Discussion in 'Article Discussion' started by Lizard, 12 Oct 2011.
Hitler finds out about the AMD Bulldozer benchmarks:
Well a few people I know have said that the FX-8150 would have to be dropped down to Â£150 - Â£180 to make it somewhat easier to consider using.
I am just waiting now for Sandy Bridge Extreme, and Ivy Bridge and then no doubt there will be a Ivy Bridge Extreme.
My 2500K does the job nicely but when I am ready to upgrade there will atleast be something better on the market by then, but at the moment there is nothing better other than the 2600K and the 2700K when released.
Commenting on reality is hardly bashing.
I was looking forward to going back to Bulldozer but it looks like that's off the cards. Unless something went horribly wrong with the testing.
That said the other gaming benchies that someone posted on here seem to put it in a better light than the BT single set of results, with regards to what I need a PC for.
I need more info though. I need to confirm it is a poor as it appears and that the power consumption is as whack as BT reports. Mind you if it was a £100 chip it'd be fantastic.
fool me once, shame on you.
fool me twice, shame on me.
Clutching at straws maybe, but maybe Bit-tech should investigate whether these chips suffer from the TLB patch problem? Given that it is an AM3+ this would not surprise me. My Phenom had this on by default and by turning it off, via BIOS and a Windows script, it improved things, even if just slightly.
Nah, probably clutching at straws.
No reason to upgrade for a while now.
Considering it plays BF3 BETA maxed out and BC2 maxed out you're just a troll!
Or even £150 but I'd still rather have the 2500k I have.
Was actually expecting something awesome after all the waiting.
You hear that Jeff? All those months waiting for Bulldozer, and all you were really doing was waiting to troll. Shame on you!
AMD marketing fail.
if they sold it as quad core and at lower price, it would be a success. it really is just quad core with a slightly more advanced HT.
Well considering Intel R n D is probably worth more than AMD it ain't that bad. Looking at leaked slides, if pile driver can increase IPC by 10-20% that'd give quite a performance boost, fingers crossed (Y)
Can I get a job at AMD as the next CPU design lead?
I have a great idea that will outperform the fx-8150 to hell and back:
- shrink the phII x6 to 32nm
- add 2 more cores
- upgrade the memory controller to what is in the current fx
- use as fast cash as sandy bridge
why would they continue with the bd design, when it is clearly rubbish?
then that would depend on how underwelming your reality is
Oh no, shame on me!!! I shall hide my self away behind the sofa and never show my face in public ever again...lol
I was hoping and it would be so good, but what we have ended up with is a server chip thats not made the cross over to desk tops particularly well. Lets see what the rest of the range is like and maybe with pile driver AMD might be able to sort out Bulldozers failings.
OMG, that is brilliant...lol I'm putting it in the general section..Bit-tech land need to see this..
Great find dude...+ rep
It clearly isn't rubbish...
Having read through and compared them to a lot of other sites reviews, id have to say they are both talking crap.
Just saw those reviews, interesting i'd like to see the performance scalling by having different ram speeds. Also the improvement from IMC overclock.
Not rubbish? Well yes, it works. However, it consumes too much power, has horrible single thread performance (seems like 25% less than Deneb) and in many well threaded applications still loses to 2500K.
OTH, it seems BD outperforms intel in x264 pass 2, pov ray and 7zip, with the rest of the benchmarks losing. In many cases, it's not faster than a thuban/x4 980.
And as a gaming cpu, i think for eyefinity systems it might be worth it, as those resolutions usually are gpu limited anyway. Perhaps a 3x24" monitor setup with 2x6990 might show better how it stacks up as a gaming cpu.
Also, what if we thought of FX-8150 as a 4core cpu, it would be good enough. Maybe we'll have to rethink it.
it would make sense considering amd-s explanations of their reasons - that hyperthreading sometimes causes negative performance, while in most cases boosts the speed by 0 - 40%.
The two integer cores are like amd-s answer to hyperthreading.
Still, isn't all this a bit deceivingourselves?
Separate names with a comma.