FX 9590 curb stomps the i7 cpus all the way. Prove me wrong. Oh that's right you can't. Suck it AMD all the way.
>>46849309
>dat power consumption
>dat need for liquid cooling
>dat lack of any overclock potential
>taking all the good and cheap 8350s off the market just to charge a premium for this shit
>still focusing on shitty 32nm and not moving down to 22
the 9xx series is a fucking abomination and all consumers should be pissed. i WAS going to buy an 8350 eventually. then they took all the high binned 8350s and rebranded them just to charge more for the same fucking chip.
fuck companies. first intel pulls this shit with their "k" varients and now AMD... fuck computers fuck companies and fuck consumers.
>>46849490
also fuck capitalism for making any products which were good/decent absolute shit just because they have to cater to the general population (aka fucking idiots with 100iq.)
>>46849490
> then they took all the high binned 8350s
Yields been getting better: five 8350's i've set up during the last four months have had almost half of the power consumption of any of the early 8350's i set up, so in that sense any of more recently produced piledrivers are of higher binning that the ones manufactured in 2012.
The FX-9* were idiotic decision really and they have sold terribly, and for a good reason.
>>46849634
And as far as performance goes: 8350@stock beats the living shit out of [email protected] in multiple make job or pretty much any multithreaded job on non-shitty operating system or with a program that has not been compiled with discriminating compiler.
Quality variances in the amd cpu's is still pretty terrible: 3 out of 10 cpu's dont report core temperature at all for example, but then again there is no real sensor in the cpu anyway (unlike any modern jewtel chip).
>>46849490
No FX8350s were taken off the market to produce the FX Centurion line. Those chips are all the functional 4M Vishera chips that failed validation at stock voltage to be either an 8350 or 8320. They are in fact a lower binned 8320 thats just clocked higher.
>>46849634
32nm Vishera chips aren't being produced any more, AMD hasn't bought any 32nm wafers for a very long time now. They're qualifying old stock as new SKUs. The FX 8370 and 8370e are the highest binned chips they had produced, some of which are absolutely incredible.
http://www.tomshardware.com/reviews/amd-fx-8370e-cpu,3929-3.html
>We measure 90 W at the appropriate rail, which should work out to somewhere between 80 and 82 W for the CPU alone, once the losses due to the voltage regulator are subtracted. Nice!
>>46849699
>And as far as performance goes: 8350@stock beats the living shit out of [email protected]
This is what piledriver fags actually believe
>>46849721
And for those of you too lazy to click the link, that FX 8370e at 4ghz has a vcore well under 1.2v.
>>46849721
>AMD hasn't bought any 32nm wafers for a very long time now.
What is a
>very long time
and care to show any source?
Either way: each and everyone of the 8350 chips i installed last year had far lower power consumption than the ones i installed during the launch and few months forward.
>>46849829
I don't feel like digging through forum posts from 2012 and 2013. Some AMD employees on various forums like OCN commented on it before they were ever released. When the FX Centurion line was just a rumor they stated outright that they had qualified existing chips into higher performance SKUs, and some into lower power SKUs.
The higher performance line was the FX Centurion.
The lower power line is the e series.
The fact about the wafers came from a conference call with Rory who stated that all wafers they were purchasing for the following year would be 28nm.
Still use my a6 3420m, I've over clocked the CPU to 2.1 GHz, and its playing war thunder at 40 fps medium settings( 1366x768).
Hope it sees the light of vulkan someday.
K10/Thuban was AMDs last decent microarchitecture and Sandy Bridge Intels.
Prove me wrong.
>>46849984
Can't prove you wrong, its true.
On the CPU side Llano per clock is faster than Kaveri, it'll likely still edge out Carrizo too.
2.9ghz K10.5 cores with no turbo vs 3.7ghz Steamroller cores with 4ghz turbo:
http://www.anandtech.com/bench/product/399?vs=1200
for the price, the performance is decent. it's not enough to complain about comparative to competing chips. there are better out there in the same bracket as the 9590, but the price jump to the next noticeably better chip by intel usually pushes a consumer back to 9590/8350
>>46849309
http://www.anandtech.com/bench/product/1289?vs=1260
Can you point out where it is curb stomping?
>>46849309
/g/uys,what about the A8-6410?
>>46850479
Its Beema.
15w quad core Jaguar based part with a 2ghz base clock and 2.4ghz turbo.
2 GCN CU clocked at 800mhz.
Its only downside is single channel memory, otherwise it would outperfom a comparatively clocked Phenom across the board.
Good enough for a little web browsing machine. Can handle full Blu-ray playback without issue, but then again so could the first generation Brazos chips. If the system has an SSD in it instead of a shitty mechanical drive it'd be decent enough.
will I ever really need an update?
this still outperforms midrange kaveri laptops
>>46849490
>32nm and not moving down to 22
Why not just move down to something smaller now that the processes are available?
>>46851903
They didn't refresh Vishera or make any other 8 core chips because they didn't sell well enough.
>>46849721
>32nm wafers
le...
>>46851930
What will the node size be of the new arch? Zen or whatever it's called
>220W TDP HOUSEFIRES
>Still slower than Haswell at STOCK speed
TOP KEK
I've heard of people running FX-9xxx chips with NH-D14s. You don't hear about it often because nobody wants to spend money on a cherry picked 8350.
>>46851962
The Zen based chips will be fabbed on GloFo/Samsung's collaborative 14nm FinFET process.
>>46851982
Will it be compatible with am3+ boards?
Or will it require new DDR4 memory?
>>46852061
>Will it be compatible with am3+ boards?
LOL of course not. You think they're going to release a 2016-2017 platform on a socket from 2012~ that hasn't gotten a new chipset in years?
They'll be using a new socket.
>>46852061
Nope... so what's coming now:
Carrizo and Carrizo-L for laptops are just around the corner (won't be offered socketed), and then ZEN on a completely new socket.
>2016
>amdfags are hyped
>amd gives shit which loses to Core i5
