AMD Ryzen 5 1600X CPU Review

The Infinity Fabric: A Blessing And A Curse

There is much speculation (and plenty of proof) that faster memory improves Ryzen's gaming performance. The theory is that the speed of AMD's Infinity Fabric is tied to the memory clock rate, and through a bit of our targeted testing below, it looks like this is true.

Again, the Zen architecture employs a four-core CCX (CPU Complex) building block. AMD outfits each CCX with a 16-way associative 8MB L3 cache split into four slices; each core in the CCX accesses this L3 with the same average latency. Two CCXes come together to create an eight-core Ryzen 7 die (image below), and they communicate via AMD’s Infinity Fabric interconnect. The CCXes also share the same memory controller. This is basically two quad-core CPUs talking to each other over a dedicated pathway: Infinity Fabric, a 256-bit bi-directional crossbar that also handles northbridge and PCIe traffic. The large amount of data flowing through this pathway requires a lot of scheduling magic to ensure a high quality of service. It's also logical to assume that the six- and four-core models benefit from less cross-CCX traffic compared to the eight-core models.

Data that has to move between CCXes incurs higher latency, so it's ideal to avoid the trip altogether if possible. But threads may be forced to migrate between CPU Complexes, thus suffering cache misses on the local CCX's L3. Threads might also depend on data in other threads running on the CCX next door, again adding latency and chipping away at overall performance.

Intel employs a dual ring bus, seen to the right on a Broadwell die and covered in depth here, which serves much the same purpose. However, it doesn't suffer the same latency penalty between cores due to its contiguous design and dual independent rings. Although AMD hasn't confirmed, we suspect that the Infinity Fabric is a single pathway.

AMD's Infinity Fabric serves a key strategic advantage for the company as it rolls out 32-core Naples processors featuring four CCXes per package. The interconnect is more scalable than Intel's ring bus, which encounters increased latency as more cores are added. But it isn't without weakness (clearly). Intel's design may be more limited in terms of scaling out, but the company already utilizes a mesh topography with its Knight's Landing products (covered here) to  solve that. We suspect we'll see something similar from Intel's next-gen desktop products.

Putting Numbers To Theory

Accurately measuring bus latency is tricky. Fortunately, SiSoftware recently introduced its Sandra Business Platinum version that includes a novel Processor Multi-Core Efficiency test. It's able to measure inter-core, inter-module, and inter-package latency in a number of different configurations using Multi-Threaded, Multi-Core Only, and Single-Threaded tests. We use the Multi-Threaded metric with the "best pair match" setting (lowest latency) for our purposes.

First, we measured core-to-core latency on Intel's Hyper-Threaded Core i7-7700K to establish a baseline for AMD's SMT-equipped Ryzen processors. The results from Intel's solution are incredibly consistent between runs. That's in stark contrast to the Ryzen CPUs, which varied from one run to the next.

Core i7-7700K Memory Data Rate
Inter-Core Latency
Core-To-Core Latency
Core-To-Core Average
1333 MT/s
14.8ns
38.6 - 43.2ns
41.5ns
2666 MT/s
14.8ns
29.4 - 45.5ns
42.13ns
3200 MT/s
14.7 - 14.8ns
40.8 - 46.5ns
43.08ns

The inter-core measurement quantifies latency between threads that are resident on the same physical core, while the core-to-core numbers reflect thread-to-thread latency between two physical cores. As we can see, there are slight gains as we increase the memory data rate on our stock Core i7, but they're only in the 4% range. This explains, at least in part, why we don't see explosive gains from overclocked memory on Intel CPUs.

Core i7-7700K Memory Data RateInter-Core LatencyCore-To-Core Latency
Core-To-Core Average
1333 MT/s
12.9 - 13.3ns
38.3 - 41.1ns
39.59ns
2666 MT/s
12.9ns
34.5 - 39.9ns
37.67ns
3200 MT/s
12.9ns
36.1 - 39.2ns
37.8ns

After increasing the CPU's clock rate to 5 GHz, we observe a quantifiable reduction in inter-core and core-to-core latency.

Ryzen 5 1600X Memory Data RateInter-Core Latency Range
Inter-CCX Core-to-Core LatencyCross-CCX Core-to-Core Latency
Cross-CCX Average Latency
% Increase From 1333
1333 MT/s
14.8 - 14.9ns
40.4 - 42.0ns
197.6 - 229.8ns
224ns
Baseline
2666 MT/s
14.8 - 14.9ns
40.4 - 42.6ns
119.2 - 125.4ns
120.74ns
46%
3200 MT/s
14.8 - 14.9ns
40.0 - 43.2ns
109.8 - 113.1ns
111.5ns
50%

Repeated tests on our stock Ryzen 5 1600X revealed more run-to-run variance, so these numbers are the average of two test sessions. The inter-core latency measurements represent communication between two logical threads resident on the same physical core, and they're unaffected by memory speed. Inter-CCX measurements quantify latency between threads on the same CCX that are not resident on the same core. We observe slight variances, but inter-CCX latency is also largely unaffected by memory speed. 

Cross-CCX quantifies the latency between threads located on two separate CCXes, and as we can see, there is a tremendous penalty associated with traversing the Infinity Fabric. Dialing in a higher memory data rate reduces the Infinity Fabric's latency, though that's subject to diminishing returns.

Doubling the data rate from 1333 to 2666 MT/s grants a 46% performance increase. However, jumping up another 20% to 3200 MT/s only yields a 4% reduction in cross-CCX latency. Moreover, we tested our 4 GHz overclock against the 1600X's stock 3.7 GHz, but latency was largely unaffected.

We also measured Inter-Core bandwidth during our tests, and that progresses nicely as smaller chunks of data populate the L1 cache. As the data size increases, it falls back into the L2 and L3 caches, and then finally encounters main memory on the chart's far right.

Though this brief experiment only looks at Ryzen 5 1600X, we're planning another deep-dive story that includes all available Ryzen CPUs. There are so many possible combinations to test, including SMT and various power profiles, that might affect Infinity Fabric performance. Stay tuned.

...And We Still Can't Measure Cache Performance

During the Ryzen 7 1800X launch, it came to light that AMD disagreed with the cache testing methodology employed by leading test software vendors. That still holds true, apparently, even after recent updates to the utilities. From AMD's Ryzen 5 reviewer's guide:

Despite recent application updates, public cache analysis tools continue to produce spurious results for the AMD Ryzen family of processors. Though further investigation is warranted, we believe the public tools are utilizing datasets larger than the available cache size(s) and/or inadvertently mixing DRAM access into cache results.
 
With specific regard to the AMD Ryzen 5 processor, we believe these applications are making erroneous assumptions about the size and/or topology of the 1500X and 1600X cache hierarchy. We continue to work with software vendors to accurize the results.

The first chart above comes from our performance data, while the second is AMD's reference results. We used Sandra's cache latency measurement, which produces similar results as AMD's numbers. As you can see in our chart, L1 latency (far left) weighs in at 1.1ns, L2 measures 4.8ns, and L3 registers 10.2ns. By comparison, it appears that Intel enjoys a considerable cache latency advantage.

We also tested bandwidth with other utilities in an attempt to match AMD's results, but noticed large disparities in L2 and L3 cache reporting. We've included AMD's in-house numbers, and are talking to other ISVs to facilitate accurate measurements. 

MORE: Best CPUs

MORE: Intel & AMD Processor Hierarchy

MORE: All CPU Content

Create a new thread in the US Reviews comments forum about this subject
This thread is closed for comments
129 comments
Comment from the forums
    Your comment
    Top Comments
  • RCaron
    Great review, but I agree there's an incorporated bias against AMD.

    I think it does well for the review that it is, a technical review. As such a technical review can't really postulate on where the programming world is going with respect to mutlicore processing.

    There has been a speed bottleneck in processors that is holding up research on 3D modelling, virtual reality, and see-and-avoid systems (object detection and avoidance) for drones. None of these technologies can be solved using Intel's single core strategy (researchers have tried around the world and gave up), multicore processing is required. (Supercomputers are too expensive of a solution)

    Games will also be going this direction. You can do more faster with more cores, even if those cores aren't amazing (like Intel's cores). There is no beating this. This means for the next year or more Intel will be behind AMD in multicore applications.. but I'd expect Intel's response to be impressive.

    What these reviews do show, and this is where the author ultimately fails, is that Ryzen is fast enough to play games at high resolution, without noticing any real difference in performance. Future games that are programmed for mutlicores will run faster on a Ryzen CPU with more cores than on the comparable Intel offering with fewer cores. This article shows this.

    So the final comment should be, would you rather spend $240 on a chip (not to mention motherboard and ram) that is going to be slow as molasses running games optimized for multiple cores next year, or do you want to pay $249 for a chip that will only get faster from year to year as more and more software is designed for multicore processing.

    At this point the decision is simple. Intel as it is right now, is selling yesterday's technology.

    If you buy a computer to last 1 year, then buy Intel. But if you want something that will last 5 years and still be able to play games at high resolution then your ONLY option is buy a Ryzen. There is no escaping this fact, it's the single common comment by most technical reviewers, including this one. He clearly states that Intel is benefiting from software designed to run single-core processors.

    Add to that that Ryzen boards are compatible with future Ryzen chips, while Intel will have to come out with a completely new motherboard and ram specifications when it answers AMD next year. If you buy Intel you're literally throwing your money in the toilet. with respect to the future performance of your new computer.

    This isn't a just a choice between chips, it's a choice between a retiring platform (single-thread) versus the adoption of a new platform (multicore, parallel processing).

    A great many people will be buying AMD for this reason.
    34
  • irish_adam
    you say that the i5 7600k comes out on top at stock but just on the gaming benchmarks i make it 4-4 with 2 draws. I wouldnt say that it came out on top at all. I would say they are pretty evenly matched at the moment. Also apart from the odd couple from both sides their frame difference was less than 10, at over 100FPS i'd pay good money that no one would be able to distinguish a difference between either system.
    25
  • ninja_warrior
    If you can reliably overclock any of the ryzen 5/7 to 4.0, why would you get the 1600x over the 1700? Comparing a 1600x at 4.0 to a 1700 at 3.0 and then concluding that it's a better CPU when the 1700 can overclock exactly the same seems pretty stupid
    19
  • Other Comments
  • ninja_warrior
    If you can reliably overclock any of the ryzen 5/7 to 4.0, why would you get the 1600x over the 1700? Comparing a 1600x at 4.0 to a 1700 at 3.0 and then concluding that it's a better CPU when the 1700 can overclock exactly the same seems pretty stupid
    19
  • dstarr3
    Well, good effort from AMD, at least.
    3
  • bloodroses
    A little disappointing for the Ryzen 5's imo. You'd think with the reduced core count you'd get better frequencies (and OC'ing) than what you get with the Ryzen 7.

    I honestly don't see a reason why to get a Ryzen 5 at this point since the i5 is definitely better for gaming and the Ryzen 7 is better for workstation use. The price alone takes it out of its own market.
    -13
  • tamban
    A CPU review with only gaming benchmarks? Tom's hardware really likes Intel's hardware.
    -8
  • FormatC
    Anonymous said:
    A CPU review with only gaming benchmarks? Tom's hardware really likes Intel's hardware.

    Try page 10 :P

    31(!) Workstation benchmarks. Too less?
    12
  • Oranthal
    How about a real world test where you play a game and run a 1080p stream then compare performance? How about 1440p? How about broadening the scope of testing? Nah just ignore the strength of more cores and focus on single thread work and a few games.
    11
  • tamban
    Haha, my bad.
    2
  • irish_adam
    you say that the i5 7600k comes out on top at stock but just on the gaming benchmarks i make it 4-4 with 2 draws. I wouldnt say that it came out on top at all. I would say they are pretty evenly matched at the moment. Also apart from the odd couple from both sides their frame difference was less than 10, at over 100FPS i'd pay good money that no one would be able to distinguish a difference between either system.
    25
  • elbert
    Great review Paul and Igor. Best review I have seen given its the only review with 2 intel cpu's in the price range of Ryzen 5. The RAM info is great which shows that Ryzen gains a real 9ns latency advantage using higher clocked RAM on the Ryzen 5. Given the Ryzen 7 has less cache per core I would expect that gain to be higher.

    An issue that does stick out here is high price of the overclocking solution. How does the 7600k fair with a stock intel heatsink compared to the 1600x wraith spiral best overclocks? I think Ryzen has a real price advantage given the cooler required for a reasonable overclock.

    Also how does the 7600K compare in games while twitch streaming against the 1600X?
    5
  • dstarr3
    Anonymous said:
    How about a real world test where you play a game and run a 1080p stream then compare performance? How about 1440p? How about broadening the scope of testing? Nah just ignore the strength of more cores and focus on single thread work and a few games.


    Maybe that's your real-world test, but that isn't mine. And am I the only one that can see the workstation benchmarks on page 10? Everyone seems to be ignoring them and then complaining that they aren't there.
    5
  • RCaron
    Great review, but I agree there's an incorporated bias against AMD.

    I think it does well for the review that it is, a technical review. As such a technical review can't really postulate on where the programming world is going with respect to mutlicore processing.

    There has been a speed bottleneck in processors that is holding up research on 3D modelling, virtual reality, and see-and-avoid systems (object detection and avoidance) for drones. None of these technologies can be solved using Intel's single core strategy (researchers have tried around the world and gave up), multicore processing is required. (Supercomputers are too expensive of a solution)

    Games will also be going this direction. You can do more faster with more cores, even if those cores aren't amazing (like Intel's cores). There is no beating this. This means for the next year or more Intel will be behind AMD in multicore applications.. but I'd expect Intel's response to be impressive.

    What these reviews do show, and this is where the author ultimately fails, is that Ryzen is fast enough to play games at high resolution, without noticing any real difference in performance. Future games that are programmed for mutlicores will run faster on a Ryzen CPU with more cores than on the comparable Intel offering with fewer cores. This article shows this.

    So the final comment should be, would you rather spend $240 on a chip (not to mention motherboard and ram) that is going to be slow as molasses running games optimized for multiple cores next year, or do you want to pay $249 for a chip that will only get faster from year to year as more and more software is designed for multicore processing.

    At this point the decision is simple. Intel as it is right now, is selling yesterday's technology.

    If you buy a computer to last 1 year, then buy Intel. But if you want something that will last 5 years and still be able to play games at high resolution then your ONLY option is buy a Ryzen. There is no escaping this fact, it's the single common comment by most technical reviewers, including this one. He clearly states that Intel is benefiting from software designed to run single-core processors.

    Add to that that Ryzen boards are compatible with future Ryzen chips, while Intel will have to come out with a completely new motherboard and ram specifications when it answers AMD next year. If you buy Intel you're literally throwing your money in the toilet. with respect to the future performance of your new computer.

    This isn't a just a choice between chips, it's a choice between a retiring platform (single-thread) versus the adoption of a new platform (multicore, parallel processing).

    A great many people will be buying AMD for this reason.
    34
  • rgd1101
    Would like to see the charts add i7 7700k/7700 for comparison
    0
  • bigedmustafa
    It was a little weird seeing the i7-7700k disappear from all of the workstation benchmarks; such a comparison might have actually highlighted the value of Ryzen 5. It was also weird seeing the FX9590 appear only on the power consumption tests when the FX8370 was used for software testing. Whichever chips are chosen for comparison, it would be nice to see those same chips shown consistently throughout the review and not dropped or replaced from section to section.
    12
  • FormatC
    I removed the 7700K from workstation, because my colleague does the same in gaming section. It is a totally other price target.

    And:
    We used the Creators Update to be fair to AMD, but only I need one day per CPU for one workstation run. This all was a decision, how to use the time optimal. Someone asked about 1440p benchmarks... What is better? Creators Update and fresh content or all CPUs and outdated crap?


    The power consumption thing is right, but my mainboard for the FX-8370 with the soldered shunts for measuring was damaged (the 5th in the last two years). So I was not able to measure the power draw on the same way and I had no time to build me the same setup again in this few hours for testing.
    0
  • ykki
    What gpu was used? Can't find it in the test setup table.
    3
  • FormatC
    As every time: GTX 1080 FE...
    (take a look at "US all")
    2
  • captaincharisma
    Anonymous said:
    How about a real world test where you play a game and run a 1080p stream then compare performance? How about 1440p? How about broadening the scope of testing? Nah just ignore the strength of more cores and focus on single thread work and a few games.


    riiiiiight, cause 6 lackluster cores can beat out 4 quality cores all day . you should know AMD's motto is always going to be quantity over quality. but hey, if a few hits in performance are good enough for people then buy it.
    -9
  • tripleX
    Anonymous said:
    Great review, but I agree there's an incorporated bias against AMD.

    I think it does well for the review that it is, a technical review. As such a technical review can't really postulate on where the programming world is going with respect to mutlicore processing.

    There has been a speed bottleneck in processors that is holding up research on 3D modelling, virtual reality, and see-and-avoid systems (object detection and avoidance) for drones. None of these technologies can be solved using Intel's single core strategy (researchers have tried around the world and gave up), multicore processing is required. (Supercomputers are too expensive of a solution)

    Games will also be going this direction. You can do more faster with more cores, even if those cores aren't amazing (like Intel's cores). There is no beating this. This means for the next year or more Intel will be behind AMD in multicore applications.. but I'd expect Intel's response to be impressive.

    What these reviews do show, and this is where the author ultimately fails, is that Ryzen is fast enough to play games at high resolution, without noticing any real difference in performance. Future games that are programmed for mutlicores will run faster on a Ryzen CPU with more cores than on the comparable Intel offering with fewer cores. This article shows this.

    So the final comment should be, would you rather spend $240 on a chip (not to mention motherboard and ram) that is going to be slow as molasses running games optimized for multiple cores next year, or do you want to pay $249 for a chip that will only get faster from year to year as more and more software is designed for multicore processing.

    At this point the decision is simple. Intel as it is right now, is selling yesterday's technology.

    If you buy a computer to last 1 year, then buy Intel. But if you want something that will last 5 years and still be able to play games at high resolution then your ONLY option is buy a Ryzen. There is no escaping this fact, it's the single common comment by most technical reviewers, including this one. He clearly states that Intel is benefiting from software designed to run single-core processors.

    Add to that that Ryzen boards are compatible with future Ryzen chips, while Intel will have to come out with a completely new motherboard and ram specifications when it answers AMD next year. If you buy Intel you're literally throwing your money in the toilet. with respect to the future performance of your new computer.

    This isn't a just a choice between chips, it's a choice between a retiring platform (single-thread) versus the adoption of a new platform (multicore, parallel processing).

    A great many people will be buying AMD for this reason.


    First time poster, almost scripted response. Thanks, AMD Spokesperson!
    -6
  • tripleX
    Anonymous said:
    How about a real world test where you play a game and run a 1080p stream then compare performance? How about 1440p? How about broadening the scope of testing? Nah just ignore the strength of more cores and focus on single thread work and a few games.


    1080p streaming argument is no bueno. No one does that on the CPU, and if they do, they use Intel's dedicated hardware (quick sync) to do it. AMD doesn't have integrated graphics, so how do you do a like comparison of that?

    So again, its not a real world scenario. That's why you see absolutely no sites testing it. Marketing fluff.
    -2
  • ykki
    Anonymous said:
    Would like to see the charts add i7 7700k/7700 for comparison

    For productivity charts, yes. For gaming, ehhhh.....
    0