So, on Wikipedia, a user added that the i3-8121U has 3.1 GHz all-core frequency. Is there an independent source to verify?
It can be seen from the actual non-AVX high load that the dual-core frequency can only be stable at 15W 2.7Ghz (real frequency on the corona result). This performance is similar to that of the SKL processor with the corresponding frequency, Corona uses 12 minutes 22 seconds, corresponding to 7700HQ about 4 minutes 40 seconds
I thought so, but where do you see that confirmed?Integrated video is present, but inactive.
Interesting.
AIDA 64 screen identifies Intel HD graphics as inactive.I thought so, but where do you see that confirmed?
Core i3-8121U has got to be a binned version of m3-8114Y.
Does the Corona bench frequency of 2.7 answer the question?Also, as expected, it gets identified as (die shrunk) Skylake.
Cinebench R15 is not quite as bad as originally suggested, but at 139/306 I guess that should pretty much be as expected given the architecture and clock speed.
No overall Geekbench 4 score provided for some reason, just the subtests.
---
Regarding all-core Turbo, I will point out that the m3-7Y32, i5-7Y54, and i7-7Y75 all go to 2.6 GHz when presented with a YES x 4 test to max out all 4 threads, according to Intel Power Gadget.
What I wondering is if the all-core turbo speed might occasionally be different in certain tests or certain conditions. I would have guessed the KBL-Y all-core turbos of the m3, i5, and i7 would be different, yet with YES x 4 they all read as 2.6 GHz. Their advertised single-core turbo speeds are 3.0, 3.2, and 3.6 GHz respectively, and in fact, Wikipedia states the i5 all-core turbo speed is 2.8 GHz, and the i7 all-core turbo speed is 3.4 GHz. (No listing for m3.) Yet they all go to 2.6 GHz in Intel Power Gadget in that test.Does the Corona bench frequency of 2.7 answer the question?
https://pic3.zhimg.com/80/v2-21d97466385347f1a4e297e55e101463_hd.jpg
I don't know much about Corona bench, but it was 12 minutes long, which suggests the chip can hold 2.7 indefinitely.What I wondering is if the all-core turbo speed might occasionally be different in certain tests or certain conditions. I would have guessed the KBL-Y all-core turbos of the m3, i5, and i7 would be different, yet with YES x 4 they all read as 2.6 GHz.
All-core turbo of 7200U is 3.1 GHz.Well, the base of the 7200u is 300mhz higher, and presumably the all core turbo is as well.
7200u does Corona 1.3 in 10:43 vs 12:22 for the 8121u, so the 8121u all core turbo has got to be lower.
https://techreport.com/r.x/2017_11_21_Intel_s_Core_i5_8250U_CPU_reviewed/corona.png
Nice try.Good find. I guess the so-called "IPC Improvements" never materialized on cannon lake. You can see here that Intel's 10nm actually reforms worse on a clock for clock basis, probably because it can't even maintain its rated turbo.
It is good for consumers that intel won't be able to sell any of this junk in the US. Hopefully they never sell a single borked 10nm CPU ever.
The i5-7200U is an appropriate comparison. It has max Turbo and all-core Turbo of 3.1 GHz. Base clock is 2.5 GHz. It has 3 MB cache.Nice try.
Anyway, the 7200U is an i5, while the 8121u is an i3.
Why does the 8121u drop to 2.7 when running the Corona benchmark, if it's all core turbo is 3.1?The i5-7200U is an appropriate comparison. It has max Turbo and all-core Turbo of 3.1 GHz. Base clock is 2.5 GHz. It has 3 MB cache.
i3-8121U has 3.2 GHz max Turbo and 3.1 GHz all-core Turbo (supposedly). Base clock is 2.2 GHz. It has 4 MB cache.
Both are 2-core, 4-thread. ie. This 8th gen i3 gets HyperThreading.
If I already gave links with the score ranges, and scores are all over the place, why pick two results?https://browser.geekbench.com/v4/cpu/compare/7582239?baseline=6246666
Geekbench comparo to 8130u, for what it's worth.
If I already gave links with the score ranges, and scores are all over the place, why pick two results?
The 8121u there has much higher memory bandwidth.
So everything works then, just no IGP.