AMD’s Ryzen Threadripper 1950X is significantly slower in games than the Intel Core i9 7900X

The first third-party gaming benchmarks for AMD’s Ryzen Threadripper 1950X CPU have been unveiled and the results are quite disappointing. Naturally, we don’t really suggest this CPU for gaming enthusiasts, however AMD’s latest offer is significantly slower in video-games than its Intel’s competitor.

The AMD Ryzen Threadripper is a beast featuring 16 cores and supports 32 threads. Its primary competitor – at least for now – is Intel’s Core i9 7900X; a CPU that packs 10 cores and supports 20 threads. And from the looks of it, Intel’s offering is better suited for gamers than AMD’s Ryzen Threadripper.

GamersNexus and Sweclockers have shared the first gaming benchmarks between these two CPUs and as we can see, Intel’s CPU offers better performance in every game.

Both the AMD Ryzen Threadripper 1950X and the Intel Core i9 7900X are priced at $999. And while the AMD Ryzen Threadripper is coming with more cores, it’s pretty obvious that no game really benefits from these additional CPU cores.

Now as we’ve already stated in the past, we are only judging this new AMD CPU based on its performance in video-games. And that’s because… well… we are a PC gaming website. We don’t expect any game to be able to use 16 cores in the foreseeable future, however we are aware of some gamers wishing to get this kind of CPUs in order to future-proof their rigs. So, if you are one of them, you should be looking at Intel’s CPUs.

76 thoughts on “AMD’s Ryzen Threadripper 1950X is significantly slower in games than the Intel Core i9 7900X”

  1. Reads like an Intel sponsored article. Yeah would love to buy a 16 core HEDT 1K+$ CPU and play muh video games @720p to get dem 15 more FPS.

      1. Wait to see the 8700K: it’s gonna be the new gaming king 🙂

        I’m between it and the 7820X for my next build.

          1. Cant wait for Coffee Lake, rocking 5820K @ 4,5Ghz now, but I have a feeling that 8700K going to overclock to at least 5Ghz and probably over 5Ghz on water and delid.
            My body is ready, I only need to get a new mobo, CPU and sell my 32Gb DDR4 KIT (x4 sticks of 8Gb @ 2666Mhz) and get dual channel 32Gb 3200Mhz kit.

          2. You dont like to upgrade very often?
            I never had Quad core Phenom, but back than AMD released Phenom 1090T, it was first mainstream 6 core with good performance, after that i had Q6600 (or before that?), than 2600K, than 4560K and now 5820k and cant wait to upgrade.
            Was considering Ryzen, since i game in 4K/60Hz i dont care about

          3. The quad-core phenom 2 955 was a pretty nice chip. I went fx-8350 after and was pretty disappointed as far as games were concerned. Great for encoding/decoding but thats it. Rocking 7700k now and love it.

          4. Talking about fanboys… Two heads arent better than one when both are empty. “Old” I7 7000 series intel processors are ripping off the i7 8000 series. LOL.

    1. This whole “CPUs should not be tested at 720p during the gaming benchmarks” sounds really like the “the human eye can’t see more than 24fps”. Both are ridiculously incorrect. It will take a few years until people realise that there shouldn’t be any GPU limitation during the CPU tests (and at 1080p there are GPU limitations on current-gen titles even with the GTX1080Ti). So unless you have a single GPU that is as powerful as three GTX1080Ti (which… well… there isn’t any) you can’t have proper CPU gaming tests at 1080p. This also explains why a lot of different CPUs are so close at 1080p

      1. While I agree with you, I think that you’re being a bit overconservative here (althought I’m all for non-GPU bottlenecked benchmarks).

      2. True. For a gaming CPU test the GPU interference should be eliminated as much as possible. The way to do that is set the resolution low enough such that the GPU isn’t running at 100% and causing the CPU to have to wait on the GPU to render frames.

        A while back on TechPowerup the owner put up a poll asking members if they wanted to keep the 720p benches in CPU tests for gaming. About half said no and from comments like “no one games at 720p these days so there’s no point in benching at 720p” you could see that they really didn’t understand that a CPU gaming bench is to test the CPU to see how many frames it is capable of.

        1. Insane comment…
          720p performance will not show anything meaningfull, 1080p on the other hand will 61 vs 62 fps on 1080p while 70 vs 78 on 720p does not matter, you will be playing at 1080p anyway, they are essentially equal.
          But the price…
          It’s not like they are testing with 4k settings. LOL

      3. 720p performance says literally nothing about CPU performance today or the future. FX-series overtook its respective i5-series over time because of more games using more threads, the 720p benchmarks NEVER predicted that. The 1080p/4K benchmarks show clear enough you shouldn’t get a TR if you’re only going to play videogames on it, no need for your pseudo-science of 720p predictions.

      4. As I recall the old arguments for such testing was to predict “future performance”, once those bottlenecks were gone.

        Problem is it doesn’t even do that, it’s been clearly shown to be a fallacy.
        The results of previous tests using this method, don’t match the actual results of that future performance.

        So it’s got no benefit for today, or tomorrow, and is useless.

        Results consistantly show AMD CPU’s “Fine Wine” improvements far outpace, these type of tests, and where they predict, that Intel should be better, in actual future testing AMD is better.

        The question then becomes, when such tests are useless to current gamers, and give incorrect indications of future performance, why do them?

        Just for those that think I’m an AMD fanboy, I use an i7, and GTX1080, no AMD in my rig at all.

    2. You sound like a butt-hurt AMD fanboi… Do you understand the concept of benchmarking games at lower resolutions is to remove the GPU bottleneck from the test. If you can’t figure that out however then you clearly have no clue what you are talking about. Why would this article sound like an Intel sponsored article? At least here John isn’t cheering on one brand over another like he has been doing on his Vega articles… Believe me… John wants AMD FTW… I would imagine it is with a VERY heavy heart he even had to publish this article…

    3. Simply
      When the GPU on full load the CPU spend most of the time waiting for the GPU
      So, you do the test on lower resolutions to make the GPU not work @ full load

      it’s not a real problem for Threadripper with current GPUs but it’ll be a problem with more powerful GPUs even on 4K

  2. Ohh, surprise surprise, I’ve been saying that ever since when TR was f- ucking announced.

    I love being right 😀

  3. At least the top of the range Threadripper performs more competitively in the game most known for effectively utilising multi-core multi-threaded CPUs – Ashes of the Singularity: Escalation.

    With games increasingly expected to take advantage of extra CPU cores and threads in the future then that bodes well for Threadripper as a workstation plus gaming CPU. Obviously there are better value choices available if the primary purpose is gaming.

    1. Try as you might you can’t spin Threadripper into being a good buy for gaming. I give you a A for trying though.

  4. Of course it doesn’t perform well in games. Did you ever point that R7 1700 overclocked perform closely to i9-7900X in games for 1/3 of price? I doubt you did.
    Judging Threadripper just for gaming performance is stupid, doesn’t matter if DSOG is just a gaming website. AMD managed to bring fair competition and because of that Intel has began to make their move, that’s why they’ve launched those Core i X series.
    You forgot to mention that 1950X mops the floor with 7900X in other applications while costing less, with less heat and power consumption.

    I really love this site, I come here everyday, but sometimes is hard to take you seriously. You sound like a Intel + Nvidia fanboy.

    Come on, you can be better than that.

    1. And you don’t sound like an AMD fanboy at all. “OMG JUST OC TO DEATH THE 1700 UNTIL IT REACHES i9-7900-LEVELS!”. Because f- uck objective and equalitarian benchmarks.

      How deluded people can be….gosh.

      Also, don’t be hypocrites. Like if AMD haven’t clearly bought articles here and there, starting by LinusTechTips and GamersNexus.

      1. Oh, yes, AMD buying GamersNexus so they can show how bad Vega and Ryzen performs.

        Most sites takes into account overclock and there’s nothing wrong about this.

        I’m not being a fanboy (FFS, I have a 2700K and a GTX 1070…), I’m just criticizing DSOG for point cons in AMD CPU’s while never doing the same for Intel’s. It’s spreading misinformation.

        1. They sure joined the “talking trash about Skylake-X without any reason to” in their time.

          And DSO complained about thermals in SL-X as well if I remember correctly, which is about the only thing that can be TRULY complained about Intel’s latest releases.

          1. You do realize that that “trash” completely shames ALL their equivalent Ryzens, right?

            Or it’s your AMD fanboy hypocrisy not letting you see reality again?

          2. My CPU is i7 6700, thank you.

            You are trying 2 hard m8. Just look at all your posts on this topic and guess who is the fanboi here.

          3. What are you reading? All I see are angry AMD fanboys mad because John said Threadripper wasn’t a good cpu for gaming.

    2. Even without overclock that ihave it on stiock 3.0ghz sinc march 2017 that i bought it it beats all intel 4 and 6 core cpus. But i play on 1440p monitor DELLP2416D with GTX 970 G1 GAMING.

  5. Totally one sided “opinion”..most of other hardware review sites saying the opposite…but here, seems like dsogaming have some sort of hate towards a certain brand/product that, dsogaming posting about their “hate” by continuously posting negative headlines…..loosing my trust about this website..been a fan for over 4/5 years ..no point writing article to gain negative attention for cheap page views. Such shame.

    1. Didn’t you hear? Last week we were paid by AMD for our articles. This week we are paid by Intel. Next week we may be paid by Sony and make it a console-only website 😛

      1. As long as you’re not on the Microsoft payroll then all’s good! Leave it to Sp4ctr0 to have a man-crush on Phil Spencer!

      2. I wasn’t joking or being sarcastic. Wrote an honest opinion, with respect. I dont care if amd or intel is superior or not..but the headlines this website posts, directly indicates a negativity towards certain consumers opinion or mind set. And this type of post getting repetitive….when i see there arent much of any news or report about gaming world in general,i see dsogaming posting something, which seems to be offensive or negative type marketing news, to catch views or simply a clickbait. This is my personal viewpoint. Btw, my pc includes parts from both intel and amd. Intel processor and amd gpu to be exact. Competition is good…but being one sided offensive hurts consumers prespective or buying capacity. Dont make facts into a fiction for clicks.

        1. how is it one sided? theyre reporting gaming benchmarks and they show intel doing better. last week they used some benchmarks for some other stuff and they showed how AMD did a lot better

          1. Let’s be honest: AMD has never “done better”.

            They’re just the brand of the poor, like always: the brand of those who can’t afford either Intel or Nvidia.

    2. I find it difficult to take somebody’s accusations seriously when they spell ‘losing’ as “loosing”.

      1. maybe because my mother tongue is not English? maybe mobile auto correction issue? This is my second post in this website if i remember correctly….reason of commenting is obvious. I run a small group and page about gaming related news….dsogaming used to be my number one source to get news…but now, i am thinking differently…..what happened to dsogaming that i used to know? … they/it used to be authentic, having balanced opinions.

    3. Go complain to LinusTechTips, and not here…I’d put my c- 0ck on fire that those were INDEED paid by AMD.

      Just incredible that AMD fanboys are so deluded that they can’t assume something as obvious like that TR was gonna completely s- uck in gaming.

      Also, more delusion: “most of other hardware review sites saying the opposite”. Most of the other hardware sites benchmarks say EXACTLY THE SAME and have the same results. Assume it.

  6. My microwave plays games significantly slower than my dedicated *gaming* rig.

    This is just getting stupid though.

    1. We are not the r- etards who said in their day “OMG WAIT FOR THREADRIPPER!”, back when Ryzen was widely regarded as a failure in gaming.

      Go complain to the deluded AMD fanboys, not to the ones who said the obvious (that TR was gonna s- uck even more), even back in the day. WCCFtech is full of them 🙂

      1. Pretty much.

        When 90% of the benchmarks say that it’s performing poorly, and 10% say otherwise, yep, that’s the definition of “cherrypicking”.

        1. Be more objective… Have you seen any benchmarks? Hardware Unboxed, gamers Nexus, OC3D TV, Paul’s Hardware… All are showing impressive results for thread ripper. My question to you why buy a 7900x for a 1000 when I can get threadripper for the same price.

  7. I read the Guru3D reviews for my PC tech and from all the test they did it is better than all the Ryzen Gaming CPU’s. The ones that you have linked in the article seem to be cherry picked to show weak benches. Just saying, and I don’t blame you guys for you did not bench the CPU your selves.

    I have been running an i7 3930K since releases and have only been updating my GPU and I play everything on Ultra 60+ FPS. Right now I can run both 1080p and 5760×1080 @ 60FPS with a 1080 Ti. I find it odd when CPU reviews talk like the new CPUs are sh|t when it is behind by 15 FPS(on over 120 FPS scores) in 720p or even 1080p tests. If you did the same tests with my i7 3930K it probably be a joke in comparison yet I play everything on Ultra 60 FPS. Really the GPU is what matters.

    Any new CPU now a days and even all of last year etc are going to be better than what most of us are running. Having said that a 32 logical core CPU for $999 that is = to AMD’s top of line gaming CPUs and can push 4ghz on all those cores is an amazing achievement and CPU. I really don’t care who the brand is, that is a crazy CPU.

    I don’t really even see the point to say anything negative about such a small FPS diff considering what type of workstation CPU it is. Really we should all be going bananas that such a CPU is available at a consumer level, this thing would of been like $4000 a few years ago or not even available let alone @ $1000.

    When I have the money it will be my next upgrade for my workstation / gaming rig.

  8. its an Intel sponsored article? I never seen some one complaining that a diesel truck its slower them a Ferrari lol, Threadripperits CAN play games, but its not its focus, if you want to just game the Ryzen 5 1600 and core i7 7700K are the CPUs to buy, not some quad-channel monster…

  9. ” however we are aware of some gamers wishing to get this kind of CPUs in order to future-proof their rigs”

    Future proof their rigs? Who the hell future proof their gaming PC 10 years in advance?

  10. Most if not all the reviews I saw on the internet are basically showing that indeed in gaming Intel get higher frame rates but in some cases the low frame rates are higher in the AMD ones giving a more “smooth” feeling while playing the games (without stuttering and such, same happened with reviews on Ryzen’s 1700/1800).
    It is funny because a lot of people mentioned this, mostly while playing and recording or streaming and rendering while playing. (which is the main factor/selling point of this type of CPU)

    Also there is one review from AdoredTV (a guy who does actual testings, not just benchmarks) which shows how changing the memory access between “Distributed” to “Load” makes a difference and got better results with it in games compared to the “Gaming Mode”. (not a massive difference but in some cases way better than “Gaming mode”)

    These CPU’s are not meant for gaming but can be used to play just fine, and at the same time do a lot of stuff in the background without even sweat. Also while having lower temperatures and same or even lower power draw compared with a CPU that has less cores and costs the same or more.
    I’ll loose 10 to 20FPS in gaming? Well if my PC was a console which does 30FPS only I would care but since I play at 60+ guaranteed I wouldn’t care so much.

    Also I laugh every time I see a GTA V benchmark…it is as valid as any other yes, but it is still using an old engine which was made for “old gen” consoles, the game is a console port…
    Now Intel has its architecture running for years, AMD made an experiment with FX processors which was not good at all, now it is trying to close the gap and in my opinion is quite close in single threading (which still Intel dominates) and crushing Intel in multi-threading if we have in mind the lovely price/core/performance ratio of course.

    People forgot how Athlon made Intel look like garbage back in the day until the “i” series came out…now that AMD is making a “coming back” it feels like people got retarded as CPU’s got way faster than a decade ago, let alone a year ago in which buying a consumer CPU with these amount of cores and performance was almost impossible.

    I say keep doing benchmarks with games that use 1 CPU mostly, keep doing them at 720p, keep making people believe that playing at 100FPS instead of 120FPS is terrible not having in mind that is better in the rest of tasks, and keep buying at a higher price so prices keep rising each year as happened with GPU’s…next year a mid end GPU will cost the same as a high end one today thanks to consumer stupidity, mining and other absurd factors.

    The place we live in ladies and gentlemen.

      1. It proves that that Samba guy posted fake results, in his results non overclocked TR is whole 6 fps more than 7900x, in actual results, 7900x is 4fps faster than non overclocked TR and 1fps more than overclocked TR BUT 7900x goes up to 5Ghz with delid and 4.7Ghz without delid on AIO cooler so do the conclusion.
        All of the Intel CPU above and below TR have enormous overclocking potential, up to 1Ghz more than stock, my 5820K does 24/7 4.5Ghz on auto, i changed the multiplayer, saved bios and done, 2 years of 4.5Ghz, if i tweak I can go higher i just dont care

  11. Now as we’ve already stated in the past, we are only judging this new AMD CPU based on its performance in video-games. And that’s because… well… we are a PC gaming website.
    Does this mean that when EPYC releases you’re going to bash it because it won’t be able to run games?

    /smh

  12. Your benchmark is fake from unknown crap web site, Guru3D is extremely trustworthy.

    Look at your RofTR benchmarks 7900x has 7FPS less than TR which is obviously fake.

  13. “however we are aware of some gamers wishing to get this kind of CPUs in order to future-proof their rigs. So, if you are one of them, you should be looking at Intel’s CPUs.”

    If you are a gamer please don’t buy either. Money would be better spent on even a Titan vs either CPU and please don’t buy that either lol.

  14. Well, its kinda sad to know things are going this way. The point is, yeah If I ignore the rest of the benchmarks done around internet showing how AMD Ryzen Threadripper is a decent processor and the fact it has 6 more cores for the same price, then I should be going for Intel, because Farcry.
    Yeah John you should at least try not only point the cons. Other gaming sites does point the pros outside the gaming sphere. Why? because sometimes we use our PCs not only for gaming, our gaming rig is our recording rig, our working rig, our multimedia rig, etc.

  15. Of this site would post an article like this.

    Doesn’t anyone notice they’re never fond of AMD hardwares, always running game tests and such using only Intel/Nvidia hardware.

  16. “We don’t expect any game to be able to use 16 cores in the foreseeable future […]”

    Then you’ll be surpriced. 🙂

Leave a Reply

Your email address will not be published. Required fields are marked *