Official Bulldozer PPD??

Doozer

Doozer

2[H]4U
Joined
May 30, 2001
Messages
2,495
Anyone have F@H performance in their Bulldozer reviews?
I looked around but didn't see any.
 
Too bad it pulls so much power at 4.5Ghz. SB is so nice because running at 4.7Ghz it uses the same power as a stock i7 920.
 
I didn't expect the power draw to be so high. I was almost shocked when I saw the results under load.
 
yeah the power draw seems insane. but meh, whats an extra 100 watt's anyways :D

you could try asking kyle if he would be willing to run musky's benchmark on their 8120/50 but i doubt we will see anything F@H related yet.
 
I want to see what Movieman does with WCG, might be stuck drawing lines with his numbers.
 
After skimming all the BD reviews, I was surprised to find not one evaluating F&H performance. I suppose that probably had to do with the limited amount of time that was available since many of those other sites do often include F&H performance info.

If BD goes on sale in retail this week, I just hope that we get some numbers soon as i'm very curious. Since BD appears to shine in massively multithreaded stuff, I hope it at least does decently. The power draw was out of this world though and is really an embarrassment considering that AMD was talking all about efficiency and power savings with the BD architecture for a while.
 
Power draw is sky high on these, but also keep in mind that all power saving states are typically off for those tests. If you aren't folding those are likely to buy you back quite a bit of that delta.

From all the multithreaded benchmarks I'd expect the PPD is just about right on with a 2600k which is certainly a big improvement for AMD, its just too bad the chip didn't launch on schedule as it trounces Nehalem versus being pretty 'meh' versus SB. :(
 
Power draw is sky high on these, but also keep in mind that all power saving states are typically off for those tests.
Click to expand...
Irrelevant. All those power saving states would also be off for the i7 CPUs compared in the reviews and will also be off if you are full-time Folding SMP, overclocked.

My upgrade decisions for my Folding machines just became much simpler.
 
i'm still building an bulldozer rig so i can finally get back to folding(i need a room heater for the winter). 200 bucks for the 8120 + 180ish for the sabertooth board is still pretty nice. plus i still can't bring myself to ever buy intel again.
 
Why don't you want Intel ever again?

I'm no "fanboy" I follow the performance crown only. And my recent intel's have been amazing. The i7-930 clocked all the way to 4.8ghz on only an H-50 cooler, and willing to OC more i just didnt have the balls lol.
 
thegtproject said:
Why don't you want Intel ever again?

I'm no "fanboy" I follow the performance crown only. And my recent intel's have been amazing. The i7-930 clocked all the way to 4.8ghz on only an H-50 cooler, and willing to OC more i just didnt have the balls lol.
Click to expand...

long list or reasons.. yeah i agree their processors are great, i recommend them to most people that i build systems for, but as a company i won't support them with my own money. the last intel processor i've personally owned was a pentium II 300mhz and it will continue to be the last one i'll ever own.
 
i was just wondering if they did something bad to you, or horrible support when you needed it most type thing.
 
If FAH uses floating point instructions (there are only four FP cores on bulldozer), FAH performance might actually be lower than Thuban. Check out the cinebench results.

http://www.legitreviews.com/article/1741/11/

I'm not sure if FAH is floating point heavy or not.

Edit. This might be bad. Bruce said FAH uses mostly floating point calculations.

More analysis:
Despite a 9% higher base clock speed (more if you include turbo core), a 3.6GHz 8-core Bulldozer is only able to outperform a 3.3GHz 6-core Phenom II by less than 2%. Heavily threaded floating point workloads may not see huge gains on Bulldozer compared to their 6-core predecessors.

There's another issue. Bulldozer, at least at launch, won't have to simply outperform its quad-core predecessor. It will need to do better than a six-core Phenom II. In this comparison unfortunately, the Phenom II has the definite throughput advantage. The Phenom II X6 can execute 50% more SSE2/3 and x87 FP instructions than a Bulldozer based FX.

Since the release of the Phenom II X6, AMD's major advantage has been in heavily threaded workloads—particularly floating point workloads thanks to the sheer number of resources available per chip. Bulldozer actually takes a step back in this regard and as a result, you will see some of those same workloads perform worse, if not the same as the outgoing Phenom II X6.
Click to expand...
http://www.anandtech.com/show/4955/the-bulldozer-review-amd-fx8150-tested/2
 
Last edited:
Well, I bought an 8120. Hopefully I can get my BIOS updated.
 
EngrChris said:
To add, the first comment. 3.6GHz gives 13-14k on SMP.
Click to expand...

Thats only 2k more than my stock clocked 1090T, what we need to see are some -bigadv numbers - preferably from a server chip when they finally release them
 
umm hes confused cause he says hes getting 13.8k on big adv. im assuming hes not taking into account the bonus point system?

even 13-14k on smp that sucks.
 
I don't hold much hope for the server bound Interlagos substantially outperforming Magny Cours. On the desktop they mitigated their IPC reduction with aggressive clock scaling and high power draw, similar to what Intel did with Netburst. However everything I'd read on IL show it clocked at 2.6ghz, this points to it performing more like a mid range Dodeca.
 
IPC per core is considerably lower than last gen AMD. I don't know why they didn't just redesign Thuban for 8 cores on 32nm. I think it could have been considerably better than this and they're already using a huge number of transistors.
 
Last edited:
Zink said:
IPC per core is considerably lower than last gen AMD. I don't know why they didn't just redesign Thuban for 8 cores on 32nm. I think it could have been considerably better than this and they're already using a huge number of transistors.
Click to expand...

They tried to keep IPC the same while enabling a 30% increase in clockspeed. What they ended up with is a 7% reduction in IPC with a 9% increase in clockspeed, hence the performance wash with considerably higher power draw.

Here's hoping the Piledriver launching next year brings the sort of improvement we saw from Phenom to Phenom II.
 
What if they spent all of that time figuring out how to use those 2 billion transistors to fit 10 of their last gen cores and 10mb L2 on one die. They could have been seriously competitive in the server market with 2.5ghz range server chips with 20 power gated cores. That would have been really neat to see and possible if that is the direction AMD had decided to go.
 
here is the full number from the comments section, i believe these were the overclocked numbers of nearly 5 gig but i could be wrong.

hokiealumnus said:
I had promised folding results to Shelnutt2, but they couldn't make the review. So, they will be posted in the first post!

Regular SMP work unit - 13698.9ppd

Bigadv work unit - 13859.2ppd

So between 13,500 and 14,000 at stock, which is right where it's positioned - around the PPD of a 2500K.
Click to expand...
 
R-Type said:
Power draw is sky high on these, but also keep in mind that all power saving states are typically off for those tests. If you aren't folding those are likely to buy you back quite a bit of that delta.
Click to expand...
The power saving features work fine, which is why idle power consumption is very low. The problem is when the CPU is under load.
 
The power saving features work fine, which is why idle power consumption is very low. The problem is when the CPU is under load.
Click to expand...
So that makes it attractive if you don't do anything with it. Yippee. If you are a Folder, then....
 
R-Type said:
Here's hoping the Piledriver launching next year brings the sort of improvement we saw from Phenom to Phenom II.
Click to expand...
It's actually really funny you said that, because that is the exact same story we've been hearing from AMD since August 2006 when the Core 2 Duo launched.

This feels just like the Barcelona launch. Cool new architecture, but the IPC is down, power consumption is too high, and clock speeds aren't high enough to compete with Intel.

And don't look now AMD, but there's an Intel beatdown on its way and it has Sandy Bridge E written all over it.

That being said, I'm still excited to get my hands on my 8120. Maybe, just maybe, with enough of an overclock, it will be able to match the 20k PPD an i7 920 gets on bigadv :rolleyes::p
 
I don't think I'm going to upgrade the Thuban I just bought. I will likely wait for Piledriver.
 
I don't think I'm going to upgrade the Thuban I just bought. I will likely wait for Piledriver.
Click to expand...
Why, sir? Why not "wait" for Sandy E or Ivy? Do you really thing AMD has tricks up their sleeve to make Piledriver into what BD should have been? I just don't see your logic.
 
Linden said:
Why, sir? Why not "wait" for Sandy E or Ivy? Do you really thing AMD has tricks up their sleeve to make Piledriver into what BD should have been? I just don't see your logic.
Click to expand...
because he's already got a Thuban rig, it will likely be cheaper & easier to upgrade to an AMD CPU in the future.
 
Well, OK, I just have a feeling it would be better to cut one's losses and move on. The system parts will all be good for another platform except the motherboard, right? I'm assuming the motherboard wasn't a $250 purchase, was it? The Thuban + MB should fetch a decent price on the trading forum.
 
Linden said:
Do you really thing AMD has tricks up their sleeve to make Piledriver into what BD should have been? I just don't see your logic.
Click to expand...
I just said I was going to wait. What is wrong with that logic? Sorry I wasn't specific enough for you. I will WAIT AND SEE what happens with piledriver. How is that, clear enough now?

What the fuck, go troll someone else.
 
Sorry I hit a nerve. My bad, I'd be sensitive, too. Yes, I understand now, I'm waiting, too.
 
Linden said:
Why, sir? Why not "wait" for Sandy E or Ivy? Do you really thing AMD has tricks up their sleeve to make Piledriver into what BD should have been? I just don't see your logic.
Click to expand...

There is a certain logic to the expectation that Piledriver will substantially improve the faults in Bulldozer, particuarly if you are familiar with project management. Bulldozer is very late, and AMD certainly knew the performance was lacking well before that. Unfortunately, once initial design is done you are limited in the scope of changes you can make without essentially restarting the project. I think it would be conservative to say they knew the BD performance was not going to hit expectations at least two years ago and began taking actions with the followup Piledriver to fix it. We now just have to see if this project was managed more effectively...
 
windows home 64. 80 to 90% of all the work i get is 6900 on 3 i7's ill assum so does everyone ealse. 2684 is 22k a day 2685 is 24k a day.
 
Last edited:
You must log in or register to reply here.
Back
Top