Sunday, October 5th 2025
Intel's "Panther Lake" Microarchitecture Deep Dive Set for October, Full Launch in 2026
The story of Intel's first 18A product, "Panther Lake," remains shrouded in mystery, at least for a few more days. According to earlier reports, Intel was expected to start shipping the first PTL-U and PTL-H SKUs by the end of 2025, with additional SKUs to follow in 2026. However, that situation could be subject to change, according to Golden Pig Upgrade. The rumored October 9 launch date is reserved only for the microarchitecture deep dive, and the launch and final specifications of specific models will occur at CES 2026. Being Intel's first big bet with the 18A node, Panther Lake is a product of years of Intel's manufacturing innovation, which needs to demonstrate that the investment in advanced manufacturing was worthwhile. Reportedly, Intel started limited shipments of its 18A nodes to U.S. customers in Q3, with these wafers already in production and initial output of its own CPUs expected in Q4.
Given that Panther Lake is the first 18A node product, we could expect that to be a large part of those shipped wafers, most likely going to OEMs for testing before integration. As a reminder, the low-power PTL-U models are designed for a 15 W TDP and are expected to come in 6-core and 8-core versions. Some SKUs will feature four high-performance P-cores paired with four LPE-cores, while others will have only two LPE-cores complementing four P-cores. Both families will utilize Xe3-based integrated graphics, with entry models featuring four GPU cores. The more powerful PTL-H line will scale up to around 16 CPU cores, comprising four P-Cores, eight E-cores, and four LPE-cores. Some H-series parts might include up to 12 GPU Xe3 cores for integrated graphics, but the final configurations will be revealed when Intel officially launches the Panther Lake product family.
Sources:
Golden Pig Upgrade on Weibo, via VideoCardz
Given that Panther Lake is the first 18A node product, we could expect that to be a large part of those shipped wafers, most likely going to OEMs for testing before integration. As a reminder, the low-power PTL-U models are designed for a 15 W TDP and are expected to come in 6-core and 8-core versions. Some SKUs will feature four high-performance P-cores paired with four LPE-cores, while others will have only two LPE-cores complementing four P-cores. Both families will utilize Xe3-based integrated graphics, with entry models featuring four GPU cores. The more powerful PTL-H line will scale up to around 16 CPU cores, comprising four P-Cores, eight E-cores, and four LPE-cores. Some H-series parts might include up to 12 GPU Xe3 cores for integrated graphics, but the final configurations will be revealed when Intel officially launches the Panther Lake product family.


30 Comments on Intel's "Panther Lake" Microarchitecture Deep Dive Set for October, Full Launch in 2026
Source: Intel® Architecture Instruction Set Extensions Programming Reference September 2025
And APX is likely to be a good performance jump, but most significantly in an area where CPUs scales very poorly to begin with; logic-heavy code. This is also an area where x86 already have an advantage over other ISAs, and this will only grow with APX. But just don't expect to see the full potential of this on launch day, while no code needs to be rewritten, applications obviously needs to be recompiled, and I don't know yet the maturity of these implementations.
Now I wonder what will AMD response and feature set of ZEN6.🧐
Hopefully by the release of DDR6 Intel’s APX and rebranded AVX10.2 will be a thing.
Zen 6 is too close, and rumored to be on AM5 for consumer variants as well so DDR5. Especially since AVX10.2 received a breaking change in its specification recently that eliminated the 256-bit only variant. All AVX10.2 CPUs, even E-cores, will be capable of 512-bit execution, backwards compatible with AVX-512.It's mobile, -U(ltraportable) and -H(igh performance) segments of it.
It is true that simply adding - say, AVX-512 - to a processor will not make it magically faster, however, if the code is optimized and specifically written to take advantage of these advanced SIMD instructions, you will see an exponential increase in performance. This has always been the case, since the MMX and 3DNow! days, and has historically been a pain point for certain architectures (such as the AMD K10 from the Phenom and Phenom II lines, which wasn't capable of doing SSSE3 or SSE4.1 and towards the end of their lifespans, were routinely running into compatibility issues).
In case you don't know what I'm talking about: en.wikipedia.org/wiki/IAPX
-- 6 Core should finally die. Windows 11 Pro is very bloated and slow on my ryzen 4650u with samsung nvme C: .I think my gentoo linux has only ~10 packages with AVX512. I have a total of only 1205 packages installed. We need a common standard. Intel should finally stick to the standard. So we get better software. Why do i care? I want fast software with fast hardware. I do not need fancy fast hardware which is never supported in software. E.g. xop instruction does not exists anymore. X86 or amd64 is not backward compatible!
time up. i'm not sure
packages.gentoo.org/useflags/cpu_flags_x86_avx512f
quick and dirty -- I'm not sure. this is only one subset of avx512. One subset the two instruction cycles avx512 capeable ryzen 7600x can do:
www.techpowerup.com/339224/intel-nova-lake-s-core-ultra-3-ultra-5-ultra-7-and-ultra-9-core-configurations-surface
It's a bigger question whether APX will have to wait for "Zen 7", but when it arrives it might even yield larger relative gains for AMD, and considering their good track record with AVX-512 and AVX2, I'd expect them to do a good implementation (even if they are a cycle behind).That's still a chip coming for entry level servers with no confirmation whether it will run on consumer grade chipsets. I don't get why people are so excited about it, as by the time it arrives it will be very irrelevant.Well, it's not like Linux distributions are going to ship such binaries anytime soon.
GCC 15 added support, but I haven't evaluated the quality of the support.