Broadwell, is there a point?

[SiliconWars]

What are you basing these expectations on? It's been years since we've seen a die shrink from intel give better overclocking results.

It's only been one node, and FinFET's may have been a one-time hit at 22nm. There's a whole bunch of other reasons why Broadwell might not clock so well though.

 

[Homeles]

It's only been one node, and FinFET's may have been a one-time hit at 22nm. There's a whole bunch of other reasons why Broadwell might not clock so well though.

Precisely. I'm banking on a positive outlook. This is all provided Intel doesn't introduce any other modifications to their process or architecture that would benefit mobile at the expense of high performance. Technically we already know they did by putting a higher priority than usual on density, but I'm still expecting better performance than 22nm.

Somewhat similarly, I expect GloFo's 20nm to substantially improve in performance over their 28nm, since they'll be going gate-last. They also won't have to deal with the one time hit from regressing from PDSOI to bulk, so it should turn out to be a good node for AMD, especially if they move to FDSOI. I just wish it would come a lot sooner.

 

[NTMBK]

Intel cares about low power consumption, not top end clocks. They need low power for servers (more cores per socket), and they need low power for mobile. They will make a process focused on power consumption, not maxing out switching speeds. I don't expect an overclocking beast to be honest.

 

[SiliconWars]

AMD won't have any high performance chips at 20nm though. I'm not actually sure what it was they taped out last quarter, been scratching my head over that one.

 

[witeken]

I have seen all these promises of improvement and then you can only get those fancy IGP advancements in embedded motherboards or AIO gear.

In Haswell, but likely not in Broadwell.

 

[Homeles]

Intel cares about low power consumption, not top end clocks. They need low power for servers (more cores per socket), and they need low power for mobile. They will make a process focused on power consumption, not maxing out switching speeds. I don't expect an overclocking beast to be honest.

I don't either. I'm just expecting things to be in line with 32nm, and I think that's a pretty modest expectation.

AMD won't have any high performance chips at 20nm though. I'm not actually sure what it was they taped out last quarter, been scratching my head over that one.

Is AMD porting the console SoCs to 20nm?

 

[ElFenix]

Here's to hoping 14nm will be the tank that mature 32nm was.

My logic says no tho - because it's a new process what's to say it will OC well?

Historicly all good OC nodes were mature ones - weren't they?

historically new nodes brought better overclocking. but, intel was also trying to get more MHz out of the nodes as well. intel hasn't been focused on upping the clocks lately, so not as much work has gone into that on both the node side and the processor design side.

 

[SiliconWars]

Is AMD porting the console SoCs to 20nm?

I don't think the mobile-optimized 20nm will be able to match 28nm HP in performance.

This is also why I'm losing belief in 20nm graphics.

 

[Homeles]

I'd forgotten it was only offered in one flavor. Well, here's hoping 14XM saves the day...

 

[TuxDave]

I'm going to make some Broadwell predictions right here in writing so I can look back and see how I did.

CPU performance vs. Haswell - Barring any new instruction based performance increases I'm betting on a 2 to 3% improvement in IPC over Haswell. This is based on Haswell just receiving an improved branch predictor, wider execution engine, 2x increase cache bandwidth, and better memory controller. Unless Intel can somehow pull a rabbit out of it's hat I think we're topping out on legacy code IPC-wise. The low hanging fruit has been gone for 5 years now and it's "blood from a stone" time.

iGPU - 50% increase in performance across the board when looking at new version of HD Graphics, 4000 series, and 5000 series. ie 4600 moves from 20EU's to 30. Perhaps some new architectural improvement will provide 10% additional performance gains over the 50% I quoted above.

I enjoy reading prediction posts. Some are good and some are off but who cares. What key benchmarks will you will use to evaluate your prediction?

 

[2is]

With desktop sales in a pretty consistent decline the last several years, and the ultra mobile market continuing to eat away at Intel's bottom line, if Intel can make Broadwell 1% more efficient at the expense of a 10% decrease in overclockability, they're probably going to do it. Not to mention, most of the hype is around the IGP and efficiency. Heck, it wasn't long ago that there was talk about Broadwell not even making it to the desktop market.

I hope I'm wrong, but I personally see zero evidence for being optimistic on Broadwell's over clocking potential.

 

[Hulk]

I enjoy reading prediction posts. Some are good and some are off but who cares. What key benchmarks will you will use to evaluate your prediction?

Here's a post I made regarding Haswell performance. I'm talking about these same apps and benches.
http://forums.anandtech.com/showthread.php?t=2324672&highlight=

For iGPU performance games and the major compute apps.

 

[tweakboy]

Wait for Skylake. gl

 

[tweakboy]

Wait for Skylake ,,,,,