AMD Bulldozer Finally!

[Helios1234]

how they convincing us that bulldozer is good? Helios?, with those slides i dont understand specially cinebench slide, 4.8 ghz bulldozer scores what was it 5,8 pts and 1090t @ 4ghz can score more then 7pts in my case 7.11pts.

No according to that leaked slide, it scored 5.95 at stock and more when overclocked to 4.8GHz. It's merely showing how overclocking scales in performance.

Anyway, those leaked slides were never meant to be out in public so they aren't supposed to be convincing anyone anything. God know how sites like Donanimhaber get hold of these information.

 

[psychas]

oh yea my bad it gets 7.8 pts on 4.8 ghz and 5.95 pts at stock. still should be bit higher

 

[gareth170]

Well you're saying reviewers can't be trusted. I'm merely saying that they are no different to a regular user and as such they are no less trustworthy than anyone else.

but users results cover awide range of rigs / hardware and if you've asked a group of users for results with similar hardware you'll clearly see whats true.

 

[Final8y]

Well you're saying reviewers can't be trusted. I'm merely saying that they are no different to a regular user and as such they are no less trustworthy than anyone else.



That's always been how marketing works hence me saying that we should probably wait for the reviews to see how it really performs. Even now, no one can really say for sure what it will be like. I've seen a couple of benchmarks myself but that's hardly anything to go by.

I have been watching AMD long enough to notice that this is stretching it for them this round as they know full well that the enthusiasts will look and take note first and they expect us not to notice the contradiction as if Intel don't offer better value than that Intel 980 chip and people could buy the cheaper Intel's instead & save a bundle and get a better gfx cards.
It should be against the 2600/2500 in regards to gfx performance in conjunction with the GPU.

Well maybe they think this marketing has worked for NV so why not give it ago.

 

[gareth170]

oh yea my bad it gets 7.8 pts on 4.8 ghz and 5.95 pts at stock. still should be bit higher

at 3.6ghz i get 6.38 pts,,, which is 8150 stock clock

 

[Martini1991]

Compare an FX6100 in Fritz at 3.3GHZ against an 1100T, the results will shock you more than likely.

 

[gareth170]

i can only think it's because they using Os X?

 

[Martini1991]

i can only think it's because they using Os X?

Or, BD simply isn't what was expected of it.

 

[opethdisciple]

What does the difference between 95w and 120w make to overclockability of a chip?

 

[Sin_Chase]

What does the difference between 95w and 120w make to overclockability of a chip?

I don't fully understand all of it myself but I believe the general consensus is that within the same family the lower the better/more overclockability.

I think

 

[Martini1991]

What does the difference between 95w and 120w make to overclockability of a chip?

It depends.
95W could denote higher binned.

But AMD give "Worst case scenario".

 

[randal]

But then again you still have not removed it and the best that you could come up with was this, clearly you don't understand.

I know exactly what you're talking about, stop being pedantic.

For what it's worth, I've asked a mod to delete the lot. So leave off.

 

[Martini1991]

I know exactly what you're talking about, stop being pedantic.

For what it's worth, I've asked a mod to delete the lot. So leave off.

 

[LeJosh]

What does the difference between 95w and 120w make to overclockability of a chip?

I imagine it will allow all modules to run at the specified clock and make it more reliable due to all modules having a higher % of watts.

I actually don't know the proper details but still, an educated guess.

 

[muon]

Compare an FX6100 in Fritz at 3.3GHZ against an 1100T, the results will shock you more than likely.

Problem is 6 genuine cores vs 3 modules.

FX-6100 isn't intended to replace 1100T and so comparison is pointless.

 

[kylew]

Problem is 6 genuine cores vs 3 modules.

FX-6100 isn't intended to replace 1100T and so comparison is pointless.

If they sit in the same price bracket, comparing isn't pointless at all.

 

[Juantjie]

Isn't the most important thing the price/performance compared to the 2500k? The rest is just cross fingers and hope for the best.

 

[muon]

If they sit in the same price bracket, comparing isn't pointless at all.

That's my point, they wont?

1100T is £150 on ocuk ($189 in states). The FX-8100 could easily end up at that.

 

[LeJosh]

If they sit in the same price bracket, comparing isn't pointless at all.

1st generation of a new architecture will always skew the results comparing it to the last.

Which gen was the phenom 2 series?

I can just see good things to come with BD especially with their new instruction set and I LOVE to emulate older games.

Plus I want to have an eyefinity set-up one day, if their CPU handles mega-resolutions better, then great.

Intel have been in the game a very long time and AMD obviously have some good engineers who I understand used to work for Intel and vice versa.

The most exciting thing about Intel is their new criss-cross transistors which should be in Ivy Bridge. But I reckon they'll go the same route as AMD or similar with modules with more than 1 thread. Just seems like natural progression of the CPU until silicon stacks are viable.

 

[Martini1991]

Problem is 6 genuine cores vs 3 modules.

FX-6100 isn't intended to replace 1100T and so comparison is pointless.

So, BD is an improvement, or isn't an improvement?
The 6100FX will still be called a hexcore.
Compare it to a 1090T, whatever, or a 1055T. If there's no IPC increase, it's still a pointless CPU.

I can't believe that people are bringing in the whole "IT'S NOT 6 REAL CORES" because BD isn't going to perform as they expected.