[THG]Core i7-4770K: Haswell's Performance-Previewed

[Homeles]

Not seeing this addressed --

I'm seeing a lot of support for intel noting that the reason for the overall small IPC increase is because of the ASSUMED higher perf/watt increase.

However, is it not true that the reason IB ran so HOT which also DECREASES perf/watt is due to it's cheaped out TIM? I am seeing threads about some 30 deg celcsius drop in temps just due to TIM change which is EYE POPPING. Would intel not have a cooler running, less wattage sucking chip with better TIM?

They would. I'm sure they crunched the numbers and determined that they'd come out ahead if they cut back.

 

[cytg111]

The question begs to differ: Coming from an i7 920 @ 3.6ghz, is it worth upgrading to a 4770k in the near future, or go straight to a 3930k now?

PC use = 2% Productivity 98% Streaming and Gaming

i7 920@3.6ghz
12 gb ddr3
gtx680 @ 2500x1440

What games do you have trouble playing now and at what settings?

 

[WaitingForNehalem]

I'm definitely waiting for the enthusiast Haswells. I sure have gotten a lot of use out of this i7 920.

 

[Enigmoid]

The question begs to differ: Coming from an i7 920 @ 3.6ghz, is it worth upgrading to a 4770k in the near future, or go straight to a 3930k now?

PC use = 2% Productivity 98% Streaming and Gaming

i7 920@3.6ghz
12 gb ddr3
gtx680 @ 2500x1440

Probably the 4770k because its a fraction of the cost (50% when you count motherboard).

 

[Riotvale]

What games do you have trouble playing now and at what settings?

I don't really have issues running stuff i'm playing now per se.

40% Starcraft 2
40% Path of Exile
20% Random flavor of the month game (DMC, Tomb Raider, Simcity, etc)

I like to play with most eye candy on, obviously more GPU intensive there.

Thing is, I like to stream on twitch 720p, 3500kbps bitrate, and even though I have a capture card (Avermedia Livegamer HD), I still feel a performance hit when streaming and playing at the same time.

I know that a 3930k will rectify some of these issues, and I've been jonesing for some more cpu power for the longest time.

 

[Idontcare]

I'm definitely waiting for the enthusiast Haswells. I sure have gotten a lot of use out of this i7 920.

Would be funny as hell if you petitioned the admins to change your username to WaitingForHaswell

 

[Riotvale]

Probably the 4770k because its a fraction of the cost (50% when you count motherboard).

Is there an idea of when the haswell equivalent of the 3930k will arrive?
I might wanna step it up to a higher level with this upgrade....

I guess what i'm saying is that I'm looking for something powerful along the lines of a 3930k..

I guess the 4770k just won't cut it.

 

[Enigmoid]

That's why you see mobile CPUs nowadays rivaling desktop CPUs, and specs that are even better than on desktop(like the presence of HD 3000 and 4000 on ALL laptop Core parts for example).

Agree 100%.

And while there hasn't been much progress in desktop cpus now most mobile i7's will tie with a 3.6 ghz i7 920 but at 45 watts vs 130+. That's where the massive increase in performance has gone.

Everyone thinks intel is slowing down but they are not. Mobile performance has increased at incredible rates (those incredible historic rates that everyone talks about).

An i7 720qm (first gen i series) is equal to a ULV i7 ivy at 17 watts vs 45 watts in threaded tasks. The ULV blows it away in singlethread. Intel also managed to throw in a fairly decent (for everything except gaming) hd 4000 as well. (looks like for cinebench r 11.5 its 3 ish for the i7 720 and about 2.9 for the i7-3667U).

At the higher level a i7-3620qm is about twice the power of an i7 720 at the same power envelope and again with a pretty decent igp. Cinebench r 11.5 ~3 points vs ~6.5 points for the 3630qm.

So for mobile, performance has just about doubled and you get a free decent igp. Amazing.

Mobile has advanced to the point where an i7-3630m can compete with an even footing (less than 15% difference) with any non-overclocked i5 in tasks like video transcoding or editing. I expect haswell to shorten the gap even more.

 

[WaitingForNehalem]

Would be funny as hell if you petitioned the admins to change your username to WaitingForHaswell

People would have to know the name of the next tock to know what my name is .

 

[WhoBeDaPlaya]

Is there an idea of when the haswell equivalent of the 3930k will arrive?

When desktop Broadwell hits.

 

[IntelUser2000]

People would have to know the name of the next tock to know what my name is .

WaitingForSkylake?

 

[WhoBeDaPlaya]

It would be kinda funny if his handle got changed to WaitingForAMD

 

[WaitingForNehalem]

WaitingForSkylake?

By the time Haswell-EP comes out that might be a possibility ^_^

It would be kinda funny if his handle got changed to WaitingForAMD

I LOLed at this one.

 

[crashtech]

Maybe Intel is trying NOT to get ahead TOO far, if they do, the accusations of monopoly might start being taken seriously. Really, it almost seems like these performance increases are being carefully calibrated...?

 

[Idontcare]

It would be kinda funny if his handle got changed to WaitingForAMD

Oh that is sooo evil :twisted: I like ^_^

Let me forward this on to some admins who might be interested in seeing if the "protection collection jar" needs a little topping off :whiste:

 

[Homeles]

Maybe Intel is trying NOT to get ahead TOO far, if they do, the accusations of monopoly might start being taken seriously. Really, it almost seems like these performance increases are being carefully calibrated...?

If you ask me, my opinion is that Skylake is going to be much more substantial. I think Intel's team in Israel makes the bigger leaps. Let's face it -- they brought us Conroe and Sandy Bridge.

 

[IntelUser2000]

If you ask me, my opinion is that Skylake is going to be much more substantial. I think Intel's team in Israel makes the bigger leaps. Let's face it -- they brought us Conroe and Sandy Bridge.

I felt that Israel team makes core-wide changes while the USA-bound team does the platform-wide changes. I think neither can be ignored, but obviously with enthusiasts they favor the core-wide changes more.

 

[cytg111]

I don't really have issues running stuff i'm playing now per se.

40% Starcraft 2
40% Path of Exile
20% Random flavor of the month game (DMC, Tomb Raider, Simcity, etc)

I like to play with most eye candy on, obviously more GPU intensive there.

Thing is, I like to stream on twitch 720p, 3500kbps bitrate, and even though I have a capture card (Avermedia Livegamer HD), I still feel a performance hit when streaming and playing at the same time.

I know that a 3930k will rectify some of these issues, and I've been jonesing for some more cpu power for the longest time.

Well, then I suppose it is not crystal clear what your best option is .. I guess it depends on a number of more specifics like budget,
how much $$ are you willing to spend?
how much of a performance hit do you take? (minimum fps, current and target).
For the live streaming stuff, more cores is apparently put to good use, I believe we had a thread here some weeks back where the AMD's FX8350 smoked the top intel offers in this very particular configuration (and about the only place amd can rub one off...).
Glad I am not making your decision

 

[Piroko]

Not running games well has not much to do with your argument though. The whole point is that you are saying that Trinity not being able to reach Base clocks is somehow justified while Core chips not being able to reach peak Turbo(which is far above Base) is a sin.

Actually I've never said or indicated such. I don't approve lower than expected performance in both camps. But that's my point, the Intel ulv chips are just as underrated in TDP/overrated in clockspeed as the AMD pendants.
Also, i can get this chip to spike-throttle the CPU to 800 MHz reproduceable even though the cooling is adequate.

Yes, intel does clock down, I'm just trying to say that your ULV part is significantly better than any amd apu (which has the same thermal problems with much worse performance). Intel isn't great but amd is worse with this.

Honestly, it should better be. There is no other justification for its price premium. Surfing gets me around 7 hours battery life, while any games, even cheap stuff like 'Where is my water' will murder battery life (lost about 30% charge in less than an hour).

A4-4355M and A6-4455M get eaten by an ivy ULV i5 (half the cpu performance) and are similarily much worse than the ULV hd 4000 (256 cores at 200-424ish mhz, though mostly runs at 200 mhz vs the 384 cores at 424-686 of the SV a10).

Problem is, while it gets eaten in benchmarks, it'll probably still play Minecraft fine. Or other OpenGL games. Or other UnrealEngine3 games like Tera.

I'll keep the Yoga 13 because I need/want its form factor and I love the tablet conversion. But the IB core i7-u isn't a particularly good chip in my book.

 

[beginner99]

The key performance part is AVX2 and everything being 256bit wide in the core.

True. But even if it is available from day 0 for newly compiled apps, it will remain niche. What are actually common usages for it? Video trans-coding?

 

[ShintaiDK]

True. But even if it is available from day 0 for newly compiled apps, it will remain niche. What are actually common usages for it? Video trans-coding?

Everything. AVX2 will (over time) be one of the greatest leaps for many years.

 

[itsmydamnation]

Everything. AVX2 will (over time) be one of the greatest leaps for many years.

It will be another increment, just like all the others. The issue you run into is you need 16 non-dependant 32bit vectors, you know how everyone loves to quote Amdahl's law around here . Manderbolt ( what the sandra multimedia benchmark is) gets perfect scaling it is in no way representative of common workloads, if it was CPU's should pack up and go home now, GPU's are king.

Also any thing that isn't SIMD wont use AVX2 just like how it doesn't use AVX or SSE , MMX , etc now. Try keep thing in reality. Its still a struggle to find things that are compiled with AVXyet alonethings that are actually written with it in mind.

 

[ShintaiDK]

It will be another increment, just like all the others. The issue you run into is you need 16 non-dependant 32bit vectors, you know how everyone loves to quote Amdahl's law around here . Manderbolt ( what the sandra multimedia benchmark is) gets perfect scaling it is in no way representative of common workloads, if it was CPU's should pack up and go home now, GPU's are king.

Also any thing that isn't SIMD wont use AVX2 just like how it doesn't use AVX or SSE , MMX , etc now. Try keep thing in reality. Its still a struggle to find things that are compiled with AVXyet alonethings that are actually written with it in mind.

No.

Also its quite easy to find applications coded and compiled for AVX and AVX2 already.

 

[Olikan]

Also any thing that isn't SIMD wont use AVX2 just like how it doesn't use AVX or SSE , MMX , etc now. Try keep thing in reality. Its still a struggle to find things that are compiled with AVXyet alonethings that are actually written with it in mind.

software word is a total mess....i won't be surprised if there is programs beeing coded in x87 even today

 

[mikk]

Valid points. But one thing I can agree with Piroko is that I have almost never seen driver-based performance improvements for Intel graphics. Perhaps they don't know how. Or maybe its just that their hardware is at a low performance point that drivers can't do more. Or that they release everything they can at launch.

Obviously you never tested. HD3000 gained 10% after a couple of months with a new driver series. HD4000 similar, depending on the game there are huge increases from launch to now. Some games suffered from the famous low clock issue. In this games like UT2004, Half-Life 2, Gothic 2, Far Cry, recently Far Cry 2 you see a doubling or more with the new series for Ivy Bridge and Haswell. Also games like Civilization 5 or Dirt series gained a couple of fps on a HD4000 since launch a year ago.