Therefore, most opted for the $200 Ryzen 5 3600, which is what we widely recommended in the past year as the best value CPU, and that’s the part we’ll be comparing the 5600X to today.
Though we’re sure AMD would have preferred we use the 3600X or 3600XT, we’ll stick to the more popular vanilla 3600. As a side note, the rumor mill has been churning out reports of a Ryzen 5 5600, a non-X version which might arrive at around $220, just 10% more than the beloved 3600. But we won’t know until we know. In the meantime, we’ve got a somewhat pricey 6-core, 12-thread Ryzen 5600X to check out and we’re very keen to see how it compares to not just the Ryzen 5 3600, but also the Ryzen 7 3700X, and how it stacks up against Intel’s Core i5-10600K.
The 5600X features a single CCD with 6 cores enabled, which means it’s limited to a 32MB L3 cache. This is still substantial compared to Intel parts, but it’s half of what you get with the 5900X, as that higher-end processor packs two CCDs with 6 cores enabled. The 5600X features a 3.7 GHz base clock with a 4.6 GHz boost clock, very similar clock frequencies to the 8-core 5800X. Now it’s time to test and then we’ll re-evaluate the value of the 5800X towards the end of the review. For testing the AMD CPUs we used the MSI X570 Godlike motherboard along with four 8GB G.Skill TridentZ DDR4-3200 CL14 memory modules for a 32GB capacity. All cooling was handled by the Corsair iCUE H150i Elite Capellix AIO. As we’ve done with all reviews in this Ryzen 5000 series, all productivity benchmarks were run using the GeForce RTX 2080 Ti, but for the gaming tests we upgraded to the more powerful RTX 3090.
Benchmarks
Starting with Cinebench R20’s multi-core test, we see that the Ryzen 5 5600X is good for 4462 pts or 19% faster than the 3600 and 24% faster than Core i5-10600K. Compared to existing 6-core, 12-thread processors, the performance uplift provided by the 5600X is substantial, so much so that we’re very close to 8-core CPU performance. For example, the 5600X is just 9% slower than the Ryzen 7 3700X and 10% slower than the Core i7-10700K, that’s a phenomenal result given it packs 25% fewer cores.
Given the strong multi-core performance, you won’t be surprised to learn that the 5600X is very fast when it comes to single core performance. Here we’re looking at a 23% improvement from the 3600 and 18% from the 3700X. In fact, when it comes to single-thread performance the 5600X is faster than even the Core i9-10900K.
In terms of clock speeds, we monitored how the 5600X clocks in each of the Cinebench R20 tests. For the multi-core test, where all cores are heavily loaded, the 5600X clocked at around 4.4 GHz, which is well above the advertised 3.7 GHz base clock frequency. AMD also advertises a max boost clock frequency of 4.6 GHz and this should be achieved in single core or lightly threaded workloads. In the Cinebench single core test the 5600X typically operated at 4.65 GHz, that’s 50 MHz over the advertised spec.
Next up we have 7-zip compression performance and here the 5600X was able to match the 3700X and 10700K, which is a great result and it translates into a 30% performance uplift over the current 6-core/12-thread processors such as the Ryzen 5 3600 and Core i5-10600K.
Decompression performance was just as good. Here the 5600X was 7% slower than the 3700X, but 22% faster than the 3600 and a whopping 41% faster than the 10600K.
The cryptographic performance of the 5600X is very strong, beating even the 10900K for AES for encryption/decryption. It was 4% slower than the 3700X and 11% faster than the 3600.
The performance improvement in Blender is what we’ve come to expect: we’re looking at a 19% uplift over the 3600, making the 5600X 15% slower than the 3700X, which is the biggest margin we’ve seen yet to the 8-core processor, but even so still an impressive result given it features 25% fewer cores.
Once again we see that Zen 3 is a beast in V-Ray as the 5600X comes in just behind the 3700X, losing by a mere 4% and that meant it was 30% faster than the 3600 and 25% faster than the Core i5-10600K.
The last rendering benchmark we’re going to look at is Corona, and again we find more evidence of the 5600X destroying the 3600, this time by a 28% margin. At that rate, the 5600X was just 6% slower than the 3700X, finding itself situated between the 2nd and 3rd gen 8-core processors.
Code compilation performance has been improved by 8% over the 3600. The 5600X was 12% slower than the 3700X here, which is one of it’s biggest losses to the older 8-core processor.
We’re looking at a 6% performance improvement in DaVinci Resolve Studio 16, so not nearly as impressive as some of the other applications we’ve tested, but still better than a typical gen on gen improvement.
The Premiere Pro results are slightly more favorable for the new 6-core Zen 3 processor as here it was 10% faster than the 3600 and just 3% slower than the 3700X and 10700K. Here we’re looking at comparable performance to existing 8-core AMD and Intel processors.
The single thread performance of Zen 3 has been massively improved, and we’re getting a good look at what that means for single threaded applications like Photoshop. The 5600X was 22% faster that the R5 3600, and 18% faster than the 3700X and 10600K, so that’s a huge performance improvement.
After Effects is another application that replies predominantly on single core performance and here we see solid performance from the 5600X. At 21% faster than the 3600 and 15% faster than the 3700X, it also easily beat Intel’s Core i5-10600K and Core i7-10700K, hell it even beat the 10900K!
Power Consumption
Like the rest of the Zen 3 range, the Ryzen 5 5600X is exceptionally good in terms of power consumption. We’re looking at a 7 watt increase from the 3600, which is excellent considering it’s 19% faster in the Blender benchmark. The new CPU also reduced total system consumption by 14% when compared to the 3700X, though that’s in line with the performance deficit to that part. When compared to competing Intel parts such as the 10600K, we’re looking at massively improved power efficiency.
As for operating temperatures, the Ryzen 5 5600X is a very cool running CPU. Peaking at just 63C in our Blender stress test means it’s 10C cooler than the 3600X, which is impressive given the similar voltage but almost 500 MHz increase in average clock speed.
Gaming Performance
Testing with Far Cry New Dawn, we see that the Ryzen 5 5600X is 18% faster than the 3600 and 16% faster than the 3700X. That’s a serious performance gain over previous-gen parts. It was also 7% faster than the 10600K and offered better frame time performance.
The 5600X also just edged out the 5950X and 5900X as this title doesn’t heavily utilize those parts and the lower latency, single CCD design is likely playing a role here. Though we’re talking about a negligible 1.5% difference, that’s overall exceptionally good performance from the 5600X in Far Cry New Dawn.
Next up we have Rainbow Six Siege and we’re finding comparable performance to higher-end Zen 3 parts. More than comparable, it’s basically identical. The 5600X was 22% faster than the 3600 and 19% faster than the 3700X, a big generational performance gain.
The Ryzen 5 5600X is also able to match the 5800X, 5900X and 5950X in the Watch Dogs: Legion. That makes it 23% faster than the 3600, and 15% faster than the 3700X. It also edged out the 10600K by a 5% margin making it the fastest 6-core processor in this title.
Moving on to F1 2020, we’re again finding similar performance across the Zen 3 range. There’s up to a 3% difference between the 5600X and 5950X when looking at the 1% low results. In other words, virtually identical performance and that means the 5600X is 9% faster than the 10600K, 20% faster than the 3700X and 28% faster than the 3600.
For the first time in our gaming tests we’re seeing the 5600X slip behind the higher core count parts. It wasn’t that much slower in Horizon Zero Dawn, trailing the 5950X by up to 10% for the 1% low result and 4% on average. Performance is comparable to the Core i5-10600K, which is 16% faster than the 3700X and almost 30% faster than the 3600.
The gains in Borderlands 3 are fairly small compared to most other titles. We’re looking at a 7% performance improvement over equivalent Zen 2 parts. The 5600X is comparable to Intel’s 10600K in this game, too.
Death Stranding can take advantage of higher core counts on the 5900X and 5950X, though it seems to max out with the 5900X. As a result, the 5600X was 11% slower than than the 5900X and 7% slower than the 5800X, though performance was comparable to the 10900K, or at least the average frame rate performance was. When compared to the 10600K, we’re looking at a solid 18% performance improvement and a whopping 23% improvement over the 3700X. So here’s an example of a core heavy, CPU utilization heavy game that still sees the 5600X obliterate the 3700X. It’s also 33% faster than the Ryzen 5 3600.
Another CPU demanding game we like to test with is Shadow of the Tomb Raider. Do note we’re not using the built-in benchmark, but an open world section that’s much more CPU demanding than most other places in the game. SoTR looks to be maxed out with the 5900X. The 5600X was 7% slower that the 12-core CPU and 4% slower than the 5800X and Core i9-10900K. That said, it was 5% faster than the 10700K, 20% faster than the 10600K, 23% faster than the 3700X, and 31% faster than the 3600.
Hitman 2 performance is interesting as this isn’t a game that benefited from more cores with Zen 2. Clearly memory and cache latency was the primary bottleneck as we do see positive core scaling with Intel processors and now with Zen 3 chips. The Ryzen 5 5600X was able to match the average frame rate of the 5800X, though it was 7% slower for the 1% low result. It also trailed the 5900X by 8% for the average frame rate and 13% for the 1% low. When compared to the 10600K it was 15% faster and then a massive 26% faster than the 3700X and 32% faster than the 3600.
Testing with Star Wars: Squadrons shows a ~23% performance boost over the 6 and 8-core Zen 2 processors and a 16% boost over the 10600K. The 5600X’s performance output was comparable to the 10900K here.
Serious Sam 4 is an NPC heavy title which does benefit from more cores. In this case, the 5800X looks to be the sweet spot as the 5600X was 16% slower, though game time performance was strong and we didn’t see any stuttering issues with the 6-core processor. We’re also talking about a 26% boost over the 3700X and 16% greater 1% low performance.
Gaming Average
Here’s an overview of the average CPU gaming performance and overall we see a small performance decline when compared to the Ryzen 7 5800X across the 11 games tested. Of course, the margin is very game dependent. When compared to the 8-core 3700X we’re looking at a 19% performance boost on average and 24% over the 3600. In other words, the mainstream Ryzen 5 5600X is a much better gaming CPU. It’s worth keeping in mind that all tests were conducted at 1080p using a GeForce RTX 3090. Now obviously we do this as the focus is on CPU and not GPU performance, but be aware this isn’t typical gaming performance.
Normally you’d be using an RTX 3090 at 1440p or 4K gaming: at 1440p you’ll be a lot more GPU limited, while at 4K you’ll be entirely GPU limited. So don’t think an upgrade from the Ryzen 3 3600 to the 5600X will net you around 20% more performance. In the coming weeks we’ll likely investigate CPU scaling with a range of GPUs, but for now we’re looking more at CPU limited gaming performance. Getting back to the results, it’s really impressive to see AMD now offering 11% more performance than Intel’s best 6-core processor, the 10600K and we’ll look at cost per frame data in a moment, before that though here are the OC results.
Overclocking
When it comes to manual overclocking we were able to push the Ryzen 5 5600X to 4.6 GHz on all cores. This boosted Cinebench R20’s multi-core performance by just 3%, so a fairly pointless overclock.
Although we’re getting a 3% boost to all core performance, we’re also seeing a 1% decline in single core.
Blender confirms the all-core gains of 3% for the overclocked 5600X. Not exactly a worthwhile gain and it doesn’t help the 5600X catch up to the 3700X here.
For that 3% performance boost we’re looking at a 4% increase in total system power usage, so hardly anything to worry about.
And here’s why we don’t recommend gamers bother with overclocking the 5600X: we ended up reducing the frame rate in Rainbow Six Siege by 4%.
As we’ve seen before, it does help to improve performance ever so slightly in Far Cry New Dawn, though certainly not to the degree that anyone is going to notice.
Price to Performance
Here’s a look at the price vs. performance ratio using Cinebench R20’s multi-core data. At the new $300 price point, the Ryzen 5 5600X isn’t exactly great value in core-heavy workloads. While it does match the Ryzen 7 3700X in value, it’s also coming in at a 26% premium over the 3600 and that’s not great.
In order to match the value of the 3600, the 5600X would need to cost no more than $240, so here’s to hoping the $220 vanilla version of the 5600 ends up being a thing. Now, the reason the 5600X costs $300 is because of the Core i5-10600K. Compared to Intel’s competing part the 5600X is 12% better value.
It’s a similar situation when we look at Adobe Premiere data: the 5600X stacks up pretty well overall, but is still much worse value than the 3600, coming in at a massive 37% premium.
You’re also paying a 14% premium over the 3600 for gaming in Shadow of the Tomb Raider, though that’s not terrible given the performance improvement, and it’s still better value than Intel’s 10600K.
If we look at the 11 game sample, we see that in terms of value the 5600X and 10600K are comparable. Overall, the new 6-core Ryzen processor is fetching a 21% price premium when compared to the 3600, so that’s not great news.
What We Learned
On one side, it’s impressive just how much better the Ryzen 5 5600X is when compared to the 10600K, often offering more value despite costing a little more. On the other side though, it’s far less impressive in terms of value when compared to the $200 Ryzen 5 3600. While we do see a number of substantial performance improvements, a 50% price hike is going to be a tough pill to swallow for most. AMD’s really competing with themselves: if you want maximum value, get the R5 3600, if you want maximum performance, get the 5600X and that leaves no room for Intel’s Core i5-10600K. It’s worth point out that the Ryzen 5 3600 sold really well relative to the 10600K because it was cheaper. And while it is technically slower in games, when gaming you’re going to be GPU bound for the most part, and therefore the reality is that gaming performance is very close between those two. This also means gamers will remain better off buying the much cheaper Ryzen 5 3600, or waiting to see if a $220 non-X 5600 becomes a thing in the not too distant future.
Speaking of gaming performance, you’re no doubt going to hear nonsense such as “the Ryzen 5 5600X is a poor choice for gamers as it only has 6 cores,” and they’ll probably try and prove that by pointing to the new consoles which feature eight Zen 2 cores. Some people also like to confuse how games and cores work. Making statements like games will require 8 cores or something to that effect. Games don’t require a certain number of cores, they never have and they never will. Games require a certain level of CPU performance, it’s really that simple. A recent example of that is the Core i5-7600K, or in other words, quad-core processors. Three years ago the 7600K was the best value gaming CPU on the market and it shredded the Ryzen 5 1600 in every game. However, back then we did say that the 7600K’s days were counted as games were becoming more demanding and soon 4-cores/4-thread processors would be inadequate, stick with me here…
That wasn’t because games would necessarily require more than 4 cores, but because there wasn’t a single quad-core processor in existence that would be powerful enough to drive the latest and greatest games without avoiding frame stuttering issues and other performance related problems. For example, if the 7600K was capable of out-scoring the Ryzen 5 1600 in Cinebench R20’s multi-core test, it would actually be a better gaming CPU today, regardless of how many cores it featured. However, that’s not the case. The R5 1600 scores 53% higher in Cinebench and even when overclocking the 7600K to 5 GHz, the Ryzen part is over 30% faster. So, when fully utilized the Ryzen 5 1600 is a more powerful processor, which is why we knew in the not too distant future it would be a better processor for games and that’s long since been proven true. However, the Core i7-7700K hasn’t yet suffered quite the same fate despite also being a quad-core processor, as it features SMT support for 8 threads. Although it features half as many cores and threads as the Ryzen 7 1700, it’s able to keep up in the latest and greatest games, though it is starting to show some weakness in most demanding titles. At some point, we do expect the R7 1700 to beat the 7700K in games and again if we look at the Cinebench R20 multi-core performance, we see that when fully utilized the Ryzen processor is almost 40% faster and that’s obviously a massive difference. But what about the 6-core, 12-thread Ryzen 5 5600X, how will it age? Our guess is extremely well as the massive IPC increase offered by the new Zen 3 architecture means the 5600X is comparable to previous generation 8-core processors such as the 3700X and 10700K, or the Zen 2 parts used in the next gen consoles, and no one expects those processors to become obsolete any time soon. Looking at Cinebench R20 as a rough guide, we see that the 5600X’s multi-core performance is just 9% lower than that of the 3700X and 10% lower than the 10700K, and that’s not a big difference. It’s also worth noting that you need to be cautious when using Cinebench to measure heavy CPU utilized gaming performance. We use it to observe how CPUs might compare when fully utilized in games, but Cinebench isn’t particularly memory sensitive, so a CPU like the 5600X which is much improved compared to the 3700X in terms of cache and memory latency, will perform better in games under heavy load than the Cinebench R20 score would suggest. The memory and cache latency improvements of the Ryzen 5 5600X means that there’s a good chance it will never end up being slower than the 3700X in games. Doesn’t matter if you’re talking average frame rates or frame time performance, it should be better by all metrics. If you have the option of the Ryzen 5 5600X or the 3700X for $300, you should absolutely get the 5600X in our opinion. It’s a lot better right now in games and we expect that still be the case in a few years’ time. Or save your money and get the R5 3600 as it still gets the most out of high-end GPUs at 1440p, or if you can wait for the Ryzen 5 5600.
