See with PCI 4.0 on the way and with all new HBM 2 , I am sure that will come handy with new GPUs. Thus , I am waiting a little longer .

If Broadwell-E : Haswell-E = Broadwell : Haswell, I won't take much notice of this.

I am definitely close to going 6 core now that DDR4 is heading towards affordable but yeah, I wouldn't if it's the same story with Broadwell-E as Broadwell. Unless Skylake-E is just round the corner, I expect Broadwell-E will be good.

From all the DX12 benchmarks and info I've read, in respect to gaming, DX12 doesnt appear to scale up past 6 cores. So getting 8 cores might just be a waste of money right now but maybe in the future it will scale to 8 cores. Also since the benchmarks were from a month or two ago they may have made changes to allow it to scale up to 8 cores. The benchmarks were performed using a pre-release so I guess its still up in the air

Honestly people are speaking more and more about overclocking, and it confuses me a little. I understand we are hardware enthusiasts who like to tweak and get higher numbers every day. We like to gloat about what we've achieved, and what we plan on next. We constantly cycle out old for new, or modify the old to work as new. But what I am looking forward to in the future roadmap, CPU's that place efficiency over overclocking. Think about, say in 2 years Intel comes out with.... Cove rail core(Fun name for a fake cpu core!), stock clocks are only 2.5 ghz and average overclock is only about 400-500mhz on air. Uses a fraction of the power compared to say Sandy Bridge, and IPC performance is near 1.5 times Sandy Bridge IPC performance. I would take that over overclocking. Now I admit my example was a bit extreme, however that's just an example. 😀 Intel is starting to hit the same wall they did before with Conroe, you can shrink it and make the core clock higher but gains are very minimal.

Honestly people are speaking more and more about overclocking, and it confuses me a little. I understand we are hardware enthusiasts who like to tweak and get higher numbers every day. We like to gloat about what we've achieved, and what we plan on next. We constantly cycle out old for new, or modify the old to work as new. But what I am looking forward to in the future roadmap, CPU's that place efficiency over overclocking. Think about, say in 2 years Intel comes out with.... Cove rail core(Fun name for a fake cpu core!), stock clocks are only 2.5 ghz and average overclock is only about 400-500mhz on air. Uses a fraction of the power compared to say Sandy Bridge, and IPC performance is near 1.5 times Sandy Bridge IPC performance. I would take that over overclocking. Now I admit my example was a bit extreme, however that's just an example. 😀 Intel is starting to hit the same wall they did before with Conroe, you can shrink it and make the core clock higher but gains are very minimal.

It's everybody's choice if they want to overclock or not... nothing keeps you from using your hardware at stock, and presumably at their most efficient settings. But efficiency in stock settings is improving anyway, and trust me, companies try to cut back on overclocking as much as possible in one way or the other anyway, as much as they can make it happen, or get away with it. Look at locked multiplyers.... no real sense behind that besides cutting down overclockability. If we all were after efficiency in the first place, we wouldn't be talking about overclocking, but undervolting 😀

so skylake is no better than my 4670k at 4.6 ghz and my z87 mobo combo? need honest opinions before i plan an upgrade. i can save mysekf a 1000 bucks and just get a 980ti or a fury instead of a full upgrade.