Friday, June 26th 2020

Intel Lakefield Core i5-L16G7 Performance Benchmarks Leak

Performance benchmarks have started leaking for Intel-s upcoming Lakefield CPUs - low-power SoCs designed with Intel's latest technology. The Lakefield family of CPUs will make use of an Arm-similar big.LITTLE design, where this particular CPU, the Core i5-L16G7, will ship with four low-power "Tremond" cores and one large, high-performance "Sunny Cove" core for peak workloads. Built using Intel's Foveros stacking technology, these are the first chips to be built on Intel's modular platform, which should allow for pairing of I/O dies, chiplet-like CPU arrangements and memory in a 3D package. Physical distance reductions impact latency and power consumption, which should allow for an interesting design result.

Notebookcheck has tested an Intel Lakefield Core i5-L16G7 CPU that's being deployed on upcoming Samsung's Galaxy Book S, and the results are sort of a mixed bag. For one, Intel's Lakefield seems to be around 67% slower than the company's previous ultra-low-power architecture, Amber Lake. Something of this might have been caused by the fact that the Lakefield CPU didn't boost towards its advertised 3.0 GHz; it only managed to reach 2.4 GHz, which obviously hampered performance. Perhaps pre-release silicon is the culprit, or perhaps it's the galaxy Book S that's been configured with more restrictive thermal and power characteristics than the chip was actually designed to run at. The chip did manage to run the FireStrike test beating the Amber Lake-based Acer Swift 7 by 23%, though, so not all is looking bleak.
Looking at the screenshots, it does seem that most load is being distributed to the low power Tremond cores, which is obviously the idea in such a low-power design such as this (this assumes that the high-performance Sunny Cove core is last on the Task manager). It remains to be seen if hardware scheduling is being optimal, or if there is some sort of workload shuffle between cores due to OS mismanagement of the SoC - it is a possibility, perhaps, that the OS is shuffling data between cores, thus affecting performance benefits of L1 and L2 caches too aggressively, in an attempt to keep hotspot operating temperatures in check.
There is a moment in the first benchmark though where the fifth core sees a surge in workload, moving up to 100% utilization, and then declines, giving way to more work on two other cores. This looks well in line with the expected use cases for the Sunny Cove core and Intel's thoughts on designing these chips: most users only require high performance in short bursts that need to provide as much system responsiveness as possible, such as web page loading and other similar workloads.
Source: Notebookcheck
Add your own comment

20 Comments on Intel Lakefield Core i5-L16G7 Performance Benchmarks Leak

#1
ppn
Sunny is 10%loaded in the multicore test. add 180%.
Posted on Reply
#2
Vayra86
OK. So this really looks like a combination of bad things, if you ask me. Yes, we have a tiny perf/watt jump if you have no substantial workload to offer the machine... :ohwell:

There is no peak or burst performance to be had beyond whatever Intel already offered. Low power core does not enable higher power other core. So effectively the ceiling has just been lowered a bit.
Posted on Reply
#5
ppn
what is the point when the only Sunny cove core sits in 90% idle mode.Apparently cinebench distributes the tasks evently, so unless this is fixed we will get some ridiculous results for Alder Lake. where the weak cores will hold the Ocean Cove cores to 50% of their capacity or something.
Posted on Reply
#6
Crackong
Tried the Book S in the shop.
Open windows's setting page takes 10 second.

The Book S itself is well built, good keyboard , feather light , good screen .
But the CPU is a total drawback.
Posted on Reply
#7
R0H1T
phanbuey
looks not great.
Yes looks like a repeat of the old Atom/netbooks disaster :shadedshu:
Posted on Reply
#8
Caring1
Wow, barely better than the Celeron, and beaten by a Pentium.
At least it can run Firestrike. :rolleyes:
Posted on Reply
#9
ShurikN
So it's a Bulldozer of low power chips. Nice.
Posted on Reply
#10
1d10t
Now we know why Apple switch to ARM :rolleyes:
Posted on Reply
#12
R0H1T
1d10t
Now we know why Apple switch to ARM :rolleyes:
Apple, ARM has had experience with Big Little for close to a decade now. Intel or MS' failure in matching that doesn't represent anything about how the upcoming Mac OS on ARM will behave.
Posted on Reply
#13
TheGuruStud
R0H1T
Apple, ARM has had experience with Big Little for close to a decade now. Intel or MS' failure in matching that doesn't represent anything about how the upcoming Mac OS on ARM will behave.
Keep prices the same and save money making a cheapo chip/board. Their customers are happy to be swindled. Pump those margins, baby!
Posted on Reply
#14
R0H1T
Not an Apple user so can't say how much they're swindled but you have to give credit to the innovation ARM, or Apple, brought in this space especially with the stagnation we saw with Intel. Without AMD on the one side & ARM on another, Intel may well have stopped caring even marginally for general users.
Posted on Reply
#15
iO
A: the scheduler still has no idea was it's supposed to do with a hybrid design and it will get patched

B: it's working as intended and the big core isn't for any kind of sustained but only for ultra short and bursty workloads
Posted on Reply
#16
ARF
Crackong
Tried the Book S in the shop.
Open windows's setting page takes 10 second.

The Book S itself is well built, good keyboard , feather light , good screen .
But the CPU is a total drawback.
R0H1T
Yes looks like a repeat of the old Atom/netbooks disaster :shadedshu:
Caring1
Wow, barely better than the Celeron, and beaten by a Pentium.
At least it can run Firestrike. :rolleyes:
1d10t
Now we know why Apple switch to ARM :rolleyes:
TheGuruStud
wccftech.com/idf13-intel-ship-10nm-chips-2015-7nm-chips-2017/

Mmmkay, Intel.
Yeah, guys, I don't know why they keep making so lacklustre CPUs.
Why don't the users object and sue Intel for making Celeron and Pentium, and this Core i5 ? :confused:

You know some people sue the coffee company for making too hot coffee that burns their hands........

Don't worry, once AMD reaches the end of road for the Zen development, AMD will be the same as Intel is today.

ARM and RISC is the future.
Posted on Reply
#17
1d10t
R0H1T
Apple, ARM has had experience with Big Little for close to a decade now. Intel or MS' failure in matching that doesn't represent anything about how the upcoming Mac OS on ARM will behave.
If I remembered correctly, the first big.LITTLE design that shape into market was LG Nexus 5X and Huawei Nexus 6p, that only 4 years ago. Considering the same year there was also Intel Atom Cherrytrail powering chinese Android tablet Chuwi, I seriously doubt that Qualcomm or Apple had a head start.
Posted on Reply
#18
Assimilator
On one hand, ouch.

On the other hand, Cinebench is hardly this chip's intended use-case.

So I'm still going to hold off judging this one until reviewers have final silicon in hand.
Posted on Reply
#19
watzupken
I feel critically, what this short review did not mention is the battery life between this and the Snapdragon version of the Galaxy Book. At the end of the day, people who buy this niche laptop is most likely one that needs the portability/ long battery life. Performance looks pretty close between the 2, so it really boils down to whether Intel can offer better performance + maintain a good battery life.
Posted on Reply
#20
Vayra86
Caring1
Wow, barely better than the Celeron, and beaten by a Pentium.
At least it can run Firestrike. :rolleyes:
Don't ask what settings & FPS though lol

This is a pretty sad display of ingenui... err desperation
watzupken
I feel critically, what this short review did not mention is the battery life between this and the Snapdragon version of the Galaxy Book. At the end of the day, people who buy this niche laptop is most likely one that needs the portability/ long battery life. Performance looks pretty close between the 2, so it really boils down to whether Intel can offer better performance + maintain a good battery life.
Yes, so then you take a low power chip, they've had that since forever and they mostly perform better. Higher performance is more time in idle, so its really more of the same.
See, the fact is, this design is also not cheaper than a single low power equivalent... so they've just added complexity for what exactly?

I mean... a full day of battery life... I might be stupid, but isn't the key then to improve the battery a little bit? This is just an awful race to the bottom just to maintain some silly form factor.
Posted on Reply
Add your own comment