Architecture:NVIDIA Blackwell Architecture Process Size:4N NVIDIA Custom Process Transistors:92.2 bilion Die Size:750 mm2 CUDA Parallel Processing cores:24,064 NVIDIA Tensor Cores:752 NVIDIA RT Cores:188 Single-Precision Performance1:125 TFLOPS Al Performance1:4,000 Al TOPS2 RT Core Performance1:380 TFLOPS GPU Memory:96 GB GDDR7 with ECC Memory Interface:512-bit Memory Bandwidth:1,792 GB/s Max Power Consumption:600W Multi-Instance GPU: Up to 4x 24GB Up to 2x 48GB Up to lx 96GB Graphics Bus:PCI Express 5.0 x 16 Display Connectors:DP 2.1 (4) Form Factor:137mm(H) x 304mm(L), XHFL Dual Slot Product Weight:1.950 kg Thermal Solution:Double Flow Through NVIDIA(R) 3D Vision(R) and 3D Vision Pro:Support via3 -pin mini DIN Frame Lock:Compatible (with NVIDIA RTX PRO Sync) Power Connector:1x PCle CEMS 16-pin NVENC | NVOEC | JPEG:4x | 4x | 4x
6000 Blackwell Max-Q Workstation Edition
Architecture:NVIDIA Blackwell Architecture Process Size:4N NVIDIA Custom Process Transistors:92.2 bilion Die Size:750 mm2 CUDA Parallel Processing cores:24,064 NVIDIA Tensor Cores:752 NVIDIA RT Cores:188 Single-Precision Performance1:110 TFLOPS Al Performance1:3,511 Al TOPS2 RT Core Performance1:333 TFLOPS GPU Memory:96 GB GDDR7 with ECC Memory Interface:512-bit Memory Bandwidth:1,792 GB/s Max Power Consumption:300W Multi-Instance GPU: Up to 4x 24GB Up to 2x 48GB Up to lx 96GB Graphics Bus:PCI Express 5.0 x 16 Display Connectors:DP 2.1 (4) Form Factor:112mm(H) x 1267mm(L), XHFL Dual Slot Product Weight:1.23 kg Thermal Solution:Blower Active Fan NVIDIA(R) 3D Vision(R) and 3D Vision Pro:Support via3 -pin mini DIN Frame Lock:Compatible (with NVIDIA RTX PRO Sync) Power Connector:1x PCle CEMS 16-pin NVENC | NVOEC | JPEG:4x | 4x | 4x
Architecture:NVIDIA Blackwell Architecture Process Size:4N NVIDIA Custom Process Transistors:92.2 bilion Die Size:750 mm2 CUDA Parallel Processing cores:12,800 NVIDIA Tensor Cores:400 NVIDIA RT Cores:100 GPU Memory:48 GB GDDR7 with ECC Memory Interface:384-bit Memory Bandwidth:1,344 GB/s Max Power Consumption:300W Multi-Instance GPU: Up to 4x 24GB Up to 2x 48GB Graphics Bus:PCI Express 5.0 x 16 Display Connectors:DP 2.1 (4) Form Factor:112mm(H) x 267mm(L), XHFL Dual Slot Product Weight:1.219 kg Thermal Solution:Blower Active Fan NVIDIA(R) 3D Vision(R) and 3D Vision Pro:Support via3 -pin mini DIN Frame Lock:Compatible (with NVIDIA RTX PRO Sync) Power Connector:1x PCle CEMS 16-pin NVENC | NVOEC | JPEG:3x | 3x | 1x
NVIDIA RTX PRO(TM)4000 Blackwell
Architecture:NVIDIA Blackwell Architecture Process Size:4N NVIDIA Custom Process Transistors:45.6 bilion Die Size:378 mm2 CUDA Parallel Processing cores:8,960 NVIDIA Tensor Cores:280 NVIDIA RT Cores:70 GPU Memory:24 GB GDDR7 with ECC Memory Interface:192-bit Memory Bandwidth:- Max Power Consumption:140W Multi-Instance GPU: - - Graphics Bus:PCI Express 5.0 x 16 Display Connectors:DP 2.1 (4) Form Factor:112mm(H) x 241mm(L), XHFL single Slot Product Weight:0.554 kg Thermal Solution:Blower Active Fan NVIDIA(R) 3D Vision(R) and 3D Vision Pro:Support via3 -pin mini DIN Frame Lock:Compatible (with NVIDIA RTX PRO Sync) Power Connector:1x PCle CEMS 16-pin NVENC | NVOEC | JPEG:2x |2x | 1x
■ 株式会社ジーデップ・アドバンスについて 株式会社ジーデップ・アドバンス(東証スタンダード、証券コード:5885)は、「Advance with you」をミッションに、GPGPUをはじめとするアクセラレーターやハイエンドワークステーション、広帯域ネットワークや高速ストレージを用いたクラスターシステム、さらにライブラリやコンパイラ、ジョブスケジューラなどの運用ツールの提供構築から運用支援まで、仕事や研究を前に進めるための手段をオンプレミスやレンタル、クラウドなどあらゆる形態で総合的に提供するAIとビジュアライゼーションのソリューションプロバイダです。NVIDIA社の「NPN(NVIDIA Partner Network)」においてエリートパートナーとして活動しています。 URL:https://www.gdep.co.jp/