NVIDIA introduced its newest GPUs, the GeForce RTX 30-series, primarily based on its ‘Ampere’ structure. The corporate revealed the worth and launch date for the RTX 3070, 3080, and 3090, with the primary of the playing cards launching in simply two weeks.
In a pre-recorded presentation streamed as we speak, Nvidia revealed the GeForce RTX 3070, 3080, and 3090 GPUs primarily based on its newest ‘Ampere’ structure. Earlier than we dive into particulars, right here’s the discharge date, worth, and fundamental specs for the ‘Founder’s Version’ of every card:RTX 3090 RTX 3080 RTX 3070 Value $1,500 $700 $500 Launch Date September 24th September 17th October CUDA Cores 10,496 eight,704 5,888 Enhance Clock (GHz) 1.7 1.71 1.73 Reminiscence 24GB (GDDR6X) 10GB (GDDR6X) 8GB (GDDR6) Connectors 1x HDMI 2.1, 3x DisplayPort 1.4a
Succeeding the ‘Turing’ structure of the prior technology, ‘Ampere’ goals to ship extra efficiency throughout three pillars: shading, ray tracing (RTX), and AI (DLSS). The corporate claims the brand new graphics playing cards signify the “best generational leap” in efficiency from one sequence of Nvidia GPUs to the following.Picture courtesy NVIDIA
For video games that make use of these options, Nvidia claims as much as 2x efficiency and 1.9x energy effectivity between 20-series and 30-series playing cards. For video games that don’t use options like raytracing, the features don’t fairly attain 2x efficiency, however nonetheless supply a pleasant soar, in line with the corporate.
Nvidia additionally introduced ‘RTX IO’ a brand new accelerated storage pipeline designed to dump recreation asset knowledge decompression from the CPU to the GPU, accelerating efficiency by “as much as 100x in contrast with arduous drives and conventional storage APIs.”
“When used with Microsoft’s new DirectStorage for Home windows API, RTX IO offloads dozens of CPU cores’ price of labor to your GeForce RTX GPU, bettering body charges, enabling near-instantaneous recreation loading, and opening the door to a brand new period of huge, extremely detailed open world video games,” the corporate says.Picture courtesy NVIDIA
RTX IO would require an NVMe SSD and an RTX GPU (so far as we perceive, this contains current 20-series playing cards in addition to the upcoming 30-series).Picture courtesy NVIDIA
Whereas the primary wave of the prior RTX 20-series GPUs included USB-C ports with help for the VirtualLink normal, to this point it appears like the primary wave of RTX 30-series will eschew a USB-C port. Whether or not the piping is there for third-party producers so as to add USB-C ports (and due to this fact VirtualLink help), isn’t clear right now however we’ve reached out to Nvidia for clarification.
The corporate additionally launched a brand new know-how known as Nvidia Reflex which goals to scale back recreation latency. The characteristic must be particularly constructed into purposes, however Nvidia says “by integrating immediately with the sport, Reflex Low Latency Mode aligns recreation engine work to finish just-in-time for rendering, eliminating the GPU render queue and decreasing CPU again stress in GPU intensive scenes. This delivers latency reductions above and past current driver-only methods, similar to NVIDIA Extremely Low Latency Mode.”
Whereas the corporate is highlighting Reflex’s means to scale back latency for aggressive non-VR esports video games, it mentioned in a touch upon its web site that “in principle, it might enhance latency in VR video games, nevertheless we have now not examined it.” We’ve reached out for extra particulars.
Go to our Digital Actuality Store
Go to our sponsor Video 360 DigicamCredit score : Supply Hyperlink