AMD Precision Boost 2 and Wraith Prism Deep Dive

When AMD presented it new Precision Boost 2 technology, I have to think AMD undersold the feature in terms of what it was actually able to accomplish. The original Precision Boost on the 1st gen Ryzen processors was anemic at best. After seeing what Precision Boost 2 can accomplish, do we even need to try to overclock our new 2700X? Blasphemy!


Synthetic Arithmetic Benchmarks & RAM Bandwidth

As always, synthetic benchmarks are a good marker for performance, but are far from the end-all be-all of the CPU measuring stick. That said, if you find more value in these, then so be it. We generally use these for finding out if anything is "wrong" rather than when something is "right."

SiSoft Sandra 2017

Article Image

Right out of the gate, we see the Prism cooler afford us an 8.36% increase in compute score. Moving from Prism to H2O, we see the water cooling afford us a 13% increase. Our 4.2GHz hand overclocking comes in neck and neck with our Level 2 Precision Boost water cooling. Our Level 3 overclocking shows no big increase and is actually negative on the increase side with the Prism.

Article Image

As you might expect, our memory bandwidth stays pretty much the same across all data points. I did find it a bit odd that the Level 3 overclocking were a bit muted, but nothing to write home about. That said, I am not sure as to what all the Level 3 setting actually changes since there is no real transparency on that.

Hyper Pi

Article Image

Paying attention to our middle grouping, we see that our Prism at Level 2 actually out-performs our H2O but this is easily inside our margin of error. However, I probably ran this set of benchmarks 50 times and what I saw here was consistent. What is apparent here is that our 4.2GHz overclock is simply slower. After seeing the chart on page 2, this is surely what we would expect. Still why our Prism would boost slightly higher than the water is beyond me. Keep in mind this is a single threaded benchmark.


Article Image

We run wPrime in a 4 thread configuration. We again see results that almost mirror what we have seen above in terms of scaling, except this time we seen our water cooling pulling slightly ahead, like you would think in this wider benchmark that is going to produce a bit more heat.