{"product_id":"tenstorrent-blackhole-p150b-ai-accelerator-32gb-gddr6-664-tflops-120-tensix-cores-passive-cooling-pcie-5-0","title":"Tenstorrent Blackhole P150b AI Accelerator | 32GB GDDR6 | 664 TFLOPS | 120 Tensix Cores | Passive Cooling | PCIe 5.0","description":"\u003cp\u003e\u003cspan\u003eTenstorrent Blackhole P150b AI Accelerator – 32GB GDDR6 | 664 TFLOPS | 120 Tensix Cores | 512GB\/s | Passive Cooling | PCIe 5.0\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eEnterprise-Grade AI Acceleration for High-Performance Servers \u0026amp; Scalable AI Infrastructure\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eExperience next-generation artificial intelligence computing with the Tenstorrent Blackhole P150b\u003c\/span\u003e\u003cspan\u003e, a high-performance, passively cooled PCIe AI accelerator designed for rack-mounted servers, AI research laboratories, enterprise computing infrastructure, and multi-accelerator deployments.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003ePowered by Tenstorrent’s innovative Blackhole Tensix architecture, the P150b combines \u003c\/span\u003e\u003cspan\u003e120 Tensix AI processing cores, 16 SiFive X280 RISC-V processor cores, 32GB GDDR6 memory, 512GB\/s memory bandwidth, and up to 664 TFLOPS of BLOCKFP8 compute performance\u003c\/span\u003e\u003cspan\u003e.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eUnlike conventional desktop AI accelerator cards, the Blackhole P150b uses a \u003c\/span\u003e\u003cspan\u003epassive heatsink engineered for server environments with high-static-pressure forced-air cooling\u003c\/span\u003e\u003cspan\u003e.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eIts four dedicated 800Gbps QSFP-DD interconnect ports enable direct communication between compatible Blackhole accelerators, making it a compelling platform for scalable AI research, distributed machine learning, and high-density server installations.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eKey Features \u0026amp; Highlights\u003c\/span\u003e\u003c\/p\u003e\n\u003cul\u003e\n\u003cli\u003e\n\u003cspan\u003e664 TFLOPS Peak BLOCKFP8 Compute\u003c\/span\u003e\u003cspan\u003e – High-throughput AI acceleration for supported machine learning applications.\u003c\/span\u003e\n\u003c\/li\u003e\n\u003cli\u003e\n\u003cspan\u003e120 Tensix AI Cores\u003c\/span\u003e\u003cspan\u003e – Dedicated processing resources for tensor-intensive computing.\u003c\/span\u003e\n\u003c\/li\u003e\n\u003cli\u003e\n\u003cspan\u003e16 SiFive X280 RISC-V Cores\u003c\/span\u003e\u003cspan\u003e – Programmable processor architecture integrated into the Blackhole chip.\u003c\/span\u003e\n\u003c\/li\u003e\n\u003cli\u003e\n\u003cspan\u003e32GB GDDR6 Memory\u003c\/span\u003e\u003cspan\u003e – High-capacity onboard memory for AI models and data processing.\u003c\/span\u003e\n\u003c\/li\u003e\n\u003cli\u003e\n\u003cspan\u003e512GB\/s Memory Bandwidth\u003c\/span\u003e\u003cspan\u003e – High-speed memory access for demanding computational workloads.\u003c\/span\u003e\n\u003c\/li\u003e\n\u003cli\u003e\n\u003cspan\u003e180MB On-Chip SRAM\u003c\/span\u003e\u003cspan\u003e – Fast local memory for efficient AI processing.\u003c\/span\u003e\n\u003c\/li\u003e\n\u003cli\u003e\n\u003cspan\u003ePCI Express 5.0 x16\u003c\/span\u003e\u003cspan\u003e – High-bandwidth connection to compatible server motherboards.\u003c\/span\u003e\n\u003c\/li\u003e\n\u003cli\u003e\n\u003cspan\u003e4 × QSFP-DD 800Gbps Interconnects\u003c\/span\u003e\u003cspan\u003e – Direct accelerator-to-accelerator connectivity.\u003c\/span\u003e\n\u003c\/li\u003e\n\u003cli\u003e\n\u003cspan\u003ePassive Server Cooling\u003c\/span\u003e\u003cspan\u003e – Designed for rack-mounted systems with sufficient forced airflow.\u003c\/span\u003e\n\u003c\/li\u003e\n\u003cli\u003e\n\u003cspan\u003e300W Total Board Power\u003c\/span\u003e\u003cspan\u003e – High-performance accelerator operating envelope.\u003c\/span\u003e\n\u003c\/li\u003e\n\u003cli\u003e\n\u003cspan\u003eUp to 1.35GHz AI Clock\u003c\/span\u003e\u003cspan\u003e – Advanced Blackhole processing architecture.\u003c\/span\u003e\n\u003c\/li\u003e\n\u003cli\u003e\n\u003cspan\u003eOpen-Source Software Ecosystem\u003c\/span\u003e\u003cspan\u003e – Access to Tenstorrent development tools and programming resources.\u003c\/span\u003e\n\u003c\/li\u003e\n\u003c\/ul\u003e\n\u003cp\u003e\u003cspan\u003ePowerful Blackhole Tensix AI Architecture\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eThe Tenstorrent Blackhole P150b incorporates an advanced AI processing architecture designed for modern machine learning and neural network applications.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eAt its core are \u003c\/span\u003e\u003cspan\u003e120 specialized Tensix processing cores\u003c\/span\u003e\u003cspan\u003e engineered to perform tensor operations efficiently.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eThe processor also integrates 16 SiFive X280 RISC-V cores, providing additional programmable computing capabilities.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003e664 TFLOPS AI Processing Performance\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eThe Blackhole P150b delivers up to \u003c\/span\u003e\u003cspan\u003e664 TFLOPS of theoretical BLOCKFP8 computing throughput\u003c\/span\u003e\u003cspan\u003e.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eThis performance makes the accelerator suitable for exploring:\u003c\/span\u003e\u003c\/p\u003e\n\u003cul\u003e\n\u003cli\u003e\u003cspan\u003eArtificial intelligence inference\u003c\/span\u003e\u003c\/li\u003e\n\u003cli\u003e\u003cspan\u003eNeural network development\u003c\/span\u003e\u003c\/li\u003e\n\u003cli\u003e\u003cspan\u003eTransformer-based AI architectures\u003c\/span\u003e\u003c\/li\u003e\n\u003cli\u003e\u003cspan\u003eGenerative AI applications\u003c\/span\u003e\u003c\/li\u003e\n\u003cli\u003e\u003cspan\u003eComputer vision processing\u003c\/span\u003e\u003c\/li\u003e\n\u003cli\u003e\u003cspan\u003eCustom AI kernel optimization\u003c\/span\u003e\u003c\/li\u003e\n\u003cli\u003e\u003cspan\u003eDistributed machine learning research\u003c\/span\u003e\u003c\/li\u003e\n\u003c\/ul\u003e\n\u003cp\u003e\u003cspan\u003ePerformance depends on supported numerical formats, model architecture, software implementation, and system configuration.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003e32GB GDDR6 High-Bandwidth Memory\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eModern artificial intelligence workloads require fast and efficient memory access.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eThe Blackhole P150b features an advanced memory subsystem with:\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003e32GB GDDR6 Memory\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eProvides onboard storage for model parameters, intermediate tensors, and supported computational workloads.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003e512GB\/s Memory Bandwidth\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eEnables high-speed transfer of data between onboard memory and processing resources.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003e256-bit Memory Interface\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eSupports efficient data movement across the accelerator’s memory subsystem.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003e180MB On-Chip SRAM\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eProvides fast local memory resources for supported processing operations.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eThis combination makes the P150b particularly attractive for advanced AI development and multi-accelerator computing environments.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eFour 800Gbps QSFP-DD Ports for Multi-Accelerator Scalability\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eOne of the defining capabilities of the Tenstorrent Blackhole P150b is its high-speed accelerator interconnect architecture.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eThe card includes \u003c\/span\u003e\u003cspan\u003efour passive QSFP-DD ports, each providing 800Gbps connectivity\u003c\/span\u003e\u003cspan\u003e.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eThese ports support direct communication between compatible Tenstorrent Blackhole P150a and P150b accelerator cards.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eBenefits of Multi-Card Connectivity\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eScalable AI Computing\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eBuild supported multi-card configurations to increase aggregate computational resources.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eExpanded Distributed Memory\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eCombine the memory resources of multiple accelerator cards for supported distributed workloads.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eHigh-Bandwidth Communication\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eDedicated interconnects support fast communication between compatible Blackhole processors.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eEnterprise AI Infrastructure\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eDeploy multiple accelerators in compatible rack-mounted servers for research and development.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eFlexible AI Research\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eExplore distributed model execution, multi-device computing, and advanced AI software optimization.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eQSFP-DD ports are intended exclusively for supported Tenstorrent Blackhole accelerator connections, not general-purpose Ethernet networking. Compatible cables and supported software are required.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003ePassive Cooling Designed for Rack-Mounted Servers\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eThe \u003c\/span\u003e\u003cspan\u003eBlackhole P150b is specifically engineered for high-density server environments\u003c\/span\u003e\u003cspan\u003e.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eUnlike the Blackhole P150a, which includes an active blower-style cooling solution, the P150b uses a passive heatsink.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eAdvantages of Passive Server Cooling\u003c\/span\u003e\u003c\/p\u003e\n\u003cul\u003e\n\u003cli\u003e\u003cspan\u003eNo dedicated onboard cooling fan\u003c\/span\u003e\u003c\/li\u003e\n\u003cli\u003e\u003cspan\u003eDesigned for centrally managed server airflow\u003c\/span\u003e\u003c\/li\u003e\n\u003cli\u003e\u003cspan\u003eSuitable for compatible high-density rack-mounted systems\u003c\/span\u003e\u003c\/li\u003e\n\u003cli\u003e\u003cspan\u003eEnables server integrators to manage cooling through chassis fans\u003c\/span\u003e\u003c\/li\u003e\n\u003cli\u003e\u003cspan\u003eSupports multi-accelerator installations with appropriate thermal design\u003c\/span\u003e\u003c\/li\u003e\n\u003c\/ul\u003e\n\u003cp\u003e\u003cspan\u003eImportant:\u003c\/span\u003e\u003cspan\u003e Passive cooling does not mean the accelerator can operate without airflow. The P150b requires sufficient high-static-pressure forced-air cooling from the host server.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eIt should not be installed in an ordinary desktop computer without a validated cooling solution.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eApplications \u0026amp; Use Cases\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003e1. Enterprise AI Servers\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eDeploy compatible Blackhole accelerators in rack-mounted infrastructure for specialized artificial intelligence workloads.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003e2. Large Language Model Research\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eExplore supported transformer architectures, language model inference, and distributed AI processing.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003e3. Deep Learning \u0026amp; Neural Networks\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eDevelop, test, and optimize supported neural network workloads.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003e4. Generative Artificial Intelligence\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eExperiment with compatible generative AI models and inference applications.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003e5. AI Research Laboratories\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eSupport universities, research institutions, and advanced engineering teams working on AI accelerator technology.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003e6. Multi-Accelerator Computing\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eInterconnect multiple P150a and P150b cards using supported QSFP-DD cables and configurations.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003e7. Custom AI Software Development\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eExplore hardware-level programming and custom AI kernel development using Tenstorrent’s open-source software stack.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003e8. High-Performance Computing Research\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eInvestigate specialized parallel computing workloads that can be implemented using the Tensix architecture.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eComplete Technical Specifications\u003c\/span\u003e\u003c\/p\u003e\n\u003ctable width=\"964.0\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n\u003ctbody\u003e\n\u003ctr\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eSpecification\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eDetails\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eBrand\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eTenstorrent\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eProduct Model\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eBlackhole P150b\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eManufacturer Part Number\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eTC-03002\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eProduct Type\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003ePCIe AI Accelerator\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eProcessor Architecture\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eBlackhole Tensix\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eTensix AI Cores\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003e120\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eRISC-V Cores\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003e16 SiFive X280\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eAI Clock Speed\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eUp to 1.35GHz\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003ePeak BLOCKFP8 Performance\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003e664 TFLOPS\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eOnboard Memory\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003e32GB GDDR6\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eMemory Bus Width\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003e256-bit\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eMemory Speed\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003e16 GT\/s\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eMemory Bandwidth\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003e512GB\/s\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eOn-Chip SRAM\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003e180MB\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eHost Interface\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003ePCI Express 5.0 x16\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eAccelerator Interconnect\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003e4 × QSFP-DD 800G\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eInterconnect Cable Length\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eUp to 2 metres\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eTotal Board Power\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003e300W\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eIdle Power\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003e75W\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eExternal Power Connector\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003e12V-2x6, 12+4-pin\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003ePSU Requirement\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eATX 3.1 Certified or Better\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eCooling Type\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003ePassive Heatsink\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eRequired Cooling Environment\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eForced-Air Server Chassis\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eCard Width\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003e42mm\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eCard Depth\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003e270mm\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eCard Height\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003e111mm\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eExpansion Form Factor\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eDual-Slot PCIe\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eSoftware\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eTenstorrent Open-Source Stack\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eIntended Deployment\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003ctd valign=\"middle\"\u003e\n\u003cp\u003e\u003cspan\u003eRack-Mounted AI Servers\u003c\/span\u003e\u003c\/p\u003e\n\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003c\/tbody\u003e\n\u003c\/table\u003e\n\u003cp\u003e\u003cspan\u003eSystem Requirements \u0026amp; Installation\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eThe Tenstorrent Blackhole P150b is intended for compatible server installations.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eRecommended Hardware Environment\u003c\/span\u003e\u003c\/p\u003e\n\u003cul\u003e\n\u003cli\u003e\n\u003cspan\u003eCPU:\u003c\/span\u003e\u003cspan\u003e Compatible x86-64 host processor\u003c\/span\u003e\n\u003c\/li\u003e\n\u003cli\u003e\n\u003cspan\u003eMotherboard:\u003c\/span\u003e\u003cspan\u003e PCIe 5.0 x16 slot without bifurcation recommended\u003c\/span\u003e\n\u003c\/li\u003e\n\u003cli\u003e\n\u003cspan\u003ePower Supply:\u003c\/span\u003e\u003cspan\u003e ATX 3.1 certified or newer\u003c\/span\u003e\n\u003c\/li\u003e\n\u003cli\u003e\n\u003cspan\u003eExternal Power:\u003c\/span\u003e\u003cspan\u003e One compatible 12V-2x6 connector per card\u003c\/span\u003e\n\u003c\/li\u003e\n\u003cli\u003e\n\u003cspan\u003eCooling:\u003c\/span\u003e\u003cspan\u003e Server chassis with sufficient high-static-pressure forced airflow\u003c\/span\u003e\n\u003c\/li\u003e\n\u003cli\u003e\n\u003cspan\u003eExpansion Clearance:\u003c\/span\u003e\u003cspan\u003e Dual-slot space per accelerator\u003c\/span\u003e\n\u003c\/li\u003e\n\u003cli\u003e\n\u003cspan\u003eOperating System:\u003c\/span\u003e\u003cspan\u003e Supported Linux distribution and Tenstorrent software stack\u003c\/span\u003e\n\u003c\/li\u003e\n\u003cli\u003e\n\u003cspan\u003eInterconnect:\u003c\/span\u003e\u003cspan\u003e Compatible QSFP-DD cables for multi-card deployments\u003c\/span\u003e\n\u003c\/li\u003e\n\u003c\/ul\u003e\n\u003cp\u003e\u003cspan\u003eFor installations involving multiple accelerator cards, verify total server power capacity, chassis airflow, PCIe slot availability, interconnect topology, and software compatibility.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eWhy Choose the Tenstorrent Blackhole P150b?\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003e1. Enterprise-Oriented Design\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eEngineered for rack-mounted servers and high-density AI computing environments.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003e2. Powerful AI Acceleration\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eDelivers up to 664 TFLOPS of theoretical BLOCKFP8 compute throughput.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003e3. High-Capacity Memory\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003e32GB GDDR6 memory with 512GB\/s bandwidth supports demanding AI development workloads.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003e4. Multi-Card Scalability\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eFour dedicated 800Gbps QSFP-DD ports support compatible accelerator interconnections.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003e5. Passive Server Cooling\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eIntegrates with properly designed server cooling infrastructure without requiring an onboard fan.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003e6. Open-Source Software\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eTenstorrent’s development ecosystem enables experimentation and custom AI software development.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003e7. Advanced RISC-V Architecture\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eCombines dedicated Tensix AI cores with programmable RISC-V processor cores.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eFrequently Asked Questions\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eQ1. What is the Tenstorrent Blackhole P150b?\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eThe P150b is a passively cooled PCIe AI accelerator designed for rack-mounted servers, machine learning research, and scalable multi-card computing.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eQ2. How much memory does the Blackhole P150b have?\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eIt features 32GB GDDR6 memory with 512GB\/s memory bandwidth.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eQ3. What is the difference between Blackhole P150a and P150b?\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eBoth offer the same core specifications, including 32GB GDDR6, 120 Tensix cores, 664 TFLOPS BLOCKFP8 performance, and four 800Gbps QSFP-DD ports.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eThe primary difference is cooling: P150a uses active cooling for desktop workstations, while P150b uses passive cooling for rack-mounted servers.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eQ4. Does passive cooling mean the P150b operates without fans?\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eNo. It requires forced airflow supplied by the server chassis.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eQ5. Can the P150b be used in a desktop PC?\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eIt is not recommended for conventional desktops unless the system has a validated high-pressure forced-air cooling solution.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eQ6. Can multiple P150b cards be connected together?\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eYes. Compatible P150a and P150b cards can be interconnected using supported QSFP-DD cables.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eQ7. Does the P150b support NVIDIA CUDA?\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eNo. It uses Tenstorrent’s software ecosystem rather than NVIDIA CUDA.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eQ8. Is the P150b suitable for large language models?\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eIt can be used for supported language model workloads. Actual compatibility depends on model architecture, memory requirements, software support, and configuration.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eQ9. What is the maximum power consumption?\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eThe specified total board power is 300W.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eQ10. Is the P150b suitable for AI data centres?\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eYes, it is designed for compatible server environments and can be deployed in properly engineered AI computing infrastructure.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eBuy Tenstorrent Blackhole P150b AI Accelerator in India\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eBuild scalable AI computing infrastructure with the Tenstorrent Blackhole P150b – a high-performance, server-oriented AI accelerator designed for advanced machine learning and artificial intelligence research.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eFeaturing 32GB GDDR6 memory, 664 TFLOPS peak BLOCKFP8 compute performance, PCIe 5.0 connectivity, four 800Gbps QSFP-DD ports, and passive server cooling, the P150b is designed for organizations developing specialized AI computing environments.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eOrder Now | Request Best Price | Enquire for Availability | Contact for Enterprise \u0026amp; Bulk Orders\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eFinal pricing, availability, warranty, shipping, and compatibility are subject to confirmation.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003ePrimary keyword:\u003c\/span\u003e\u003cspan\u003e Tenstorrent Blackhole P150b India\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan\u003eSecondary keywords:\u003c\/span\u003e\u003cspan\u003e Buy Tenstorrent P150b India, Blackhole P150b Price India, Tenstorrent 32GB AI Accelerator, Passive Cooled AI Card, Server AI Accelerator India, 664 TFLOPS AI Processor, PCIe 5.0 AI Accelerator, RISC-V AI Hardware, Tenstorrent Data Centre Accelerator, QSFP-DD AI Card.\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e \u003c\/p\u003e\n\u003cp\u003e \u003c\/p\u003e","brand":"DealCliq.com","offers":[{"title":"Default Title","offer_id":68073715171627,"sku":null,"price":264999.0,"currency_code":"INR","in_stock":true}],"thumbnail_url":"\/\/cdn.shopify.com\/s\/files\/1\/0705\/8240\/3371\/files\/IMG_3263.png?v=1794395402","url":"https:\/\/dealcliq.com\/products\/tenstorrent-blackhole-p150b-ai-accelerator-32gb-gddr6-664-tflops-120-tensix-cores-passive-cooling-pcie-5-0","provider":"DealCliq.com","version":"1.0","type":"link"}