PRODUCTS GPU髢「騾」陬ス蜩・/span>GPU繧ウ繝ウ繝斐Η繝シ繝・ぅ繝ウ繧ー繝槭す繝ウ

DeepLearning BOX 縺ッ譌・譛ャ縺ァ蛻昴a縺ヲ繝・ぅ繝シ繝励Λ繝シ繝九Φ繧ーGPU繝医Ξ繝シ繝九Φ繧ー繝サ繧キ繧ケ繝・Β縺ョ縲君VIDIAR DIGITSTM繧ス繝輔ヨ繧ヲ繧ァ繧「縲阪・豁」隕上Λ繧、繧サ繝ウ繧ケ繧貞叙蠕励@縲∝酔繧ス繝輔ヨ縺後・繝ャ繧、繝ウ繧ケ繝医・繝ォ縺輔l縺溘が繝シ繝ォ繧、繝ウ繝ッ繝ウDeepLearning髢狗匱繧ュ繝・ヨ縺ァ縺吶€・br> 螟ァ螟蛾屮縺励>豺ア螻、蟄ヲ鄙堤皮ゥカ縺ォ縺翫>縺ヲ縲・ォ伜柑邇・〒鬮倥>繝代ヵ繧ゥ繝シ繝槭Φ繧ケ繧堤匱謠ョ縺ァ縺阪k繧医≧縲.eepLearning BOX縺ョ驛ィ蜩√・TITAN X GPU縺ッ繧ゅ■繧阪s縺ョ縺薙→縲√Γ繝「繝ェ縺九iI/O,髮サ貅舌↓縺・◆繧九∪縺ァ縲∝・縺ヲ縺梧怙驕ゥ蛹悶&繧後※縺・∪縺吶€・br> 縺輔i縺ォG-DEP DL繝√・繝縺ァ縺ッ繝上・繝峨え繧ァ繧「縺ョ謠蝉セ帙□縺代↓縺ィ縺ゥ縺セ繧翫∪縺帙s縲りア雁ッ後↑DL繧「繝励Μ繧ア繝シ繧キ繝ァ繝ウ繧・ヤ繝シ繝ォ縺ョ髢狗匱邨碁ィ薙r蜈・↓縺励◆縲轡L繧ケ繧ソ繝シ繝医い繝・・繧オ繝昴・繝医€阪r縺皮畑諢上€ょョ牙ソ・@縺ヲ蟆主・縺・◆縺縺代k豺ア螻、蟄ヲ鄙・DeepLearning)遐皮ゥカ縺ョ繧ケ繝斐・繝峨い繝・・繝励Λ繝・ヨ繝輔か繝シ繝縺ァ縺吶€・


 

 

繝・ぅ繝シ繝励Λ繝シ繝九Φ繧ー遐皮ゥカ縺ォ譛€驕ゥ蛹悶↑繧ス繝輔ヨ繧ヲ繧ァ繧「迺ー蠅・/h4>
OS縺ッUbuntu繧偵う繝ウ繧ケ繝医・繝ォ縺励€√ョ繧」繝シ繝励Λ繝シ繝九Φ繧ーGPU繝医Ξ繝シ繝九Φ繧ー繝サ繧キ繧ケ繝・Β縺ョ縲轡IGITS繧ス繝輔ヨ繧ヲ繧ァ繧「縲・豁」隕上Λ繧、繧サ繝ウ繧ケ迚・繧偵・繝ェ繧、繝ウ繧ケ繝医・繝ォ縺励∪縺励◆縲・br /> 繝・ぅ繝シ繝励・繝九Η繝シ繝ゥ繝ォ繝サ繝阪ャ繝医Ρ繝シ繧ッ縺ョ髢狗匱縺ァ繝・・繧ソ繝サ繧オ繧、繧ィ繝ウ繝・ぅ繧ケ繝医d遐皮ゥカ閠・′蠢・ヲ√→縺吶k繧ス繝輔ヨ繧ヲ繧ァ繧「縺ッ縲√☆縺ケ縺ヲ繝励Μ繧、繝ウ繧ケ繝医・繝ォ縺輔l縺ヲ縺・∪縺吶€・br /> 豺ア螻、蟄ヲ鄙偵ヵ繝ャ繝シ繝繝ッ繝シ繧ッ縺ィ縺励※荳€逡ェ莠コ豌励・Caffe縲ゝheano縲ゝorch縲√∪縺溘€¨VIDIA縺瑚ェ・k蝣・欧縺ェGPU繧「繧ッ繧サ繝ゥ繝ャ繝シ繝・ャ繝峨・繝・ぅ繝シ繝励Λ繝シ繝九Φ繧ー繝サ繝ゥ繧、繝悶Λ繝ェ縲…uDNN 2.0繧ゅ・繝ェ繧、繝ウ繧ケ繝医・繝ォ縺輔l縺ヲ縺・∪縺吶€・br /> 縺輔i縺ォ繧ェ繝励す繝ァ繝ウ縺ァ縲∝ー主・縺九i60譌・髢薙・縲窪-mail Q&A繧オ繝昴・繝・蟆主・繧オ繝昴・繝亥性繧・蝗・縲阪→縲悟ー主・譎りィェ蝠上し繝シ繝薙せ縲阪b縺皮畑諢上€・eeplearning縺ョ繧ケ繧ソ繝シ繝医r繝輔Ν繧オ繝昴・繝医@縺セ縺吶€・br /> 縲€

 

NVIDIA縺梧署蜚ア縺吶kGPU繧ウ繝ウ繝斐Η繝シ繝・ぅ繝ウ繧ー譁ー繧ケ繧ソ繧、繝ォ

蝨ァ蛟堤噪縺ェ蜀キ蜊エ諤ァ閭ス縺ィ諡。蠑オ諤ァ繧貞y縺医◆繧ュ繝・繝シ繝悶ョ繧カ繧、繝ウ縺ョ繧キ繝」繝シ繧キ繧呈治逕ィ縲・br /> 繧ア繝シ繧ケ蜀・Κ繧貞キヲ蜿ウ縺ォ繧サ繝代Ξ繝シ繝医@縺溘ョ繝・繧「繝ォ繝・Ε繝ウ繝占ィュ險医〒辭ア貅舌↓蟇セ縺励※逶エ謗・逧・↑繧ィ繧「繝輔Ο繝シ繧剃ク弱∴縲√h繧雁柑譫懃噪縺ォ繧ア繝シ繧ケ蜀・Κ繧貞・蜊エ縺励∪縺吶€・br /> 蜉ケ邇・噪縺ェ蜷ク謗呈ー励・螳溽樟縺ォ繧医j繝上う繧ケ繝壹ャ繧ッ縺ェGPU繧ォ繝シ繝峨€隈P 100縲阪b螳牙ョ壹@縺ヲ蜍穂ス懊&縺帙k縺薙→縺悟庄閭ス縺ァ縺吶€ゅ€€

 

NVIDIA Quadro GP100 謳ュ霈・Deep Learning BOX逋コ螢イ髢句ァ・/h4>
遨カ讌オ縺ョ繧ー繝ゥ繝輔ぅ繝・け繧ケ

NVIDIA Quadro GP100 縺ッ縲・NVIDIA Pascal邃「 繧「繝シ繧ュ繝・け繝√Ε繝シ, 繧呈治逕ィ縺励◆遨カ讌オ縺ョ繧ー繝ゥ繝輔ぅ繝・け繧ケ繧ォ繝シ繝峨〒縺吶€ゅ←縺ョ繧医≧縺ェ逶ョ逧・↓菴ソ逕ィ縺吶k蝣エ蜷医〒繧ゅ€√%縺ョ逕サ譛溽噪縺ェ Quadro GP100縺後≠繧後・縲√≠繧翫∴縺ェ縺九▲縺溘%縺ィ縺碁#謌舌〒縺阪k繧医≧縺ォ縺ェ繧翫∪縺吶€・/p>

 

NVIDIA PASCAL

荳也阜譛€蜈育ォッ縺ョ GPU 繧「繝シ繧ュ繝・け繝√Ε
Pascal 繧呈治逕ィ縺励◆ GeForce GTX 10繧キ繝ェ繝シ繧コ縺ョ繧ー繝ゥ繝輔ぅ繝・け繧ケ繧ォ繝シ繝峨・縲∝燕荳紋サ」縺ョ繧ー繝ゥ繝輔ぅ繝・け繧ケ繧ォ繝シ繝峨↓豈斐∋ 譛€螟ァ3蛟阪・繝代ヵ繧ゥ繝シ繝槭Φ繧ケ繧堤匱謠ョ縺励€・擠譁ー逧・↑譁ー縺励>繧イ繝シ繝溘Φ繧ー繝・け繝弱Ο繧ク縺ィ鬟幄コ咲噪縺ォ騾イ豁ゥ縺励◆ VR 菴馴ィ薙r繧ゅ◆繧峨@縺セ縺吶€・

蝨ァ蛟堤噪縺ェ繝代ヵ繧ゥ繝シ繝槭Φ繧ケ
縺薙・ GPU 縺ォ縺ッ縲∝庄閭ス縺ェ髯舌j譛€螟ァ縺ョ閭ス蜉帙r隧ー繧∬セシ縺ソ縺セ縺励◆縲2uadro GP100縺梧政霈峨☆繧季P100縺ッ・・584蝓コ縺ョFP32逕ィCUDA Cores繧帝寔遨阪@縺檬PU繧ウ繧「縺ォ・檎キ上Γ繝「繝ェ繝舌せ蟶ッ蝓溷ケ・′720GB/s縺ォ驕斐☆繧狗ゥ榊ア、繝。繝「繝ェ謚€陦薙€粂BM2縲搾シ・igh Bandwidth Memory 2・峨r邨・∩蜷医o縺帙◆GPU縺ァ縺吶€・/p>

繧ケ繝壹ャ繧ッ
CUDA繧ウ繧「:3584
繝。繝「繝ェ縲€・・6GB HBM2
繝槭Ν繝;PU蟇セ蠢懶シ哢VLink・遺€サ譛€螟ァ2-way)
繝・ぅ繧ケ繝励Ξ繧、蜃コ蜉・1xDVI-D/4xDP 1.4
繝・ぅ繧ケ繝励Ξ繧、繧オ繝昴・繝・4x4K/120Hz/4x5K/60Hz

 

 
 
 
 
 
 
 

 

G-DEP縺ョDeepLearning繝「繝・Ν縺ェ繧牙ー主・蠕後b螳牙ソ・/h4>
DL繧ケ繧ソ繝シ繝医い繝・・繧オ繝昴・繝医・縲.L繧「繝励Μ繧ДL繝・・繝ォ縺ョ髢狗匱縺ェ縺ゥ邨碁ィ楢ア雁ッ後↑謚€陦薙せ繧ソ繝・ヵ縺・span style="line-height: 20.7999992370605px; text-align: justify;">蟆主・譎ゅ↓險ェ蝠上@縲・/span>縺雁ョ「讒倥・縺碑ヲ∵悍縺ォ蜷医o縺帙※繝輔Ξ繧ュ繧キ繝悶Ν縺ォ謚€陦灘ッセ蠢懊r陦後≧繧オ繝シ繝薙せ縺ァ縺吶€・/span>

縲€蟇セ雎。・・span style="border: 0px; outline: 0px; font-weight: inherit; font-style: inherit; vertical-align: baseline; list-style: none; margin: 0px; padding: 0px;">驛ス蜀・ソ鷹リ縺ョ莨夂、セ縲∫皮ゥカ謇€讒倥・縺ソ縺ィ縺ェ繧翫∪縺吶€・/span>窶サ蟄ヲ譬。繧・蕗閧イ讖滄未縺ク縺ョ譛ャ繧オ繝シ繝薙せ謠蝉セ帙・陦後▲縺ヲ縺翫j縺セ縺帙s縲・br /> 縲€譎る俣・・譎る俣
縲€蜀・ョケ・咼eepLearning縺ォ髢「縺吶k縺碑ウェ蝠丞・闊ャ
縲€(萓具シ・br /> 縲€繝サ貍辟カ縺ィ縺励◆DL縺ョ繧「繧、繝・ぅ繧「縺ョ逶ク隲・↓荵励▲縺ヲ谺イ縺励>
縲€繝サ邏譚宣寔繧√°繧丑N讒狗ッ峨∪縺ァ蜈ィ闊ャ逧・↑逶ク隲・↓荵励▲縺ヲ谺イ縺励>
縲€繝サ邁。蜊倥↑繧ウ繝ウ繧オ繝ォ繧偵@縺ヲ縺サ縺励>
縲€遲峨€・€∫オ碁ィ楢ア雁ッ後↑繧ケ繧ソ繝・ヵ縺後♀螳「讒倥・縺碑ヲ∵悍縺ォ譟碑サ溘↓縺雁ソ懊∴縺吶k繧オ繝シ繝薙せ縺ァ縺吶€・/p>

隧ウ縺励¥縺ッ縺雁撫縺・粋繧上○荳九&縺・€・/p>

 

 

 
 
 
 
 
 
 
 
CPU Core i7 5930K 6core/12thread 3.5GH・夲シ・urbo Boost譎よ怙螟ァ3.7GHz)
Core i7 6900X 8core/16thread 3.2GH・夲シ・urbo Boost譎よ怙螟ァ3.7GHz)
Core i7 6950X 10core/20thread 3.0GH・夲シ・urbo Boost譎よ怙螟ァ3.5GHz)
Xeon E5-1620 v4 4core/8thread 3.5GH・夲シ・urbo Boost譎よ怙螟ァ3.6GHz)
Xeon E5-1680 v4 8core/16thread 3.4GH・夲シ・urbo Boost譎よ怙螟ァ4.0GHz)
GPU GP100 16GB
GeForceGTX 1080Ti 11GB DVI
Quadro M6000
Tesla K40
譛€螟ァ4譫壽政霈会シ医げ繝ゥ繝輔ぅ繝・け繧ケ繝懊・繝峨↓萓晏ュ倥@縺セ縺呻シ・
繝槭じ繝シ繝懊・繝・/th> X99Chipset
繝。繝「繝ェ 8繧ケ繝ュ繝・ヨ
DDR4-2133/1866/1600/1333/1066MHz Quad-Channel non-ECC un-bufferedmemory 8GB/16GB/32GB/64GB
DDR4-2133/1866/1600/1333/1066MHz Quad-Channel ECC Registerd 128GB
HDD S-ATA竇。 500GB縲€・槭€€S-ATA竇。 6TB
SSD SSD intel陬ス縲€180GB MLC縲€・槭€€SSD intel陬ス縲€1.2TB MLC
RAID 繧ェ繝励す繝ァ繝ウ縲€RAID-0/1/5/6/10
蜈牙ュヲ繝峨Λ繧、繝・/th> DVD-ROM x16x52
DVDツアR/RW 2螻、蟇セ蠢懊せ繝シ繝代・繝槭Ν繝√ラ繝ゥ繧、繝・br> Blu-Ray繝峨Λ繧、繝・BD-R/RE/ROM DVDツアR/RW/DL 蟇セ蠢・
繝阪ャ繝医Ρ繝シ繧ッ Onbord Dual GigabitEther
繧ア繝シ繧ケ DeepLearning BOX 繧サ繝代Ξ繝シ繝医く繝・繝シ繝・・・400 x H480 x D520mm)
髮サ貅・/th> 1500W 80PLUS GOLD隱崎ィシ髱咎浹髮サ貅・100V蜈・蜉・GPU繝ッ繝シ繧ッ繧ケ繝・・繧キ繝ァ繝ウ逕ィ・・/td>
縺昴・莉・/th> WEB繧ォ繝。繝ゥ
譌・譛ャ隱槭く繝シ繝懊・繝・br> 繝槭え繧ケ
OS Ubuntu + CUDA + NVIDIA DIGITS Interactive Deep Learning GPU Training System
DL繧オ繝昴・繝・/th> DL繧ケ繧ソ繝シ繝医い繝・・繧オ繝シ繝薙せ・医が繝励す繝ァ繝ウ・・br> 窶サ蟇セ雎。縺ッ驛ス蜀・ソ鷹リ縺ョ莨夂、セ縲∫皮ゥカ謇€縺ョ縺ソ縺ィ縺ェ繧翫∪縺吶€りゥウ縺励¥縺ッ縺雁撫縺・粋繧上○荳九&縺・€・
菫晏ョ井ス灘宛 1蟷エ髢薙そ繝ウ繝峨ヰ繝・け譁ケ蠑擾シ域ィ呎コ厄シ・/td>
GeForceGTX 1080Ti 謳ュ霈峨Δ繝・Ν
繝「繝・Ν CPU 繝。繝「繝ェ BOOT SSD HDD GPU
・・TX1080)
萓。譬シ・育ィ手セシ縲・€∵侭霎シ・・/th>
GTX 1080Ti繝「繝・Ν1 Core i7 6900K 8core 32GB SSD 480GB 2TB 1譫・/strong> ツ・758,484
GTX 1080Ti繝「繝・Ν2 Core i7 6900K 8core 64GB SSD 480GB 4TB 2譫・/strong> ツ・933,444
GTX 1080Ti繝「繝・Ν3 Core i7 6900K 8core 64GB SSD 800GB 4TB 4譫・/strong> ツ・1,201,392
GTX 1080Ti繝「繝・Ν4 Core i7 6950X 10core 64GB SSD 480GB 4TB 1譫・/strong> ツ・926,100
GTX 1080Ti繝「繝・Ν5 Core i7 6950X 10core 64GB SSD 800GB 4TB 2譫・/strong> ツ・1,086,048
GTX 1080Ti繝「繝・Ν6 Core i7 6950X 10core 128GB SSD 800GB 6TB 4譫・/strong> ツ・1,382,400
GTX 1080Ti繝「繝・Ν7 Xeon E5-1620V4 4core 32GB SSD 480GB 2TB 1譫・/strong> ツ・633,312
GTX 1080Ti繝「繝・Ν8 Xeon E5-1620V4 4core 64GB SSD 800GB 2TB 2譫・/strong> ツ・833,112
GTX 1080Ti繝「繝・Ν9 Xeon E5-1620V4 4core 64GB SSD 800GB 2TB 4譫・/strong> ツ・1,079,028
GTX 1080Ti繝「繝・Ν10 Xeon E5-1680V4 8core 64GB SSD 800GB 4TB 1譫・/strong> ツ・997,704
GTX 1080Ti繝「繝・Ν11 Xeon E5-1680V4 8core 64GB SSD 800GB 4TB 2譫・/strong> ツ・1,105,704
GTX 1080Ti繝「繝・Ν12 Xeon E5-1680V4 8core 128GB SSD 800GB 6TB 4譫・/strong> ツ・1,402,056
窶サ繝。繝「繝ェ縺ョ螟画峩縲?DD縺ョ霑ス蜉縲∵政霈烏PU縺ェ縺ゥ螟画峩蜿ッ閭ス縺ァ縺吶・縺ァ縲∽ク玖ィ倩ヲ狗ゥ阪b繧贋セ晞シ縺九i縺雁撫縺・粋繧上○縺上□縺輔>縲・/strong>


蜷・Δ繝・Ν蜈ア騾夐・岼
繝槭じ繝シ繝懊・繝・/th> Intel X99 ExpressChipset WS逕ィ繝。繝「繝ェ繝シ128GB蟇セ蠢・/td>
蜈牙ュヲ繝峨Λ繧、繝・/th> DVDツアR/RW 2螻、蟇セ蠢懊せ繝シ繝代・繝槭Ν繝√ラ繝ゥ繧、繝・/td>
繧オ繧ヲ繝ウ繝・/th> onboard HighDefinition Audio
繝阪ャ繝医Ρ繝シ繧ッ onboard Dual GigabitEther
繧ア繝シ繧ケ DeepLearning BOX 繧サ繝代Ξ繝シ繝医く繝・繝シ繝・・・400 x H480 x D520mm)
髮サ貅・/th> 1500W 80PLUS GOLD隱崎ィシ髱咎浹髮サ貅・100V蜈・蜉・GPU繝ッ繝シ繧ッ繧ケ繝・・繧キ繝ァ繝ウ逕ィ・・/td>
繧ュ繝シ繝懊・繝・/th> 譌・譛ャ隱・Logicool 繧ケ繧ソ繝ウ繝€繝シ繝峨く繝シ繝懊・繝・K120 鮟・/td>
繝槭え繧ケ Logicool 繧ェ繝励ユ繧」繧ォ繝ォ繝槭え繧ケ M100r
WEB繧ォ繝。繝ゥ HD1080 WEB繧ォ繝。繝ゥ
OS Ubuntu
繝輔Ξ繝シ繝繝ッ繝シ繧ッ Caffe / CUDA / CuDNN / cuda-convnet2 / Chainer / DIGITS / G-DEP MCL(RC) / Python / Pycuda / Torch7 / Theano / TensorFlow
繧オ繝ウ繝励Ν繝励Ο繧ー繝ゥ繝 繝・ぅ繝シ繝励Λ繝シ繝九Φ繧ー縲€繝上Φ繝峨し繧、繝ウ隱崎ュ倥ョ繝「繝励Ο繧ー繝ゥ繝蜿翫・讒狗ッ画ク医∩繝九Η繝シ繝ゥ繝ォ繝阪ャ繝医Ρ繝シ繧ッ
DIGITS譌・譛ャ隱槭ぎ繧、繝・/th> G-DEP縲€DIGITS譌・譛ャ隱槭ぎ繧、繝・/td>
繝ェ繧ォ繝舌Μ繝シ繧、繝。繝シ繧ク 閾ェ蜍募セゥ譌ァ繝ェ繧ォ繝舌Μ繝シUSB繧ケ繝・ぅ繝・け
菫晁ィシ 1蟷エ髢薙そ繝ウ繝峨ヰ繝・け菫晁ィシ・域ィ呎コ厄シ・/td>