Expected, but is way higher than BR by a mile and is near on Haswell levels.CB R15 low single thread score, it seems.
Expected, but is way higher than BR by a mile and is near on Haswell levels.CB R15 low single thread score, it seems.
Expected, but is way higher than BR by a mile and is near on Haswell levels.
Hopefully you can add another stick.
- Number of Memory Sticks Included
1
So how does it compare to Intel and the MX150?
Or do we really care?
8650U is 175 / 506 in CB R15 for comparison.
That doesn't seem quite accurate. The 8250u in actual devices is median of 145/543. The 8550u is median of 167/553.
Must be some sort of estimate?
https://www.notebookcheck.net/Intel-Core-i7-8650U-SoC-Benchmarks-and-Specs.242726.0.html
Yes, but unfortunately most people would casually look at the individual results instead of the graph that you linked, and then the eternal debate would ensue on internet forums.Individual Cinebech results are actually very useful and relevant, as long as both the reviewer and the reader knows what to look for.
Individual Cinebech results are actually very useful and relevant, as long as both the reviewer and the reader knows what to look for.
Case in point, the Dell Inspiron 17 773 review, a 17" notebook equipped with a 8550u and MX150. In terms of cooling and configurable TDP it doesn't get any better than this for a 15W TDP CPU.
Here's data for continued CB15 runs on the 8550u:
Here's the power characteristics of 8550u as configured in the Dell Inspiron - graph bellow is from the Prime 95 stress test:
It's important to note this CPU is configured with a PL2 Power Limit of 44W. For the limited turbo time window, as long as thermals and max current are within acceptable parameters (temps lower than 85C most likely), this CPU will go up in frequency as high as turbo bins allow.
Isn't PL1 Tau the same as Turbo Time Parameter (Tau)? If so it is altered in many if not most units, and AFAIK PL2 is a very meaningful power limit for burst performance in modern Intel 15W TDP CPUs.Unless the PL1 Time Limit (PL1 Tau) is altered, the PL2 limit doesn't basically matter at all.
By default the PL1 Tau value is 1 second, so any meaningful benchmark will soak that additional boost without any issues.
Individual Cinebech results are actually very useful and relevant, as long as both the reviewer and the reader knows what to look for.
Case in point, the Dell Inspiron 17 773 review, a 17" notebook equipped with a 8550u and MX150. In terms of cooling and configurable TDP it doesn't get any better than this for a 15W TDP CPU.
Here's data for continued CB15 runs on the 8550u:
Isn't PL1 Tau the same as Turbo Time Parameter (Tau)? If so it is altered in many if not most units, and AFAIK PL2 is a very meaningful power limit for burst performance in modern Intel 15W TDP CPUs.
IIRC my Haswell-U came with a ~30s turbo time limit at stock settings.
Here's what Intel had to say about Turbo Time Parameter (TAU) in the Haswell-U Datasheet - Chapter 5.3I'm not sure about Haswell, but "PL1 Tau" defines how long the CPU can exceed PL1 (PL2 being the actual power limit instead).
This applies at least to Skylake and newer parts, and like I said unless the "PL1 Tau" is changed from the default value (1 second) the fact that PL2 is significantly off makes no pratical difference.
It makes sense for Intel to ship the product with PL2 effectively disabled since it's up to the OEM to properly configure Turbo Boost based on available cooling and power delivery. The final product will likely use a considerably higher value.This value is a time parameter that adjusts the algorithm behavior to maintain time averaged power at or below PL1. The hardware default value is 1 second, but 28 seconds is recommended for most mobile applications.
Though it requires certification by AMD this time, since it sounds like mXFR to me. Or maybe its a different system?I'm not sure about Haswell, but "PL1 Tau" defines how long the CPU can exceed PL1 (PL2 being the actual power limit instead).
This applies at least to Skylake and newer parts, and like I said unless the "PL1 Tau" is changed from the default value (1 second) the fact that PL2 is significantly off makes no pratical difference.
AMD parts are no different in that regard, eventhou the operation is more complex than on Intel. AMD 15W TDP SKUs are allowed to consume 25W until the simulated "skin temperature" of the system reaches certain ODM configurable limit.
The feature is called STAPM and it is more advanced in Raven than it was in the design it was first introduced in (Carrizo).
Here's what Intel had to say about Turbo Time Parameter (TAU) in the Haswell-U Datasheet - Chapter 5.3
Yes, and it was the same with 4th gen CPUs. My point was Intel ships the default value and OEMs adjust based on application. Intel's recommendation for most applications back when 4th gen was built was 28 seconds, in the current documentation they no longer offer a straight (and simplified) example.Reading the datasheet for the 7th Gen Core, PL1 Tau is indeed as TheStilt describes - time it can stay in PL2.
Actually... default PL2 value for KBL-R 4C is 1.25*15=18.75W (page 93). It was the OEM who set a custom value of 44W.PL2 is the burst spec, which can be set to 44W on the Kabylake-R 4 core CPUs.