Warning: "continue" targeting switch is equivalent to "break". Did you mean to use "continue 2"? in /homepages/9/d659054407/htdocs/app660029368/wp-content/plugins/magneto-sidebar-generator/html.php on line 83

Warning: session_start(): Cannot start session when headers already sent in /homepages/9/d659054407/htdocs/app660029368/wp-content/plugins/wpmu-dev-facebook/lib/external/facebook.php on line 60

Warning: session_start(): Cannot start session when headers already sent in /homepages/9/d659054407/htdocs/app660029368/wp-content/plugins/wpmu-dev-facebook/lib/external/facebook.php on line 60

Warning: session_start(): Cannot start session when headers already sent in /homepages/9/d659054407/htdocs/app660029368/wp-content/plugins/wpmu-dev-facebook/lib/external/facebook.php on line 60

Warning: session_start(): Cannot start session when headers already sent in /homepages/9/d659054407/htdocs/app660029368/wp-content/plugins/wpmu-dev-facebook/lib/external/facebook.php on line 60
Ryzen 7 1800X, AMDs entry into the HEDT market: Part 1 - Page 3 of 7 - Tech Altar

Ryzen 7 1800X Compute Performance

We’re still ensuring we have the best and most accurate results for our benchmarks, but we do have some numbers to share at this time. We’re being very careful in how we test, so you can have a good idea of how Ryzen actually performs. Below you’ll find a number of compute benchmarks to show just how well Ryzen does for compute-focuses tasks. We tested on Cinebench, which is a good indication of IPC gains, SiSoft Sandra’s Monte Carlo simulation, a Tensorflow benchmark of an Alexnet implementation as well as an x265 encoding benchmark and POVRay.

What’s interesting is that it also comes very close to the 4.2GHz Kaby Lake-based i7-7700K, which does have a slightly improved IPC compared to Skylake and also operates at a higher base, and boost frequency. Broadwell-EP, with eight-cores, isn’t quite able to keep up. For AMD, this bodes very well. For Cinebench R15 when we normalize frequencies, we see that Ryzen is actually faster. We disabled boost on the i7-7700K for the normalization test.

Moving on to POVRay, Ryzen is distinctly faster. Better cache speed and just an overall better design allow it to ray-trace much more quickly.

For some reason we see that it does quite a bit worse on the x265 benchmark, and we’re not quite sure why. We would expect the Ryzen 1800X to be close to the i7-6900K, though it isn’t. In the coming week we’ll be checking that result against actually encoding utilizing handbrake with H.265.

When calculating a Monte Carlo financial analysis using a synthetic benchmark, there is still a lead over both Kaby Lake and Broadwell-E in a multi-threaded scenario, though it isn’t nearly as high. In this benchmark, single-threaded performance is also slower than Kaby Lake, though not nearly by as much as I would have thought.

The last compute-focused test we have is Tensorflow. This is a very well optimized Tensorflow is likely boosted by the very fast, large and low latency L1 and L2 cache since a lot of the data is held there and repeated often in this type of workflow. It too, is faster on Ryzen.

But what of games?