AMD Ryzen 7 1800X, 1700X and 1700 Processor Review

Jump To:

AMD Zen Architecture, SenseMI and XFR Technologies


With AMD’s Zen microarchitecture used on Ryzen processors, AMD has been able to get greater than a 52% IPC improvement over AMD’s previous desktop design. This is thanks to a 1.75x larger instruction scheduler window, a 1.5x grater issue width and resources as well as a new micro-op cache that allows the L2 and L3 cache to be bypassed with frequently-accessed micro operations. To top that off, AMD came up with a great SMT and even made a neural network-based branch prediction unit to prepare the right instructions and pathways are being used on future workloads.

AMD Zen CCX Unit

The AMD Zen architecture is comprised of what they call a CCU Complex (CCX), which is a 4-core, 8-thread module. AMD can place more than one of these CCX units together on a package, which is exactly what AMD did to get the 8-core, 16-thread processors in the Ryzen 7 processor series. Each CCX has 64K L1 I-Cache, 64K L1 D-Cache, 512KB dedicated L2 cach per core and then finally 8 MB of L3 cache shared across all the cores. In processors where more than one CCX is used, AMD connects the CCXes together with their high-speed Infinity Fabric (this fabrid also handles system memory, I/O, PCIe and more). From what we have heard AMD has been tinkering around with connecting up to four of these complexes together, so at some point in the future a 16-core, 32-thread processor is entirely possible and likely going to make its way to the desktop market when the time is right.

AMD Neural Net Prediction

Each AMD Ryzen processor features AMD SenseMI technology, which is basically a ton of interconnected sensors that are accurate to 1mA, 1mV, 1mW and 1C with a pooling rate of 1000/sec that allows for real time adjustments to the processors behavior. This data allows for improved power use of the cores, management of the speculative cache fetches and even to perform AI-based branch prediction.

Ryzen XFR 1800X

Thanks to the real time power, temperature and load data from the Infinity Fabric, AMD precision boost can adjust the processors clock speeds on the fly in 25MHz steps. This ‘fine’ clock tuning means that you are able to stay at the highest possible clock speed for that particular workload. AMD also has a new feature called Extended Frequency Range (XFR) on all Ryzen processors that have the -X suffix at the end of the model number.


XFR increases the maximum Precision boost frequency on systems that are using high-end CPU cooling solutions (think high-end air cooling or most any liquid cooling loop). This additional headroom is given automatically thanks to the SenseMI data that allows the processors distance to the pre-set junction thermal limits and giving you additional clock frequency when you have some headroom available.

What is the rough temperature threshold/range at which XFR kicks in?

It starts to kick in below 60C Tj and gradually allows faster frequency down to ~25C Tj

What is the rough temperature threshold at which XFR remains stable at 4.1GHz (e.g. no more “dithering” between 4.0/4.1)?

Roughly 25C Tj but that will depend on the heatsink performance and the system ambient temperature and the workload being run.

Let’s take a look at the test system and then move onto the benchmarks!

Jump To:
  • Charles Mather

    I got 156 single core on cinebench R15 for my 7700k at 4.7GHZ

  • cryptoflower

    Intel: Anti-Competitive, Anti-Consumer, Anti-Technology

  • Michael Carroll

    Is the only difference between a 1700 and a 1800x the clock speed?

  • ja_1410

    Close but no cigar this time. Great effort, but I would still pick 7700 based on the average of the presented benchmarks and price. Hopefully, unlike Intel mature architecture, Ryzen is new and might have some low hanging fruit to be improved soon with upcoming next generations by larger amount than Intel can muster from its line. Also it would be interesting to see performance, when the software will get some optimization toward Ryzen. Currently most software has been compiled having only Intel optimization in mind. With this new CPU being very good at some multi threading tasks and good for number crunching we might have some market share for AMD. That in turn will cause new software to be optimized better for both chips and it might improve the Ryzen position.

  • Brian

    try with windows 8.. scheduler works much better compared to 10

  • msnengco

    SMT & memory lattice issue?

    Thank you for the balanced and professional review of the AMD Rysen CPU’s. The real and synthetic benchmarks of Ryzen CPU are essentially “twins” compared to Intel’s i7 6900K. I am puzzled thou. The gaming frames per second lags against the Intel CPU’s. One would expect there would be not much difference between the CPU’s during game testing which leaves one to think that this maybe is an game engine optimization issue (drivers etc.,). Did your testing evaluate the SMT on and off during game testing for AMD’s Rysen CPU, and to be fair likely wise for Intel’s HT for Intel’s CPU’s? I am wondering if the Rysen CPU’s are hampered because of combination of bios (UEFI), lack of game engine optimization, windows driver updates, and memory lattice issues. If there is a significant improvement with SMT turned off this would indeed unfortunately invalidate the gaming testing of your report. Could you please repeat some of the game testing for comparison. Thank you.


  • koenshaku

    I was more interested in seeing how these performed against intel’s 4+ core processors in gaming. We all know games are optimized for quad-core processors and performance falls off for processors with more than 4 cores…

  • Coach

    Ryzen is a beast in multi-core workloads and it will only get better. Here are my Geekbench 4 numbers comparing my i7-4790k @4.9 oc and Ryzen 7 1800X @3.99 oc with SMT disabled (8c/8t only)

    4790K=5631 single core, 18227 multi-core (I was in the high 80s C)
    1800X=4349 single core, 18600 multi-core (I maxed out @72 C)

    This is 8 cores vs 8 threads which puts the 4790k at a bit of a disadvantage, I know, but it is also almost a whole gig faster clock-wise. Other Ryzen scores on Geekbench 4 show it scoring around 27450 @4ghz, with SMT enabled I assume.

  • Eric Marshall

    Are those vengeance DIMM’s used for the test setup single or dual rank? Reason I ask is; previous generations of AMD chips have benefited from rank interleave a lot more than Intel counterparts. In other-words, if the test was performed without rank interleave on any of the systems, then Ryzen may have been at a greater disadvantage than if all the systems were tested with dual rank DIMM’s. Also, seems a bit odd to be comparing the X99 system with only a dual channel kit. Not like anyone building that sort of system would skimp on the quad channel arrangement anyway.

    • Coach

      My bios gives me the option (Gigabyte-AMI) to enable channel interleave and rank interleave. I believe I have them enabled. I’m guessing other board partners have those options.

  • i am grad you guys didn’t join the 720p and 480p low settings benchmarks hype i have seen this famous tech review website they only posted 640 x 480 gaming benchmarks showing 7700k beating ryzen 1700 by like 50 60 fps

  • Coach

    Some of my initial Ryzen 7 1800X (w/RX 480gpu, 16gb ddr4 3000, Gigabyte Aorus Gaming 5 mobo)
    **Note: I’m not sure if the ram is running @ correct speeds–bios says 2933, but cpu-z shows 2133 (which is the default) I have it set on XMP. This BIOS is new to me, but it seems quite limited in ability to tweak memory. On my Asrock mobo, I can change each and every memory clock & setting. This is very limited in comparison. **
    Cinebench R15 @3.6 stock=1540, @3.8ghz=1613.

    TimeSpy=4543 (compared to my i7-4790k @4.4ghz=4339)
    (TimeSpy gpu=4241, cpu=7623–CPU not recognized by Futuremark) I got the Steam achievement for “mystery machine” because of that. 😉

    CPU-z @3.6ghz, single core bench=2137 (i7-4790k@4.4=2003)
    CPU-z multi @3.6=14504 (i7-4790k=8387)

    SuperPi Mod 1.5XS 1M=11.903s @3.6ghz (stock)
    SuperPi Mod 1.5XS 32M=10m 21.343s @ stock

    I cannot wait for optimizations, etc. As an experienced builder and mediocre overclocker with AM3 and Intel, this is new and I need to learn more about my system to get a good and stable setup. I believe there is a new bios out for my mobo, even though the bios is very new. I have not yet flashed the newer version, but will do so and re-test as I go. The CPU generally runs pretty warm. I’m not sure why, etc. I’m using a Corsair H110i to cool the system. Regardless of thermal paste setup/burn-in, I would expect cooler temps. I’m @ 52C most of the time at idle but when I ran Cinebench R15 @3.8ghz, I hit 73C max, which isn’t necessarily crazy hot, stock was 60-something(?). Again, time will tell us all more.

    • Coach

      SMT was left at stock setting (enabled) for tests above.

  • Disable SMT and run the gaming benchmarks again. Until Microsoft issues the Windows 10 gamers update, this is the only way around the issues right now. AMD is also working on releasing patches over the next few months to fix other issues such as system memory support and buggy turbo boost. It seems you were one of few reviewers who had the Precision Boost and XFR features working properly, scoring 163 cb in Cinebench R15 ST; the highest I’ve seen thus far at stock.

  • After reading the whole review I do see how price/performance is a bit better in most cases with ALL the cpus and really shines with the 1800X due to the price point being half of Intel’s counterpart. I do not understand why the driver was not ready at launch. I mean, people see the first numbers and go, ok it’s not bad, when with the drivers it could be a HUGE game changer…maybe…..who knows if it is just a lot of hot air until they (the drivers) are actually released. It is unfinished yet pushed out to meet a deadline, that’s a shame. Also the 1080P gaming debacle is just sad. Sweet spot for gaming is still 1080P – – so shouldn’t they have focused on that?

    I would still be happy buying one of them as I don’t game much on the computer anymore (ps3 and ps4 for most of my gaming) and it is a nice price point for performance that should get better when the drivers (which should have already been ready) are released that fix the problems (maybe, hopefully).

    I quoted myself from the forums – – hop on in and discuss it further.

  • Riptor

    I want to know how this is a disappointment? The Ryzen 1700 is a $330 CPU that is 40% faster than the i7-7700k, all while using less 30% power. You can overclock it and have a CPU within spitting distance of a $1,000 intel CPU. An excellent processor which in my opinion even exceeded the hype.

  • David Hall

    Seems pretty darn good for me on price and performance. I’d happily sacrifice some fps today for gains later as the use of multigpu rises, drivers and BIOS mature. Can’t wait open play with the RAM timings. Who would have thought you could get a fast 8C16T cpu and mobo combo for such reasonable prices. It’ll make a nice addition to my Intel, Nvidia and AMD family.

  • jojigzaw

    Thanks for putting in KeyShot. It will be great if SolidWorks and Inventor was on the test to.

  • Ed

    It’s launch day…can’t expect everything to be perfect on an entirely new CPU. I am a bit disappointed in the gaming numbers but feel better seeing the feedback from Oxide and CA both saying the CPU optimizations are coming. I think I will hold off 60-90 days an see how things shake out with Ryzen before building a new gaming rig…

  • Vanesa Fernandez

    Intel is altering those test-ers.

  • cris levin

    at 1080p, as long as it is >60fps, does it really matter if its 200 or 300fps?

    • Pui Ho Lam

      144 or even 240Hz monitors exist

      • Coach

        fps and hz are two different things.

        • Pui Ho Lam

          fps means frames per second. Hertz means per second without specifying the unit of what you’re counting. On monitors, it’s apparently the refresh rate of pictures, ie. frames.

      • Coach

        I don’t have time to educate you, so here…read this link. They ARE different.

        • Pui Ho Lam

          LMAO, you think I don’t know when the output is not perfectly in sync with the refresh rate?
          But that’s not the point in this case.

  • All right, everybody can relax.
    AMD fanboys: Ryzen is miles better than FX.
    Intel fanboys: Intel is still better overall.

    Btw, the i3 7350k, currently selling at $178, doesn’t count as a low-cost processor (bottom of page 11 of the review). That would be the Pentium G4560 ($62).

  • Dusan

    This is good, this is competition. It will make Intel work harder next time to give us more than just 1% performance increase. Go AMD!

    • Nathan Kirsch


      • gamersglory

        I look at this and think most of the gaming setbacks can be fix with mature chipset drivers and BIOS. I am curious though if ryzen work better with the latest windows insider build as the most up to date codebase would be in it

        • Nana Hermawan

          look at power consumption, ryzen not realy work hard in gaming test, maybe 20% lost potential

    • Toss

      -1% tbh.

    • John

      Or you could just buy the AMD solution and not worry about Intel’s next offering.

  • tacominnie

    It is even slower then an i5 7600k on games, ofc if you bench at 4k you will be gpu bound and you wont notice a difference even when compared to a G4560.

  • Mahesh Walatara

    Fairly Disappointment- but the 1800x is great for Photoshop Premier Pro use.

    • Mark

      Puka nodi palayan ponnaya … english grammer danneth na tho, wesi …

  • Andrew S

    A little disappointed considering the hype.

    Seems that at 1080p these CPUs can’t keep up with even the likes of the 7700k.
    But at 1440p/4k it’s much better. Maybe the extra cores/threads make the difference there.

    • Yes Indeed

      nothing keeps up with the 7700k. clock speed rules in most games and that cpu gets 5ghz.

      • Clock speed combined combined with great IPC, you mean. If just clock speed were the primary determinant, the FX-9590 would be king.

    • al

      No, extra cores/threads are not making the difference there. The reason is that the system goes from cpu-bottleneck to gpu-bottleneck.
      The sad part for me is that the Windows CPU drivers, UEFI bios were not ready, and in a month most people will not see that review.
      That the games were not optimized was expected, most of them are probably using at max 2-4 cores, and maybe some Intel compiler cheat magic on some libraries is still there.

      • Nathan Kirsch

        You hit the nail on the head on all your comments. I really wish the Windows drivers were ready before we were given the parts to review. Instead we got a statement 24 hours before launch from AMD saying that they’ll be coming in 30 days if all goes well. Game optimizations will be hit or miss when they come, but they appear to be coming. That takes time though and we’ll see what happens.

        • Tyrann

          where did they state that? i wanna read it 🙂

        • Nathan Kirsch

          I put it in the conclusion on the last page. The quote came direct from AMD’s John Taylor. There was talk of it coming with Ryzen 5 and then they said ‘in the next month’ in an official statement that was e-mailed out last night.

        • Tyrann

          thank you

      • epigramx

        > intel compiler cheat
        That’s a stupid meme. Most benchmarks are compiled on either gcc or vstudio. Let alone even if that meme was even true – which it isn’t – those CPUs are 99% identical on instructions anyway so even if it’s optimized well for an Intel it’s de facto optimized 99% well for an AMD.

      • tacominnie

        Amd fanboys are famous for excuse finding…That’s a good sign right?

        • disqus_GB8lUuziuG

          intel fanbots game at 1080p with TitanXP then bitch on the internets that their 7700k is uber fastest cpu of all time..

    • disqus_GB8lUuziuG

      only if you play 1080p with a gtx 1080 or titan.. these are misleading reviews. look at 1440p or 4k if you game with a 1080 or titan. If you game at 1080p look at reviews with 1070 or less for more accurate results.

      • Andrew S

        Yep completely agree with that statement.
        I’m personally waiting to see the 1300, 1400 and 1400x perform with something like a gtx 1060/RX480.
        That’s where the market is the most interesting in my opinion.

      • haralampi nedelin

        yeah, it does better comparatively at 1440p/4k, due to GPU bottlenecking. an i3 gets similar framerates on most titles here, so no point in spending $350 or more on ryzen. The 1080ti will widen the bottleneck and reveal, once again, that higher IPC and clock is what makes the difference, not core count

        • disqus_GB8lUuziuG

          Exactly buy the R5 or R3 Ryzen if you need a gaming chip, good point. Nah 1080ti won’t do that at all, it will be just as things are now.

    • haralampi nedelin

      both intel and AMD run into GPU bottlenecks at those resolutions. The 1080ti will widen that bottleneck, and maybe we’ll see true performance differences.

  • well that was quick, ,6.09 Am Pacific Time Vancouver Canada, and i am reading your report, exciting, thanks,