What others?
![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
AMD launches Zen+ 12nm Ryzen and X470 motherboards
- Thread starter _mockingbird
- Start date
What others?
Just for one example, the i7 6700K to i7 7700k. Same performance, slightly less power, no IPC gain.
What about for tasks that aren't games since games aren't the only thing to get a computer for? Also if you really want to impact gaming performance, the CPU is the second or third thing to change considering the overall difference between the highest and lowest price Intel CPU and the Highest and lowest price AMD CPU in games is largely inconsequential. The GPU is always most important and has the most impact in frames per second scores anyway. Also they tested this on a potentially performance limiting motherboard, (I know, XFR is some sort of AMD mandated evil overclock or something and must be stomped into the ground) so I have to wonder on what the scores are like on a sensible motherboard for the product.
A320 supports XFR 2. This is in the slides that were leaked days ago.
I might be confusing XFR with various precision boost, but there is no way the a320 will maximize stock protocols of Ryzen+ like the x470 or b450.
Some sites as videocardz are having trouble with the math of percentages. 175% is only a 8% faster than 161%, not 14% faster. The graph is correct.
You cannot do 175% - 161%, you have to do 175/161
The new Precission boost overdrive requires 400 series mobos. But what have been I saying since past year? What full performance would be only possible on new 400-series mobos? Funny that some people tried to crush me for saying that!
In any case overclocking in france esitmates a X470 mobo would add only 1--2% performance to Canard review results.
Ce qui implique que les CPU en question avaient leur Precision Boost désactivé, cette technologie n’étant que sur les cartes série 400. Cela ne change pas beaucoup les performances, 1 à 2% grand maximum, et ne change pas le ratio perf/conso en demi-teinte.
Guru3D is making the same mistake. And since Guru3D wrongly computes performance, they also get wrong efficiency and are claiming that efficiency has improved a bit when it is just the contrary. Efficiency is worse in this '12nm' series.
It is particularly worse in the 2600X case. The 2600X is nearly 9% faster than 1600X on applications, but consumes 20% more power.
_mockingbird
Gawd
- Joined
- Feb 20, 2017
- Messages
- 992
What I want to know is why CanardPC is testing the processors on an A320 motherboard.
...too cheap to buy an X370 motherboard for testing?
...too cheap to buy an X370 motherboard for testing?
I think they were worried about false MCE-like presets in the early review bioses. Gigabyte did this in their F2 review bios of the CFL, which had MCE on by default with no option of even disabling it.
It' a weird game that MB manufactures have been playing lately to make their motherboards look faster.
CPC Hardware explained before publication why are using A320 instead X370/X470: To avoid the omnipresent cheats in the X370/X470's BIOS ("Pour éviter les cheats à la con omniprésents dans les BIOS des X370/X470").
The overclockingmadeinfrance magazine also mentions that using a 400-series mobo wouldn't really change the performance or efficiency results, performance would change only 1 or 2% top maximum ("Cela ne change pas beaucoup les performances, 1 à 2% grand maximum").
Gideon
2[H]4U
- Joined
- Apr 13, 2006
- Messages
- 3,557
It's CanardPC their results rarely reflect reality, also using translate on a tech site sucks. One of the main reasons I came here so many years ago to the [H] was that they did reviews that actually reflected reality.
_mockingbird
Gawd
- Joined
- Feb 20, 2017
- Messages
- 992
Was Core i7-7700K also tested using the cheapest H110 motherboard that CanardPC can find?
Of course not, omnipresent cheats only affect AMD products, on Intel boards they are just features.
How would they know since they never tested it? So they must be just guessing, wonder how much more guessing is done in their reviews.
So exactly how much guessing is done on that CanardPC site, or shall we just deflect and redirect as always.
Master_shake_
Fully [H]
- Joined
- Apr 9, 2012
- Messages
- 17,794
sirmonkey1985
[H]ard|DCer of the Month - July 2010
- Joined
- Sep 13, 2008
- Messages
- 22,414
And roughly 20-25w of that is the SOC.
So glad we can ignore the comparison review right here on [H] showing around 1% difference between 6700k and 7700k. All hail the CPC Overlords of truth.
https://www.hardocp.com/article/2016/12/09/intel_kaby_lake_core_i77700k_ipc_review/4
https://www.hardocp.com/article/2016/12/09/intel_kaby_lake_core_i77700k_ipc_review/4
Master_shake_
Fully [H]
- Joined
- Apr 9, 2012
- Messages
- 17,794
Yes according to tech power up whom I believe is more reputable than cpc.
Master_shake_
Fully [H]
- Joined
- Apr 9, 2012
- Messages
- 17,794
_mockingbird
Gawd
- Joined
- Feb 20, 2017
- Messages
- 992
So, is there going to be an announcement (for Ryzen 2nd gen) soon or is AMD waiting until the last second, hoping not to disrupt sales of Ryzen 1st gen?
jenosmaverick
Limp Gawd
- Joined
- May 24, 2017
- Messages
- 142
Uhm so zen+ will consume more power than Coffee lakes at similar clocks? How bout heat?
OrangeKhrush
[H]ard|Gawd
- Joined
- Dec 15, 2016
- Messages
- 1,673
Wait what? so a budget board with less chokes and VRM's which are also lower quality is not going to make a difference, yet you sit here and tell me with a straight face it was okay to bench a 8700K using a board that is 7.5K in my country with MCE. on a A320 basically all the benefits of a X chip are going to be negated, no overclocks, no turbo states including XFR and 2.0 stages and you are going to have Controller issues.
A320's will give a fair baseline but it is going to come at the cost of a lot of performance. I have been in the testing environment and motherboards are not created equally.
OrangeKhrush
[H]ard|Gawd
- Joined
- Dec 15, 2016
- Messages
- 1,673
What state? in Speed step idle Intel aggressively throttles to 800mhz which is why our testings turn off all intel throttles and we go for EPS12V readings, the result is doubling the idle readings.
DuronBurgerMan
[H]ard|Gawd
- Joined
- Mar 13, 2017
- Messages
- 1,340
I'm waiting for the [H] review anyway. zOMG it's a few percent more/less efficient! Yeah, who cares? I want to see how this thing OCs. If 4.4GHz+ I'm interested. If it's a dud and does like 4.1... then shame on AMD. We'll see soon enough.
(i) TPU didn't test Zen+. CPC did.
(ii) I commented before on the weird results that TPU measures for Zen and other chips.
(iii) I can add now TPU is one of those tech sites that fail at basic math
Last edited:
Overclocking is a non-issue because CPC hardware tested all chips on stock settings.
300-series motherboards support both XFR 2 and Precision Boost 2. The only feature wasn't activated during the review was the new overdrive/enhancement boost, because that requires a 400-series motherboard. But as mentioned before
Last edited:
Master_shake_
Fully [H]
- Joined
- Apr 9, 2012
- Messages
- 17,794
Oh juan...
Sorry man it's 152 watts.
It really is. That what they measured that's what it is.
And yes that's more watts than the 2700x too.
Watts that about?
Incorrect measurements aren't valid. And those measurements coming from a site that doesn't know basic math isn't helping much.
Master_shake_
Fully [H]
- Joined
- Apr 9, 2012
- Messages
- 17,794
it doesn't fit my world view so it is wrong.
math isn't needed when all you are doing is looking at the screen of a multi-meter and recording the number.
which is 152 by the way.
Digital Viper-X-
[H]F Junkie
- Joined
- Dec 9, 2000
- Messages
- 15,116
Wouldn't it depend on the comparison?
Since their baseline is the 7600k @ 100%, Zen is 150% and Zen+ is 170%, then one can say, in this test, Zen+ is 20% faster than Zen. (Since we are comparing against the 100% baseline) which is why I don't like those comparisons at all
Ce qui implique que les CPU en question avaient leur Precision Boost désactivé, cette technologie n’étant que sur les cartes série 400. Cela ne change pas beaucoup les performances, 1 à 2% grand maximum, et ne change pas le ratio perf/conso en demi-teinte.
This implies that the CPUs in question had their Precision Boost [Overdrive] disabled, this technology being only on the 400 series motherboards. This does not change much the performances, 1 to 2% as top maximum, and does not change the ratio perf / consumption a bit.
How exactly is this not a guess? Since they admit to this being a guess and not actually tested, why should it be taken seriously. How do they know? Where is the scientific basis of guessing about hardware they didn't test on? Since they are fine with just pulling numbers out of thin air, what other numbers do they just pull out of thin air? This will lead to questions like, do they actually test hardware or just guess at testing hardware? How many other of their reviews have numbers published that weren't actually tested? Where does the guessing stop then?
More info directly from CPCHardware.
They confirm that tested 400-series mobos, and repeat again why rejected the results with the X470
Don't expect miracles from new features active only on 400-series mobos such as "XFR2 enhanced"
Confirmation that X470 mobo with stock specs would increase performance by 1% or 2% over the results that they published in the magazine. The performance can increase by 5--8% when using auto-overclocking features
They confirm that tested 400-series mobos, and repeat again why rejected the results with the X470
Don't expect miracles from new features active only on 400-series mobos such as "XFR2 enhanced"
Confirmation that X470 mobo with stock specs would increase performance by 1% or 2% over the results that they published in the magazine. The performance can increase by 5--8% when using auto-overclocking features