Tech News
← Home  ·  All topics

Crescent Island

3 GoKawiil briefs on this topic · RSS for crescent island

Intel's D270 GPU claims up to 5.9x lead over Nvidia RTX Pro 6000 in FP64 finance benchmarks

Intel has published its own benchmark results for the data-center GPU D270, formerly known as Crescent Island, pitting it against Nvidia's RTX Pro 6000 Blackwell Server Edition on four double-precision financial pricing workloads shown at the SC26 supercomputing conference. Intel's card posted gains of 1.92x to 5.9x over Nvidia's, with its largest advantage on a European Monte Carlo simulation, while drawing 350W versus Nvidia's 600W rating. In AI-focused tests, Intel reported only modest gains, including a 12% throughput edge on one language model.

Intel renames Crescent Island accelerator as GPU D-Series, reveals specs at OCP 2026

Intel announced at OCP 2026 that its previously code-named Crescent Island chip will ship as the GPU D-Series, disclosing a theoretical peak of 1.3 PFLOPS FP8 performance and 1.5 TB/s memory bandwidth. The air-cooled PCIe card runs at 350W TDP with up to 480GB of LPDDR5X memory, built on Intel's new Xe3P architecture, and can scale to eight GPUs per server.

Intel details Crescent Island data center GPU: 32 Xe3P cores, up to 480GB memory

At Hot Chips 2026, Intel disclosed full specifications for its Crescent Island AI accelerator, including 32 Xe3P cores, 256 XMX engines, 32MB of L2 cache, PCIe Gen5 x16, and a 350W TDP. Shipping units will carry 160GB of LPDDR5X memory, though ODM designs can scale up to 480GB of VRAM.