• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

[coolaler] Devils Canyon: 4.0 base/ 4.4 turbo @ stock

Page 13 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
It's all about total cost for me, we have to compare between 6 cores vs 6 cores
In my country

Plan A

i7 4930K : USD 600
16Gb DDR3 1600 : USD 160
Board X79 : USD 350 ( only SATA 6GB & USB 3.0 )
Water Cooler : USD 100
Total Cost : USD 1210

Plan B
i7 5820K : USD 400
16GB DDR4 2133 : USD 250
X99 Board : USD 350 ( X79 plus SATA Express and maybe USB 3.1 )
Water Cooler : USD 100
Total Cost : USD 1100

For me 6 cores running @ 4 Ghz is enough for me
So, the point is clear for me
 
I won't be doing that for the same reason I don't run my CPU outside of Intel spec. I spend half my life trying to troubleshoot issues on stock builds already, I don't care for throwing potential IMC stability issues into the mix.

Just prime95 it 😎 You'll be just FINE!
 
https://www.youtube.com/watch?v=qM6tBoP0AWU

Pack your bags kids. 82C at stock with the stock heatsink. Serious overclockers will still need to delid and lap. It's an improvement at least and it uses LESS power at least, but still, not living up to the hype.

Per the comments, they're also struggling to hit 5GHz and are even suspecting that the Computex chips were not only highly binned, but only hit 5.5GHz on a single core. Ouch. Devil's Canyon is not a noteworthy improvement. Enthusiasts should be looking at Haswell-E.
 
Last edited:
https://www.youtube.com/watch?v=qM6tBoP0AWU

Pack your bags kids. 82C at stock with the stock heatsink. Serious overclockers will still need to delid and lap. It's an improvement at least and it uses LESS power at least, but still, not living up to the hype.

Per the comments, they're also struggling to hit 5GHz and are even suspecting that the Computex chips were not only highly binned, but only hit 5.5GHz on a single core. Ouch. Devil's Canyon is not a noteworthy improvement. Enthusiasts should be looking at Haswell-E.


Maybe, but it still running cooler and is a good cpu for people that don't overclock or people that only do moderate overclocks like me.

Plus Haswell-E is a lot more expensive. Not really comparing like for like.
 
7 degrees coolers while running 500mhz faster. Pretty good improvement.
My thoughts exactly. If it was running at the same speed as the 4770K you'd probably be looking at 10C-15C less.

Also it uses 18W less power under load, too.

Considering it isn't soldered, these are great results. :thumbsup:
 
Which board did you choose an why?
The Asus Z97-K, sitting on my table right now all lonely because it's waiting for the 4970K to arrive.

I was going to get the 'A' but I decided I didn't need SATA Express, or the extra cost.

I used Gigabyte all the time in the past but Asus' UEFI features are far ahead of them now, especially with fan controls.

I always go for quality no-frills motherboards and this board looks like it'll deliver. I also needed at least one PCI slot for my excellent X-Fi.
 
https://www.youtube.com/watch?v=qM6tBoP0AWU

Pack your bags kids. 82C at stock with the stock heatsink. Serious overclockers will still need to delid and lap. It's an improvement at least and it uses LESS power at least, but still, not living up to the hype.

Per the comments, they're also struggling to hit 5GHz and are even suspecting that the Computex chips were not only highly binned, but only hit 5.5GHz on a single core. Ouch. Devil's Canyon is not a noteworthy improvement. Enthusiasts should be looking at Haswell-E.

That shows it running cooler at significantly higher speeds with a higher TDP. Looks fantastic to me?
 
1401896045RWLb4BbL6g_1_1.jpg


If I can get 5ghz on air with decent temps, I'm in.
 
I think we need more thorough testing. So far I'd say 4770k users have no real need to upgrade to a 4790k but if you are using something older then its a good upgrade. Those are my impressions so far when it comes to 4770k vs 4790k. Like I said though we need more thorough testing.
 
I did a thorough investigation of each of the -E platforms, and each time I came to the conclusion that they were worthless. The traditional 4/8 platforms are not only cheaper but also faster for 99% of the tasks people use them for, including gaming.

Also -E may have more bandwidth, but the quad-memory controller has more latency, which can hurt games. Lower clocked Haswells often beat them for games because of this. Witness the 4820K losing against the 4770K despite the former being two speed bins higher (this is a recurring trend):

IVB-4820-62.jpg


The 4790K would absolutely dominate that chart if it were tested.

The -E platform is for people who need >32GB RAM, need 6/12 for very specific applications (not games), and/or who need more PCI lanes for tri/quad GPU. If you fall under any of those categories you know exactly why you're getting that platform. For everyone else, the platform is useless.

Thanks. Can we sticky this? Because I don't want to hear about E platforms for gaming anymore. It's like constantly being hit with noise.
 
Intel has actually done this in the past (1.13GHz P3 for example), but the people who made those decisions are no longer in the position to make those kinds of decisions anymore.

So what one could argue is that precisely because they have made that mistake in the past, they are hell-bound to not repeat it. And so they are not likely to be launching DC chips that are effectively at their OC limit. These will be cherry-binned chips for sure.

I like how DC is clearly nothing more than what was supposed to come as part and parcel to the Haswell Refresh cycle, but they delayed it and renamed it something cooler for branding purposes.

How much you want to bet that every single Haswell Refresh chip has the exact same "new and improved next-gen TIM" under the IHS?

I'd be absolutely surprised if that wasn't the case.

Well they effectively operate as a monopoly in this market so they do what they can to milk it. But I definitely am in the "better late than never" camp. After all what choice do we have?
 
Thanks. Can we sticky this? Because I don't want to hear about E platforms for gaming anymore. It's like constantly being hit with noise.

Come on, one old, poorly threaded game tested at 720p hardly proves a point. If you want the best bang for the buck, for sure it is not about the E platform. But if you want to eke out the most performance in recent, highly threaded games, a case could be made for 6 cores.

I think it is too soon to tell really, but new games like watchdogs and the mordor game coming out make hex core look more attractive. What remains to be seen is whether these games are a trend of things to come or just poorly coded aberations.
 
Come on, one old, poorly threaded game tested at 720p hardly proves a point. If you want the best bang for the buck, for sure it is not about the E platform. But if you want to eke out the most performance in recent, highly threaded games, a case could be made for 6 cores.

I think it is too soon to tell really, but new games like watchdogs and the mordor game coming out make hex core look more attractive. What remains to be seen is whether these games are a trend of things to come or just poorly coded aberations.

Well for one lower resolutions take the GPU out of the equation in terms of bottlenecking factor. But I can concede one game does not tell the whole story.

I'm sympathetic to the 6 core cause. Today's consoles have 8. However I haven't seen any evidence suggesting that any games need more than 4 cores yet.



The arma3 chart shows only a 2fps difference for 2 extra cores plus 100MHz extra clock speed. The game is probably GPU bottlenecked there but I'd need to see someone run that test with SLi'd Titans. And that test shows an old Sandy Bridge at 4.8GHz.

Games will likely be coded for 8 cores in the future but that time is not upon us. By that time there may well be a 6 core mainstream part. There isn't any reason we don't already have that beyond Intel holding a monopoly on performance CPUs.

It seriously took long enough for games to properly use 4 cores. An still no evidence that they do. I haven't seen a game that pegged all my cores at 100% yet.
 
Well for one lower resolutions take the GPU out of the equation in terms of bottlenecking factor. But I can concede one game does not tell the whole story.

I'm sympathetic to the 6 core cause. Today's consoles have 8. However I haven't seen any evidence suggesting that any games need more than 4 cores yet.



The arma3 chart shows only a 2fps difference for 2 extra cores plus 100MHz extra clock speed. The game is probably GPU bottlenecked there but I'd need to see someone run that test with SLi'd Titans. And that test shows an old Sandy Bridge at 4.8GHz.

Games will likely be coded for 8 cores in the future but that time is not upon us. By that time there may well be a 6 core mainstream part. There isn't any reason we don't already have that beyond Intel holding a monopoly on performance CPUs.

It seriously took long enough for games to properly use 4 cores. An still no evidence that they do. I haven't seen a game that pegged all my cores at 100% yet.
well I don't know what games you are playing but I saw nearly 100 percent usage on my overclocked 2500 K in several cases. And you sure as heck aren't playing Crysis 3 as that will eat 4 cores for lunch in parts. there is no chance of staying above 40-45 fps in some areas with an i5 no matter how much gpu power you have. I see cpu usage over 80 and even 90% on my 4770k in that game. Watch Dogs really needs more than 4 cores too depending on where you are testing and what's going on.
 
Last edited:
well I don't know what games you are playing but I saw nearly 100 percent usage on my overclocked 2500 K in several cases. And you sure as heck aren't playing Crysis 3 as that will eat 4 cores for lunch in parts. there is no chance of staying above 40-45 fps in some areas with an i5 no matter how much gpu power you have. I see cpu usage over 80 and even 90% on my 4770k in that game. Watch Dogs really needs more than 4 cores too depending on where you are testing and what's going on.

Ok I can concede crysis 3 is a killer on the CPU. That game was the one that made me think my CPU couldn't max it out. I checked resource monitor and it seems to leave a core or two below maximum usage. However this in itself doesn't prove it may need more cores. Which is why I said I'm sympathetic to the 6 core "cause".

But still I haven't come across anything conclusive that indicates that upgrading to an "E" platform would help. I run everything at 1080p maxed out and AA maxed out. Over clocking the GPU makes a big difference no doubt.

Are there any benchmarks that show the CPU as the bottleneck at completely maxed out settings?
 
Plenty of sites out there that compare the 4770k to 4960x and in most cases there's very little difference between the two. In some cases 4770k does slightly better and in others the 4960X a $1000 processor does slightly better. Most of the time they are pretty much evenly matched. Also I realize one is ivy and the other is haswell but by the logic of many the 4960x should be leaving the 4770k in the dust and it does not. At this time your still better off going with a i7 4790k and just overclocking it vs shelling out hundreds more for the cpu along with a premium for ddr4 if all you do is play games when haswell e comes out. At this current time the 4930K is still going for almost $600 which is $300 more than you can get a 4770/90k for. I'm sure pricing on haswell e wont be much different. That $300 is much better spent just getting a better GPU especially if you are on a budget because it will have a significantly larger impact than spending the extra cash on a cpu that might in some cases be slightly better.

http://us.hardware.info/reviews/476...se-benchmarks-hd-7970-crysis-3-1920x1080-high

http://www.anandtech.com/show/7255/intel-core-i7-4960x-ivy-bridge-e-review/5
 
Last edited:
well I don't know what games you are playing but I saw nearly 100 percent usage on my overclocked 2500 K in several cases.
Just because something is showing 100% CPU usage, it doesn't mean it's CPU limited. You can't make such an inference solely from CPU load unless you have actual code profiling to go with it.

And you sure as heck aren't playing Crysis 3 as that will eat 4 cores for lunch in parts. there is no chance of staying above 40-45 fps in some areas with an i5 no matter how much gpu power you have.
c3-fps.gif


http://us.hardware.info/reviews/476...-benchmarks-hd-7970-crysis-3-1920x1080-medium

If you want to drop a thousand bucks on a 4960X for those results, be my guest. But it doesn't change the reality of things.
 
Last edited:
Back
Top