• 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!

Martini1991 said:
Two different things?
What he said is a wall of text, what I was analysing were actual results.. And it's basically a little faster clock for clock than Deneb, 8 cored and 32nm.
And on the O/S front? That's funny, I was running an old Athlon x2 6000+ on my SSD, and noticed zero difference with my 4.8GHZ 2500k or 4.3GHZ Phenom II 1055.

The way to see which is the best architecture, is a face off with the 4 core BD CPU and the 2500k, put them at the same clock, then overclock both to the maximum.
Click to expand...

His 'wall of text' explained exactly how and why SuperPi is totally irrelevant in the real world in terms of performance. You weren't really analysing any results at all.

I suspect you're exaggerating rather heavily re: noticing no difference in general responsiveness and speed between an Athlon x2 6000+ and either of the latter.
 
pmc25 said:
His 'wall of text' explained exactly how and why SuperPi is totally irrelevant in the real world in terms of performance. You weren't really analysing any results at all.

I suspect you're exaggerating rather heavily re: noticing no difference in general responsiveness and speed between an Athlon x2 6000+ and either of the latter.
Click to expand...


In the O/S, no, I'm not over exaggerating.. When doing something, such as extracting, the difference was painfully noticeable.

I never even said SuperPi was good, but there's many of us who need higher clock performance, emulators etc all only use X amount of cores, so do games for christs sake. That BD sample will get outpaced by a 2120 in Crysis for example.
 
Martini1991 said:
The way to see which is the best architecture, is a face off with the 4 core BD CPU and the 2500k, put them at the same clock, then overclock both to the maximum.
Click to expand...
True that tells you what the best architecture is but to see which is the best CPU to buy, clock-for-clock performance is irrelevant. You need to compare two CPUs at the same price at both stock and maximum expected overclock.

It's pretty obvious that Zambezi will beat Sandy Bridge at multi-threaded stuff in this price-for-price case (since the FX-8130P is meant to be around the same price as the i7-2600K) judging by recent overclocking results. Single-threaded is the thing we don't know much, if anything, about.
 
Martini1991 said:
That only makes BD look crap lol.


Single core benchmark, BD, unlike Phenom II has the same instruction sets as Intel CPU's, why shouldn't it be viable?
Click to expand...

SuperPI using Legacy code which Intel is much better at then AMD & always has been.
Even when Athlon was king over the P4 the P4 would always wipe the floor with AMD in SuperPI.

Having the same instruction set for the most part but because of architecture differences the CPU does the same job but goes about it differently will always lead a CPU to be better at specific instructions over another.

And would be no different to a AMD GPU system & a NV GPU system running enhances PhysX games, both systems can run the PhysX instructions but the NV GPU system will swipe the floor with the AMD GPU system even when the AMD GPU system is significantly more powerful because the NV GPU system handles the PhysX instructions differently.
 
Last edited:
Martini1991 said:
Two different things?
What he said is a wall of text, what I was analysing were actual results.. And it's basically a little faster clock for clock than Deneb, 8 cored and 32nm.
And on the O/S front? That's funny, I was running an old Athlon x2 6000+ on my SSD, and noticed zero difference with my 4.8GHZ 2500k or 4.3GHZ Phenom II 1055.

The way to see which is the best architecture, is a face off with the 4 core BD CPU and the 2500k, put them at the same clock, then overclock both to the maximum.
Click to expand...

The problem is you cant compare the two as the fetch,decode is common between the cores in one module but the execute instruction splits into two separate branches, so you would have to disable half of one module on each core and then compare the half disabled quad module processor to the 2500k
 
fish said:
The problem is you cant compare the two as the fetch,decode is common between the cores in one module but the execute instruction splits into two separate branches, so you would have to disable half of one module on each core and then compare the half disabled quad module processor to the 2500k
Click to expand...
Why can't you compare the two? That's just an excuse.
4 v 4.
 
Wow, fanbois-a-plenty in this thread. I'm looking forward to seeing some useful benchmarks (in things that actually matter to me, not random bits of software that I don't and won't ever use).

I would imagine this is going to be the same as the usual AMD thing. It won't be quite as good as Intel's best but it'll a lot cheaper for 95% of the performance. I'd love to be wrong though.
 
Martini1991 said:
Why can't you compare the two? That's just an excuse.
4 v 4.
Click to expand...

Read my post, the architecture Between the two are different. BD is not like a 2600k where you can turn off ht and compare it to a 2500k, to compare apples to apples you must disable on execution unit per module. If you don't do this it is not an even comparison. One cycle is made up of fetch decode execute having a shared first two steps will not result in a 4 v 4, but essentially closer to 4 v 3.
 
marms said:
4 V 4 would be 4 Intel cores that can run 2 threads each against 4 AMD Modules that can run 2 threads each, simple, you just don't like the results.
Click to expand...

But it doesn't work that way.
fish said:
Read my post, the architecture Between the two are different. BD is not like a 2600k where you can turn off ht and compare it to a 2500k, to compare apples to apples you must disable on execution unit per module. If you don't do this it is not an even comparison. One cycle is made up of fetch decode execute having a shared first two steps will not result in a 4 v 4, but essentially closer to 4 v 3.
Click to expand...

Which makes its design it's own worst enemy due to software.
Semantics aside, a program will see a 2 module BD as a quad core, and a 2500k as a quad core, which will win? The 2500k. Nuff said.
 
Martini1991 said:
But it doesn't work that way.


Which makes its design it's own worst enemy due to software.
Semantics aside, a program will see a 2 module BD as a quad core, and a 2500k as a quad core, which will win? The 2500k. Nuff said.
Click to expand...

Hmmm I'm not sure I agree here, the only possible fair way to do a apples to apples test is take the 2 corresponding CPU's whatever that maybe' BD sandy, run every single real world benchmark we can get our hands on (except that sys mark thing that's been written by intel for intel - also remove any benchmarks written by amd for amd)' and then we will end up with sayy 20 odd benchmarks and award a percentage point for every point one CPU beats the other, so if intel does 9sec super pi, amd gets 19sec, that's + X % increase which can be converted to a numerical figure for intel, or amd gets 2000 cpu Mark points and intel gets 1500, x points for amd and so on.

Anyways whatever way you want to work out a fair points scheme and total all the points up, one will win OVERALL, quoting just one benchmark is silly as its only exploiting one tiny part if the CPU's ability and what it has to offer.
 
The 2500k will outplay the 2 module BD CPU in practically everything.
The 8130p will hold its ground in applications that can make use of the remaining 4 threads.

Until AMD can get a module to run a thread, then throwing modules into the equation won't solve anything, if they go get a thread to run on a module, AMD will kill intel.
 
MANFACE said:
I'm only interested in this chip for gaming.
Click to expand...

Me to. And going to get this no matter what. My only concern is do I have enough cash for the most powerful FX model. Hope so.

Intel would be great also for gaming but I want something .. hmm.. different. And after all the hype and speculations bulldozers fits my requiments. And eh.. fits also for my natural updating cycle lol.
 
Mollari said:
Me to. And going to get this no matter what.
Click to expand...

Why? Makes no sense.

When you are ready to 'buy' your system, at that point weigh up which is best "bang for buck" and buy accordingly. Don't just buy because you are excited about it, or want to support AMD, buy sensibly :-) (Even if that ends up AMD being the best purchase option :-) )
 
Compleatly new to the building your own computer and buying parts scene.

When can we expect a release for this CPU? I'm in the works of making first desktop and i dunno if i should wait it out for the bulldozer or get a AMD Phenom II X6 1100T.

Thanks.
 
eyyaz said:
Compleatly new to the building your own computer and buying parts scene.

When can we expect a release for this CPU? I'm in the works of making first desktop and i dunno if i should wait it out for the bulldozer or get a AMD Phenom II X6 1100T.

Thanks.
Click to expand...

No point getting the 1100T regardless, you can get an SB set up for the same price.
BD is due anytime in august, 1st-30th
 
Martini1991 said:
No point getting the 1100T regardless, you can get an SB set up for the same price.
BD is due anytime in august, 1st-30th
Click to expand...

Alright thanks, im working on building a gaming rig and just wanna get myself a good CPU.
Got a Radeon HD 6790 graphics card, 4gb DDR3 ram and ASUS Crosshair IV Extreme Desktop MB.
 
You must log in or register to reply here.
Back
Top Bottom