AMD FX-8150 Bulldozer On Ubuntu Linux

[edgar_wibeau]

Shining Arcanine:
Funny, in Germany the prices are:
2500k: ?182
8150: ?214
2600k: ?260

That's 2500k + ?32 = 8150 + ?46 = 2600k
So here the price is closer to 2500k. And it's not really in the market by now, only two dealers have them in stock. Usually prices trickle down once products are in stock widely.

raj7095: It's a comparison of CPUs of a certain price range at stock speeds. Nothing unfair here. If you want an overclocking review/comparison, that would be a different review.

The power draw is a problem to some, though. And to some, it's not

[crazycheese]

Could be, yes. But you should also mention, that power consumption increases disproportionally.
Look at this bench for example:

http://www.hardocp.com/article/2011/10/11/amd_bulldozer_fx8150_desktop_performance_review/9

numbers under load:
i7 2600K @stock: 181W
i7 2600K @4,8GHz: 275W
FX-8150 @stock: 258W
FX-8150 @4,6GHz: 452W
Impressive, isn't it?

Thats not complete graph to be impressive:

--IDLE--
FX-8150 @ 112W
FX-8150/4.6Ghz @ 120W

i7 2600K @ 102W
i7 2600K/4.8Ghz @ 144W

i7 920 @ 169W
i7 920/4Ghz @ 257W
Note: Amd power management is actually more efficient!

--LOAD--
FX-8150 @ 238W
FX-8150/4.6Ghz @ 442W

i7 2600K @ 181W
i7 2600K/4.8Ghz @ 275W

i7 920 @ 265W
i7 920/4Ghz @505W!

It has something to do with core construction.
i7 2600 is particulary very efficient (less waste) with each Hz+, compared to BOTH i7 920 and BD.
Notice, 920 consumed MORE than BD, in terms of efficiency it is worser than BD.

Also, previous tests in that review under link you posted, show that not only BD scales in terms of watt waste per Mhz a lot worser than 2600, yet similar to 920;
BUT Bulldozer overclock to 4.6Ghz brings very LITTLE performance advantage, yet MAJOR BOOSTS power consumption. This means unfortunately BD is very inefficient for overclocking. Do not overclock your BD CPU.

Compared with i920, which not only required more voltage, but also has boosted performance (unlike BD), it looks like bulldozer has major bottleneck somewhere.

[septimus]

important comparison missing from initial review

The review compared the Bulldozer to an Opteron 23xx. Any chance of seeing a comparison against an Opteron 4100 or 6100? This would be the most relevant comparison for those of us considering Opteron 42xx or 62xx machines. I haven't seen any 'real' benchmarks of an Interlagos system, so the FX8150 is closest thing available...

[raj7095]

WTH is wrong with you people? AMD clearly has a more advanced, forward thinking architecture, and Intel fanboys think if that translates into doing well in any benchmark, that it's "unfair".

Unfortunately, the engineering they did doesn't translate into a clear win in every possible category, but this whining is just unbelievable.

I suppose if Llano or Trinity just stomp Ivy Bridge into the ground in OpenCL benchmarks with their vastly superior GPUs, that OpenCL benchmarks will be unfair in general?

I am sorry, but why should I care about if the architecture is forward thinking in the first place? Isn't that supposed to be the corporation's job? Also, I didn't even make fun of the architecture. I am just saying that it would be nicer if we knew how sandybridge would do with hyperthreading. Also, 400 mhz isn't a too low in this case. Because, 3.8 ghz is about 10 percent less than 4.2 ghz. And it wasn't an unfair benchmark, I just said it was somewhat unfair to not see 2600k in there. Actually, lemme make it clearer, it was not unfair, just a little disappointing for me to not see 2600k in there. I would have really loved that. Also, everyone knows amd integrated gpus are way better than intel integrated gpus. nobody denying that.

[schmalzler]

Thats not complete graph to be impressive:

--IDLE--
FX-8150 @ 112W
FX-8150/4.6Ghz @ 120W

i7 2600K @ 102W
i7 2600K/4.8Ghz @ 144W

i7 920 @ 169W
i7 920/4Ghz @ 257W
Note: Amd power management is actually more efficient!

I don't know, but something probably went wrong, the systems probably were not idling? AntiVirus, perhaps?
Tomshardware tested sandy bridge Overclocking abilities, and they did not see any increase in the idle consumption - which is understandable, as SpeedStep should lower the voltage to a fixed value (was it 0,8V?) when idling, regardless which multiplicator was set.

[nepwk]

You're absolutely right. It's just that me and a handful of other nutjobs tend not to buy "a more advanced, forward thinking architecture", but just go for the best bang for the buck. Now.

Right. This article tells us that if you use modern software that scales well with semi-real cores, that Bulldozer just may be the best bang for the buck. Of course, you chose to take one tiny part of my post completely out of context, rather than to argue my actual point, which pretty much says it all.

I am sorry, but why should I care about if the architecture is forward thinking in the first place? Isn't that supposed to be the corporation's job?

I never said you should. Stop trying to take things I say out of context.

Also, I didn't even make fun of the architecture. I am just saying that it would be nicer if we knew how sandybridge would do with hyperthreading. Also, 400 mhz isn't a too low in this case. Because, 3.8 ghz is about 10 percent less than 4.2 ghz. And it wasn't an unfair benchmark, I just said it was somewhat unfair to not see 2600k in there. Actually, lemme make it clearer, it was not unfair, just a little disappointing for me to not see 2600k in there. I would have really loved that. Also, everyone knows amd integrated gpus are way better than intel integrated gpus. nobody denying that.

No, you said that this article was "unfair" because it didn't have a hyperthreaded CPU, as if that would suddenly make Intel shine, you didn't say it would be "nice". Big difference. Now you retract it, and that's fine, but I stand by my point about hyperthreading, it's not going to change the fact that AMD caught up quite a bit with this new CPU.

I'm not sure why you feel so strongly about Bulldozer, surely you've never even been in the same room as one before to have an opinion. For that matter, you quite possibly have never used a Sandy Bridge machine either, I'm not sure why your panties are in a wad just because Phoronix/OB.org, quite possibly the most scientific and unbiased benchmark site in the world, has declared that Bulldozer beat Sandy Bridge in a few benchmarks, and still loses in many. Considering how open and automated OB.org is, Michael would have a really hard time pulling off sleazy benchmarketing propaganda like Tom, Anand and others do.

[ciplogic]

(...)I'm not sure why you feel so strongly about Bulldozer, surely you've never even been in the same room as one before to have an opinion. For that matter, you quite possibly have never used a Sandy Bridge machine either, I'm not sure why your panties are in a wad just because Phoronix/OB.org, quite possibly the most scientific and unbiased benchmark site in the world, has declared that Bulldozer beat Sandy Bridge in a few benchmarks, and still loses in many. Considering how open and automated OB.org is, Michael would have a really hard time pulling off sleazy benchmarketing propaganda like Tom, Anand and others do.

Phoronix benchmarking had issues at least in the past by using a broken methodology (comparing perfomance without having states the expectations, the methodology or making it in a bogus way, testing OSes some with composite enable, some don't, and so on). Even actual article said about 8 cores when most people will likely say: 4 modules of 2 cores.
Anyway, those benchmarks were good in some ways as:
- Linux is traditionally a platform that works right for multicore because it comes from server/enterprise area, so Bulldozer CPU may be better used
- BD CPU may give good performance if all cores are used. In fact it happen to other sites like this one when that site claims: "Video transcoding is well suited for systems that have more CPU cores. Encoding/transcoding to x.264 format is one of the most intensive tasks a processor can perform. As such this is one of the better tests in the entire review. "
It just happen to have SB Mobile CPU (to not have critics, also I have an I7, at a time I was considering between that i7 or AMD X6, yet the Intel System was very Linux friendly hardware, so I did not had any issues with installing Linux on it, and was one of the reasons of buying the Intel one).
I think most of BD backslash was not because BD is bad per-se, but because: it come late, in some benchmarks it not behave stellar, which is to be expected, not having a partnership with MS to put a patch that AMD CPUs were detected as hyperthreaded cores (which reflects most of the performance issues in medium threaded benchmarks), the hope that as being a new architecture, like Athlon was, will simply trash the Intel one. In fact even Athlon was a great CPU, being a "Celeron" of the Alpha CPUs, it was just like 5-10% faster than Intel Pentium 3, and polishing Pentium 3 was making today that Intel looks so aggressive. Bulldozer will need just applications to take in account to use all 8 threads. Can we find that many applications to use all those 8 threads? The likely answer is: NO. But considering that in 1-2 years most bought of all CPUs will have minimum 4 Cores/modules or at least dual-core HT, software will get eventually have to spawn more threads. How many? I cannot give a precise number but the direction is certainly that Bulldozer CPU, if someone buys the 8120 CPU, will be the PC that will work for another 10 years, people will have to add RAM and disk drive.

[Kivada]

I don't know, but something probably went wrong, the systems probably were not idling? AntiVirus, perhaps?
Tomshardware tested sandy bridge Overclocking abilities, and they did not see any increase in the idle consumption - which is understandable, as SpeedStep should lower the voltage to a fixed value (was it 0,8V?) when idling, regardless which multiplicator was set.

They're forgetting that when they are OCing the CPU they are pegging it at full speed instead of letting PowerNow/Cool and Quiet/TurboCore(They are all the same thing, TC just adding a few more stages) underclock it on the fly. So of course power consumption goes through the roof, since it's running full bore without stepping down to idle, hence why most OCers run F@H or the like so that extra power isn't wasted while watching porn.

Some sites that actually gave the 8150 a good whack:

http://www.legitreviews.com/article/1741/19/

http://benchmarkreviews.com/index.php?option=com_content&task=view&id=831&Itemid=63&limit=1&limitstart=15

http://www.madshrimps.be/articles/article/1000220/AMD-FX-8150-Bulldozer-CPU-Review/3

http://www.overclock3d.net/reviews/cpu_mainboard/amd_fx8150_cpu_review/3

[blackiwid]

T

http://www.overclock3d.net/reviews/c...0_cpu_review/3

Their Result site sounds not very possitive

[nepwk]

Phoronix benchmarking had issues at least in the past by using a broken methodology (comparing perfomance without having states the expectations, the methodology or making it in a bogus way, testing OSes some with composite enable, some don't, and so on).

I agree with a lot of what you say, of course there are inconsistencies. However, compared to other benchmark sites, like Anandtech continuing to use Futuremark even after the entire industry cast shame on it as being Intel-biased, etc... OB.org is actually pretty reasonable, as long as you understand that there may be inconsistencies. Unlike most of the other benchmark sites, the inconsistencies don't always favor Intel,<sarcasm> by coincidence </sarcasm>, of course.