Overclock.net › Forums › Industry News › Hardware News › [Various] Intel Haswell Reviews
New Posts  All Forums:Forum Nav:

[Various] Intel Haswell Reviews - Page 35

post #341 of 858
Quote:
Originally Posted by sdlvx View Post

The two are related. MS doesn't have enough people with powerful hardware to add large amounts of features to Windows, because if they do it'll be "bloat." So, instead, they try and reach a new market.

Intel then turns around and says, "no one needs this much power anymore!" and then they don't release chips.

I think a lot of this "these CPUs are fast enough!" mentality is like saying 640k is enough. Yeah, it was at the time, but software will use additional hardware if enough people have it.

Ask yourself what you think Windows 8 would look like if Intel quads were priced like i3s, hexes were priced like i5s, and 8 cores were at 3930k price levels. My guess is that we'd be seeing some sort of fancy new 3d GUI with cool features (like a modern Linux distro), DirectX coming up with some sort of solution to use all those cores through an easy to use API, etc.

It's a real shame, it feels like they're both conspiring against enthusiasts. I feel like they look at how big Apple has become by selling consumerist crap instead of developing world class processors and the most popular operating system (hate it or not) of all time, so they're trying to emulate Apple.

It's a shame really, because MS and Intel are going to alienate the hardcore geeks and enthusiasts, and they've always been the ones there to recommend those products to other people.

But to say we have enough performance now is crazy. The software just simply isn't there yet. I could take a Pentium 4, run DOS, and then go "look, this P4 is so strong that we don't need anything better!", but I'm pretty sure I"m not alone in thinking that I enjoy a modern OS and the software that we run now as opposed to just having a fast, ancient OS running programs in 640k of ram.

And it's never going to be there, Intel wants to sell cheap, disposable crap that no one can upgrade to be like Apple, and Microsoft wants to have the big OS that's competing with iOS. AMD is just as guilty, but at least it seems like they're not giving up on enthusiasts.




Strongly agree with this sentiment.

It's the pursuit of short term profit in the case of Microsoft (to preserve their huge margins on Office and Windows) and the desire to keep the status quo (Intel with it's margins on servers and desktop processors). There's a lot of things that could be done, perhaps to enhance our productivity, what computer gaming could feasibly offer, etc with more processing power. Ironically this pursuit of short term profits may end up hurting the companies themselves and the very margins that there were trying to protect (ex: look at Windows 8 and the failed attempt to unify the desktop/mobile).

Looking into the future, I see this computer hardware ending up like cars ... a very mature industry with a handful of enthusiasts. There will be enthusiast PCs like there are enthusiast aftermarket car parts. Year over year gains will be much less and more emphasis will be put on other things like aesthetics. For most, the computer will be as mundane as say, a refrigerator - an appliance, very important yes, but just an appliance.

Of course, the other limitation looking in the future is that we are getting closer to the atomic scale. The laws of physics do dictate what can happen I fear. Leakage and heat in particular are problems that will only worsen as the die shrinks continue to come in, which means diminishing returns. Also, there are upper limits, like the Heisenberg Uncertainty Principle. However right now, in my opinion, the issue is not so much the technical issues as it is the priorities - namely if society were to value such performance, well, the needed R&D would get done.

Look at the rapid advancement of ARM processors in the mobile sector for example, very competitive, and lots of performance gains in a short period of time. Yes they are at the point where they can make these gains (and I would imagine that in the future, the rate of performance gains will begin to diminish generation over generation), but it is an example of a sector where well, performance still matters and what is possible.
Trooper Typhoon
(20 items)
 
  
CPUMotherboardGraphicsGraphics
5960X X99A Godlike MSI 1080 Ti Lightning MSI 1080 Ti Lightning 
RAMHard DriveHard DriveHard Drive
G.Skill Trident Z 32 Gb Samsung SM843T 960 GB Western Digital Caviar Black 2Tb Samsung 850 Pro 
Hard DriveOptical DriveCoolingCooling
Samsung SV843 960 GB LG WH14NS40 Cryorig R1 Ultimate 9x Gentle Typhoon 1850 rpm on case 
OSMonitorKeyboardPower
Windows 7 Pro x64 Korean 27" 2560x1440 Ducky Legend with Vortex PBT Doubleshot Backlit... EVGA 1300W G2 
CaseMouseAudioOther
Cooler Master Storm Trooper Logitech G502 Proteus Asus Xonar Essence STX Lamptron Fanatic Fan Controller  
  hide details  
Reply
Trooper Typhoon
(20 items)
 
  
CPUMotherboardGraphicsGraphics
5960X X99A Godlike MSI 1080 Ti Lightning MSI 1080 Ti Lightning 
RAMHard DriveHard DriveHard Drive
G.Skill Trident Z 32 Gb Samsung SM843T 960 GB Western Digital Caviar Black 2Tb Samsung 850 Pro 
Hard DriveOptical DriveCoolingCooling
Samsung SV843 960 GB LG WH14NS40 Cryorig R1 Ultimate 9x Gentle Typhoon 1850 rpm on case 
OSMonitorKeyboardPower
Windows 7 Pro x64 Korean 27" 2560x1440 Ducky Legend with Vortex PBT Doubleshot Backlit... EVGA 1300W G2 
CaseMouseAudioOther
Cooler Master Storm Trooper Logitech G502 Proteus Asus Xonar Essence STX Lamptron Fanatic Fan Controller  
  hide details  
Reply
post #342 of 858
Where are those temperatures coming from?

http://www.techpowerup.com/reviews/Intel/Core_i7-4770K_Haswell/9.html

My 4.5ghz 3570K runs at a max of 82C under full load. $30 air cooler..
Edited by demoship - 6/1/13 at 11:31pm
post #343 of 858
Quote:
Originally Posted by PostalTwinkie View Post

You are forgetting the IPC improvement over Ivy, which was about the same as Ivy over Sandy. A rough idea for you....

4.8Ghz Sandy = 4.6Ghz Ivy = 4.4Ghz Haswell.

(Hold off on the flame throwers guys, I am more illustrating a point than carving fact into stone)

What I am getting at is it isn't all about clock speed, but what you do with it.

I know that, kinda early to tell but I wonder what is the better considering it's faster clock for clock but seem to oc less. Overall performance would be less? temps are high but is it because of the Igpu, can you disable it? I've seen review they hardly had time to overclock it, so have to take em with a grain of salts.
post #344 of 858
Quote:
Originally Posted by demoship View Post

Where are those temperatures coming from?

http://www.techpowerup.com/reviews/Intel/Core_i7-4770K_Haswell/9.html

My 4.5ghz 3570K runs at a max of 82C under full load. $30 air cooler..

Maybe it's because you've newer batch of the chip? I saw a picture at TPU the i7-3770K is Costa Rica built whilst the i5-3570K ain't (if that matters at all).
post #345 of 858
Quote:
Originally Posted by TPU 
'Temperatures—the well-known culprit of Ivy Bridge chips. We were anxious to find out if Intel solved this problem with the Haswell series, which thankfully seems to be the case.
Quote:
Originally Posted by candy_van View Post


.__.

Indeed.

However, it seems that TPU's puzzling declaration directly stems from their decision to test @ 4.2GHz max, as that's just before temps really start to ramp up on Haswell...
Julia
(11 items)
 
Centurion (Dead)
(11 items)
 
 
CPUMotherboardGraphicsRAM
AMD Athlon 64 3700+ EPoX EP-9NPA+ Ultra EVGA GeForce 6800 GS (256-P2-N389-AX) CORSAIR VS1GB400C3 
Hard DriveOptical DriveOSMonitor
Western Digital Caviar SE WD1600JS NEC ND-3520A Windows XP Dell M780 
PowerCaseOther
FSP Group SAGA+ 450R Cooler Master Centurion 5 w/ Window (CAC-T05-WW) Samsung Floppy Drive (SFD321B/LBL1) 
  hide details  
Reply
Julia
(11 items)
 
Centurion (Dead)
(11 items)
 
 
CPUMotherboardGraphicsRAM
AMD Athlon 64 3700+ EPoX EP-9NPA+ Ultra EVGA GeForce 6800 GS (256-P2-N389-AX) CORSAIR VS1GB400C3 
Hard DriveOptical DriveOSMonitor
Western Digital Caviar SE WD1600JS NEC ND-3520A Windows XP Dell M780 
PowerCaseOther
FSP Group SAGA+ 450R Cooler Master Centurion 5 w/ Window (CAC-T05-WW) Samsung Floppy Drive (SFD321B/LBL1) 
  hide details  
Reply
post #346 of 858
Quote:
Originally Posted by PostalTwinkie View Post

What on gods green earth are you talking about? It was made public that Intel released to OEMs the first week of April. Who cares about an NDA? An NDA on the K series has nothing to do with the fact that Intel released to OEMs almost two months ago.

You are literally trying to base the foundation of your argument, that Intel is playing some "Games", around Intel leading their launch with the K series. When it is fact that they didn't lead the launch with the K series, but with the non-K and mobile variants back in April, almost 2 months ago!

How can you sit and say "Intel is playing games because........K series first!", when that couldn't be the farthest from the truth?

Seriously!

Launch = available product. K series will be Haswell first available product.


Rough paraphrase - "Hey guys Haswell is a mobile chip any disappointment in the enthusiast line up is just sour grapes. You enthusiasts are tiny and insignificant."

Well guess what, they didn't lead the retail availability with Haswell notebooks. K series is the one being reviewed and it's what will be sold next week. Makes sense to then see how it stacks up to previous K chips. "Playing games" was in reference to the Saturday NDA lift for minimal traditional press coverage and to a lesser extent transitioning some Intel reps a few weeks before launch (good excuse to not give a company backed response to Haswell K impressions).
Everyday PC
(13 items)
 
  
CPUMotherboardGraphicsRAM
i5-3470 (+4 Turbo bins) Biostar T77XE3 Radeon 7950 16GB DDR3 1600 
Hard DriveOSMonitorKeyboard
Samsung F3 1TB Windows 7 Pro x64 Samsung 24" TN MS Natural Elite 4000 
PowerCaseMouse
Antec HPG 900W Lian-li K57W Logitech G500 
  hide details  
Reply
Everyday PC
(13 items)
 
  
CPUMotherboardGraphicsRAM
i5-3470 (+4 Turbo bins) Biostar T77XE3 Radeon 7950 16GB DDR3 1600 
Hard DriveOSMonitorKeyboard
Samsung F3 1TB Windows 7 Pro x64 Samsung 24" TN MS Natural Elite 4000 
PowerCaseMouse
Antec HPG 900W Lian-li K57W Logitech G500 
  hide details  
Reply
post #347 of 858
Quote:
Originally Posted by faiyez View Post

Relax guys. There's still Ivy Bridge-E.


That's what everyone said about Haswell last year.  wink.gif

post #348 of 858
Off topic, does anyone know how long it is before ek comes out with blocks for the z87 mobos?

I dont want to order a mobo and end up with a brand new board that I cant watercool.
post #349 of 858
Quote:
Originally Posted by PostalTwinkie View Post

You are forgetting the IPC improvement over Ivy, which was about the same as Ivy over Sandy. A rough idea for you....

4.8Ghz Sandy = 4.6Ghz Ivy = 4.4Ghz Haswell.

(Hold off on the flame throwers guys, I am more illustrating a point than carving fact into stone)

What I am getting at is it isn't all about clock speed, but what you do with it.

The turbos do not work the same, and you can see that here:


3770k gain in x264 with turbo enabled: 5.3%
4770k gain in x264 with turno enabled: 10.8%.

3770k gain in winrar with turbo: 11.7%
4770k gain in winrar with turbo: 5.7%
(I have no idea, so turbo works worse with Haswell in winrar?)

3770k Sandra turbo: 8%
4770k Sandra turbo: 11%

3770k Super Pi turbo: 10.8%
4770k Super Pi turbo: 10.8%

http://in.inpai.com.cn/thread-5651-1-1.html

The only way to accurately tell a chip's IPC increase is to completely disable turbo and set them to the same clocks. None of these reviews seem to be doing that, and it's entirely possible that these larger gains you think you're seeing are actually the chip turboing better.

It is bad for us enthusiasts, because people seem to think that it's just flat out IPC increase and that a 4.8ghz IB is the same as a 4.6ghz Haswell. The problem is that if Haswell is more lenient with turning turbo down, the IPC is not what you think it is.

My point, basically, is that when turbo is enabled, you have no way of telling (in modern reviews at least) what frequency the chip is running at. I'm having a hard time tonight because I"m tired, but it's basically like this:

3770k advertised speeds: 3.5ghz base 3.9ghz boost
4770k advertised speeds: 3.5ghz base 3.9ghz boost

Everyone assumes that means the turbo works the same way, it clearly doesn't from the link I posted.

So, people basically take 4770k running at 3.7ghz and 3770k at 3.5ghz on the same benchmark, ignore the speed difference completely, and then chalk it up to IPC increase.

Haswell is not the only one that does this, Ivy Bridge has had IPC regressions when locked to the same clock speed:



But if you even mention that IB has situations where it doesn't offer any better IPC, it's like people swarm on you. They somehow think you're some sort of insane fanboy because you're questioning the effect of turbo on a CPU review and the merits of using that chip in a setup where turbo isn't enabled (overclocked) and they just get mad as hell.

I do realize there have been IPC increases with IB and there are some with Haswell. But we don't know what that number is without seeing a comparison with them locked at both frequencies. Until then, I refused to ut an IPC number on anything. Maybe overall performance, including turbo boost, but to simply say that two CPUs with the same advertised clock speeds and differently working turbos are actually working at the same speeds is a massive jump in logic, specially considering there are a lot of pstates between 3.5ghz and 3.9ghz and CPU boost is a lot more complex than

if TDP < 84 clockspeed = 3.9ghz else clockspeed = 3.5ghz (sorry for lame pseudocode)
Tyrant Kuma
(13 items)
 
Starscythe
(13 items)
 
Mobility
(6 items)
 
CPUCPUCPUCPU
Opteron 8431 Opteron 8431 Opteron 8431 Opteron 8431 
MotherboardGraphicsRAMHard Drive
Super Micro h8qme-2+ Sapphire 4870 Samsung DDR2 ECC 667mhz western digital caviar blue 
CoolingCoolingOSPower
FX 8350 stock cooler FX 6300 stock cooler Gentoo Linux 2x Silverstone PSU 
Case
Custom Fabbed Steel Case 
CPUMotherboardGraphicsRAM
AMD FX 8350 gigabyte 990FXA-UD5 7970 Mushkin Enhanced Blackline 16GB 
OSOSMonitorKeyboard
Gentoo Linux Windows 7 x64 Yamakasi Catleap Q270 Corsair K90 
PowerCaseMouseAudio
Silverstone Strider Gold Evolution 850W 80 PLUS... Antec 1200 Logitech G9x Asus Xonar D2X 
Other
XSPC RS360 Raystorm with custom radiator mounts 
CPUGraphicsRAMOS
a4 5000 Radeon HD 8330 8GB G. Skill DDR3 1600 9-9-9 Gentoo Linux 
OSOther
Windows 7 Lenovo Thinkpad x140e 
  hide details  
Reply
Tyrant Kuma
(13 items)
 
Starscythe
(13 items)
 
Mobility
(6 items)
 
CPUCPUCPUCPU
Opteron 8431 Opteron 8431 Opteron 8431 Opteron 8431 
MotherboardGraphicsRAMHard Drive
Super Micro h8qme-2+ Sapphire 4870 Samsung DDR2 ECC 667mhz western digital caviar blue 
CoolingCoolingOSPower
FX 8350 stock cooler FX 6300 stock cooler Gentoo Linux 2x Silverstone PSU 
Case
Custom Fabbed Steel Case 
CPUMotherboardGraphicsRAM
AMD FX 8350 gigabyte 990FXA-UD5 7970 Mushkin Enhanced Blackline 16GB 
OSOSMonitorKeyboard
Gentoo Linux Windows 7 x64 Yamakasi Catleap Q270 Corsair K90 
PowerCaseMouseAudio
Silverstone Strider Gold Evolution 850W 80 PLUS... Antec 1200 Logitech G9x Asus Xonar D2X 
Other
XSPC RS360 Raystorm with custom radiator mounts 
CPUGraphicsRAMOS
a4 5000 Radeon HD 8330 8GB G. Skill DDR3 1600 9-9-9 Gentoo Linux 
OSOther
Windows 7 Lenovo Thinkpad x140e 
  hide details  
Reply
post #350 of 858
Quote:
Originally Posted by sdlvx View Post

The turbos do not work the same, and you can see that here:


3770k gain in x264 with turbo enabled: 5.3%
4770k gain in x264 with turno enabled: 10.8%.

3770k gain in winrar with turbo: 11.7%
4770k gain in winrar with turbo: 5.7%
(I have no idea, so turbo works worse with Haswell in winrar?)

3770k Sandra turbo: 8%
4770k Sandra turbo: 11%

3770k Super Pi turbo: 10.8%
4770k Super Pi turbo: 10.8%

http://in.inpai.com.cn/thread-5651-1-1.html

The only way to accurately tell a chip's IPC increase is to completely disable turbo and set them to the same clocks. None of these reviews seem to be doing that, and it's entirely possible that these larger gains you think you're seeing are actually the chip turboing better.

It is bad for us enthusiasts, because people seem to think that it's just flat out IPC increase and that a 4.8ghz IB is the same as a 4.6ghz Haswell. The problem is that if Haswell is more lenient with turning turbo down, the IPC is not what you think it is.

My point, basically, is that when turbo is enabled, you have no way of telling (in modern reviews at least) what frequency the chip is running at. I'm having a hard time tonight because I"m tired, but it's basically like this:

3770k advertised speeds: 3.5ghz base 3.9ghz boost
4770k advertised speeds: 3.5ghz base 3.9ghz boost

Everyone assumes that means the turbo works the same way, it clearly doesn't from the link I posted.

So, people basically take 4770k running at 3.7ghz and 3770k at 3.5ghz on the same benchmark, ignore the speed difference completely, and then chalk it up to IPC increase.

Haswell is not the only one that does this, Ivy Bridge has had IPC regressions when locked to the same clock speed:



But if you even mention that IB has situations where it doesn't offer any better IPC, it's like people swarm on you. They somehow think you're some sort of insane fanboy because you're questioning the effect of turbo on a CPU review and the merits of using that chip in a setup where turbo isn't enabled (overclocked) and they just get mad as hell.

I do realize there have been IPC increases with IB and there are some with Haswell. But we don't know what that number is without seeing a comparison with them locked at both frequencies. Until then, I refused to ut an IPC number on anything. Maybe overall performance, including turbo boost, but to simply say that two CPUs with the same advertised clock speeds and differently working turbos are actually working at the same speeds is a massive jump in logic, specially considering there are a lot of pstates between 3.5ghz and 3.9ghz and CPU boost is a lot more complex than

if TDP < 84 clockspeed = 3.9ghz else clockspeed = 3.5ghz (sorry for lame pseudocode)
http://www.hardocp.com/article/2013/06/01/intel_haswell_i74770k_ipc_overclocking_review/1#.Uar0PkCyAzQ
http://www.anandtech.com/show/7003/the-haswell-review-intel-core-i74770k-i54560k-tested/6

there is an IPC difference, but i still think this is intels "bulldozer". Quite a disappointment to enthusiasts.
Edited by th3illusiveman - 6/2/13 at 12:32am
Bender
(18 items)
 
  
CPUMotherboardGraphicsGraphics
Core i5-2500k @ 4.6Ghz Gigabyte Z68X-UD3H-B3 Sapphire AMD R9 290X Tri-X (1050/1300mhz)  EVGA GeForce GTX 1080 (How do you OC?) 
RAMHard DriveHard DriveOS
8GB 1600Mhz OCZ Vertex 3 240GB SSD OCZ Vertex 3 240GB SSD windows 10 64bit 
MonitorKeyboardPowerCase
Asus MG279Q Logitech G510  Corsair TX750 CM II 690 Advanced 
MouseMouse PadAudioAudio
CM Sentinel Advance  some big corsair one beyerdynamic DT 770 Pro 250 Ohm SoundBlaster Z Soundcard 
  hide details  
Reply
Bender
(18 items)
 
  
CPUMotherboardGraphicsGraphics
Core i5-2500k @ 4.6Ghz Gigabyte Z68X-UD3H-B3 Sapphire AMD R9 290X Tri-X (1050/1300mhz)  EVGA GeForce GTX 1080 (How do you OC?) 
RAMHard DriveHard DriveOS
8GB 1600Mhz OCZ Vertex 3 240GB SSD OCZ Vertex 3 240GB SSD windows 10 64bit 
MonitorKeyboardPowerCase
Asus MG279Q Logitech G510  Corsair TX750 CM II 690 Advanced 
MouseMouse PadAudioAudio
CM Sentinel Advance  some big corsair one beyerdynamic DT 770 Pro 250 Ohm SoundBlaster Z Soundcard 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Hardware News
Overclock.net › Forums › Industry News › Hardware News › [Various] Intel Haswell Reviews