Teamgroup unveils 128GB DDR5 CQDIMM for AI workloads
Teamgroup has announced a 128GB DDR5 CQDIMM for AI and creative workloads, rated at 8,000 MT/s at 1.1V, but has not disclosed pricing or retail availability.
Each module supports 8,000 MT/s at 1.1V without overclocking, according to TechRadar. Teamgroup combines an independent clock driver with its patented integrated circuit grading verification technology to support that speed. Two compatible modules can provide 256GB of memory, which the company positions for running AI models locally and for demanding creative applications.
The module is designed for Intel Z890 platforms using a two-slot, single-processor channel configuration, so buyers need to check compatibility carefully, TechRadar reported. The intended workloads include local language models, 4K and 8K video production and complex 3D rendering, which can consume large amounts of system memory.
Teamgroup says the module uses an aluminum-alloy heat spreader with a perforated structure to improve heat dissipation during demanding operations. Its slim profile is intended for AI PCs, while the cooling design is meant to manage rising temperatures during extended workloads.
Pricing is the central unanswered question, especially in an expensive memory market, according to TechRadar. The company has not disclosed a price, but its existing T-CREATE EXPERT 128GB DDR5-6400 kit offers a comparison. B&H Photo Video lists that two-module kit at $2,339.99, while Tom's Hardware recently reported comparable 128GB DDR5 pricing at $3,399, TechRadar said.
Because the new CQDIMM packs the same capacity into one module and uses a different memory design, its eventual price is difficult to predict. TechRadar said a provisional estimate of $3,000 to $4,000 for one module is plausible but remains speculation rather than confirmed pricing. Such spending may be difficult to justify for routine desktop upgrades, though AI developers and professional workstation users needing substantial memory capacity may consider it.
Teamgroup's intended users include AI developers, 3D animators and content creators working with computationally demanding applications. The company has not published benchmark results establishing how much faster those applications would run. Capacity can help when workloads exceed available memory, but it does not automatically improve performance when other hardware is the limiting factor, according to TechRadar.