Intel Meteor Lake “Core Ultra” CPUs Specs, Performance, Price, & Availability – Everything We Know So Far

Intel Meteor Lake CPUs will mark a major shift in Chipzilla's product lineup, featuring a brand new core architecture, brand new design methodology, brand new features & even a brand new family nomenclature. The CPU lineup is expected to be introduced this fall and we are going to talk about what you can expect in terms of specifications, performance, and prices.

Intel Meteor Lake CPUs - Chipzilla Goes All Onboard The Chiplet Train For Clients

Intel first teased its Meteor Lake CPUs back in 2021 during the IDM 2.0 keynote where the chip was confirmed to have taped-in its compute tile based on the latest process technology. The chipmaker followed the momentum by gradually revealing more information and details of Meteor Lake CPUs at various events including one major session during Hot Chips 2022 where the next-gen roadmap was laid out, focusing on a disaggregated and a chiplet-heavy future for Intel's next-gen CPUs. So as of today, we are about a week away from Intel's Innovation 2023 event where the company is going to give us the full disclosure on Meteor Lake CPUs so let's begin right from the start.

What Is Meteor Lake?

Intel Meteor Lake CPUs are the follow-up to the company's Raptor Lake CPUs. Intel has dropped its Tick-Tock model but Meteor Lake in a sense is both a Tick and a Tock since it implies a new process node technology and also comes with brand new core architectures for virtually the whole chip. It's also Intel's first full-on Client CPU family that will be utilizing the chiplet architecture thanks to latest packaging technologies such as Foveros.

Following are some of the main features you can expect from the 1st Gen Core Ultra "Meteor Lake" family:

• Triple-Hybrid CPU Architecture (P/E/LP-E Cores)
• Brand New Redwood Cove (P-Cores)
• Brand New Crestmont (E-Cores)
• Up To 14 Cores (6+8) For H/P Series & Up To 12 Cores (4+8) For U Series CPUs
• Intel 4 Process Node For CPU, TSMC For tGPU
• Intel 'Xe-MTL' GPU With Up To 128 EUs
• Up To LPDDR5X-7467 & DDR5-5200 Support
• Up To 96 GB DDR5 & 64 GB LPDDR5X Capacities
• Intel VPU For AI Inferencing With Atom Cores
• x8 Gen 5 Lanes For Discrete GPU (Only H-Series)
• Triple x4 M.2 Gen 4 SSD Support
• Four Thunderbolt 4 Ports

From a naming point of view, Meteor Lake is more reminiscent of the Comet Lake and Rocket Lake lineups rather than Alder Lake and Raptor Lake. But these are just fancy names that Intel selected for its product families. All you should know is that Meteor Lake is a brand new CPU family that is bringing some major updates as we have detailed below.

Meteor Lake Is For Laptops, Not Desktops

If you have been following the developments of Meteor Lake, you must know that the family is only going to be offered on notebooks. This will mark the fourth major notebook-exclusive CPU family in Intel's portfolio. The company previously had Cannon Lake, Ice Lake, and Tiger Lake CPUs exclusive to the client notebook platform with intermediary solutions & refreshes serving the client desktop space.

Yes, it is indeed true that Intel had desktop Meteor Lake CPUs in the works at one point and we also got to see various SKUs leak out by they are now a thing of the past and have been replaced by the 14th Gen Raptor Lake Refresh Desktop CPU family that launches in the coming month. But one will wonder what led Intel to cancel a Meteor Lake desktop launch.

So far there have only been rumors with some pointing out yields on the Intel 4 process node being a major reason. The Intel Meteor Lake-S lineup was anticipated to utilize a variant of the mobile die but Intel saw the use of those chips on the notebook segment being more crucial. It's a trade-off of whether you want to cater to both segments or focus on one segment. Intel obviously chose the latter.

Intel Meteor Lake Family Nomenclature: Meet 1st Gen Core Ultra CPUs

Another big update for Intel Meteor Lake CPUs will be the family branding and nomenclature. The company reveals that it will no longer place generational messaging in front of the Intel Core brand in marketing or in silicon but would instead be called out in the processor number. That's a lot similar to how AMD is managing the new branding for its Ryzen 7000 mobile SKUs which feature several different generations of Zen CPUs that are then differentiated by their SKU numbering. The following changes will be introduced starting with Meteor Lake CPUs:

• Introducing Intel Core Ultra processor brand for the most advanced client processors.
• Simplifying the Intel Core processor brand for mainstream client offerings.
• Moving to Intel 3/5/7/9 processor tiering starting with upcoming next-generation processors.
• Evolving the Intel Evo Edition platform brand for Evo-verified designs.
• Introducing Intel vPro Enterprise and Intel vPro® Essentials device labels for relevant commercial systems.

2 of 9

The new naming scheme will be applied across all client products including desktops and mobility but the Meteor Lake CPU family will be the first to utilize it starting in 2H 2023. The most significant change that the new branding brings is dropping the "i" which has been part of Intel's naming scheme for over a decade.

So instead of using the traditional Core i3, Core i5, Core i7, & Core i9 branding, the new chips will shift to Core 3, Core 5, Core 7, and Core 9. Once again, these changes will be reflected in the brand strings, text, and on the badge of the product.

• Old Naming Scheme: Core i3, Core i5, Core i7, Core i9
• New Naming Scheme: Core 3, Core 5, Core 7, Core 9

2 of 9

So introducing its new Meteor Lake branding, Intel has announced two lines of products. The client mainstream SKUs feature the Core 3, Core 5, & Core 7 SKUs, and the client premium SKUs feature the Core Ultra 5, Core Ultra 7, & Core Ultra 9 SKUs. These processors will drop the "i" from their naming entirely and once again, the generation will be embedded within the numbering. I think it would have been even better if Intel had used an example SKU to further elaborate their new naming scheme like AMD did but that isn't the case.

“Our client roadmap demonstrates how Intel is prioritizing innovation and technology leadership with products like Meteor Lake, focused on power efficiency and AI at scale. To better align with our product strategies, we are introducing a branding structure that will help PC buyers better differentiate the best of our latest technology and our mainstream offerings.”

–Caitlin Anderson, Intel vice president and general manager of Client Computing Group Sales

To make things more clear, here's what is being replaced and with what:

• Core i9 -> Core 9 Ultra
• Core i7 -> Core 7 Ultra or Core 7
• Core i5 -> Core 5 Ultra or Core 5
• Core i3 -> Core 3

If we translate that to products that already exist on the market, we get something like the following:

• Current Desktop Branding: Core i9-13900K
• New Desktop Branding #1: Core 9 Ultra 13900K
• New Desktop Branding #2: Core 9 13900K

And for mobile, it will look something like this:

• Current Laptop Branding: Core i7-1370P
• New Laptop Branding #1: Core 7 Ultra 1370P
• New Laptop Branding #2: Core 7 1370P

2 of 9

The Building Blocks of Meteor Lake CPUs

Starting with Meteor Lake CPUs, Intel is bringing a brand new chip layout that includes various tiles or chiplets (as you like to refer to them) with various IPs. On the vast majority of Meteor Lake CPUs, you will find a total of four tiles. The quad-tile layout includes the CPU Tile, Graphics Tile, SOC Tile, and IOE Tile.

All four tiles will utilize internal and external fabrication processes which means that some of the tiles will be made by Intel while the rest will be made by 3rd party fabs such as TSMC. The main CPU tile will be using the "Intel 4" or 7nm EUV process node while the SOC Tile and IOE Tiles will be fabricated on TSMC's 6nm process node (N6). Intel calls Meteor Lake the first step into the chiplet ecosystem in the client segment. The tGPU which is a new name for the iGPU (Tiled-GPU) is also a major component of Meteor Lake CPUs and utilizes TSMC's 5nm process node. So to quickly sum it all up:

• Intel Meteor Lake Compute Tile: Intel 4 "7nm EUV"
• Intel Meteor Lake Graphics Tile: TSMC 5nm
• Intel Meteor Lake SOC Tile: TMSC 6nm
• Intel Meteor Lake IO Tile: TSMC 6nm

So starting with the dissection of each tile, first up, we have the Compute Tile which is fully scalable across various core counts, core generations, nodes, and cache. Intel can mix and match not just different core architectures within its Foveros 3D package CPUs such as Meteor Lake but they can also scale up or down to a different node.

The same is true for the graphics tile which can be scaled in terms of core count, node, and cache too. These diagrams are just for illustration purposes but they show a tGPU block scaling from 4 Xe cores (64 EUs) up to 12 Xe Cores (192 EUs). However, based on the die shot in the same picture, we can see 8 Xe Cores (128 EUs).

2 of 9

The SOC Tile can also be scaled up or down depending on the SKU. The main blocks here are the Low-Power IP (referring to the VPU), SRAM, IO, and a scalable voltage design. The same is true for the last Tile, the I/O Extender, or IOE in short. The tile is fully scalable in terms of the number of lanes, bandwidth, protocols, and speed.

With the tiles out of the way, it's time to put together everything and for that, Intel has shown a breakdown of how the CPU dies are arranged together. The top layer has a back-side metallization and is also where the Foveros passive die sits. Right below these are the known good tiles that we've just discussed above. These tiles are connected to the Base Tile using a 36um pitch (die-to-die) inter-connect. The Base Tile comes with large capacitance and has metal layers for IO/power delivery and D2D routing.

2 of 9

Intel also provides a close-up of the Base Tile's metal layer which features 3D capacitors and die-to-die power delivery plus package I/O routing. Each metal layer is modular with active silicon for logic and memory. The top and bottom have package bumps for interconnecting with the top & bottom layers.

The configuration shown here is also a mobile-specific chip with a 6+8 (6 P-Cores + 8 E-Cores) layout. You can also note that there are two D2D (Die-To-Die) links between the CPU/IOE Tile and the Graphics Tile leading into the SOC Tile. This is part of the Foveros 3D Packaging and the blue team states that there's a passive interposer on top of the main chiplets which is based on a 22nm (FFL) process from Intel itself. This interposer currently serves no purpose but the company plans to use active chiplets within it in the future with more advanced packaging technologies. The Intel Meteor Lake CPUs don't utilize EMIB technology.

The FDI (Foveros Die Interconnect) technology offers:

• A Low Voltage CMOS Interface
• High Bandwidth, Low Latency
• Synchronous and asynchronous Signaling
• Low Area Overhead
• Operation @ 2GHz, 0.15-0.3 pJ/bit

The Interconnects between the CPU and the SOC have a mainband width of around 2K (2x IDI), the Graphics and SOC tiles have an interconnect mainband width of around 2K too (2x iCXL) while the SOC and IOE tiles have a mainband width around 1K (IOSF, 4x Display Port).

Another key area where Intel's Meteor Lake CPUs have improved a lot is the maximum turbo power capability. Since its inception and with the help of co-optimizations, the Meteor Lake chips can achieve higher turbo power capabilities than the previous generation Alder Lake CPUs while utilizing the brand new "Intel 4" process node. The total capacitance has also touched 500 for the Meteor Lake Base Tile.

Intel gives us an old-school comparison between a Haswell and Meteor Lake CPU as far as their I/O capabilities are concerned. Another aspect that was touched upon by Intel is pricing. With the costs of next-gen wafer prices going up with every new node, the cost of developing a monolithic die is also going to go up.

If you were to take Meteor Lake as it is and design it monolithically on a leading process node, I would say it actually is extremely competitive with that if not actually cheaper.

via Intel

Intel shows that a disaggregated design like Meteor Lake with Tiled-architecture can deliver higher performance, higher transistor performance uplifts, & better IP refresh pace across various process nodes, all at higher power efficiency versus a monolithic solution.

2 of 9

Intel revealed that its Meteor Lake CPUs will scale from <10W to over 100W SKUs, offering the CPU performance of a monolithic design in a disaggregated package.

Intel Mobility CPU Lineup:

The Latest & Greatest CPU/GPU Cores

So with the building blocks out of the way, it's time to dive into the two main tiles, the Compute & Graphics Tile.

Redwood Cove & Crestmont Are A Meteoric Combo

The Intel Meteor Lake Compute Tile is once again making use of a hybrid architecture layout. The chip will utilize the brand new Redwood Cove architecture for its P-Cores and the Crestmonth CPU architecture for its E-Cores. Once again, each Redwood Cove core will feature SMT support while the Crestmont cores will ship without SMT support which is a common occurrence with E-Cores. In terms of IPC improvement, we don't have enough data yet but the "Intel 4" process node itself is said to deliver a 20% performance per watt increase.

Die-shot decoder, @Locuza_, provided a dissection of a Meteor Lake CPU die-shot a while back. The Meteor Lake CPU pictured here featured 2 P-Cores and 2 E-Core clusters comprised of 8 E-Cores in total. Each of the P-Cores features 3.0 MB of L3 cache while each of the E-Core clusters features 3.0 MB of L3 cache. As for the L2 cache, the Redwood Cove cores seem to carry 2 MB of it versus the 1.25 MB featured on Golden Cove while each Crestmont cluster seems to have 3 MB of L2 cache.

A Major Leap In GPU Design With Arc Alchemist

On the GPU side, Intel will be leveraging its Arc Alchemist graphics architecture in its Xe-LPG variation. This chip will pack a total of 128 EUs within the GPU title which utilizes TSMC's 5nm process node. It is reported that the iGPU can clock up to 2.2 GHz which is pretty much what Alchemist discrete parts are able to achieve with much higher TDPs.

This should put the FP32 compute performance of the chip around 4.5 TFLOPs. This is a 2.25x increase in compute power over the Iris Xe 96 EU graphics found on the Raptor Lake-H CPUs. Comparison cannot be made with AMD's RDNA GPUs since they have a different architecture & also compute the FP32 TFLOPs differently but it looks like Intel may take a lead over AMD's Radeon 700M "RDNA 3" iGPUs by a fair amount. Manufacturers have already cited up to RTX 3050 performance out of this iGPU.

Having the Alchemist architecture onboard the Meteor Lake CPUs also means that these will enable enhanced support for DirectX 12 Ultimate and XeSS amongst a range of features that have so far only been limited to Intel's discrete lineup of Arc chips. We also cannot forget about the rumored & highly anticipated "Adamantine" cache that is rumored to be featured on some Meteor Lake SKUs in up to 128 MB capacities.

This cache can be a major boost for iGPUs which are usually deprived of bandwidth but Intel has also recently shown some Meteor Lake SKUs with on-package DRAM packages utilizing Samsung's latest LPDDR5x memory solution. The secondary solution may be designed for more size-constrained platforms with the memory serving as the main system DRAM while the Adamantine Cache will be on-die and will only be accessible to either the GPU or Compute Tile (if need be).

Since the Adamantine L4 cache can also be used by the Compute Title (CPU Cores), which are made up of Redwood Cove (P-Core) & Crestmont (E-Core) hybrid configurations, it can lead to faster boot times and overall lower latencies compared to moving data to the primary DRAM. The patent states that:

Next generation client SoC architectures may introduce large on-package caches, which will allow novel usages. Access time for the L4 (e.g., “Adamantine” or “ADM”) cache may be much less than the DRAM access time, which is used to improve host CPU and security controller communications. Embodiments help to protect innovations in boot optimization. Value is added for high end silicon with higher pre-initialized memory at reset, potentially leading to increased revenue. Having memory available at reset also helps to nullify legacy BIOS assumptions and make a faster and efficient BIOS solution with a reduced firmware stage (e.g., pre-CPU reset stage, IBBL stage and IBB stage) for modern device use cases like Automotive IVI (in-vehicle infotainment, e.g., turn on rear view camera within 2 sec), household and industrial robots, etc. Accordingly, new market segments may be available.

Intel Patent (Early Platform Hardening Technology For Slimmer And Faster Boot)

Meteor Lake Configurations Amasse!

So that leads us into the next point of discussion and that's the configurations. With Intel Meteor Lake CPUs, the chipmaker would offer a diverse portfolio of SKUs in three key segments which include Meteor Lake-U ranging from 7-15W, Meteor Lake-P which includes 28W chips, and Meteor Lake-H which includes up to 45W chips.

The Intel Meteor Lake-U lineup will consist of the most diverse range of configurations including the following:

• MTL-U 12 Core - 4P+8E (7/8 Xe Core iGPU) @ 15W
• MTL-U 10 Core - 2P+8E (4 Xe Core iGPU) @ 15W
• MTL-U 6 Core - 2P+4E (3 Xe Core iGPU) @ 15W
• MTL-U 10 Core - 2P+8E (4 Xe Core iGPU) @ 9W
• MTL-U 10 Core - 2P+8E (4 Xe Core iGPU) @ 9W
• MTL-U 6 Core - 2P+4E (3 Xe Core iGPU) @ 9W
• MTL-U 9 Core - 1P+8E (4 Xe Core iGPU) @ 7W
• MTL-U 9 Core - 1P+8E (4 Xe Core iGPU) @ 7W
• MTL-U 5 Core - 1P+4E (4 Xe Core iGPU) @ 7W

Moving over to the Meteor Lake-P lineup, we have a total of three die configurations starting with the top 14-core and moving down to 12 and 10 cores.

• MTL-P 14 Core - 6P+8E (8 Xe Core iGPU) @28W
• MTL-P 12 Core - 4P+8E (8 Xe Core iGPU) @28W
• MTL-P 10 Core - 2P+8E (4/7 Xe Core iGPU) @28W

Lastly, for the top Intel Meteor Lake-H lineup, there are also three configurations, and each features 8 Xe iGPU cores:

• MTL-H 14 Core - 6P+8E (8 Xe Core iGPU) @45W
• MTL-H 14 Core - 6P+8E (8 Xe Core iGPU) @45W
• MTL-H 12 Core - 4P+8E (8 Xe Core iGPU) @45W

Once again, we would also point out the now canceled Meteor Lake-S Desktop CPUs which were going to be compatible with the LGA 1851 platform:

• Meteor Lake-S 22 (6P + 16E) / 4 Xe Cores / 125W TDP
• Meteor Lake-S 22 (6P + 16E) / 4 Xe Cores / 65W TDP
• Meteor Lake-S 22 (6P + 16E) / 4 Xe Cores / 35W TDP
• Meteor Lake-S 14 (6P + 8E) / 4 Xe Cores / 65WTDP
• Meteor Lake-S 14 (6P + 8E) / 4 Xe Cores / 35WTDP

Here Are The 1st Gen Core Ultra CPUs We Know So Far

Coming to actual SKUs, we have so far seen a total of 7 SKUs leak out so far and all of them are part of the MTL-H family, featuring 14 & 12 core configurations. You can add an additional two E-Cores if you count the ones featured on the SOC tile.

The Intel Core Ultra 9 185H seems to be the fastest SKU that we have seen so far with 14+2 cores, 20+2 threads, 24 MB of L3 cache, and up to 5.1 GHz boost clocks. Talking about clocks for a second, Meteor Lake being the first iteration of the "Intel 4" process node is bound to have some frequency regression versus the much optimized "Intel 7" process node. But with that said, Intel's fastest 12th Gen Core i9-12900HK CPU also peaked out at 5.0 GHz when it launched while its successor boosted the clock speeds by an additional +400 MHz (Core i9-13900HK). So 5.1 GHz and anything above that is a great start for "Intel 4" and we can only see things get better as the process technology & Intel's EUV implementation improves over time.

• Intel 10 - Core i9-11980HK - 5.0 GHz (Max Clocks)
• Intel 7 - Core i9-12900HK - 5.0 GHz (Max Clocks)
• Intel 7 - Core i9-13900HK - 5.4 GHz (Max Clocks)
• Intel 4 - Core Ultra 9 185H - 5.1 GHz (Max Clocks)

The rest of the SKUs are mentioned in the chart below along with their respective specifications (Do note that these aren't final specs by any means until announced by Intel themselves).

Intel 14th Gen Meteor Lake CPU Family "Preliminary":

A Major Test For Intel's Client-Side AI Endeavors

One big marketing push by Intel for Meteor Lake has been its AI prowess. This push comes from the fact that AMD has beaten Intel in bringing out a formidable AI solution in the form of its Ryzen AI solution which features AVX-512 support and various inferencing capabilities. So for Intel to get back up to speed, the company will be introducing its brand new VPU engine and new capabilities powered by AI that not only make user experiences better on a software level but also on a deep-down hardware level.

For this purpose, Intel is bringing forth its VPU (Versatile Processing Unit) which is designed to offer a richer and faster experience in Windows 11. Intel has been working together with Microsoft to optimize the AI capabilities of Meteor Lake's VPU. At Computex 2023, the company gave us a brief look at the AI capabilities of Meteor Lake within Generative AI content and Stable Diffusion.

Intel has two main advantages of having a dedicated VPU onboard the Meteor Lake CPU. The first tier advantages come in applications on the client side which include:

• Enhanced Audio Effects
• Creator And Gaming Effects
• AI Assistants (Microsoft 365 Copilot)
• Bringing AI locally to client machines.

The second-tier advantages come in pure and dedicated AI workloads which include:

• AI Compute Required For Enhancements  GOPS
• AI Compute Required For Large Language Models (TOPS) - ChatGPT

Intel showed two demos, one with Stable Diffusion AI processing and the other with Advanced Blue processing, The Blur processing achieved with the VPU not only looked better than without the VPU but also consumed 1/5th of the power.

The Stable Diffusion demo was run within GIMP using an open-source plugin developed by Intel. The VPU generated images in a very short time compared to what it usually takes on a standard CPU or iGPU. Although we were unable to get a comparison between VPU off and VPU on, it's still a big deal for consumers.

Furthermore, Intel will be leveraging the AI capabilities of Meteor Lake to handle the chip's own power and thermal envelopes, making it much more power-efficient and faster.

Intel Meteor Lake CPU Performance

Performance of the Intel Meteor Lake "Core Ultra" CPUs is anticipated to be good with the addition of new cores. We have seen a few leaks where the chips are slightly faster than the existing Raptor Lake CPUs and these are early engineering samples that we are looking at. Once the final CPU and GPU performance numbers start rolling in, we will see an even bigger uplift. Efficiency is also going to be a major deal on the Meteor Lake CPUs since they are designed around the mobility platforms.

Following are the recently leaked Core Ultra 7 155H and Core Ultra 5 125H benchmarks with the former exceeding the Core i9-13900H and the latter exceeding the Core i7-13700H in multi-thread performance tests.

Intel Meteor Lake Family Pricing & Availability

Intel is hosting a grand unveiling for its 1st Gen Core Ultra "Meteor Lake" CPU family at the Innovation 2023 event which takes place next week on 19th September. We can expect specs, benchmarks, various products, and more. The actual launch will take place a bit later with few reports highlighting that Intel Meteor Lake Core Ultra 5 SKUs will be the first to hit the market whereas Core Ultra 7 and Core Ultra 9 SKUs won't appear until early 2024.

Price-wise, the first laptops might come at a premium since these utilize the latest fabrication process, and making chips on advanced technologies is not just harder but expensive too these days.

WccftechContinue reading/original-link]