Intel Rocket Lake-S CPU Rumors: Core-i9 chips can turbo all the way up to 5.4-5.5 GHz | Engineering Samples already reaching 5.3GHz

[BiG StroOnZ]

According to Twitter user @MebiuW, who has posted some accurate leaks in the past, the Core i9 variant of Rocket Lake-S will hit a turbo clock of 5.4GHz to 5.5GHz, while at least one of the Core i7 SKUs will turbo to 5GHz to 5.2GHz. The leaker claims Rocket Lake-S will handle an all-core overclock to 5GHz without any trouble.

 

Quote

Rumor has it Intel's upcoming Rocket Lake-S processors will bring some fast turbo clocks to the table. Raw clockspeed gains could be a missing piece of the puzzle, in Intel's attempt to regain the performance crown. The upcoming chips will be labeled as 11th generation Core processors.

 

So brace yourself for faster boost and overclocking speed capabilities, if the information on Rocket Lake-S proves to be accurate. In a separate leak on Facebook, another leaker who has been reliable in the past, ITCooker, says that even though Rocket Lake-S is another 14nm series, it is a major architectural upgrade with gaming performance that is "a lot stronger" than Comet Lake-S (based on Facebook's translation of the post). It's also said that AVX512 will be added to the mix.

 

It's hearsay, and not much more. But Twitter user MebiuW claims to have information about the upcoming i7 and i9 CPUs. Both the Core i7 and i9 are said to have 8 cores and 16 threads, combined with 16MB of L3 cache. The i9 has to distinguish itself by incorporating higher clock speeds, for a higher price. With a boost clock around 5.5 GHz, the Core i9 would be positioned against the Ryzen 7 5800X at the same price of $449. The i7 has to do with a maximum boost clock of 5 to 5.2 GHz, and should compete against a future Ryzen 7 5700X. According to the Twitterer, the Intel CPU will change hands for less than $400. The Core i5 Rocket Lake SKUs will retail around $250-$300 US or slightly lower than the AMD Ryzen 5 5600X. However, AMD also has the Ryzen 5 5600 being prepped for a 2021 launch so that might be the real opponent for the Core i5-11600K.

 

Source 1: https://www.guru3d.com/news-story/rocket-lake-s-could-be-substantially-faster-for-games-than-the-current-comet-lake-s.html

Source 2: https://hothardware.com/news/intel-rocket-lake-s-11th-gen-cpus-5ghz-all-core-turbo

 

What's interesting the most about this rumor / leak, are the claims that the ES (Engineering Samples) CPUs are reaching 5.3GHz already. While 5.4-5.5GHz seems pretty outlandish / far-fetched for 14nm, typically ES chips aren't very close to retail clock speeds / clock rates. Usually the clockspeeds introduced with ES chips can be way off the retail mark (typically the retail sample's clocks are much higher). Regardless, for context, here's the CL-S 10th gen clock speed breakdown; Core i9-10900K: 3.7GHz base, 5.1GHz single-core, 5.2GHz max turbo 3.0, 4.8GHz all-core turbo. Core i9-10900: 2.8GHz base, 5GHz single-core, 5.1GHz max turbo 3.0, 4.5GHz all-core turbo. Core i7-10700K 3.8GHz base, 5GHz single-score, 5.1GHZ max turbo 3.0, 4.7GHz all-core turbo. Lastly, please note that Intel's Thermal Velocity Boost feature permits the i9-10900K and i9-10900 to hit single-core boosts of 5.3GHz, and all-core boost of 4.9GHz, when conditions are ideal. Therefore, there's a possibility that these 5.4-5.5GHz rumor numbers are in actuality regarding Thermal Velocity Boost specifications for the 11th Gen Rocket Lake chips. This makes the claim / rumor a bit more realistic. Personally looking forward to seeing the outcome of this rumor.

[manikyath]

accurate depiction of how they got 5.5GHz out of 14nm, and the thermals to expect.

also explains the name:

 

also, something something 2kw chiller to acutally match the claims?

[Bombastinator]

High clocks doesn’t always mean a fast chip.  It might though.  My understanding is this will compete against zen3 and perhaps zen4?  Not sure what comes after zen3

[Fasauceome]

47 minutes ago, BiG StroOnZ said:

Core i9-10900: 2.8GHz base

low TDP means they're gonna be trying to bill these CPUs as no major thermal uptick over 10th gen

 

47 minutes ago, BiG StroOnZ said:

there's a possibility that these 5.4-5.5GHz rumor numbers are in actuality regarding Thermal Velocity Boost specifications

and this would mean almost no real performance gain, sad. the 10th gen CPUs are on the redline already, and moving the redline 200MHz or so up from there is not terribly impressive.

 

Quote

Rocket Lake i7 has the same 8C16T 16M L3 as i9

am I reading this right? another 8 core i9? that would be extraordinarily confusing.

 

Not sure how Intel can pull off a 10 core i9 to compete with the 5800X though, price wise. unless all of their yield problems are solved, and their IPC increase is actually noticable. otherwise, we're looking at a mere 5% or so increase in performance.

[williamcll]

Isn't the surface area of the Rocket lake-S going to be bigger? Might help with the cooling.

[tim0901]

1 hour ago, Fasauceome said:

am I reading this right? another 8 core i9? that would be extraordinarily confusing.

This rumour has been bouncing around for a while - that the top 11th gen chip will only have 8 cores. Probably has to do with the backported Willow Cove architecture they're using for Rocket Lake - those cores are likely significantly bigger (transistor count wise) than their Skylake++++++ cousins. That new architecture will hopefully result in far higher IPC gains than we've seen from them recently, but of course the 14nm-based node won't help that at all.

 

My question - if this is all true - is will there be any point in buying the i9 if it has the same thread count? or will it be the prime candidate for the 2021 GN Disappointment PC? Given the situation they're in at this point, I wouldn't blame Intel for pulling a Radeon and giving up on even trying to go for the performance crown, instead focusing on price/performance with a <$400 8C/16T i7. Maybe claw back some marketshare (and reputation) before coming back swinging when they finally get their manufacturing problems sorted. If and when that ever happens...

[BiG StroOnZ]

3 hours ago, Fasauceome said:

and this would mean almost no real performance gain, sad. the 10th gen CPUs are on the redline already, and moving the redline 200MHz or so up from there is not terribly impressive.

 

am I reading this right? another 8 core i9? that would be extraordinarily confusing.

 

How do you figure "no real" performance gain? I recently tapered my CPU clocks (Per-Core overclocking), and I definitely notice performance improvements across the board, even if the fastest core doesn't represent the all-core overclock (it still means I have cores that are running faster than before).

 

Also, I think it's beneficial (the faster 8-core / 16-thread i9 alternative) because some people prefer not to buy extra unneeded cores, in order to attain the highest clock speeds. It's a relatively understandable compromise. 

[Bombastinator]

29 minutes ago, BiG StroOnZ said:

 

How do you figure "no real" performance gain? I recently tapered my CPU clocks (Per-Core overclocking), and I definitely notice performance improvements across the board, even if the fastest core doesn't represent the all-core overclock (it still means I have cores that are running faster than before).

 

Also, I think it's beneficial (the faster 8-core / 16-thread i9 alternative) because some people prefer not to buy extra unneeded cores, in order to attain the highest clock speeds. It's a relatively understandable compromise. 

We don’t know what they’re going to be like so everyone is guessing.  If you run a program that doesn’t utilize multiple threads well, a many core cpu is of course going to run more slowly.  Conversely a program that can utilize many threads will run more quickly on a cpu that can utilize them.  I think this is more of a difference in the meaning of performance than a disagreement in concept.

[justpoet]

I have doubts about the claims.  But, if they can make it happen, or be competitive with it, I'm all for it.

Bring on the CPU wars and the prices that will follow!

[Fasauceome]

51 minutes ago, BiG StroOnZ said:

How do you figure "no real" performance gain? I recently tapered my CPU clocks (Per-Core overclocking), and I definitely notice performance improvements across the board, even if the fastest core doesn't represent the all-core overclock (it still means I have cores that are running faster than before).

 

not that I think gains in single core speed aren't gains, what I mean is that it's a small increase to the single core speed over current gen (taking TVB into account.) Stock speeds are "a little bit above 5GHz" again.

 

51 minutes ago, BiG StroOnZ said:

Also, I think it's beneficial (the faster 8-core / 16-thread i9 alternative) because some people prefer not to buy extra unneeded cores, in order to attain the highest clock speeds. It's a relatively understandable compromise. 

they currently have the 10900K and the 10850K, so I think there's a market for same core/thread count and different clocks. However, both should be an i7, not one i7 and one i9. it will confuse consumers easily, and at least make some feel cheated.

[schwellmo92]

If true, Intel will be taking back the gaming and single-threaded crown quite convincingly. 

[schwellmo92]

5 hours ago, Fasauceome said:

and this would mean almost no real performance gain, sad. the 10th gen CPUs are on the redline already, and moving the redline 200MHz or so up from there is not terribly impressive.

With Willow Cove cores that are a 15-20% IPC increase it is actually substantial.

[Arika]

forever and always, wait for independent reviews, if they overheat, it will not be a secret.

 

obligatory: don't pre-order.

[jagdtigger]

12 minutes ago, schwellmo92 said:

If true, Intel will be taking back the gaming and single-threaded crown quite convincingly. 

Wont mean anything if their prices will be through the roof...... (and it will be because these kind of clocks will require a very strict binning IMO)

[Bombastinator]

13 minutes ago, schwellmo92 said:

If true, Intel will be taking back the gaming and single-threaded crown quite convincingly. 

Maybe. Maybe not.  Iirc there were some very high clock AMD chips at one point that still weren’t particularly fast when it came to actual use.  We’ll see what’s actually up when they are tested

[schwellmo92]

1 hour ago, Bombastinator said:

Maybe. Maybe not.  Iirc there were some very high clock AMD chips at one point that still weren’t particularly fast when it came to actual use.  We’ll see what’s actually up when they are tested

FX-9590, that was because it was a dog shit architecture. This will be using WIllow Cove cores which have 15-20% more IPC than Skylake cores.

[AldiPrayogi]

It's pretty exciting to see Intel pushing hard to innovate (bigger package, ipc improvements, etc). Here's hoping that It'll be competitive against AMD.

 

Also 5.5 seems like it'll need some hefty cooling (not to mention the power consumption).

[RejZoR]

1 hour ago, schwellmo92 said:

FX-9590, that was because it was a dog shit architecture. This will be using WIllow Cove cores which have 15-20% more IPC than Skylake cores.

Zen1 was on par with Skylake on IPC front. Zen 2 brought such IPC uplift and Zen 3 another up to 19%. Intel's only saving grace is their stupid high clocks, really. They just feel like they are in absolute panic mode and every meeting ends with "lets just raise the clock some more". Where AMD just keeps smacking 20% IPC boosts with every major Zen iteration. I wonder what will Zen 4 bring. If they can pull 5GHz and also another big IPC uplift, they'll be unstoppable.

[ne0tic]

I feel like Rocket Lake-S will destroy ZEN 3 in gaming. But I don't really think that's enough. ZEN 3 will have so many more cores that Rocket Lake-S will just be stupid for anything other than gaming. I'm getting a Ryzen 5900X since it is good enough for gaming and will probably be more future-proof because of all the extra cores. 

[porina]

8 hours ago, Bombastinator said:

High clocks doesn’t always mean a fast chip.  It might though.

If we assume no change from Skylake we still have a tiny gain from clock. If we assume the rumoured core update is correct, you get best of both worlds.

 

7 hours ago, tim0901 said:

That new architecture will hopefully result in far higher IPC gains than we've seen from them recently, but of course the 14nm-based node won't help that at all.

14nm process is a double edge sword. It is well understood so they're likely to get good results from it, putting aside it is "still" 14nm.

 

7 hours ago, tim0901 said:

My question - if this is all true - is will there be any point in buying the i9 if it has the same thread count?

If someone already owns an Intel (or even AMD) 8 core CPU, probably not, but not everyone upgrades every generation. It will have to be considered amongst other offerings when people do decide to buy.

 

7 hours ago, tim0901 said:

Given the situation they're in at this point, I wouldn't blame Intel for pulling a Radeon and giving up on even trying to go for the performance crown, instead focusing on price/performance with a <$400 8C/16T i7. Maybe claw back some marketshare (and reputation) before coming back swinging when they finally get their manufacturing problems sorted. If and when that ever happens...

Two things here: one is the so called "halo" product. This is a product that shows your best. It doesn't have to be good value, it only has to exist at all. It does have significant marketing value. Secondly, I'd argue Intel still has the better branding. I know, in this enthusiast forum this may not be the view, but in the wider world, I think it is still the case, and they can get away with charging a premium. Actually, with Zen 3, it might even be AMD making the bigger move of pricing going up, rather than Intel's going down. This will be something to watch as more models are released.

 

4 hours ago, Fasauceome said:

they currently have the 10900K and the 10850K, so I think there's a market for same core/thread count and different clocks. However, both should be an i7, not one i7 and one i9. it will confuse consumers easily, and at least make some feel cheated.

I think people read too much into the marketing segments (the 3/5/7/9 thing). It is only to indicate relative "goodness" of a product, and shouldn't be taken to necessarily imply a certain configuration.

 

3 hours ago, schwellmo92 said:

With Willow Cove cores that are a 15-20% IPC increase it is actually substantial.

The cores were originally designed to go on their 10nm process, so changes may be required to make it work on 14nm. One recent rumour suggested it was going to be more Sunny Cove like in cache size, so wont have the full IPC uplift expected from Willow Cove. That's not to say there wont be an IPC uplift, just not as much as if it was the whole experience. Sunny Cove is still a good uplift from Skylake.

 

19 minutes ago, RejZoR said:

Zen1 was on par with Skylake on IPC front. Zen 2 brought such IPC uplift and Zen 3 another up to 19%.

I'd argue Zen 1 IPC was still behind Skylake IPC if you consider mixed workloads, even if Zen 1 was faster in some multi-thread workloads *cough*cinebench*cough*. Zen 2 was when they more clearly passed Skylake overall and removed most of the edge cases. We're still waiting on an architectural description of the changes in Zen 3, which apparently will be released the same time they go on sale.

 

Below is my previous testing of Zen, Zen+, Zen 2 relative to (essentially) Skylake. I don't have any plans to repeat this for Zen 3.

 

 

19 minutes ago, RejZoR said:

Intel's only saving grace is their stupid high clocks, really. They just feel like they are in absolute panic mode and every meeting ends with "lets just raise the clock some more". Where AMD just keeps smacking 20% IPC boosts with every major Zen iteration. I wonder what will Zen 4 bring. If they can pull 5GHz and also another big IPC uplift, they'll be unstoppable.

There is no question that Intel's inability to execute 10nm process in a timely manner has cost them severely in the last 3 years or so, but it is not for the lack of trying. The products they offer are what they can offer, not what they want to offer. They have the designs, but have been unable to make them. If, and it is a big if, they get their process sorted out, they could come back rapidly.

 

AMD have also steered them towards the clock race since AMD started that also. Not absolute clock, but relative to the limits of the design. Ryzen CPUs since Zen+ boost pretty close to their limits already, much to the cries of overclockers. Intel historically held more headroom which they are eating into to compete.

 

And finally, as the old investment advice goes, past performance should not be taken as an indicator for future performance. The world is unpredictable. It will be great if AMD continue on their trajectory, but do not assume it will always be the case.

[LAwLz]

If it's true, and power consumption isn't extremely high then this will be fantastic. Rumors has it that Rocket Lake will be based on Willow Cove which offers roughly 18% higher IPC over Skylake.

Couple that with this ~10% frequency increase and we should see a very healthy performance increase.

 

But as always with leaks and rumors, I'll believe them when I see it in independent third party reviews of real products. I strongly recommend everyone else do that too.

[leadeater]

49 minutes ago, porina said:

If someone already owns an Intel (or even AMD) 8 core CPU, probably not, but not everyone upgrades every generation. It will have to be considered amongst other offerings when people do decide to buy.

Still running a 4930K, almost any CPU is an upgrade for me now lol

[RejZoR]

13 minutes ago, leadeater said:

Still running a 4930K, almost any CPU is an upgrade for me now lol

If overclocked, not quite. I only have 1 gen newer CPU than you and sure, it's on HEDT side, but Zen and Zen+ didn't particularly excite me. Zen 2 was okay-ish, but spending 1000€ for whole platform replacement is a stretch. Only Zen 3 became interesting for me and it's likely I'll go with this. I'm just not sure if I want to go with mature DDR4 platform or wait for Zen 4 and potential DDR5 platform. It would be the "bleeding edge", but also a very new tech full of problems and more expensive. I'm guessing buying most mature last "year" tech is the way to go as every new DDR has always proven to be barely any better than last gen's peak and only gets better with time.

[leadeater]

5 minutes ago, RejZoR said:

If overclocked, not quite. I only have 1 gen newer CPU than you and sure, it's on HEDT side, but Zen and Zen+ didn't particularly excite me. Zen 2 was okay-ish, but spending 1000€ for whole platform replacement is a stretch. Only Zen 3 became interesting for me and it's likely I'll go with this. I'm just not sure if I want to go with mature DDR4 platform or wait for Zen 4 and potential DDR5 platform. It would be the "bleeding edge", but also a very new tech full of problems and more expensive. I'm guessing buying most mature last "year" tech is the way to go as every new DDR has always proven to be barely any better than last gen's peak and only gets better with time.

4930K is pretty damn slow now, it's not a problem but like I mentioned any current CPU is significantly faster but right now it does not matter as I have matched GPU power to the CPU aka my system sucks ass hahaha. But yea what I have would be at the bottom of any CPU or GPU review chart today.

 

Still not going to buy anything until Ryzen 5000 reviews come out but I get the feeling I'll end up going GPU only and waiting longer, I'm a single player graphics fidelity person who also plays a lot of RTS, TBS and City Builders as well so I'm fine with just maxing all the settings and living with "lower" FPS, even if it's CPU limited.

[Fatih19]

@RejZoR@leadeaterThat's crazy skipping a whole generation of memory lmao.