The more cores the better for me. 6 or 8 is better then 4 for me, even if they are individually slower
-
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!
- Thread starter AceTK
- Start date
The more cores the better for me. 6 or 8 is better then 4 for me, even if they are individually slower
Caporegime
- Joined
- 18 Sep 2009
- Posts
- 30,521
- Location
- Dormanstown.
AMD in market position mistake shocker. Anyone would think this semiconductor business was hard or something
I'm not calling it a mistake.
If the benchies above are true then BD is a mistake IMO. Phenom II seems to be faster per clock bar heavily threaded workloads (I.E. not desktop). Would a die shrink Phenom X4/X6 with a much higher stock frequency and maybe a more aggressive turbo be better all round?
Only a few days to find out though. I hold a little hope that these are all wrong and BD will be a monster of a chip. Only a little hope, mind
If for gaming and no need for heavily threaded apps. Then yeh, the 2500k is an awesome chip, complete with an IGP.
If you do lots of work that involve many threads, then depending on price, the bulldozer could be your choice.
My question is, how would a 2500k/2600k perform video encoding with Lucid Virtu using the IGP compared to a bulldozer thats missing the IGP?
Would the SBs still be the video encoding choice?
Caporegime
- Joined
- 18 Sep 2009
- Posts
- 30,521
- Location
- Dormanstown.
I think it'll shine in the future, but right now? Not really.
Isn't it that sync Z68 feature that uses the GPU?
Associate
- Joined
- 2 Feb 2010
- Posts
- 345
- Location
- London
If this is true then it doesn't mean AMD made a mistake...
I am no fan boy and have actually have a 2500k. But I can see that AMD needed bulldozer to set up a new generation of architecture.
You have to remember that:
A) These things overclock far better that 1090t's (almost 1ghz higher if the leaks are to be believed).
B) The Bulldozer architecture is only going to get sweeter as more revisions generation are released.
I imagine my next CPU could likely be an AMD for this reason....
I am no fan boy and have actually have a 2500k. But I can see that AMD needed bulldozer to set up a new generation of architecture.
You have to remember that:
A) These things overclock far better that 1090t's (almost 1ghz higher if the leaks are to be believed).
B) The Bulldozer architecture is only going to get sweeter as more revisions generation are released.
I imagine my next CPU could likely be an AMD for this reason....
Caporegime
- Joined
- 18 Sep 2009
- Posts
- 30,521
- Location
- Dormanstown.
The architecture should make it easier to "Throw cores at it" for the future.
However, I'm sure Intel aren't going to keep releasing quads as mainstream.
Once hex,octo become mainstream, you can bet Intel will be on it too.
If Software gets a massive boost in utilising cores, it could catch Intel underhand, and AMD could possibly ramp out some CPU's with x amounts of modules quite easily you'd hope.
However, I'm sure Intel aren't going to keep releasing quads as mainstream.
Once hex,octo become mainstream, you can bet Intel will be on it too.
If Software gets a massive boost in utilising cores, it could catch Intel underhand, and AMD could possibly ramp out some CPU's with x amounts of modules quite easily you'd hope.
Well that's just marketing isn't it, 8 core obviously sounds a lot better than 4 core if you're doing your shopping in PC World.
8 core BD much closer resembles a quad core with a type of 'Hyperthreading on steroids' than it does a true 8 core.
Caporegime
- Joined
- 18 Sep 2009
- Posts
- 30,521
- Location
- Dormanstown.
No it doesn't.
BD's cores much more resembles cores than an Intel 2600k core resembles two cores.
The only thing about these cores, is they share some resources.
But they're not 8 individual true cores if that's what you're getting it.
Also, a module isn't limited to two cores IIRC.
I think they should have called it a 4 core personally (Module = core). But it's not, and so thinking of it as a quad core is weird.
B) does not help the user who parts cash for the first revision though
I see what you allude to though
Caporegime
- Joined
- 18 Sep 2009
- Posts
- 30,521
- Location
- Dormanstown.
Lol, that one obviously means nothing 
Lucid Virtu lets you use a Sandy Bridge IGP for things like QuickSync whilst using a separate GPU. So if you have a Z68 motherboard and a separate GPU you get to have an awesome gaming and number crunching machine (overclocking) and also a mean encoder (QuickSync).
Sure it does,especially if you are upgrading like I'm from socket 939 CPU
,end of the day there is always something faster down the road,that's nature of technology.
Caporegime
- Joined
- 18 Sep 2009
- Posts
- 30,521
- Location
- Dormanstown.
Cheers for the elaboration.
Kind of helps for a 2500k to stay on top of the 8150 in apps that work with QS.
It is a bit of a reversal for AMD. Remember all the original marketing from AMD that they had the only true quad core design, whereas Intel just stuck 2 dual cores together to make quad core Kentsfield. Trouble for AMD was, Intel's approach actually worked very well indeed. Now AMD themselves don't seem to have "true" 6 core and 8 core designs anymore with Bulldozer. All depends on what you are actually doing with the CPU
Caporegime
- Joined
- 18 Sep 2009
- Posts
- 30,521
- Location
- Dormanstown.
Yeah, they've basically made an 8 core consisting of 4 dual cores
.Although, the "True quad core" was just marketing, the Q6600 destroyed the Phenom's.
Yeah, they've basically made an 8 core consisting of 4 dual cores
But they aren't even true dual cores either, they've been cut down into some kind of single/dual core hybrid.
Caporegime
- Joined
- 18 Sep 2009
- Posts
- 30,521
- Location
- Dormanstown.
I don't see why people are getting their knickers in a twist. There's obviously something funky going on in some benchmarks where performance isn't correct.
Given that for the CH V board, newest BIOS is 3 versions ahead of publicly released (and half these results are on 2 versions ahead), AMD are obviously still working feverishly to optimise things in certain usage scenarios and generally tune it.
The fact that AMD claim (in the leaked slides) very quick x264 & Handbrake performance (similar kinds of instructions / parts of the processor used to - for example - Cinebench) tends to suggest that once either the software or BIOS are tuned, you'll see much better performance than some of these anomalies indicate. That said, the SuperPi numbers might be as good as it gets ... but who cares.
If 'IPC' was so much lower in all usage scenarios, you wouldn't see decent performance in games (which they also claim).
Given that for the CH V board, newest BIOS is 3 versions ahead of publicly released (and half these results are on 2 versions ahead), AMD are obviously still working feverishly to optimise things in certain usage scenarios and generally tune it.
The fact that AMD claim (in the leaked slides) very quick x264 & Handbrake performance (similar kinds of instructions / parts of the processor used to - for example - Cinebench) tends to suggest that once either the software or BIOS are tuned, you'll see much better performance than some of these anomalies indicate. That said, the SuperPi numbers might be as good as it gets ... but who cares.
If 'IPC' was so much lower in all usage scenarios, you wouldn't see decent performance in games (which they also claim).