Bulldozer – a funny review but appalling power efficiency.

All about them.

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee

HFat
Posts: 1753
Joined: Thu Jul 03, 2008 4:27 am
Location: Switzerland

Re: Bulldozer – a funny review but appalling power efficienc

Post by HFat » Mon Oct 24, 2011 1:51 am

quest_for_silence wrote:a market isn't just its supply side.
Of course not. In this case, there are many buyers alongside will all the other prerequisites. In this case, the market fails on the supply side.
As you point out, there's little that's extraordinary about it. Anti-competitive patent priviledges play a more important role in this industry than in some others but markets do need regulation, yes.
quest_for_silence wrote:There's a market whenever there's a demand of some goods
A potential for a market maybe (in some contexts, that's what people mean when they say market) but we're not talking about potentialities but the actual situation which isn't going to change fundamentally any time soon. You can also imagine a potential market where there is no demand by the way.

andyb
Patron of SPCR
Posts: 3307
Joined: Wed Dec 15, 2004 12:00 pm
Location: Essex, England

Re: Bulldozer – a funny review but appalling power efficienc

Post by andyb » Thu Oct 27, 2011 5:27 pm


andyb
Patron of SPCR
Posts: 3307
Joined: Wed Dec 15, 2004 12:00 pm
Location: Essex, England

Re: Bulldozer – a funny review but appalling power efficienc

Post by andyb » Thu Oct 27, 2011 5:30 pm

Second Ouch - or should I say third........ I am talking about "Trinity" of course :roll:

http://www.xbitlabs.com/news/cpu/displa ... rease.html

You will understand why if you read the details rather than the headline.


Andy

andyb
Patron of SPCR
Posts: 3307
Joined: Wed Dec 15, 2004 12:00 pm
Location: Essex, England

Re: Bulldozer – a funny review but appalling power efficienc

Post by andyb » Thu Oct 27, 2011 5:35 pm

I am not reading this now its 2:30 AM FFS - bed time.

http://techreport.com/articles.x/21865

Does look interesting though and this has been discussed in this thread already so this is spot on subject.


Andy

Mr Spocko
Posts: 229
Joined: Sun Jul 10, 2011 1:56 pm
Location: UK/Eire

Re: Bulldozer – a funny review but appalling power efficienc

Post by Mr Spocko » Fri Oct 28, 2011 12:09 am

Trinity isn't that interesting it's another new socket (FM2) promising BD design (which is doubtful as it is) and higher clock speeds.
I'm dead set against a dual socket approach.

Back to FX it's a turkey power consumption wise as the article shows they even said if AMD used the older processor design on 32nm that would have easily delivered better power consumption and performance. Meantime no signs of a major price cut so AMD are expecting folks to buy underperforming overpriced processors and yet again "wait" for piledriver.

I'm just not convinced with the FX processor design it's deeply flawed and reliant on high clock speeds to try to boost performance. AMD can show as many slideshows as they want if the design is poor it's not going to help. We want more work done per core, lower clock speeds and less power consumption.

quest_for_silence
Posts: 5275
Joined: Wed Jun 13, 2007 10:12 am
Location: ITALY

Re: Bulldozer – a funny review but appalling power efficienc

Post by quest_for_silence » Fri Oct 28, 2011 12:44 am

Mr Spocko wrote:lower clock speeds

Lower? They're already under Westmere and Sandy Bridge!!!

Mr Spocko
Posts: 229
Joined: Sun Jul 10, 2011 1:56 pm
Location: UK/Eire

Re: Bulldozer – a funny review but appalling power efficienc

Post by Mr Spocko » Fri Oct 28, 2011 1:18 am

I'm looking back to Athlon XP Athlon 64 days.
Yes they should have more performance per clock cycle with reduced clock speed v their current speeds

AMD should know only too well that FX is not optimal and is poorly designed it simply doesn't have the legs to run for the next few years. The performance increase on FX by overlclocking is frankly pathetic so doing that won't give them the boost they need. A 4 core FX overclocked to hell can't even match an Athlon II x4 running stock speeds in many applications.

andyb
Patron of SPCR
Posts: 3307
Joined: Wed Dec 15, 2004 12:00 pm
Location: Essex, England

Re: Bulldozer – a funny review but appalling power efficienc

Post by andyb » Fri Oct 28, 2011 11:09 am

I have read all of the article and the posts on forum about disabling one core in each of the Bulldozer modules, and comparing it in benchmarks with half of the modules disabled - the difference is as expected - it does show some promise for lowly threaded apps, and almost as important (from the technical standpoint) it highlights an area that could increase the BD line of CPU's dramatically in the future if that "problem" was sorted out.

Here is the article in question.

http://techreport.com/articles.x/21865

XtremeSystems thread is interesting.

http://www.xtremesystems.org/forums/sho ... eaded-Perf

And here is an article in French that is also interesting and looks at the same problem with a different set of benchmarks.

http://www.hardware.fr/articles/842-9/e ... e-cmt.html

Converted to English via Google Translate.

http://translate.googleusercontent.com/ ... FCrUfkpGQA

The "problem" appears to be "the scheduler" or possibly the entire "front end" of each module. In a perfect world the scheduler (and other front end components) would allow the FPU and both integer units to be as fast with full load on all 3 computational units simultaneously as each one is individually with no use at all on the other two units - in reality this is not possible, often because of the cache and moving threads around (cache data movements), however as you will see from the previous links there are some serious gains to be made from using one core per module vs two cores per module but half the amount of modules even though the latter employs full turbo-boost and has an obvious clock-speed advantage - this is the point to be highlighted, basically when 2-cores on a module are working hard they are a lot slower than 2cores in 2 different modules.

As you would imagine this is an obvious problem with the design of BD, and is why some people say that the BD design is basically hyper-threading - HT it is not, it has full-blown computational units - but shares the front end, as these tests have proven, the front end is simply not adequate to keep both integer cores (and the FPU as well.?) working efficiently - that much is obvious from all of the tests - but it does look like the BD architecture was "originally designed" to save die space by removing front end components whilst NOT reducing the per-core performance - this is what all of their announcements over the years stated until about 12-months ago - this may be wishful thinking (probably is) but it does stand to reason that there is a problem with the CPU that was encountered that was not expected and has as yet not been overcome.

If AMD closed the gap that has been shown from running a single core per module vs two cores per module by half, that would be an instant gain of 5% when running on only half of the cores, so in theory could get a 10% gain when all 8-cores are active - that is a pretty big gain to be made without making a gigantic change to the design (at least I don't think that it is/should be a massive design change - more of an iron out the bugs, or beef up the front end a little).

There is one other possibility that has popped up a few times (on forums), there might be (not a shred of evidence as yet - only speculation) a colossal cache cock-up that is constantly moving data from the L1 cache of one core in a module to the other core in the same module via the L2 cache - as you can imagine this is hammering the L1 and L2 cache that will affect both cores (and due to the cache abuse the FPU as well.???), this would also is meaning that the cores spend more time waiting for data and less time processing it (whilst still appearing to be running at 100%). There is as yet no evidence - only speculation, if however this turns out to be true, this should also leverage a large performance bonus when it is rectified - and should be doable without a great deal of work on the CPU design itself and could be fixed with a newer stepping (Phenom 1 anybody).

If any of the above is true and AMD remove a digit from their anus and deals with this, BD might still be a contender in performance terms, a price drop will still be needed for price/performance, and it might even make a drop in power consumption possible as well if it does turn out to be a cache-thrashing issue as cache can suck a lot of power (and would explain why there is such a huge difference in power use between idle and load), clock-speeds will still need to go up, and the single threaded performance would hopefully be brought to the level of the PII's - fingers crossed - which overall would make a future BD CPU an attractive option for many people.


Andy

Mr Spocko
Posts: 229
Joined: Sun Jul 10, 2011 1:56 pm
Location: UK/Eire

Re: Bulldozer – a funny review but appalling power efficienc

Post by Mr Spocko » Thu Nov 03, 2011 10:04 am

The CPU was in development for years so AMD should have been well aware of potential problems.
IMO the design is flawed so really they'll have to make significant modificantions to that to deliver better performance and power efficiency

No real excuses on this one from AMD if it were a 6 months mad dash I'd be more willing to accept that

1337
Posts: 41
Joined: Sat Sep 11, 2010 6:47 am
Location: EU

Re: Bulldozer – a funny review but appalling power efficienc

Post by 1337 » Sat Nov 05, 2011 4:12 am

The new B3 stepping will likely address some of the issues but it won't be a game-changer. Sadly. It's hard to recommend it over SB.

Mr Spocko
Posts: 229
Joined: Sun Jul 10, 2011 1:56 pm
Location: UK/Eire

Re: Bulldozer – a funny review but appalling power efficienc

Post by Mr Spocko » Sat Nov 19, 2011 1:47 pm

No it might not be a game changer but B3 will have to be quite a decent revision because if it's not then that will not go down well!
If they can tweak it to improve performance (at least a decent bit) but more importantly reduce power consumption a good bit they're in with a shout.
It's a bit frustrating to be "FX ready" build wise only to discover FX Mk I was a lot less than I expected.

I honestly didn't expect it to beat up sandybridge, but I did expect good power consumption.

1337
Posts: 41
Joined: Sat Sep 11, 2010 6:47 am
Location: EU

Re: Bulldozer – a funny review but appalling power efficienc

Post by 1337 » Tue Nov 22, 2011 5:54 pm

Mr Spocko wrote:No it might not be a game changer but B3 will have to be quite a decent revision because if it's not then that will not go down well!
If they can tweak it to improve performance (at least a decent bit) but more importantly reduce power consumption a good bit they're in with a shout.
It's a bit frustrating to be "FX ready" build wise only to discover FX Mk I was a lot less than I expected.

I honestly didn't expect it to beat up sandybridge, but I did expect good power consumption.
The server parts based on the Bulldozer tech have recently been released. And according to the datasheets that I have seen, are very power-efficient. Especially the "Energy Efficient" series.

In my opinion, 50% of the FX problems lie in manufacturing (i.e. GLOBALFOUNDRIES). The FX should get better as the manufacturing matures. By how much? time will tell :)

Until then, we're buying Thuban based chips :D

zefiris
Posts: 19
Joined: Sat Jun 05, 2010 5:51 pm
Location: USA

Re: Bulldozer – a funny review but appalling power efficienc

Post by zefiris » Sun Nov 27, 2011 4:42 pm

my opinion about new CPUs is that the first one in the series is usually pretty crappy while it gets improved with its successors.

Almost always the previous gen chip can usually match or almost match the processing power of the new chip.... (ie Intels first Dual core vs the late P4... not much of a difference, same with AMD)

So I expect AMD to pick up with their with their successors of Bulldozer, and being first on the market to have that many cores I expect AMD to start kicking butt soon or at least matching Intel until they get familiar with their new chip.

Mr Spocko
Posts: 229
Joined: Sun Jul 10, 2011 1:56 pm
Location: UK/Eire

Re: Bulldozer – a funny review but appalling power efficienc

Post by Mr Spocko » Mon Jan 02, 2012 8:57 am

More odd stuff from AMD
This time they're supposed to be working on Phenom II branded CPU's that are cut down FX processors

http://www.xbitlabs.com/news/cpu/displa ... ssors.html

Unless they have some decent performance improvements and better power consumption (and a low price) I can't see how this is going to help AMD

lodestar
Posts: 1683
Joined: Fri Aug 05, 2005 3:29 am
Location: UK

Re: Bulldozer – a funny review but appalling power efficienc

Post by lodestar » Mon Jan 02, 2012 12:00 pm

Even odder are the new Athlon II X4 631 and X4 651 socket FM1 CPUs. This are nothing more than Fusion APUs with the graphics unit disabled. Disabled that is in terms of producing any graphics but not in consuming power for no good purpose at all. So the 631 and 651 are 100w TDP parts like their APU counterparts. In the UK at least, the X4 631 2.6Ghz is priced at about the same price as the Intel G850 2.9Ghz Sandy Bridge Pentium. And the G850 does of course have a working graphics unit.

That said, I suspect that some of the Bulldozer CPUs are now selling reasonably well although this will be in part because the older Athlon and Phenom ranges have almost, if not totally disappeared from the retail channel. So I can understand in a way why AMD would have wanted to hold onto the Athlon and Phenom names, but it does seem a rather strange way of doing it.

Post Reply