News for hackers. Stuff that matters.
Re:iMac Pro 18-core Follow Up Review by saagarjha on Saturday February 03, 2018 @06:14PM (#16300254)
> Though I haven't mentioned it, if you look back through the various benchmark results, you'll see that the 18-core iMac Pro shows no disadvantage for single-core performance, despite running at a lower clock speed (2.3GHz/4.3GHz) than the 10-core iMac Pro (3.0GHz/4.5GHz). Often times, the price of scaling a CPU architecture to more cores is a loss of single-core performance, but no such penalty seems to exist here. The 18-core iMac Pro brings 8 more cores to the table on the high end with no loss of performance on the low end.

I remember this being an issue on the older Mac Pros; it's interesting that this isn't an issue anymore. What's the reason behind it? Better thermals? Turbo Boost?

Re:iMac Pro 18-core Follow Up Review by rsynnott on Saturday February 03, 2018 @06:48PM (#16300370)
Turbo boost, yeah. Modern Intel chips generally have a much higher clock cap if only a core or two is active. I think to an extent this should be true of the 2013 Mac Pro, too, but less so for the cheesegrater, which used much older chips.
Re:iMac Pro 18-core Follow Up Review by 013a on Saturday February 03, 2018 @07:18PM (#16300491)
I wonder if per-core performance is impacted if every core is being blasted. I imagine most benchmarks do single-core runs by running one process of work, and do multi-core runs by maxing out the cores and then totaling the performance. Do any benchmarks max out the cores, but only report the performance of one process of work?
Re:iMac Pro 18-core Follow Up Review by dboreham on Saturday February 03, 2018 @07:35PM (#16300545)
>I wonder if per-core performance is impacted if every core is being blasted.

Yes.

>I imagine most benchmarks do single-core runs by running one process of work, and do multi-core runs by maxing out the cores and then totaling the performance.

Aggregate throughput, yes.

>Do any benchmarks max out the cores, but only report the performance of one process of work?

Sort of. You can get that figure by dividing the aggregate throughput by the # cores. Sometimes it is reported explicitly.

Re:iMac Pro 18-core Follow Up Review by boomskats on Saturday February 03, 2018 @07:02PM (#16300437)
But it is an interesting point - why does the bigger chip perform better with a lower turbo clock speed? Does it dissipate heat better & so is capable of sustaining boost for longer?
Re:iMac Pro 18-core Follow Up Review by wtallis on Saturday February 03, 2018 @07:04PM (#16300443)
It isn't too surprising that a 24.8MB L3 cache vs 13.8MB L3 cache can make up for a 4.4% clock speed deficit.
Re:iMac Pro 18-core Follow Up Review by digi_owl on Saturday February 03, 2018 @07:55PM (#16300619)
Yeah the Hz race is long dead. These days it is all about avoiding cache misses. Perhaps why Intel got egg on their face with Meltdown, as that seems to have all to do with avoiding said misses.