Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Does this mean I won the "cpu lottery"?
- Thread starter Camarillo Brillo
- Start date
- Sort by reaction score
seems you got a keeper if I compare with the list:
Such notions as winning the CPU lottery refer only to overclocking potential, which you shouldn't do on a Mac, not sure if it actually can be done on a Macbook Pro, but I wouldn't even recommend it on a Mac Pro. Your CPU is very likely performing well thanks to the presence of proper cooling and heat dissipation found in the 16" Macbook Pro, but there is no lottery to be had. I'm not trying to bum you out, but yeah. Nice score though!
My Geekbench CPU test
Cinebench
Geekbench Metal Compute with 5500M 4 GB
Geekbench Compute with OpenCL
As far as your success with beating a 28-core Mac Pro, that isn't really surprising. Most multi core CPUs are geared towards multi-threaded performance for relevant workloads. Single core isn't really relevant in most situations, but among the most prominent usage cases for higher single thread performance being relevant is with video gaming. But you did beat my 2.3 GHz 16" Macbook Pro! That's cool.
Cinebench
Geekbench Metal Compute with 5500M 4 GB
Geekbench Compute with OpenCL
As far as your success with beating a 28-core Mac Pro, that isn't really surprising. Most multi core CPUs are geared towards multi-threaded performance for relevant workloads. Single core isn't really relevant in most situations, but among the most prominent usage cases for higher single thread performance being relevant is with video gaming. But you did beat my 2.3 GHz 16" Macbook Pro! That's cool.
Last edited:
Here is my geekbench with the 2.3ghz CPU, I'm pretty happy to see I'm very close in single core and just as good in multicore as a good 2.4ghz!
It seems kind of sad or silly (at the same time) to be trying to "win' the various lotteries on the MBP. Sad, because for a premium device, people are worried about getting a bad machine. Silly, because the cpu is probably way faster than the majority actually need.
My thinkpad isn't as fast, but it is a year old and a generation behind, but I've not seen much talk about a geekbench score showing slightly slower numbers and thus the owner is complaining about losing the cpu lottery. I suspect that most PC owners just use their machine, and I may go out on a limb and say the majority of MBP owners are doing the same.
tl;dr Who cares, and just enjoy your machine
My thinkpad isn't as fast, but it is a year old and a generation behind, but I've not seen much talk about a geekbench score showing slightly slower numbers and thus the owner is complaining about losing the cpu lottery. I suspect that most PC owners just use their machine, and I may go out on a limb and say the majority of MBP owners are doing the same.
tl;dr Who cares, and just enjoy your machine
The silliest part of all that is that Geekbench is an unreliable benchmark that is very poorly suited for estimating the performance of the CPU. It's scores are all over the place.
DDR stands for double data rate. So you need to double the 1333mhz to get the actual speed, 2600mhz.
Not to put a damper on your excitement, but keep in mind that mbp has a ridiculously high ratio from base clock to turbo boost frequency. (2.4 to 5.0)
So, while that will show high scores in a brief benchmark, in real extended heavy use, you’ll get no where near the performance those benchmarks suggest.
I’d rather have a computer with a higher base clock (like a 3.6 that turbos to 4.1) than one that can briefly strain to more than twice its base frequency in order to achieve a geekbench result. Thats not sustainable in real world use.
So, while that will show high scores in a brief benchmark, in real extended heavy use, you’ll get no where near the performance those benchmarks suggest.
I’d rather have a computer with a higher base clock (like a 3.6 that turbos to 4.1) than one that can briefly strain to more than twice its base frequency in order to achieve a geekbench result. Thats not sustainable in real world use.
Last edited:
Still not indicative probably. But my use case may be an outlier, I often edit and render fairly long (20 minute plus) Videos. In that use, turbo boost doesnt mean squat, its base frequency that really counts.
Cinebench is not too bad, but then again it only tests a specific scenario (raytracing). At least with Cinebench you get consistent scores. To get best estimate for your use case you’d ideally benchmark your use case
[automerge]1578333279[/automerge]
Base frequency has very little practical relevance. It is just a “guaranteed” minimal performance threshold, more of a marketing parameter than anything else. The 16” for example is able to maintain 3.1-3.3 ghz under full load pretty much indefinitely - that’s 800mhz over its base frequency.
Cinebench is not too bad, but then again it only tests a specific scenario (raytracing). At least with Cinebench you get consistent scores. To get best estimate for your use case you’d ideally benchmark your use case
[automerge]1578333279[/automerge]
Base frequency has very little practical relevance. It is just a “guaranteed” minimal performance threshold, more of a marketing parameter than anything else. The 16” for example is able to maintain 3.1-3.3 ghz under full load pretty much indefinitely - that’s 800mhz over its base frequency.
What I'm saying, and this is not in debate, is that, for instance, a 1.3ghz processor that turbo's to 3.1ghz, is NOT equivalent to a 2.6ghz processor that turbos to 3.1 - assuming they're both from the same chip generation.
I real world use, the one with the higher base clock will be notably faster.
Unfortunately, this is where you are wrong. If we assume that these CPUs are the same core, and their only configuration difference is the base clock, with TDP and power management parameters set up the same, their real world performance will be very similar as well. The base clock is pretty much meaningless. It is just a minimal performance estimate from the manufacturer in the style if “if your CPU runs at or above this clock it works as advertised”. Base clocks nowadays have little to do with real CPU capabilities. Again, for the 16” MBP the “real” base clock is over 3.0 ghz.
You’re wrong sir. Turbo only works for very brief spurts, ESPECIALLY in laptops! If you want sustained performance over a strenuous task, like video rendering, a higher base clock is clearly better.
Theres a reason why Intel charges more for processors with higher base clock speeds - because they’re better.