Broadwell-E (X99) predictions

[pj-]

You know thre are 4 quarters in a year right? Broadwell-E is slated to drop in Q1'16. I said Skylake-E in Q4'16. That's a difference of 3 quarters. Dunno where you got 6.

I misread your other post pretty stupidly, I thought your other post was saying Skylake-E is 2-3 quarters from now.

I definitely agree that upgrading to x99 will be a bad move when Broadwell-E comes out, unless it's super cheap

 

[tenks]

I misread your other post pretty stupidly, I thought your other post was saying Skylake-E is 2-3 quarters from now.

I definitely agree that upgrading to x99 will be a bad move when Broadwell-E comes out, unless it's super cheap

Its cool, thats what I figured.

Yea totally agree. Just super irritated with Intel. Even if you got on x99 early with Haswell-E, I dont see many users trading up to get broadwell-e for 10% improvement... unless you're going for the 8core.

 

[Dave2150]

Yeh, the more I think about it, jumping on the Broadwell-E x99 bandwagon will be a bad move at that point in time.

I think I'll see how Skylake Z170 does, probably upgrade to that. I'll be sure to get a quad channel DDR4 kit, just incase I decide to sidegrade to Skylake-E, when it launches next year. A 32GB DDR4 kit should last me 5-8 years, just like my current 12GB DDR3 kit lasted me 6 years (X58 920 system).

 

[moonbogg]

I got nowhere to go until Skylake-E. If Haswell gives me about 15% performance increase, that means if I'm getting 50fps now, that I COULD be getting a whole 57fps for the cost of a whole new platform. Wow. That blows. I need at least 30% more to be worth while, I mean come on man.

 

[AdamK47]

My prediction is that Broadwell-E will be faster than Haswell-E.

 

[guskline]

My prediction is that Broadwell-E will be faster than Haswell-E.

And my prediction is that AdamK47 will have the Broadwell-E the day it is released if not sooner!:biggrin:

 

[AdamK47]

And my prediction is that AdamK47 will have the Broadwell-E the day it is released if not sooner!:biggrin:

That's the most accurate prediction in this thread.

 

[Ajay]

My prediction is that Broadwell-E will be faster than Haswell-E.

Yes, hopefully BDW-E will benefit from the improvement made to Intel's 14nm node and clock high - in which case we should get at least a 5% boost in throughput. The main benefit of BDW-E will be the improved thermals (some of which may be eaten up in hitting higher base clocks, especially for the 'X' version.

The one caveat is that SKLs presumed better clocks on 14nm could have come, in part, for architectural improvements.

 

[Sweepr]

By the time Broadwell-E arrives we will have >4GHz Skylake-S parts so I hope they raise the base clocks compared to current Haswell-E parts by at least 300MHz.

 

[cbn]

By the time Broadwell-E arrives we will have >4GHz Skylake-S parts so I hope they raise the base clocks compared to current Haswell-E parts by at least 300MHz.

Here is the complete listing on Ark for HEDT:

http://ark.intel.com/products/family/79318/Intel-High-End-Desktop-Processors

Historically 4GHz has been the top turbo available (although it didn't happen with Haswell presumably due to the core count increase)

Hopefully we see 4.0 Ghz turbo return for Broadwell.

 

[escrow4]

I run my 5930K @ 3.7GHz which thanks to the falling dollar is now $100+ more than I paid for it (and no, it doesn't cost that much to ship a box of CPUs around the world). If a 6930K runs @ 4.0GHz I'll pass. It it cracks 4.2GHz maybe. Its slack that the 5930K doesn't turbo to the previous 4930K of 3.9GHz.

 

[YBS1]

I don't really expect much of a default frequency increase for the top end part. It's obvious the 5960X wasn't clocked so low because it was incapable, it was clocked the way it was to fit within the same TDP window as the six core parts. Certainly Broadwell will be more power efficient, but we have no reason to believe (do we?) they won't lower the TDP goal as well. Gotta keep the OEM's happy, at least they still let us play if we are willing to feed and cool them.

 

[crashtech]

I'm presuming that some socketed Broadwell parts are still going to be released; at that time we should be able to do some better extrapolations of what we can expect from Broadwell-E. In fact, I think that is when I'm going to decide whether to wait for Skylake mainstream desktop or go X99.

 

[Ajay]

A few of us in CPU&OC think that Intel will up the clocks on BDW-E precisely because OEM want to sell us to upgrade. If BDW-E offers ~5% more perf/clock and 10% higher clocks - then there will be more demand. Demand for in place upgrades (from HW-E), new mobos and (while we are at it) new GPUs, SSD, RAM, etc. That's what needs to happen for HEDT to exist as a viable niche.

The only reason Intel wouldn't do that is if there were issues with bumping up the clocks (or because they found a better way to milk us for profits ). Which would be a bummer, since I'm hoping to upgrade to BW-E. So, I'm expecting to see 200-300Mhz bumps in base and turbo clocks - and it seems like Intel's 14nm process should have the headroom to do that and keep the thermals the same or lower.

Once Skylake is released, we'll have a better idea on just how good Intel's 14nm is (or isn't).