Monday, July 7th 2025
Intel Plans "Arrow Lake Refresh" for H2 2025 with Higher Clocks and Upgraded NPU
According to ZDNet Korea's sources familiar with the matter, Intel will release the "Arrow Lake Refresh" CPU family in the second half of this year. While no exact timeline is given, we can expect to hear more about these CPUs in the coming weeks, and for the rumor mill to intensify. The new refresh generation will introduce slight frequency adjustments, including an increased base clock and maximum boost clock, while also incorporating a faster NPU. A base clock increase, paired with a boost to maximum core clock, will give Arrow Lake Refresh a significant advantage in gaming performance. Keeping the same LGA1851 socket will extend the support for the current platform for one more generation too, until "Nova Lake-S" arrives.
Given that Intel attempted an IPO for a limited set of gamers, we could see a more widespread adoption of a similar approach. As a reminder, on the CPU side, IPO fine-tunes P-core and E-core frequencies, ring-bus speeds, the UPI interconnect, D2D links between tiles, and both PL1 and PL2 power limits. For RAM, IPO raises transfer rates and tightens timings, pushing modules beyond their factory XMP or EXPO profiles. If Arrow Lake Refresh ships with an IPO-equivalent out of the box, gamers will enjoy a significant performance increase, ready to offset some of the early disappointment with the base Arrow Lake. Additionally, Intel will place a new NPU on its IO tile. Using the NPU3 on base Arrow Lake, Intel was unable to meet the basic 40 TOPS requirement for Copilot+ AI PCs. And with NPU4, that TOPS requirement is met, allowing Intel to brand these new processors as AI PC ready.ZDNet Korea also noted that Intel is preparing Nova Lake-S for the second half of next year, meaning that we are still months away from hearing more details about its platform. For now, we can only note that the Nova Lake-S, rumored to feature up to 52 cores, will be compatible with an LGA1954 socket and will include a mix of P, E, and LPE cores.
Source:
ZDNet Korea
Given that Intel attempted an IPO for a limited set of gamers, we could see a more widespread adoption of a similar approach. As a reminder, on the CPU side, IPO fine-tunes P-core and E-core frequencies, ring-bus speeds, the UPI interconnect, D2D links between tiles, and both PL1 and PL2 power limits. For RAM, IPO raises transfer rates and tightens timings, pushing modules beyond their factory XMP or EXPO profiles. If Arrow Lake Refresh ships with an IPO-equivalent out of the box, gamers will enjoy a significant performance increase, ready to offset some of the early disappointment with the base Arrow Lake. Additionally, Intel will place a new NPU on its IO tile. Using the NPU3 on base Arrow Lake, Intel was unable to meet the basic 40 TOPS requirement for Copilot+ AI PCs. And with NPU4, that TOPS requirement is met, allowing Intel to brand these new processors as AI PC ready.ZDNet Korea also noted that Intel is preparing Nova Lake-S for the second half of next year, meaning that we are still months away from hearing more details about its platform. For now, we can only note that the Nova Lake-S, rumored to feature up to 52 cores, will be compatible with an LGA1954 socket and will include a mix of P, E, and LPE cores.

78 Comments on Intel Plans "Arrow Lake Refresh" for H2 2025 with Higher Clocks and Upgraded NPU
Ok to be fair, Arrow Lake is actually better than the terror many 'experts' and forums spread. If those optimisations manage to significantly improve latency(-ies), it could become highly recommended.
2) Dead platform is... dead platform.
No point in getting it, if you don't already own Arrow Lake (at least from my point of view).
PS. Also, Nova Lake sounds even more cursed than Arrow Lake (with three types of x86 cores inside, not counting NPU and other "AI features").
Can't Intel just figure out that all E-core CPUs for Ultra 3 (and lower) + all P-core CPUs for Ultra 5/7/9, are just fine ?
if you do enjoy tinkering, the Intel platform has a very vast potential of options to work on, on the AMD Platform, you're much an average commoner joe, everyone has the same settings and almost same latency..lol..(no pun intended)
I believe in what you said about not going through this if you don't own it, or if you don't feel like a tinkerer yourself, personally, I like the Intel more.
Will more likely be Raptor - Nova Laks-S for me.
Arrow Lake will profit hugely in gaming from a clock boost (and then also from high end DDR5). I saw what a max overclocked 265K can do, and that was coming very close to the top end CPUs in gaming and most mixed load situations in productivity (single core CB2024 around 150).
So, let's hope, Intel cranks it up a notch without forgetting about efficiency (because higher clocks are most certainly not great for efficiency).
But if we are talking about a minor update like the Raptor Lake Refresh,so only 100-200mhz on base clock and max boost, trading a little bit of power and heat to reach that, then the whole thing is bogus. Because it can be achieved with the current gen already.
I've not been super impressed with my 9800x3d on either the B650 or the Asrock B850.To be fair - this same thing was said about arrowlake... Never underestimate intel's and rumormill ability to overhype a launch.
I essentially downgraded my overall gaming experience, and my productivity in the name of +30% average FPS and +100% highs - but WAY more noticeable lows :/. Meh is exactly the result I had as well - I thought I was crazy and ended up swapping the board to rule out BIOS issues or PCIE bottlenecks - nope.
people advice getting the X870E Apex, though it won't help much as I have a turd chip, if in any case this chip of mine was golden, I would have spent more on that, otherwise on ARL, even if its a turd chip you're guaranteed on that 8k RAM speed on any A Tier board..
245K 6P 8E $330 cdn. Z890MB for $275
Add in some 7200 CL34 ram for $149
$754 CAD for CPU/MB/Ram
9800x3d is $629
7800x3d id $540
That's a pretty decent price for a solid foundation for a build
The 285k was fine when I had some hands on with it just expensive and gaming was solid if unspectacular on it.
Does MS CoPilot spec not allow GPU AI TOPS to count at all? Why do we even have dedicated NPUs and also NPU accelerators in the GPU still?
with cudimm reaching 10K MT/s what would be needed to start using part of RAM as GPU VRAM?
My radeon 780M igpu in my minipc has 16GB of vram, from my pool of 96GB 5600 ddr5
Too bad AI is such worse to answer simple question. How many mainboard changes for intel platform vs amd platform for a given time-frame from year to year.