Intel Amber Lake

Page 3 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Mar 10, 2004
28,284
180
126
#55
Not much different according to you.
Well, you said it was 7W and not 5W. I don't know where you got that number, though.
So that's technically your claim, not mine. :)

It's all preliminary info anyway.

I doubt there will be much difference between the two in reality.
 
Mar 10, 2004
28,284
180
126
#56
How often do mobile chips run at base clock under load?

I know desktop chips hardly ever run at base clock under load.
 

mikk

Platinum Member
May 15, 2012
2,396
25
126
#57
Well, you said it was 7W and not 5W. I don't know where you got that number, though.
So that's technically your claim, not mine. :)

I said 7W because 8510Y is a 7W SKU and this number comes from Intels graphics driver. What is your problem exactly?
 
Mar 10, 2004
28,284
180
126
#58
I only questioned the 40% more power number. I don't know where you went.
 

mikk

Platinum Member
May 15, 2012
2,396
25
126
#59
Driver inf 24.20.100.6222, it's accurate therefore and explains the high base clock.
 
Mar 11, 2000
22,585
0
106
#60
Has anyone seen a review of an Amber Lake machine yet. I know there are a few models like the Dells out there, but it seems just about nobody actually has tested ones for a real review.
 
Mar 10, 2004
28,284
180
126
#65
Is there a proper comparison to go by?
 
Mar 11, 2000
22,585
0
106
#67
Is there a proper comparison to go by?
What do you mean by proper?

The MacBook Kaby Lake i7-7Y75 gets around 8700/4500 in OS X, but that's with TDP up with 1.4 GHz base clock.

There is this Windows test of an i7-7Y75 with a normal 1.3 GHz base clock that scores 7156/4086:

https://browser.geekbench.com/v4/cpu/6135564

So, I guess the Google one for the i7-8500Y, if using the normal TDP, would represent an improvement of maybe 12% multi-core. The score is worse single-core though for whatever reason.
 
Mar 10, 2004
28,284
180
126
#68
Well, it's Android 9. So I guess a proper comparison would be to another Android bench.
 
Mar 11, 2000
22,585
0
106
#69
Here we go:

Celeron 3965Y Google Nocturne: 3476/2012 <-- Ouch!
m3-8100Y Google Nocturne: 7264/3885
i5-8200Y Google Nocturne: 8164/3909
i7-8500Y Google Nocturne: 8064/3764

https://browser.geekbench.com/v4/cpu/search?utf8=✓&q=nocturne

The best score is actually the i5-8200Y. It seems they ran the i7-8500Y benchmarks repeatedly and they always came up short compared to the single i5-8200Y score. Maybe throttling is an issue.

Peak speeds for the m3, i5, and i7 are 3.4 GHz, 3.9 GHz, and 4.2 GHz respectively.
 
Last edited:
Mar 10, 2004
28,284
180
126
#70
Yeah, the 8500Y and 8200Y are essentially the same chip when you put them in a tablet.
Cooling makes all the difference.
 

IntelUser2000

Elite Member
Oct 14, 2003
5,760
78
126
#71
Perhaps its throttling due to it being in a tablet? or that its running Android.
Not really the same device. Geekbench page shows "Google nocturne", which is the codename for Google's next generation Pixel device, Pixel Slate.

Performance gain nonexistent in single thread is a big issue. The nocturne and the XPS 13 is a pre-release device though, so it still has a chance to get better. We'll see.

It has 1w 1080P panels as an option, perhaps the ones Intel showed at computex?
Yea it is. The regular 1080p display is 2W. Having an option means the 1W display is more expensive.
 
Last edited:

ZGR

Golden Member
Oct 26, 2012
1,786
34
106
#72
How often do mobile chips run at base clock under load?

I know desktop chips hardly ever run at base clock under load.
Hardly ever as well. It is kind of annoying actually. My laptops love to maintain max turbo even if my CPU is hot. I usually end up turning off turbo or lowering turbo since I like to stay cool and quiet.
 

jpiniero

Diamond Member
Oct 1, 2010
5,995
125
126
#73
Clock speeds reported for one of the 8500Y (on the MT test?) was a median of 3053 and max of 3082. The max clock speed might be set to 3.1.

Meanwhile a Pixelbook with the 7y75 I saw had a max of 3558, which makes sense given the max turbo of 3.6.
 

IntelUser2000

Elite Member
Oct 14, 2003
5,760
78
126
#74
Clock speeds reported for one of the 8500Y (on the MT test?) was a median of 3053 and max of 3082. The max clock speed might be set to 3.1.

Meanwhile a Pixelbook with the 7y75 I saw had a max of 3558, which makes sense given the max turbo of 3.6.
How did you get the clock speed number? Its not from Geekbench is it?

Sidenote: Intel never gave up on x86 Android development even after cancelling Broxton. Their efforts continued elsewhere. Brief search shows Android Oreo as the latest branch. Makes sense Google Nocturne is on Android 9.

Also, performance on Android should be pretty good for Geekbench. Its generally better than on Windows for Intel chips.
 


ASK THE COMMUNITY

TRENDING THREADS