks of AMD's FX-Series Bulldozer desktop CPUs that launched on Tuesday. Specifically, it's Gentoo Linux performance results for an AMD FX-8150 Bulldozer.

The AMD FX-8150 Linux benchmark results can be found on OpenBenchmarking.org. It's an eight-core AMD FX-8150 on an ASUS Sabertooth 990FX motherboard with 4GB of RAM. Gentoo Linux was used with the Linux 3.0.6 kernel and GCC 4.5.3. Unfortunately, this system is not under my control and there's no direct comparisons available for this hardware system to any other AMD processors.

While there may not be any direct comparisons and these Bulldozer Linux benchmarks are coming in from an independent user running the Phoronix Test Suite and uploading the results to OpenBenchmarking.org, you can compare your system to this FX-8150 Gentoo desktop by running phoronix-test-suite benchmark 1110131-LI-BULLDOZER29 from the latest Phoronix Test Suite client.

Though thanks to the unique OpenBenchmarking.org feature-set, the OpenBenchmarking.org Performance Classifications (OPC) and OpenBenchmarking.org Performance Classification Index (PCI), you can see how this eight-core AMD Bulldozer compares to other Linux systems. Visit this link for the performance classification of this new octal-core processor.

With the OPC results, the "Processor Tests" are the important ones. The FX-8150 results overlayed on the OPC heat-maps indicate that the performance is high-end compared to all of the other systems on OpenBenchmarking.org that have run these tests in the past 120 days. The 7-Zip, NPB, OpenSSL, Tachyon, and Smallpt results highlight this processor the best while the performance in Crafty and EP.B NPB is not as desirable.

[videogamer323]

Bullcrap$ ...

Why do you feel the need to comment and boost your thread count?
This is a thread for information to find out if the motherboard issue is in fact part of the performance loss or none at all. Granted it isn't going to be 10-20% difference. If you don't have a bulldozer and can't test this or know anything about it then sit and watch patiently.

[The Stilt]

It is EXTREMELY hard to make any kind of difference with a different motherboard only.
Unless certain manufacturers are doing something what they should not be doing (e.g tampering with PStates or Turbo i.e overclocking at "stock settings") it is impossible the make a large difference. Even if some of the manufacturers would have done better job tweaking certain memory related timings than other, the difference would be hardly measurable.

AGESA code version for Asrock, Asus and Gigabyte review bioses was also the same, so no difference in there either.

Also do you think AMD chose Asus Crosshair V as their "reference board" by simply pulling a name out of a hat?
I would imagine they did alot comparison between the boards from different vendors (Asrock, Gigabyte, MSI) and possibly even their own reference board.

[Hondacity]

I'd love to see board comparison from chew*

but that would burn him out or fire him up

[Olivon]

Why do you feel the need to comment and boost your thread count?
This is a thread for information to find out if the motherboard issue is in fact part of the performance loss or none at all. Granted it isn't going to be 10-20% difference. If you don't have a bulldozer and can't test this or know anything about it then sit and watch patiently.

What do you want ? Change crap into gold ?

If AMD has choosen a borked mobo for official presskit, that means they're totally stupid, "shoot in the foot" style ...
Bulldozer desktop is a deception and an overhyped product ...

[undone]

http://translate.google.com/translat...postcount%3D49

As some added together .. At least after the first, current indications:

~ 7% of the Asus board
0-60% by bandwidth limitation
5-15% by the scheduler update (were a few benches under the current Win8 trial)
? % Is generally achieved by the appropriate compilation of programs with current compilers


Surely you will not be able to see the stack, but I think we're talking about a good 15% good 10% OS-/Software-seitig and again by hardware - In the Middle! Something like XOP and FMA4 times out of it.

[The Stilt]

http://translate.google.com/translat...postcount%3D49

Kill me.

[Olivon]

Kill me.

[Mechanical Man]

http://forums.extremeoverclocking.co...=358508&page=3

[Sgt.McRuff]

I was looking for ud5 bios update and noticed latest one is missing from the website when it used to be there, same with ud7, wonder why?
http://www.gigabyte.us/products/prod...?pid=3880#bios
http://www.gigabyte.us/products/prod...?pid=3891#bios

PS: not sure if relevant, or just some sorta bug so they pulled it?
I also have ud5 fx-8120 on its way.

[gOtVoltage]

I was looking for ud5 bios update and noticed latest one is missing from the website when it used to be there, same with ud7, wonder why?
http://www.gigabyte.us/products/prod...?pid=3880#bios
http://www.gigabyte.us/products/prod...?pid=3891#bios

PS: not sure if relevant, or just some sorta bug so they pulled it?
I also have ud5 fx-8120 on its way.

tweaking UD5 isnt too bad, waiting to see what happens Bios wise is kilin me.

ATM

If i had a FX id be trying to make it scream already regardless.

[Tha Last Meal]

come on guys this is bull , Some of you Speaks unreasonably . AMD Given to sites reviews a Bulldozer package and this package Included :

ASUS Crosshair V Formula motherboard
AMD 8-core FX-8150 processor

like this site legitreviews
and Other sites have the same package from AMD. so AMD Given to sites reviews this motherboard (ASUS Crosshair V Formula) I want to ask these people that talk without logic..... How do you say that this motherboard (ASUS Crosshair V Formula) is the reason of this poor performance in Bulldozer processor??? AMD put the Bulldozer and And put motherboard (ASUS Crosshair V Formula) in one package
so How, then, became the reason is ASUS Crosshair V Formula motherboard If AMD Given to you this package with this motherboard?
this is Madness guys. hardware site have this review and he test the Bulldozer with this motherboard ASUS M5A99X EVO......What was the result ? the same performance . Another site xbitlabs have this review and he test the Bulldozer with this motherboard Gigabyte 990FXA-UD5......What was the result ? the same performance Again. Another site vortez have this review and he test the Bulldozer with this motherboard Gigabyte 990FXA-UD7......What was the result ? the same performance Again. this is Means that the problem is not the motherboard the problem is Bulldozer processor.

[demonkevy666]

come on guys this is bull , Some of you Speaks unreasonably . AMD Given to sites reviews a Bulldozer package and this package Included :

ASUS Crosshair V Formula motherboard
AMD 8-core FX-8150 processor

like this site legitreviews
and Other sites have the same package from AMD. so AMD Given to sites reviews this motherboard (ASUS Crosshair V Formula) I want to ask these people that talk without logic..... How do you say that this motherboard (ASUS Crosshair V Formula) is the reason of this poor performance in Bulldozer processor??? AMD put the Bulldozer and And put motherboard (ASUS Crosshair V Formula) in one package
so How, then, became the reason is ASUS Crosshair V Formula motherboard If AMD Given to you this package with this motherboard?
this is Madness guys. hardware site have this review and he test the Bulldozer with this motherboard ASUS M5A99X EVO......What was the result ? the same performance . Another site xbitlabs have this review and he test the Bulldozer with this motherboard Gigabyte 990FXA-UD5......What was the result ? the same performance Again. Another site vortez have this review and he test the Bulldozer with this motherboard Gigabyte 990FXA-UD7......What was the result ? the same performance Again. this is Means that the problem is not the motherboard the problem is Bulldozer processor.

the issues here is that thuban can read more instruction pre-clock. 64Kbytes instruction cache x 6 vs bulldozers 64Kbytes 4x.
thread pre-thread i never expected to be 100% at all, because of the sharing. The high clocks (3.9ghz turbo) on all cores it is worse
My main disappointment is that fact that Instruction cache is still the same as phenom II 64Kbytes 2 way, instead of something like 4 way or anything else. Other fixing the the low writes form all caches, I wonder if a increase in L1 instruction cache would change anything ?

[videogamer323]

come on guys this is bull , Some of you Speaks unreasonably . AMD Given to sites reviews a Bulldozer package and this package Included :

ASUS Crosshair V Formula motherboard
AMD 8-core FX-8150 processor

like this site legitreviews
and Other sites have the same package from AMD. so AMD Given to sites reviews this motherboard (ASUS Crosshair V Formula) I want to ask these people that talk without logic..... How do you say that this motherboard (ASUS Crosshair V Formula) is the reason of this poor performance in Bulldozer processor??? AMD put the Bulldozer and And put motherboard (ASUS Crosshair V Formula) in one package
so How, then, became the reason is ASUS Crosshair V Formula motherboard If AMD Given to you this package with this motherboard?
this is Madness guys. hardware site have this review and he test the Bulldozer with this motherboard ASUS M5A99X EVO......What was the result ? the same performance . Another site xbitlabs have this review and he test the Bulldozer with this motherboard Gigabyte 990FXA-UD5......What was the result ? the same performance Again. Another site vortez have this review and he test the Bulldozer with this motherboard Gigabyte 990FXA-UD7......What was the result ? the same performance Again. this is Means that the problem is not the motherboard the problem is Bulldozer processor.

I do understand what you are saying, but the fact it is out there I think it now needs to be shown as true or false for people to stop talking about it. Basically all that is needed is 1 person with 1 bulldozer to try on both boards. I for one just want to see what goes on with the chip in non-reviewers hands that weren't rushed to get the reviews out. Also if there will be any changes to the BIOS, OS patches or whatever else that may increase performance a little (what a % or 2?). All of this for me is just curiosity more than anything as I am not looking to get a chip that will beat Intel because that is just unrealistic at this point.

The other thing I find a bit disturbing is the fact even on some sites that tested a 8150 and 8120 OCed to the same speeds, sometimes one or the other beats the opposing one... This shouldn't be happening between 2 identical chips that only have a MHz difference should it? Or did I miss something? Is this due to the scheduling problem? Overall most of what I have seen is very inconsistent data between reviews sites and even within reviews...

I have my case and watercooling rig all setup and waiting to see what gear to put inside and I'm leaning towards bulldozer just for the shear fun of overclocking it. On the other hand I know I may sacrifice some performance in a couple things I do if I went with AMD and if they say come out with a fix or whatever that gives it a few % then it will make my decision a bit easier. I currently run Intel and I am no Fanboy, just looking for something fun/good to put in my rig.

[Tha Last Meal]

I do understand what you are saying, but the fact it is out there I think it now needs to be shown as true or false for people to stop talking about it. Basically all that is needed is 1 person with 1 bulldozer to try on both boards. I for one just want to see what goes on with the chip in non-reviewers hands that weren't rushed to get the reviews out. Also if there will be any changes to the BIOS, OS patches or whatever else that may increase performance a little (what a % or 2?). All of this for me is just curiosity more than anything as I am not looking to get a chip that will beat Intel because that is just unrealistic at this point.

The other thing I find a bit disturbing is the fact even on some sites that tested a 8150 and 8120 OCed to the same speeds, sometimes one or the other beats the opposing one... This shouldn't be happening between 2 identical chips that only have a MHz difference should it? Or did I miss something? Is this due to the scheduling problem? Overall most of what I have seen is very inconsistent data between reviews sites and even within reviews...

I have my case and watercooling rig all setup and waiting to see what gear to put inside and I'm leaning towards bulldozer just for the shear fun of overclocking it. On the other hand I know I may sacrifice some performance in a couple things I do if I went with AMD and if they say come out with a fix or whatever that gives it a few % then it will make my decision a bit easier. I currently run Intel and I am no Fanboy, just looking for something fun/good to put in my rig.

For me The problem is obvious , this FX-8150 processor Failed for one reason and the reason is ? bulldozer Architectural . This is the only reason man , We can say thet AMD Failed with this bulldozer Architectural , very Bad performance with games , 4 core processor (I7 2500/2600K) can beats your 8 core processor (FX-8150) , Power consumption is very high , 8 core processor (FX-8150) with 1 core Tests? Bad performance , 8 core processor (FX-8150) with OC? No significant difference , and more .

Did you know the worst thing is that AMD Came to us with the First The best The strongest The fastest 8 core In the world and...........After all this 4 core can beats your 8 core Easily

[STEvil]

Olivion, i'm tired of reading your thread crapping replies. Stop thread crapping, this is your only warning.

The Last Meal - the processor is not a true 8 core. Yes it consumes too much power, yes it has issues, and no the boards are not going to be exempt from causing performance issues just because AMD issued them to the reviewers in the review kits.