Originally Posted by 12Cores
The FX-83XX series are not bad gaming chips, my rig runs most games silky smooth even Crysis 3. If you just want to game at 1080p, a fx-8320 and r9-280/gtx 770 will be more than enough for today's AAA " next gen " console ports. The biggest issue with these CPU's is the heat, that I can live without. That being said I will probably get an fx-9590 later this year, I am sure my CPU is well above 220w already so who cares.
This right here.
My, and most other people's, FX8320/FX8350's are fine. A blast to Overclock even (Certainly more agreeable than this Haswell Chip I've been torturing myself trying to stabilize a decent Clock on for the last two weekends), simple to set up, and a beast as far as certain loads go.
The problem with these chips ultimately is that they are not "3-4 years as the only flagship offering fine."
Which is a problem made even more frustrating and incomprehensible when you realize that any decent revision being made to this architecture which would/could offer a significant increase in IPC and competitiveness is now relegated to the lowend/midrange market. Counting improvements from Steamroller, Excavator is technically going to be at least a 30-40% improvement in IPC over Piledriver.
Unfortunately, most of us will never actually know the specific's of that benefit, because like the aforementioned Steamroller, AMD does not want to push those improvements into the enthusiast market.
It's been a sad state of affairs for both companies over the last few years honestly in my opinion.
Intel, the undisputed leader in this market, carrying nearly unlimited R&D resources and tons of money to throw around, have decided that annual updates bringing the enthusiasts less than 5% performance increases across the board is the way to go. That mindset is pretty awful for us without any threat of competition emerging nowadays.
AMD, the undisputed unleader, has decided they are okay with this as well, since they are sitting this whole cycle out. The fact that they would rather not have to put much effort into revising their flawed architecture for the highend seems to mean to me that "Hey if they aren't good enough none of these enthusiasts and system builders really care, since they ain't using these APU's." In their mind, they may as well just sit on Piledriver for the rest of this year into next while waiting for Jim Keller's team to pull a miracle architecture out of their hat.
Now certainly some of you would argue that sitting the rest of Bulldozer out and waiting for a new architecture to save the day is the smartest decision AMD has made in years. While I can see why you'd think that, I disagree wholeheartedly. I think a 6-8 core Steamroller/Excavator with even the most conservative of those IPC improvement's being realized, would have given AMD a chance to keep a lot more of it's fan base through out this time. Heck, the revisions, alongside their newly found "We own the Consoles's" clout, might even allow them to pick up a few Multithreaded benchmark and performance wins.
Or at the very least, they could have used these highend revisions to help them push a new highend chipset into consumer's hands ASAP. If for no other reason then to increase the chance that they get a positive response to Skybridge and beyond when these chips finally drop, by having the largest socket user base they could have possibly obtained beforehand.
It is in that regard especially where I really think they missed the boat.
But nonetheless, the next two years are going to be a super interesting time for both companies...- Can/will Intel bother to push a true, unique, and innovative revision of their architecture at some point?-
- Can/will AMD catch up in IPC with a guru like Jim Keller given essentially free reign in how he develop's their next two architectures?-
-Will either company make an impact with an ARM chipset?-
-Is AMD going to be right enough about HSA development in the future that all of those who end up buying Kaveri/Excavator chips will suddenly be topping out benchmarks and frame rates in games, and raw performance numbers in productivity workloads within HSA influenced tasks?-
-Will AMD being the only hardware inside of the current consoles result in the first slew of Console to PC ports that heavily favor users running AMD hardware?-
-Is Intel doing the right thing after all the progress they made with their power saving architectures to suddenly drop a 6-8 core mainstream chip with 125-140+ Watt TDP's?-
-Is Intel going to regret not going super all out with performance increases and heavily padding their lead in IPC these last three years?-
All this stuff is certainly a lot to think about guys. I realize that. But these all of these questions (and a lot more I'm sure I've forgotten to put up there) add up basically to mean that our expensive hobby is, as of right now, the most interesting it has ever been. For that reason alone, I'm excited.