Well just tested my i7 920 (Bloomfield) @ 4Ghz on CineBench and score 6.86, so Haswell should be a nice boost for me
And OpenGL test 76.29fps on a HD6970
Last edited:
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
***Official Intel Haswell Thread***
- Thread starter Yaayuh!
- Start date
Well just tested my i7 920 (Bloomfield) @ 4Ghz on CineBench and score 6.86, so Haswell should be a nice boost for me
And OpenGL test 76.29fps on a HD6970
If you are just Gaming a better GPU would be a better upgrade. I have a 4770k system I have just built running 4.6Ghz and it is no quicker than my 2500k system at 4.6Ghz both using a 780 GTX at 2560x1440. In gaming I cannot tell the difference.
I get 6.73 with a 4670k @ 4ghz (4 real cores 0 hyper threads)
5 on my old x6 1055t @ stock ( 6 real cores 0 hyper threads)
If you are upgrading you will want to skip the 4670k and go for the 4770k with hyper threading
Guess your hyper-threading makes a decent difference in programs that max all cores.
I'd imagine if you used a single thread I'd be faster
Same for me coming from a x6 1055t @ stock the only games that show any real difference are ones that only use 1-2 threads which is mostly old and f2p games, I get a massive boost in WOT which only uses 1 thread my fps jumped from around 40 to 70 in that game
For gaming I think a 4670k is enough as games don't really need hyper threading.
for rendering/transcoding/video editing then 4770k is a must really.
Heres 2 cpu usage shots I took and posted in another thread I might as well post there here to
skyrim @ max settings , no enb/fxaa/sweetfx but high res textures and a bunch of normal mods
so low cpu usage across the cores
Project cars on ultra with down sampling 4x AA method , 63 AI cars , wet weather ( game has a cpu demanding physics engine with ambient temperatures, altitude etc effecting the power of the cars and even their idle RPM! , real drive train simulation , real gearbox simulation etc all tied in to the physics and audio engines )
This can be considered a next gen game (releases in 2014) and the cpu usage is pretty high with so many AI cars but theres still a lot of headroom
Bioshock infinite @ ultra
Btw my cpu goes all the way upto around 60-64c in stress tests check how much lower is is during gaming lol
idling at 31-34c right now (hot day hot room)
Last edited:
the 4770k with hyper-threading sure.
the ones without hyper-threading like the 4670k are a lot slower at rendering , transcoding etc but perfectly fine for gaming
cool thing, i had upto 9c difference between cores on AIDA64 test, redid tests with my pc tower lying flat on its side and there's a 2c difference between cores. I thought i had the cooler on tight as it would go but apparently mavity still makes slight difference. Also 3c cooler.
Just got my Haswell system up and running and at default clock it score 8.01 in CineBench!!!, so faster than my i7 920 Bloomfield already!!!
give us some single thread results
4670k @ 3990mhz stupid blk is set to 100 but actually is 99.7 or something stupid
BTW MP ratio is how efficient your cpu is at multi coring so running 4 cores I lose 0.09 efficiency
the cpu usage graph is from the single core test, intel seems to round robbin between cores to keep even temps or something weird is happened like intel taking a single threaded program and splitting it 50% over to cores as best it can? in task manager cinebench was never ever over 25% cpu usage though which is correct for 1 of 4 cores
also the single core score *4 is slightly higher than my 4 core result hmm
No hyper-threading sucks
, not like you notice it in gaming though coming from a 6 core cpu I wouldn't say I regret not getting a 4770k but considering I could have afforded to maybe I should.
I don't do any video transcoding like I used to though so meh
Last edited:
How are people finding the temperatures as STOCK for Haswell over Ivy Bridge. I do not own a desktop so putting a new build together in a mini itx set up. Just worried about the heat issues I have been reading about with Haswell - do the heat issues only apply for overclocked CPUs? I do not intend to overclock....
4670k stock with sr1 pro heatsink had me around 50c max load.
clocked to 4ghz im about 10c higher max load but during gaming its only mid 40's as you can see from the screenshots in my post a few replies up
http://forums.overclockers.co.uk/showpost.php?p=24421040&postcount=1684
I'm guessing 4770k will be hotter because of hyper threading which the 4670k lacks
clocked to 4ghz im about 10c higher max load but during gaming its only mid 40's as you can see from the screenshots in my post a few replies up
http://forums.overclockers.co.uk/showpost.php?p=24421040&postcount=1684
I'm guessing 4770k will be hotter because of hyper threading which the 4670k lacks
Last edited:
that's a pretty damn good overclock at those volts the test to see if you have a good ocing cpu according to asus is 46x @ 1.25v I bet yours would do that without breaking a sweat
65c is well within safe temperatures the cpu doesn't start throttling until something stupid like 100c
some people are happily running 4.8ghz @ 80c + lol anything over 65 scares me
I just don't want that extra heat in my caseBtw your temps will be much less gaming, im 60-64C max load across the cores on mine but im gaming it's typically around the mid 40's
Last edited:
that's a pretty damn good overclock at those volts the test to see if you have a good ocing cpu according to asus is 46x @ 1.25v I bet yours would do that without breaking a sweat
65c is well within safe temperatures the cpu doesn't start throttling until something stupid like 100c
some people are happily running 4.8ghz @ 80c + lol anything over 65 scares me
I'm just happy I've got at least 1GHz over stock right now, will probably do more tweaking later in the year or so. I forsee I'll probably be replacing hardware in a few years anyways so that overclock should be plenty till then
My case is quite cool inside with airflow, using 2 Akasa black Apaches as intakes on full at the front, a black Apache as a out take at the back, the Enermax PSU drawing heat from the top out and I replaced the standard Akasa Nero heatsink fan with a Noctua NF-F12 PWM focused flow fan.
Case is pretty quiet now but that might be because I replaced my Radeon HD 6950 with a GTX 660 Ti
well personally I consider 3.8ghz to be stock as my cpu did that on stock volts and it's what the cpu gets boosted to if the cpus under 50% load so it's not a huge stretch to expect the other 2 cores to hit 3.8ghz..
I think my bios was set so when 4 cores are in use it only boosted to 3.6ghz but its still not a huge ask for 200mhz more on stock volts really.
BTW anything under 1.25v should be fine as long as you have a good heatsink , over 1.25v you need a massive heatsink or watercooling
I think my bios was set so when 4 cores are in use it only boosted to 3.6ghz but its still not a huge ask for 200mhz more on stock volts really.
BTW anything under 1.25v should be fine as long as you have a good heatsink , over 1.25v you need a massive heatsink or watercooling
Just got my Haswell system up and running and at default clock it score 8.01 in CineBench!!!, so faster than my i7 920 Bloomfield already!!!
With my Stable OC of 4.6Ghz, just scored 10.10 on CineBench!!
Hi, did another run of CineBench at with the old girl > 920:4.4ghz -
CineBench 11.5 -
OpenGL: 85.53FPS Ref. Match 99.3% / CPU: 7.61 / CPU[SINGLE CORE] 1.53 MPRatio: 4.96x
WoW, she showed a 4670K a clean pair of heel,s,
Oldphart.
CineBench 11.5 -
OpenGL: 85.53FPS Ref. Match 99.3% / CPU: 7.61 / CPU[SINGLE CORE] 1.53 MPRatio: 4.96x
WoW, she showed a 4670K a clean pair of heel,s,
Oldphart.
his hyper threading is giving him just over a whole cores worth of extra performance thats why.
look at the difference between 4670k and 4770k and the only difference at the same clock rate is 2mb cache and hyper threading.
single thread 4670k @ 4ghz = 1.72
single thread i7 920 @ 4.2ghz =1.52
hyperthreading doesn't scale in games like it does in synthetic benchmarks/rendering/video transcoding though AFAIK., still no point to go haswell from an 4.3ghz i7 920 though ^_^ was barely any point for me I only did it because core for core haswell it about twice as fast as my old 1055t was even with both cpus @ stock
anyway I came from a 1055t which gets 4.9 even at stock 3.2ghz in cinebench , I bet if I whacked it up to 4ghz it wouldn't be to far from my 4670k@4ghz cinebench score
Btw if he turned of hyper threading I bet his mp ratio would drop to around 3.91 efficiency our of a possible 4 thats just how much of a difference hyperthread is making
look at the difference between 4670k and 4770k and the only difference at the same clock rate is 2mb cache and hyper threading.
single thread 4670k @ 4ghz = 1.72
single thread i7 920 @ 4.2ghz =1.52
hyperthreading doesn't scale in games like it does in synthetic benchmarks/rendering/video transcoding though AFAIK., still no point to go haswell from an 4.3ghz i7 920 though ^_^ was barely any point for me I only did it because core for core haswell it about twice as fast as my old 1055t was even with both cpus @ stock
anyway I came from a 1055t which gets 4.9 even at stock 3.2ghz in cinebench , I bet if I whacked it up to 4ghz it wouldn't be to far from my 4670k@4ghz cinebench score
Btw if he turned of hyper threading I bet his mp ratio would drop to around 3.91 efficiency our of a possible 4 thats just how much of a difference hyperthread is making
Last edited:
his hyper threading is giving him just over a whole cores worth of extra performance thats why.
look at the difference between 4670k and 4770k and the only difference at the same clock rate is 2mb cache and hyper threading.
single thread 4670k @ 4ghz = 1.72
single thread i7 920 @ 4.2ghz =1.52
hyperthreading doesn't scale in games like it does in synthetic benchmarks/rendering/video transcoding though AFAIK., still no point to go haswell from an 4.3ghz i7 920 though ^_^ was barely any point for me I only did it because core for core haswell it about twice as fast as my old 1055t was even with both cpus @ stock
anyway I came from a 1055t which gets 4.9 even at stock 3.2ghz in cinebench , I bet if I whacked it up to 4ghz it wouldn't be to far from my 4670k@4ghz cinebench score
Hi arknor, lol just winding you up, watch them temp's if you go higher, enjoy.