Intel Amber Lake

Page 4 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

IntelUser2000

Elite Member
Oct 14, 2003
8,686
3,785
136
Put .gb4 at the end of the link url and you will get the raw data.

Thanks. I couldn't find a correlation between the clock speeds and the benchmark results. Maybe if we find out how the clocks were obtained then it'd be possible to analyze.

I'm interested in more package shots. The on-package PCH on Amberlake seems noticeably larger in comparison to the CPU die than the Skylake-gen Y + 200 series on-package PCH shots.
 

Eug

Lifer
Mar 11, 2000
23,586
1,000
126
It's a 7W SKU from Amber Lake.
Of an existing chip or a new one as jpiniero suggests?

The i5-8200Y at 5 W for example is 1.3/3.9 with UHD 615. The only UHD 617 I have found is that i7-8510Y, which presumably has a turbo boost speed well over 4 GHz.
 

Eug

Lifer
Mar 11, 2000
23,586
1,000
126
Interestingly, the new MB Air has Thunderbolt USB-C x 2 ports. I guess that must be using an additional Thunderbolt controller.
 

Eug

Lifer
Mar 11, 2000
23,586
1,000
126
AnandTech: https://www.anandtech.com/show/13531/apple-2018-macbook-air-announced

Under the hood details are a little harder to come by. Curiously, Apple is only offering a single CPU option here, an Intel Core i5. And while the company never names the specific processor models they use, the specifications here – 3.6GHz turbo with Intel UHD Graphics 617 – do not match any known Intel chip, even when factoring in various cTDP options. We’ve heard rumors of Intel putting together a Core i5-8210Y, and we’ve reached out to Intel to try to confirm.

At a minimum, it’s clear that this is one of Intel’s 5 Watt Y-series chips – almost certainly Amber Lake. Launched back in August, Amber Lake is Intel’s latest-generation 5W chips and goes under the 8th Gen Core branding. Compared to the Broadwell (5th Gen Core) chips in the previous MacBook Air, these chips represent a big step up in capabilities and performance, incorporating a newer CPU core design as well as a newer GPU design. However it’s also notable that the new chips, even in cTDP up mode, are also much lower power than the older 15W U-series chips Apple used, which means that processor power consumption should be significantly reduced – and the chips thinner as well – though the total performance gain won’t be quite as much as if Apple had stuck with U-series chips.

Using Amber Lake also means we’re once again looking at LPDDR3 memory for an Apple laptop. Apple outfits the standard model with 8GB of the stuff, with 16GB being an option. Amber Lake doesn’t support LPDDR4, so Apple’s options are to either stick with LPDDR3 or use less power efficient DDR4, which is something the company did do for the 15-inch MacBook Pro.
 

Eug

Lifer
Mar 11, 2000
23,586
1,000
126
Considering the Retina MacBook Air has a fan, the sustained performance is going to be significantly faster than the 12” MacBooks.

Also, no native 5K external screen support and no native Thunderbolt support.

Apple has been waiting for a chip like this for years, hanging onto the Broadwell chips in the line until they could get a Y chip that beat Broadwell U in performance.
 

asendra

Member
Nov 4, 2012
156
12
81
Considering the Retina MacBook Air has a fan, the sustained performance is going to be significantly faster than the 12” MacBooks.

Also, no native 5K external screen support and no native Thunderbolt support.

Apple has been waiting for a chip like this for years, hanging onto the Broadwell chips in the line until they could get a Y chip that beat Broadwell U in performance.

It doesn't have a fan I'm afraid. Watch the intro video again.
 

Eug

Lifer
Mar 11, 2000
23,586
1,000
126
It doesn't have a fan I'm afraid. Watch the intro video again.
Ah yes. I just watched that video again. The spaceship fan I remembered was actually in the Mac mini. It seems some articles have reported it has retained the fan, which is another reason why I was confused, but that appears incorrect.

So, if no fan, the MB Air will likely throttle significantly under sustained load. Someone (Notebookcheck) will have to do the repeated Cinebench test. For the record, my Core m3 7Y32 MacBook 12" drops only 7.7% after 10 runs of Cinebench on a wood table, and drops 4.3% on a granite table.

Someone who tried a Core i7 7Y75 with 10 runs of Cinebench saw a performance drop of 6.6% from peak.

macbook2017-cinebenchr15-m3-wood-png.727191


BTW, in that test, there was only a 3.7% difference in peak performance between the i7 and m3, and only a 4.9% difference in lowest performance between the i7 and m3, almost negligible for sustained performance in that test.
 

IntelUser2000

Elite Member
Oct 14, 2003
8,686
3,785
136
Is this real? If so, why is the performance so underwhelming?

The top 8500Y scores 4814/8692. Windows 64-bit result on HP Spectre Folio Convertible.

The top Windows score is 4375/8223 with Kabylake-Y so Amberlake is 10% better in ST. Might see it reach 5200-5300 in ST under Linux.

Peak non-overclocked 9900K is 6600, so there's only a 20-30% difference between the 9900K and 8500Y.
 
Last edited:

Eug

Lifer
Mar 11, 2000
23,586
1,000
126
The top 8500Y scores 4814/8692. Windows 64-bit result on HP Spectre Folio Convertible.

The top Windows score is 4375/8223 with Kabylake-Y so Amberlake is 10% better in ST. Might see it reach 5200-5300 in ST under Linux.

Peak non-overclocked 9900K is 6600, so there's only a 20-30% difference between the 9900K and 8500Y.
The 1.4/3.6 GHz Core i7-7Y75 MacBook gets 4559/8744 in Geekbench 4.

The i5-8210Y is 1.6/3.6 GHz with the same 4 MB L2 cache. In fact, all the specs seem to be exactly the same except for the base clock and the TDP. And for the TDP it doesn’t even seem really different because the i7-7Y75 can be configured at 7 W 1.6/3.6 GHz too. IOW, the i5-8210Y is spec’d EXACTLY the same as the i7-7Y75 with TDPup. The memory specs are the same too. Even the GPU specs are exactly the same (aside from the name). So is the i5-8210Y really even a new chip?

If the i5-8210Y were to reach 5250/10000 in macOS, it would be doing extremely well, but I’m starting to think that may be a tad optimistic.
 

Eug

Lifer
Mar 11, 2000
23,586
1,000
126
Uh huh. And I said the 8500Y. We're talking about ST.
That’s fine but I specifically wasn't talking about the 8500Y because that may not be the appropriate part to compare.

The i7-8500Y is a 1.5/4.2 GHz part at 5 Watts, and 1.6/4.2 GHz at 7 Watts.

The i5-8200Y doesn't seem like a proper match either, since it's a 1.3/3.9 GHz part, and 1.6/3.9 GHz at 7 Watts.

The point of my post was that it seemed the best match is in fact the i7-7Y75, which at 7 Watts is 1.6/3.6 Watts, which happens to be the exact same specs as the i5-8210Y. The only difference I can see is the max memory speed, so there's that.
 
Last edited:

Eug

Lifer
Mar 11, 2000
23,586
1,000
126
It doesn't have a fan I'm afraid. Watch the intro video again.
The internet is convinced it has a fan. We shall see, as some tech sites still say it has no fan, and other tech sites say it does have a fan based on the intro video (which shows a round structure in one corner).

This is important of course if we want to compare the performance vs. Kaby Lake Y.
 

IntelUser2000

Elite Member
Oct 14, 2003
8,686
3,785
136
The engadget review has a picture of the presentation for the Macbook Air. One shows a single fan on the left side.

I see why you wanted to compare using the 8210Y. It's what the Macbook Air uses.
 

asendra

Member
Nov 4, 2012
156
12
81
Nice to see that it has a fan, although it seems it’s not connected directly to the CPU heatshink?

I still think they should have used the 15w ones and ditch the non touchbar macbook pro
 

Eug

Lifer
Mar 11, 2000
23,586
1,000
126
Nice to see that it has a fan, although it seems it’s not connected directly to the CPU heatshink?

I still think they should have used the 15w ones and ditch the non touchbar macbook pro
There are no 15 Watt ones to use, unless they go quad-core, or else go with i3 chips.
 

Eug

Lifer
Mar 11, 2000
23,586
1,000
126
Please say it isn't true... Cuz if true, then ouch!

https://browser.geekbench.com/v4/cpu/10565877

Single-core 4248
Multi-core 7828

System Information
Operating System macOS 10.14.1 (Build 18B2084)
Model MacBookAir8,1
Motherboard Apple Inc. Mac-827FAC58A8FDFA22 MacBookAir8,1
Memory 16384 MB 2133 MHz LPDDR3
Northbridge
Southbridge
BIOS Apple Inc. 220.220.100.0.0 (iBridge: 16.16.1578.0.0,0)
Processor Information
Name Intel Core i5-8210Y
Topology 1 Processor, 2 Cores, 4 Threads
Identifier GenuineIntel Family 6 Model 142 Stepping 9
Base Frequency 1.60 GHz
Package
Codename
L1 Instruction Cache 32.0 KB x 2
L1 Data Cache 32.0 KB x 2
L2 Cache 256 KB x 2
L3 Cache 4.00 MB x 1

---

That's basically the performance of the Core i7-7Y75.
 

Dayman1225

Golden Member
Aug 14, 2017
1,152
974
146
Please say it isn't true... Cuz if true, then ouch!

https://browser.geekbench.com/v4/cpu/10565877

Single-core 4248
Multi-core 7828

System Information
Operating System macOS 10.14.1 (Build 18B2084)
Model MacBookAir8,1
Motherboard Apple Inc. Mac-827FAC58A8FDFA22 MacBookAir8,1
Memory 16384 MB 2133 MHz LPDDR3
Northbridge
Southbridge
BIOS Apple Inc. 220.220.100.0.0 (iBridge: 16.16.1578.0.0,0)
Processor Information
Name Intel Core i5-8210Y
Topology 1 Processor, 2 Cores, 4 Threads
Identifier GenuineIntel Family 6 Model 142 Stepping 9
Base Frequency 1.60 GHz
Package
Codename
L1 Instruction Cache 32.0 KB x 2
L1 Data Cache 32.0 KB x 2
L2 Cache 256 KB x 2
L3 Cache 4.00 MB x 1

---

That's basically the performance of the Core i7-7Y75.

Not really surprising since the biggest difference between the two is the TDP (4.5w vs 7w) and the base clock (1.3Ghz vs 1.6Ghz), plus the better graphics and higher memory clock speed support, but that's about it... I imagine with the fan though, it will have better sustained performance.

It's the same Skylake Architecture from 2015... but slightly refined.
 

Eug

Lifer
Mar 11, 2000
23,586
1,000
126
Not really surprising since the biggest difference between the two is the TDP (4.5w vs 7w) and the base clock (1.3Ghz vs 1.6Ghz), plus the better graphics and higher memory clock speed support, but that's about it... I imagine with the fan though, it will have better sustained performance.

It's the same Skylake Architecture from 2015... but slightly refined.
Graphics specs are the same as the Kaby Lake Y i7-7Y75. Furthermore, even the CPU clock speed specs are the same. The i7-7Y75 is spec'd for 1.6 GHz / 3.6 GHz when using TDP-up 7 W.

I wonder if the reason it performs the same as the i7-7Y75 is because the sub-tests are short enough that the i7-7Y75 is essentially running at max speed and doesn't throttle significantly. The main benefit of the Air will be sustained performance, so the repeated Cinebench test might reflect that MBA's fan better.

So that would mean that for light usage (surfing and office applications), the i5-8210Y may not do any better than its brother the i7-7Y75. Really the benefit of the new machine is sustained performance, but if that really mattered, you'd be better off getting the MacBook Pro non-Touch Bar. It weighs all of 0.25 lb more, and has the exact same footprint, but a significantly faster U series CPU.
 

IntelUser2000

Elite Member
Oct 14, 2003
8,686
3,785
136
Please say it isn't true... Cuz if true, then ouch!

https://browser.geekbench.com/v4/cpu/10565877

Single-core 4248
Multi-core 7828

That's basically the performance of the Core i7-7Y75.

I'm surprised at why you are surprised.

It's the same architecture, and Geekbench tests take a few seconds to run so quite bursty. Max clock speeds are same between the two. The saying that most Intel chips don't run at Base is exaggerated when looking at their Y chips - it almost never runs at base. The 1.4 vs 1.6GHz doesn't matter.

Amberlake doesn't even use updated 14nm process either, its same as Kabylake.

So that would mean that for light usage (surfing and office applications), the i5-8210Y may not do any better than its brother the i7-7Y75.

I don't think the i5-8210Y will be better than the i7-7Y75 in anything.