Skip to main content

AMD Ryzen 5 3600 Review: Non-X Marks the Spot

Editor's Choice

Rendering

Image 1 of 9

Image 2 of 9

Image 3 of 9

Image 4 of 9

Image 5 of 9

Image 6 of 9

Image 7 of 9

Image 8 of 9

Image 9 of 9

As we can see throughout this series of tests, AMD's Ryzen processors undoubtedly sit atop the multi-threaded benchmark throne. The six-core -9600K and -9400F, which both lack Hyper-Threading, are outclassed in these tasks. Meanwhile, the 12-threaded Ryzen 5 chips dominate in these types of workloads with convincing wins across the board. The Ryzen 5 3600X at stock settings is comparable, or faster, than the overclocked Ryzen 5 3600 in most of these heavily-threaded workloads, meaning it does offer some benefit for the extra $50 investment, particularly if you're not interested in overclocking.

Encoding

Image 1 of 5

Image 2 of 5

Image 3 of 5

Image 4 of 5

Image 5 of 5

The SVT-AV1 encoder is an Intel- and Netflix-designed software video encoder that became available earlier this year. This new encoder is more scalable than other encoders, thus offering faster performance paired with efficient compression. While it may seem counter-intuitive to use an Intel-designed encoder for testing AMD processors, consider that most encoders are inherently reliant upon per-core performance, which is a strength of Intel, while SVT-AV1 exposes the power of threading, a strength of Ryzen. At stock settings, the Ryzen 5 3600X beats the overclocked Ryzen 5 3600 by a slim margin, but overclocking the X-series model opens a wider gap. The Ryzen 5 3600 fares better against the -9400F, which lags by a decent margin.

Our LAME and FLAC tests, like many encoders, rely heavily upon per-core performance. That means Intel's frequency advantage comes into play, allowing the -9600K to take the lead. Conversely, the lower-clocked -9400F suffers. The -9600K has the advantage at stock settings, but overclocking propels it to the top of the chart.

Intel processors traditionally leverage high frequencies to dominate the HandBrake x265 test, which relies heavily on AVX instructions, and the x264 test. But Intel's higher clock speed isn't too much of an advantage in these tests when the similarly-priced competition has twice the number of threads, so the Ryzen 5 chips carve out nice leads in both x265 and x264 encoding. We also noticed that the Wraith Stealth-cooled 3600's overclocked configuration often trails the stock settings with the same cooler, which we confirmed with repeated testing. It looks like AMD might have some tuning left to do with its PBO algorithms in thermally-limited situations.

Compression, Decompression, Encryption, AVX

Image 1 of 10

Image 2 of 10

Image 3 of 10

Image 4 of 10

Image 5 of 10

Image 6 of 10

Image 7 of 10

Image 8 of 10

Image 9 of 10

Image 10 of 10

Our threaded compression and decompression 7-Zip and ZLib tests work directly from system memory, removing storage throughput from the equation. The combination of Ryzen 5 3600's improved memory subsystem and generous helping of cores helps it take an easy lead over the -9600K and -9400F while the Ryzen 5 3600X does provide at least some advantage over the 3600 via overclocking, although it isn't enough to justify a $50 premium for value-seekers.

We can also see the vast improvement in Ryzen's AVX performance in the y-cruncher tests: That's a massive generational leap, particularly in single-threaded performance. You'll notice that the overclocked 3600 with the Stealth cooler again suffers during a heavily-threaded AVX workload, furthering our suspicion that the PBO algorithms aren't fully optimized. In either case, the improved gen-on-gen AVX performance is truly impressive and benefits a wide range of applications.


MORE: Best CPUs


MORE: Intel & AMD Processor Hierarchy


MORE: All CPUs Content

  • RodroX
    Three things to comment,
    1. Why aren't any temps readings on the article ?

    2. I did some Cinbench testing with stock cooler and only PB enable (no PBO nor AutoOC) and I usually get around 359X to 360X. After changing the stock cooler for a better tower cooler, temps went down, frecuency went up and cinbench results landed at even better 365x to 366x.
    This was on Windows 10 (1903) + Avast antivirus and hwinfo running, with AGESA 1003 ABB and the latest AMD chipset drivers on a Gigabyte B450 Gaming X, is funny and strange to see that a very expensive motherboard like the one used is getting such lower results, wonder if the motherboard could be affecting other benchmark results aswell.

    Once again, this numbers I wrote are with stock BIOS settings, no PBO nor AutoOC, no manul oc, no vcore offset, nothing (Ive checked, had the 3 PBO options disable on BIOS).

    3. Also why not more recent and better optimized games like Shadow Of Tomb Raider or Battlefield V, or anyof the Assassin Creeds ?

    Cheers
    Reply