Introduction

It doesn’t seem that long ago that Intel launched its 10th Generation Core Desktop Processor Comet Lake-S.  Indeed, it has been less than a year since the launch of Comet Lake-S at the end of May in 2020.  You can check out our reviews of the Intel Core i5-10600K and Intel Core i9-10900K CPUs as a refresher.  With that launch, Intel also released the Z490 chipset.

Here toward the end of March in 2021 Intel is releasing a new generation of CPUs and chipsets.  Today’s article is simply an overview of what to expect, this is not a hardware review today, that is to come later.  Intel will be launching its 11th Generation Core Desktop Processors known as Rocket Lake-S.  In addition, there will be a new 500 series of Chipsets, most notably the Z590.  In this article today we will share the entire press deck with you, and information we have learned, and give our opinions.  A full review is expected upon launch.

Rocket Lake-S

Rocket Lake-S is Intel’s 11th Gen Intel Core Desktop Processors on 14nm.  However, this is a new core architecture and a whole new processor core codenamed Cypress Cove.  This is very different from the previous generation Comet Lake-S. 

For 11th Gen Rocket Lake-S Intel has taken the 10nm Sunny Cove core IP from Ice Lake notebook CPUs and re-engineered it backwards to 14nm.  Why 14nm? With the mature process, Intel can target very high frequencies that are needed for desktop users.  Due to the integration of new integrated graphics, and other goodies, Intel named these “new” cores Cypress Cove in Rocket Lake-S. 

Intel states that it wanted to target the IPC (Instructions Per Clock) increase, and also maintain a high frequency, and sticking with the mature 14nm process allowed this to happen.  Therefore, Intel took the improved IPC of the 10nm Sunny Cove IP and put it on a process where they could get higher frequencies that are needed for desktop users.  Therefore, the new Rocket Lake-S CPUs have a 19% IPC improvement over 10th Gen CPUs.  That’s a gen-to-gen IPC improvement comparison by Intel. 

The integrated Intel UDH graphics now features the new Intel Xe graphics architecture from Tiger Lake which brings a 50% improvement to better graphics performance.  Intel has also integrated new Intel Deep Learning Boost AI into the CPU.  The architecture now supports DDR4-3200 as the standard RAM speed.  In fact, Rocket Lake-S is using a new integrated memory controller. 

Due to the improved Intel Xe graphics, it can now support 10bit AVI, 12bit HEVC decompression and E2E compression.  The architecture supports up to 20 CPU PCIe 4.0 lanes now, and HDMI 2.0 and HBR3.  There are also new overclocking features, and Resizable BAR is also supported across the board.

What about the Cores?

The downside for Rocket Lake S Processors is that there will be no 10 core CPU this generation like the i9-10900k.  Instead, the highest core/thread count will be the i9-11900K with 8 cores/16 threads.  The reason for the reduction is because of the ported 10nm IP over to a 14nm process, which is bigger.  Because the IP is larger on 14nm it would mean higher power requirements, more die area, and cost, and overall lower frequencies across the board to increase core count.  Intel wanted to maintain the improved Intel Xe integrated graphics component from Tiger Lake for its corporate users, so that was another component that had to be integrated.    

The 14nm process offers them the frequencies they need for desktop CPUs, but to fit in the budget that Intel wanted the process just doesn’t allow for more scale-up in core count.  Instead, Intel wanted to shoot for those high frequencies and potential single-core IPC performance.  Therefore, it’s a mix of a 10nm IP ported over to 14nm for higher frequencies.  The loss though, as we can see, is that it tops out at 8 cores and 16 threads. 

The SKUs

Similar to what we’ve seen in the past, there will be i9, i7, and i5 SKUs.  All of the i9 SKUs and i7 SKUs will have 8 cores and 16 threads.  The i5 SKUs will have 6 cores and 12 threads.  There will be K SKUs that are unlocked, as well as KF and T SKUs. 

At the top-end is the i9-11900K which will have an Intel Turbo Boost 2.0 clock speed frequency of Up to 5.1GHz and a Turbo Boost Max 3.0 frequency of Up to 5.2GHz and an Intel Thermal Velocity Boost Single-Core Frequency of 5.3GHz or an all-core Turbo Frequency of 4.8GHz.  Yes, the frequency business here gets very confusing, and we won’t list everyone, but you can see in the chart how they are configured.  The TDP on the top-end i9 and i7 CPUs is 125W.  The i9’s and i7’s have up to 44 PCIe lanes and DDR4-3200 and 16MB of Intel Smart Cache.  There is 2MB of cache per core. 

The top-end i5 SKU will be the i5-11600K.  It has an Intel Turbo Boost 2.0 frequency up to 4.9GHz and an Intel all-core Turbo Frequency of up to 4.6GHz.  The i5’s do not support Intel Turbo Boost Max or Intel Thermal Velocity.  It will have 12M of Smart Cache and a TDP of 125W.  All the i5 CPUs have up to 44 PCIe lanes and run at DDR4-3200.

Below that, the i3 SKUs are simply going to be refreshed 10th Gen Intel Core Desktop Processors Comet Lake-S.  Therefore, to get on the new Rocket Lake gravy train, you will need to get 11th Gen i5, i7, or i9.

Brent Justice

Brent Justice has been reviewing computer components for 20+ years, educated in the art and method of the computer hardware review he brings experience, knowledge, and hands-on testing with a gamer oriented...

Join the Conversation

6 Comments

  1. It will be interesting to see the performance reviews vs the AMD chips.

    The prices seem to be pretty decent. If it doesn’t end up a paper release.

  2. It will be interesting to see the performance reviews vs the AMD chips.

    Heat is the biggest one on my mind. If larger dies are in play, these should be easier to cool the same way AMD CPUs are easier to cool with their two- and three-die configurations that spread the heat out under the heatspreader itself.

    The prices seem to be pretty decent. If it doesn’t end up a paper release.

    I’ll be focusing on overall platform capability myself, but for strictly gaming purposes or more likely for ‘gaming first’ builds, Intel may have a ringer!

  3. This too has piqued my interests. I love it when there’s true competition between Intel and AMD. We all win when that happens!
  4. 11th Gen I guess is AMD’s FX 9590 — Guess the question becomes will Intel coming roaring back when move to 10nm or 7nm happens like AMD was able to with Ryzen — or have we reached a fork in the road where Intel short of break-through will remain in the passengers seat.

    I’m bit curious though to see how Intel’s 11th Gen does with top end 6900 XT in games at 2560×1440 – single and multi-thread games.

  5. 11th Gen I guess is AMD’s FX 9590

    This is a pretty good analogy in terms of Intel having to ‘crank it’ to produce a product that’s even competitive on paper, as was the issue with the FX / ‘dozers, though there is a bit of a difference in perspective; however far ‘behind’ Intel is, it’s not overall in the way that FX was non-competitive right out of the gate.

    Guess the question becomes will Intel coming roaring back when move to 10nm or 7nm happens like AMD was able to with Ryzen

    All indications point to them having some pretty stellar architectures that… they are unable to manufacture in volume. While the question must remain open, that if Intel cannot make the transition to full-scale production at smaller nodes the stumbling giant might just fall, the chances of that situation coming to pass seem relatively remote. If other companies have been able to ‘crack that nut’, surely Intel can.

    Right?

    or have we reached a fork in the road where Intel short of break-through will remain in the passengers seat.

    This relates to the former point about manufacturability, but it’s worth pointing out that Intel remains in the ‘driver’s’ seat out of sheer production volume at 14nm today. Most estimates I’ve come across put Intel at ten times the volume of CPUs supplied of what AMD is able to supply, and that’s when AMD is being viewed optimistically.

    Yes, there are many reasons one might prefer a Zen-based CPU right now over anything Intel has available, but the truth of the matter is that the Intel CPUs are simply ‘more available’. That’s not as much of an issue for individual enthusiasts, but it is most certainly an issue for large OEMs as well as for enterprises.

    And that doesn’t touch on the platform, driver, and software support differences and even quirks and nuances that might push a particular purchasing decision one way or the other!

    I’m bit curious though to see how Intel’s 11th Gen does with top end 6900 XT in games at 2560×1440 – single and multi-thread games.

    I’m definitely looking forward to in-depth frametime analysis of emerging platforms myself, but to be honest, I don’t see the < 4MP of 2560×1440 as being that big of a challenge for any top-end CPU these days. That’s what I’m running now myself, and with a 5.0GHz 9900K, I can’t say that I’m left wanting for CPU performance when it comes to gaming specifically.

    I’m much, much more concerned about the general market price and availability of GPUs and the price and lack of quality and quality control of ‘gaming’ monitors!

Leave a comment