IBM powers fastest Summit supercomputer
Smallest, has NV100 and Power9 combo
It’s been a while since China had the fastest supercomputer and all thathas just changed with the launch of US government owned Summit computer that is currently the world’s most powerful and smartest artificial intelligence (AI) supercomputer.
Intel Xeon AWS AI can read 4X faster than Nvidia
Software optimizations to blame
There's an unexpected twist in the tangled tale as Intel managed to prove that for machine translation, which uses RNNs, the Intel Xeon Scalable processor outperforms NVidia V100 based systems by four times.
Nvidia Xavier sampling in Q1 18
Volta graphics, eight custom cores for automotive
Nvidia CEO announced that Drive Xavier world's first autonomous machine processor is up and running as the silicon got back from the Fab two weeks ago, and of course, it works flawlessly.
Nvidia pulls Titan V out of its hat
Fully enabled GV100 Volta GPU for US $2,999
Nvidia has rather surprisingly announced and launched the new prosumer Nvidia Titan V graphics card at the Neural Information Processing Systems 2017 conference, which is based on a fully-enabled GV100 GPU and available for $2,999.
Uber orders 24,000 Volvo XC90s
Self-driving Nvidia powered fleet
Uber just ordered up to 24,000 Volvo XX90 self-driving cars for its driverless, hail ride/share - post taxi fleet.
Nvidia Tensor RT 3 speeds up inferencing by 100X
4.8 FPS on CPU to 550 FPS on Volta
Nvidia has come up with a new optimized complier called Tensor RT 3 and demonstrated how Volta obliterates a modern server CPU.
Nvidia has record Q2 quarter
Stocks down as it misses on automotive
Nvidia reported record revenue for the second quarter ended July 30, 2017, of $2.23 billion, up 56 percent from $1.43 billion a year earlier, and up 15 percent from $1.94 billion in the previous quarter.
Nvidia just started shipping Volta
Going faster than last year
Despite the fact that Jensen introduced the Volta architecture and the V100 GPU, the company has only just managed to ship the first batch of products.
New Geforce will be incremental
Clock is tocking not ticking
The upcoming Geforce GPU architecture will be more incremental than revolutionary as Nvidia appears to be using Intel’s tick-tock strategy.
Nvidia officially announces the Tesla V100 card
12nm FinFET, PCIe interface, and 5120 CUDA cores
After talking about the new Volta GPU at the GPU Technology Conference earlier last month, Nvidia has now officially unveiled the PCIe-based Tesla V100 GPU accelerator, based on the GV100 GPU and offering 14 TFLOPs of single-precision compute performance.