Pure Overclock – Computer Hardware News, Reviews and More


AMD Ryzen 7 1800X Preview: Initial Thoughts with Preliminary Tests

Comments Off on AMD Ryzen 7 1800X Preview: Initial Thoughts with Preliminary Tests
Posted March 2, 2017 by Josh Jackson in CPU & Motherboards

Early Testing

My go to test is CPU Mark. Even though I believe there might be some room for optimization for Ryzen, I still like the baseline results it gives me. While I checked results against the 7700K I have, I really wanted to see how far we came from the 8350. That was the last major architectural improvement AMD released until Ryzen today. I expected to see major improvements and I did. The initial results showed about a 35% improvement on the the single-thread test with a massive 60% improvement on multi-thread results. This wasn’t a clock-for-clock test, but more of a stock-for-stock test. I ran three runs since each test has a bit of variance, which you can see below.

Ryzen 7 1800X



Next up, I did some runs of Cinebench R15. This benchmark seems to be very consistent, so I stayed with one shot per CPU. The 8350 was blown away, with the 1800X gaining about a 60% single-thread lead and a massive 160% multi-thread lead.

Finally, I wanted to run a gaming bench to see how much improvement we had on a single-thread front. Battlerite is my latest guilty pleasure, but it seems to rely heavily on a single CPU core. It’s important to realize that my Ryzen system had a 7200 RPM drive, with my other systems running off of solid state. I feel like performance is mostly hindered in loading screens, but my official gaming benches will be sure to run off of solid state as well just to make sure gaming isn’t hindered in the review. Even then we managed about a 30% improvement from the FX era.

Ryzen Battlerite

Looking at the results, considering that clocks vary a bit, and understanding that the systems aren’t quite as consistent as they need to be, I think AMD’s claim of a 52% IPC gain is pretty accurate. That will vary from various applications, but it seems like a good average, rather than a cherry-picked figure. I’ll need to confirm that with more benchmarks, but once again, we’re looking at a preview before we get to our official review.

Find us on Google+