German Retailer Sells Core i7-11700K Stock Before Launch, Benchmarks Arrive

News
By Aaron Klotz
last updated

Some Rocket Lake chips launched ahead of schedule over the weekend.

Mindfactory Rocket Lake i7-11700K Listing
(Image credit: Hexus)

Intel has a new problem on its hands: Reports indicate that German retailer Mindfactory has sold its entire stock of 120 Rocket Lake i7-11700K's to customers well before Rocket Lake's official debut, presumably late this month.

The big sell-off of 'unreleased' Rocket Lake 11700Ks has resulted in many customers benchmarking the chip and sharing their findings online. Plus, a few scalpers (naturally) are selling their own stock of i7-11700K's at inflated prices.

Reports say that Mindfactory was selling its 11700K's for a full three days before they ran out of stock, and employees of the company were told by managers that these Rocket Lake chips could be sold to customers. So it seems this sell-off was fully intentional by Mindfactory HQ.

Hardwareluxx has fully updated its 11700K article with benchmark findings from its forum users who have received the CPU and tested it for themselves. These i7-11700K's were tested in Cinebench R20.

Cinebench R20 Single Threaded Benchmark

  • Ryzen 9 5950X = 633
  • Ryzen 7 5800X = 622
  • Core i7-11700K = 600
  • Core i9-10900K = 530
  • Core i7-10700K = 506

Cinebench R20 Multi-Threaded Benchmark

  • Ryzen 9 5950X = 10146
  • Ryzen 7 5800X = 6022
  • Core i9-10900K = 5999
  • Core i7-11700K = 5749
  • Core i7-10700K = 4947

If these results are true, then the 11700K is 8% slower than the 5950X in single-threaded tests, and 15% faster than its predecessor the 10700K in the same benchmark.

In the multi-threaded test,  the 11700K is almost on par with the i9-10900K, despite the reduced core count, just 4% slower. The 10700K is about 15% slower than its Rocket Lake equivalent, similar to the single-threaded results.

But, these results can vary. We don't know what system specs, BIOS and drivers that were used in these benchmarks. So we'll still have to wait until the official reviews come out for the i7-11700K before we see how these chips really perform.

Still, these 11700Ks are the full-blown production units, not engineering or qualifying samples. So these results are most likely in the same ballpark as the final product with the finalized drivers.

We have yet to hear from Intel regarding the matter or if the tech giant is considering any kind of legal action against Mindfactory for violating NDAs and/or sale dates. But at least for us consumers, we now have a better idea as to how the 11700K will perform.

Aaron Klotz
Aaron Klotz
Freelance News Writer

Aaron Klotz is a freelance writer for Tom’s Hardware US, covering news topics related to computer hardware such as CPUs, and graphics cards.

See more CPUs News
11 Comments Comment from the forums
  • mac_angel
    typo in your article. You put 11700k for the 10700k in both single and multi. So it looks like there are two benchmarks for 11700k in each, and no 10700k.
    Reply
  • islandwalker
    mac_angel said:
    typo in your article. You put 11700k for the 10700k in both single and multi. So it looks like there are two benchmarks for 11700k in each, and no 10700k.

    Thanks. This should be fixed momentarily.
    Reply
  • spongiemaster
    Admin said:
    German Retailer intentionally sells its full stock of 120 Core i7-11700K Rocket Lake CPUs well ahead of launch. Buyers leak benchmark results online.

    German Retailer Sells 120 Core i7-11700Ks Before Launch, Benchmarks Arrive : Read more
    Assuming the leaked specs are correct and the 11700k has a 5Ghz single core boost and the 11900k has a 5.3Ghz single core boost, the extra 300Mhz will basically put the 11900k in a dead heat with the 5000 series in per core performance. Extra 200Mhz all core boost would tie it with the 10900k in multithreaded. How appealing these are will come down to price and all core overclocking headroom.
    Reply
  • TMTOWTSAC
    Did the mislabeling also mix up the ST calculations?

    8% slower ST (600) than the 5950x (633) but if I divide 600 by 633 I get .947

    15% faster ST (600) than the 10700K (506) but if I divide 600 by 506 I get 1.185
    Reply
  • CalienW2
    islandwalker said:
    Thanks. This should be fixed momentarily.


    Ryzen 9 5950X = 633
    Ryzen 7 5800X = 622
    Core i7-11700K = 600
    Core i9-10900K = 530
    Core i7-11700K = 506 <- 10700k?
    Reply
  • TechyInAZ
    Sorry about that guys! We'll get the typo sorted out ASAP.
    Reply
  • Redneck5439
    I could be wrong here, but it seems like Intel is trying to get as many processors out there as they can before independent reviews. I find it hard to believe that Mindfactory in this world economy would risk the ire of Intel, risk their supply chain and sell these processors two weeks before their official launch without Intel's permission. I was already trying to wrap my mind around Intel officially releasing Rocket Lake two weeks ahead of their review embargo lifting... Intel for awhile now has been trying to sell the idea that benchmarks are not important anymore (but only once they fell behind AMD's Ryzen in benchmarks) and now they seem to be trying to convince people that reviews aren't important anymore either. Just buy it because we say its great and you should blindly trust us seems to be their new marketing...

    From what I have seen of the benchmarks of these retail processors prematurely sold they are not overly impressive. Yes they have improved in single core compared to their own offerings, but they won't have a skew above 8 cores and they are still behind AMD's Zen IPC. Most single core scores I've seen for the 5800X put the processor at 630 and above single core in Cinebench R20. I know pure stock my 5900X scores 636 single core CB R20, and with a few slight tweaks I have it boosting to 4.99Ghz and 5.024Ghz on its two best cores of each CCD and it improved the score to 650. It is still early, and I could be wrong but with this being a new core design I don't think Intel is going to be getting much more overclocking potential than ~ 5Ghz for the 11700K and I think they already pushed the 11900K as far as it can go at 5.3Ghz single core. If that is the case then the 5800X will maintain a lead over the 11700K and the 11900K will be very close to the single core performance of the 5900X.

    From what I have seen the 11900K is going to be retailing for $600, which is $50 more than the MSRP of the 5900X. That is going to be hard to justify if its single core performance is nearly the same but it gets absolutely destroyed in multi-core oriented tasks. Yes, the 5900X isn't easy to find right now and scalpers are asking a huge mark up for them, but they can still be found at MSRP. It just seems to me that Intel is taking advantage of supply issues and once again price gouging.
    Reply
  • cryoburner
    Redneck5439 said:
    I could be wrong here, but it seems like Intel is trying to get as many processors out there as they can before independent reviews. I find it hard to believe that Mindfactory in this world economy would risk the ire of Intel, risk their supply chain and sell these processors two weeks before their official launch without Intel's permission. I was already trying to wrap my mind around Intel officially releasing Rocket Lake two weeks ahead of their review embargo lifting... Intel for awhile now has been trying to sell the idea that benchmarks are not important anymore (but only once they fell behind AMD's Ryzen in benchmarks) and now they seem to be trying to convince people that reviews aren't important anymore either. Just buy it because we say its great and you should blindly trust us seems to be their new marketing...
    Supposedly the sales embargo is March 30th, so more like 4 weeks from now. Which left me wondering why they would even be in the hands of retailers already, let alone why a retailer felt it was fine to sell them a month in advance. Either there was some miscommunication and someone thought the sales embargo was for the end of February, or Intel knowingly gave them the go-ahead to start selling them well ahead of reviews.

    Why are the benchmarks here limited to Cinebench though? That might be alright if someone is interested in how Cinema4D runs on the processor, but not necessarily representative of how performance will be in the software the vast majority of people will be running. Cinebench isn't always an accurate indicator of IPC in other software, after all. While I suspect we will see improved performance over the 10700K in most software, it's possible that there may have been performance regressions at some tasks due to the different architecture.
    Reply
  • TerryLaze
    cryoburner said:
    Supposedly the sales embargo is March 30th, so more like 4 weeks from now. Which left me wondering why they would even be in the hands of retailers already,
    Right?! Why would a company provide stock all over the world way before release date when the releases of all the other companies these years went so well...

    Mindfactory is the company that always pops up on all the articles that state that AMD outsells intel by like 90% or such, it's the only company where this happens and AMD supply is terribly low lately so they might had to sell them without intel's go ahead just to make some of the revenue they normally have.
    Reply
  • Redneck5439
    cryoburner said:
    Supposedly the sales embargo is March 30th, so more like 4 weeks from now. Which left me wondering why they would even be in the hands of retailers already, let alone why a retailer felt it was fine to sell them a month in advance. Either there was some miscommunication and someone thought the sales embargo was for the end of February, or Intel knowingly gave them the go-ahead to start selling them well ahead of reviews.

    Why are the benchmarks here limited to Cinebench though? That might be alright if someone is interested in how Cinema4D runs on the processor, but not necessarily representative of how performance will be in the software the vast majority of people will be running. Cinebench isn't always an accurate indicator of IPC in other software, after all. While I suspect we will see improved performance over the 10700K in most software, it's possible that there may have been performance regressions at some tasks due to the different architecture.

    The sales embargo lifts of March 30th, but from what I've been hearing Intel is planning a release date near the 15th for Rocket Lake (unless they changed direction on that). A March 15th (or close to 15th) launch date would make sense for stores getting inventory in now. It would also mean that Intel fully plans on releasing Rocket Lake two weeks before the review embargo lifts. It makes no sense to release a processor that no one is able to review for two weeks, unless they would rather have potential customers depend on their cherry picked numbers and don't want unbiased reviews.

    Benchmarks being mostly limited to Cinebench could simply be the result of no real independent reviews being done yet. Cinebench is widely available (for free) and easy to compare across platforms. Saying these "leaks" are from actual customers (non-professionals) of actual retail processors it may be as simple as that was the benchmark that was available to them. Although Cinebench is far from the "be all end all" of benchmarks I do actually find it beneficial. I tend to trust it a lot more than CPU-Z or Geekbench. I have seen incredibly unstable systems pass a CPU-Z validation / benchmark even though it is running such a massively unstable overclock that it will crash if tying to watch a youtube video. Geekbench is also a relatively quick run benchmark that unstable systems running high overclocks can pass. Cinebench seems to put more strain on a system and most of the time unstable overclocks are not going to be able to pass multi-core CB R20, and realistically a system has to be near rock solid (stability wise) to pass a CB R23 run.
    Reply