Intel

Intel Core Ultra 200V “Lunar Lake” SoC to offer 100+ platform TOPS for AI

Intel confirms Lunar Lake NPU will be 3x faster than Meteor Lake

More than enough for next-gen AI consumer processor.

Lunar Lake with 100 TOPS (AI-upscaled), Source: Intel

During Vision 2024, Intel CEO Pat Gelsinger announced details about the upcoming Core Ultra 200V series, codenamed “Lunar Lake.” These processors will boast a faster AI processor compared to the Core Ultra 100 series, with the next-gen NPU (Neural Processing Unit) delivering three times higher performance.

The Lunar Lake series will belong to the low-power segment and will be powered by Lion Cove and Skymont architectures. They will be offered alongside the Arrow Lake series (also part of the Core Ultra 200 lineup), which shares the same CPU architecture but with a less powerful GPU. It’s worth noting that the Core Ultra 200V will utilize the Xe2-LPG architecture, a low-power variant of Battlemage, while Arrow Lake will retain the Xe-LPG (Alchemist) architecture seen in the current Core Ultra 100 “Meteor Lake” series.

Intel estimates that the full platform TOPS (tera-operations per second) for Lunar Lake will exceed 100 TOPS, making it the highest figure among all known upcoming processors. Tom’s Hardware, who reported on this news have also reached out to AMD to inquire about the performance of their NPU (AMD XDNA2) on the upcoming Zen5 series, known as Strix Point.

Consumer APU/SoC AI Performance
VideoCardzCPUGPUNPUTotal
Intel Lunar Lake~5 TOPS~50 TOPS~45 TOPS
~100 TOPS
Intel Meteor Lake5 TOPS18 TOPS11 TOPS
34 TOPS
AMD Strix Point TBC TBC ~48 TOPS TBC
AMD Hawk Point TBC TBC 16 TOPS
39 TOPS
AMD Phoenix TBC TBC 10 TOPS
33 TOPS
Qualcomm Snapdragon X Elite TBC TBC 45 TOPS
75 TOPS

According to a representative, the upcoming AMD XDNA2 NPU is expected to be three times faster than its predecessor, XDNA. It would feature a performance of 48 TOPS compared to Ryzen 8040 “Hawk Point’s” 16 TOPS. It’s important to note that this improvement only applies to the NPU, and details about the full platform, which combines Zen5 cores and RDNA 3.5 architecture, are still to be confirmed.

“We believe an AI PC requires strong CPU, GPU and NPU engines, which is what AMD has been delivering for more than a year with our Ryzen 7040 and now 8040 Series. At our December Advancing AI event, we disclosed our next-gen “Strix Point” mobile processors with XDNA 2 architecture would have up to 3x the generative AI performance of the current generation. We believe this performance will position us to remain the leading choice for next-gen AI PCs.”

— AMD to Tom’s Hardware.

The competition for the fastest AI acceleration in consumer CPUs is intensifying, with both AMD and Intel pushing the envelope. However, a new player has entered the arena: Qualcomm. Their upcoming Snapdragon X Elite series, set to launch this year, claims to offer performance, boasting 45 TOPS from the NPU alone and a total of 75 TOPS for the entire chip. With Qualcomm’s entry, the race for AI-acceleration leadership is set to become even more competitive.

Source: Tom’s Hardware