Intel Arrow Lake processor models and frequencies revealed. And we already know when they will go on sale

Unfortunately, August is coming to an end, and so is the time when the new generation of Intel processors for the desktop, labeled Arrow Lake, will be released. It will bring new, significantly improved CPU architectures, advanced chiplet construction, as well as a new LGA 1851 platform. Last but not least, it will also replace the faulty Raptor Lake processors, so hopefully reliability will improve. Now we have learned from the leakers when these processors will be released, as well as detailed specifications.

This information came from two sources. The parameters of the processors were published both by a leaker with the nickname Jaykihn, who appeared on Twitter in June and who in a short time revealed a large amount of data about the upcoming Intel processors, apparently from insider sources, and in addition also by the website Benchlife, which may be the second independent source. The information comes from the QS parameters of the samples whose frequency and other properties correspond to the serial pieces.

It should be said that this is only a part of the processors, namely those with “steppin B0”. This indicates, as with Raptor Lake, a version with a maximum of 8 P-Core cores + 16 E-Core cores. The CPU chiplet with 8+16 cores is manufactured on TSMC’s 3nm process (it should be the basic N3B version). Later, Intel will start selling yet another version with a CPU chiplet with 6+8 cores, which will be manufactured in its own factories on a 2nm process (Intel 20A).

Release date

Benchlife also lists the date when the first Arrow Lake desktop processors will go on sale. It’s supposed to be in October, to be exact 10. 10. (and this should be the day of the actual launch, i.e. also the day of availability in stores). In this first wave, only the 125W models of the K series (respectively, probably also their KF versions with graphics) will go on sale, along with boards with Z890 chipsets. The K and KF models will again be unlocked for overclocking, the others will not.

Model

You can see an overview of the models that will be released in the October autumn term in the table. However, the leaks give a bit more detailed information on the processor boost. The most powerful model Core Ultra 9 285K with 8+16 cores, it will have a maximum E-Core core boost of 4.6 GHz. The maximum clock speed of P-Core large cores will vary depending on the number of cores loaded. With 3-8 P-Cores active, the maximum will be 5.4 GHz. With a maximum of two, the clock will be able to be 5.5 GHz (Turbo Boost 2.0), or on preferred cores 5.6 GHz (that is, when using Turbo Boost Max 3.0) and finally, the processor also has Thermal Velocity Boost, which will allow a clock of up to 5.7 GHz on one or two cores, but this will be conditional on the temperature not exceeding a certain given level.

According to this, it appears that Arrow Lake will have different degrees of turbo boost just like previous generations of Intel processors. So, in the end, there may not have been a reworking of clock management, which we expected so slightly from the Lion Cove architecture, and the combination of different types of boosts into one single mechanism with more flexible functioning, including both temperature dependence and preferred cores.

A slightly cheaper model Core Ultra 7 265K it will have 8+12 cores and it will no longer use Thermal Velocity Boost (which will therefore be the privilege of Core Ultra 9 models, as it was with Core i9), but only Turbo Boost Max 3.0, or the use of preferred cores. The E-Core will also max out at 4.6GHz, but the P-Core will see the maximum clocks drop by 200MHz. It will be possible to have a maximum of 5.2 GHz on 3-8 P-Cores. A maximum clock speed of 5.4 GHz will be possible on 1-2 cores, or 5.5 GHz on preferred cores.

Next comes the model Core Ultra 5 245Kwhich has 6+8 cores, and this time it doesn’t have any of the extended boosts. The E-Core clock is also 4.6 GHz (also for all cores), the P-Core clock is a maximum of 5.0 GHz for 3-6 active cores and a maximum of 5.2 GHz for 1-2 cores.

All three K models have full-fat graphics with 512 shaders (64 EU / 4 Xe Core). The first two models run at 2.0 GHz, but the Core Ultra 5 model has the frequency reduced to 1.90 GHz. There will also be a model Core Ultra 7 265Kwhich will be the same as the 265K but without integrated graphics, and then the model Core Ultra 5 245Kwhich in turn will be 245K with graphics off.

Arrow Lake will also bring an upgrade in memory support, the processors will officially support DDR5-6400.

Unfortunately, we have not yet confirmed the maximum turbo consumption for serial models. Preliminary information based on ES samples indicates that the 125W Core Ultra 245K and 245KF models will have PL2 159W, the Core Ultra 7 265K (and KF) model will have 250W. The most powerful Core Ultra 9 285K model will have a maximum turbo consumption of probably 250W at those boards that use the “Performance Profile”. But boards that would use the “Extreme Profile” can set it to a maximum consumption of up to 295W.

Regression in clocks of large cores

So it is confirmed that Arrow Lake will reach lower clocks than what Intel managed with its 7nm process. It will limit its performance a bit. The P-Core cores, which provide single-threaded and gaming performance, should have 14% higher IPC according to Intel, but since the clock will be reduced, the performance increase will be less. Fortunately, the performance of the E-Core cores is also a few hundred MHz higher, and their IPC is also expected to increase dramatically (more than the P-Core), so multi-threaded performance can improve a lot, even though the P-Core cores lost HT.

Other models: 65W and 35W series

We already have part of the models and specifications for the locked models and the energy-saving T models, but not all of them – there is no information about the models that will have the 2nm chiplet from Intel. There is one exception: the cheap 6+4 core Core Ultra 5 225 (and its derived model Core Ultra 5 225F). They will have both versions with a 2nm chiplet and versions with a 3nm B0 chiplet. So it will be exactly like the Core i5-13400(F) and Core i5-14400(F) models, the processor is obviously their successor.

Thanks to it, Intel will be able to use partially defective silicon, including pieces with a defect in the GPU part, because the 225 model will have a chopped GPU with only 2 Xe Core / 256 shaders, clocked at 1.80 GHz. Its P-Core clock will be 4.7 GHz for 3-6 cores and 4.9 GHz for 1-2 cores, with E-Core it will have a maximum clock of 4.4 GHz.

Furthermore, the Core Ultra 5 235 (similar to the i5-14500, has a GPU with 384 shaders) and Core Ultra 5 245 (similar to the i5-14600, GPU with 512 shaders) models should also be on offer, but these are not yet confirmed. These will have 6+8 cores and a 2nm CPU chiplet, but their other parameters are not yet clear. The maximum turbo consumption should probably be set to 121W for all Core Ultra 5 with 65W TDP, but this is not yet confirmed. At the bottom of the menu should be the Core Ultra 3 215 (or maybe 205 and there will probably also be an F variant without graphics) with 4+4 cores and a 256-shader GPU and a maximum consumption of 76 W. Again, there are no parameters for it yet, because it will use 2nm chiplet

However, we already know the parameters for the higher 65W Core Ultra 7 and Core Ultra 9 models, because they will have a 3nm B0 stepping chiplet. Higher model Core Ultra 9 285 it has 8+16 cores. For E-Core cores, the maximum clock speed of 4.6 GHz will still be active for all cores. But there will be complications again with P-Core. Basic maximum boost for all active P-Core cores is 4.6 GHz, 5.3 GHz is possible for 3-6 cores and 5.4 GHz (Turbo Boost 2.0) or 5.5 GHz is possible for 1-2 cores on preferred cores and using Thermal Velocity Boost up to 5.6 GHz. With the exception of the all-core boost, the clock speed of the P-Core is reduced by only 100 MHz compared to the 125W version.

It will be a cheaper alternative Core Ultra 7 265. It again has 8+12 cores. For E-Core, the maximum is again 4.6 GHz even for all 12 active ones. For P-Core, the clock speed for 3-8 cores is 5.1 GHz maximum, for 1-2 cores 5.2 GHz or 5.3 GHz on preferred cores (Turbo Boost Max 3.0).

The graphics core is unchanged compared to the 125W models – it has 512 shaders at 2.0 GHz. But there will be a version from the 265 model Core Ultra 7 265F without GPU. The maximum turbo consumption of all three models should be 182 W (this is preliminary information). As you would expect, compared to the 125W models, the base clocks of both P-Core and E-Core have been reduced quite a bit for these processors. But as the all-core boost of the 285 model shows, the reduction of PL2 from 250 W to 128 W significantly limits the ability to boost on all P-Core cores.

Part of 35W models

In addition, the parameters of two economical models that use B0 stepping have also been leaked. These processors will probably again be labeled as the T series and have a 35W TDP. But this will not mean much, because according to as yet unverified information, the maximum turbo consumption (which in practice is the value on which it really depends) should be 112 W.

The most powerful model Core Ultra 9 285T it has 8+16 cores. For E-Core cores, the maximum clock speed will be 4.6 GHz, but only up to 12 active cores, for 13 and more active E-Cores, it will be only 4.5 GHz. P-Core will no longer have Thermale Velocity Boost. Basic maximum boost for all active P-Core cores is 4.7 GHz, 5.1 GHz is possible for 3-6 cores and 5.3 GHz (Turbo Boost 2.0) or 5.4 GHz is possible for 1-2 cores on preferred cores. With the exception of the all-core boost, the clock speed of the P-Core is reduced by 300 MHz compared to the 125W version. However, the processor will probably be limited mainly by the 112W value of PL2 when boosting multiple cores, it will probably not be able to reach the maximum all-core boosts in many tasks.

Cheaper Core Ultra 7 265T it has 8+12 cores. With E-Core, the maximum is again 4.6 GHz for 1 to 8 cores, but only 4.5 GHz for 9 to 12 active cores. For P-Core, 7-8 cores max 4.6 GHz, 3-6 cores 5.0 GHz, 1-2 cores 5.2 GHz or 5.3 GHz on preferred cores (Turbo Boost Max 3.0) . The maximum single-thread performance will also be 300 MHz below the 125W version, but the multi-thread performance will probably be limited by consumption.

Much will indicate the base frequencies, which are only 1.4 GHz for these two models, respectively 1.5 GHz for the P-Core, and only 1.2 GHz for the E-Core (the base frequencies are so low to allow running at these clocks with all cores active at consumption not exceeding TDP, i.e. here 35 W).

The T series will probably have a full range of models including Core Ultra 5 and Core Ultra 3, but the other processors are likely to have 2nm silicon from Intel, so we don’t know about them yet.

35W and 65W models only in January

As already mentioned, the 65W and 35W models will not be released in the first wave, they will not hit the market until about January during CES 2020 or after. Boards with cheaper H810 and B860 chipsets will also come out with them.

So, apparently, we already know the most important thing. The 10/10 release date means that there are just over 50 days until Intel’s new generation of CPUs and new platform become a reality. Less than three years after the release of Alder Lake and LGA 1700 in 2021, Intel will again bring new architecture and features and will have the opportunity to respond to AMD’s Ryzen 9000.

Resources: BenchLife, Jaykihn