• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

AMD Bulldozer Finally!

AMD will need a miracle chip to recover from this to try and appeal to it's gamers/performance orientated crown but unfortunately that will never happen I don't think.
 
If Drunkenmaster was telling the truth, my understanding is that BD was never a volume chip. This was just an interim CPU, to be succeeded very quickly by the version2 (pile driver or Trinity...I really don't know what names these CPUs are).

IF the above is correct, then the fact that BD is a little slow, isn't too big a deal. They should be able to make sales with correct marketing (ie. focus on the "8 cores...more cores than any other CPU). When they are ready to release BD2, then I'm sure they can get back on the pace. HOWEVER, they might be in danger of running into Ivybridge (which could be a major problem).

IMO, if they drop the prices of BD, they can sell good quantities of this CPU. Price, as always though, is one of the most important factors when it comes to sales. If they can lower their BD prices below 2500K, I think this would be a pretty good CPU.

My understanding is that the only Intel CPU which can beat the 8150 is their own Phenom and the 2500K (and upwards). The lower end Intel CPUs are inferior to the 8150.
 
Yes, they intended for this fail launch ;). All that marketing hype, the bringing back the FX name the "AMD unlocked" lark, all leading up to Piledriver, give me a break. I suppose Llano bad yields are also a lead up to Piledriver?
PileDriver isn't going to pull a conroe I don't think, at least given AMD's own performance increase figures which IIRC, are 20% over Zambezi. For all we know, it's all clock speed increase, or all IPC increase, probably a bit of both.

On the flipside, Ivybridge is only going to be a few percent higher clock than SB and a few percent better IPC IIRC, it's not going to be any substantial improvement.

Beating the 8150 depends on the situation.
An E5400 will wipe the floor with the FX 8150 in PS2 emulation for example.. The ever popular Rome Total War will perform better on an E5400 etc.
People have a lot of uses etc, but it's not just the 2500k that beats it, or the Phenom II's at all.

DM cares too much for theory, fair play to him, but knowing the theory is completely different to knowing the practicality. Take BD for example, the modules, I said about two threads being ran off a module would cause a performance hit, DM said I wasn't understanding it right. BD launches? Oh look, the two threads being executed from BD does give a performance loss in scaling between the module. The scheduling "patch" should improve situations in applications that utilise 2-6 threads.

Knowing/Speaking the theory like DM does, gives the advantage of being able to, I suppose, describe what you're saying, which is where I fall down, I can't tell you why something is the way it is, just that it is that way.

Although, the scaling was worse than I'd suggested.
 
Last edited:
If anything was an interim CPU it was the Phenom X6, BD was just delayed for far too long and even if it had been released 2 years ago the low IPC would have meant that it still lost to the CPU's at the time in a lot of tests.
 
mmj_uk said:
If anything was an interim CPU it was the Phenom X6, BD was just delayed for far too long and even if it had been released 2 years ago the low IPC would have meant that it still lost to the CPU's at the time in a lot of tests.
Click to expand...

Indeed.
The FX4 gets ravaged by the Q6600 at the same clock, and that CPU is edging, on over 4 years old.
I always thought that Thuban was meant to be the stop gap to Bulldozer, as Bulldozer was meant to be out a while ago.
 
Martini1991 said:
And their bizarre opinions :p.
So the FX8150 is faster than the Thuban in highly threaded things, check.
The FX6 and FX4 which actually replace current Phenom II products at the price section perform worse.
Click to expand...
not if clock for clock

anything over 4.5ghz than yes in highly threaded apps
 
Last edited:
The cinebench is an anomaly I think (As I said before, I think it's a case of the architecture falling flat on its face, simply to how bogged down Cinebench is making it), as you noticed slightly faster encoding?
 
Thuban was actually a pretty decent idea, the MOAR COARS played well for AMD considering the 980x's cost even if performance was in another league.

BD is just embarrassing being unable to out sprint the 1100T in a number of cases. Last time I remember that was the initial launch of the P4 where the P3 had the upper hand on the slower clocked cpu's.
 
Completely gutted this release to be honest.

I do not understand what happened, did they just fail at some point in development massively and decided 'release it anyway' as the optimal way to cut losses, trying to ride it out on the 8-cores slogan for the masses ?

Perhaps some unexpected problem they could not overcome (to acheive the desired performance) delays and all. 'Hush-hush' everything of course.

Sad times.
 
Last edited:
Martini1991 said:
The cinebench is an anomaly I think (As I said before, I think it's a case of the architecture falling flat on its face, simply to how bogged down Cinebench is making it), as you noticed slightly faster encoding?
Click to expand...
not clock for clock
http://i559.photobucket.com/albums/ss33/CAT-THE-FIFTH/HandBrakeChart17-10-2011.png

these tests below i did for CAT-THE-FIFTH
CAT-THE-FIFTH said:
Thanks!

The Tron trailer encode speed is 37.41FPS which puts in between a Phenom II X6 1100T(33.74FPS) and a Xeon E3-1270(39.4FPS). The Xeon E3-1270 is essentially the same CPU as a Core i7 2600K but with the IGP switched off. However,looking at table a Phenom II X6 at similar speeds would be around the same.
Click to expand...

CAT-THE-FIFTH said:
Thanks again!!

23.976FPS with the Splice Trailer. This puts it slightly above a Phenom II X6 1090T. Again it looks like a 3.6GHZ Phenom II X6 would produce a similar result.
Click to expand...

CAT-THE-FIFTH said:
The Big Buck Bunny encode is 37.11FPS which puts it half way between a Phenom II X6 1100T and a Xeon E3-1270. So again it looks like a 3.6GHZ Phenom II X6 should be around the same speed.

Thanks again.
Click to expand...
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom