Early iMac Pro Benchmarks [Chart]

POST

MAIL

Posted December 28, 2017 at 5:00pm by iClarified

Here are some early benchmarks of Apple's new iMac Pro.

Geekbench is reporting single-core and multi-core scores for the 8, 10, and 14 core models. The iMac Pro also comes with an 18-core processor but scores aren't available for that model yet.

In single-core performance, the 8-core Intel Xeon W-2140B @ 3.2 GHz model scores 4982, the 10-core Intel Xeon W-2150B @ 3.0 GHz model scores 5270, and the 14-core Intel Xeon W-2170B @ 2.5 GHz scores 5282.

In multi-core performance, the new iMac Pro really shines. The 8-core Intel Xeon W-2140B @ 3.2 GHz model scores 30240, the 10-core Intel Xeon W-2150B @ 3.0 GHz model scores 35412, and the 14-core Intel Xeon W-2170B @ 2.5 GHz scores 41120.

For comparison the top of the line mid-2017 iMac scores 5681 in single-core performance and 19353 in multi-core performance.

Check out the charts below and hit the link for more details...

Follow iClarified

Federico Campanale - December 29, 2017 at 9:34am

So the top of the line mid2017 iMac (with 5681) is better than all the 3 new models in single core performance.... quite a downer! Or is there a typo here...?

Reply · Like

dj lytic - December 29, 2017 at 10:35am

I didn't think this needed to be said again... but chips with >4 cores are usually clocked lower to fit in the same thermal envelope. In other words, 10 cores at 3ghz might generate as much heat as 4 cores at 4ghz, quite possibly more. What is truly impressive about the Xeon-W line is the high boost speed relative to the base clock speed. If you look at a pre-purley xeon, you'll notice the increased multi-core score comes at a large cost of a lower single core score. With the W-chips, this trade-off seems to be almost negligible, even as core counts ramp up and base clock speeds drop. I am interested to see the effects of Turbo3.0 if the feature makes its way to the Kaby-based workstation Xeons. Now, I'd like to test your assertion by actually forcing a single-core transcoding of a large 4k file or other high-cpu intensive task in both a 2017 iMac and iMac pro: it's entirely possible the thermal throttling might cause the Kaby lake chip to throttle in real world loads that benchmarks may not emulate fully. It is also possible that the new cooling system in the iMac Pro was barely enough to make up for the TDP difference in the 2 processor lines, and the Xeon chip will throttle heavily under full load too, in which case, we can blame the case. ;) Jokes aside, the Xeon-W line makes a huge improvement in single core speed for high core count chips. You are right that peak speeds may still be slower, but even in gaming, how much of an fps difference will it make? I'm not sure, I stopped gaming a few years ago, but I can say that the time it could save me every time I compile code or bounce an audio track will actually pay for itself.

matrixmaniac - December 29, 2017 at 4:23am

I'm a (superexpensive) Mac. And how do I now compare to (superexpensive) PC?