More Nvidia Ampere details leak – Rumoured RTX 3080 Ti detailed
More Nvidia Ampere details leak – Rumoured RTX 3080 Ti detailed
Before we start talking about Ampere, remember that rumours must always be taken with a huge grain of salt. Even the most reliable of sources can be given incorrect information, and vital details about future products could always be missing, or subject to change, until their official unveiling. The information below comes via the YouTube Channel Moore’s Law is Dead.Â
Previously, Nvidia’s next-generation Ampere architecture was said to deliver higher clock speeds and notable IPC increases when compared to Turing, delivering performance increases thanks to both clock speeds boosts and boosted performance per clock cycle. For Nvidia’s GA102 silicon, which will likely be used to create an RTX 3080 Ti graphics card, clock speeds are said to reach above 2.2GHz, representing a significant clock speed boost over Turing, especially on the high-end.Â
Nvidia’s RTX 3080 Ti is said to deliver a 40% boost in rasterisation performance when compared to today’s RTX 2080 Ti, and that is in “unfavourable games”. This means that higher performance gains are possible in other titles, with the best case likely being games with raytracing support. Most games are said to offer gamers a 50% increase in performance, with 70% boosts being possible in games where Nvidia’s architectural enhancements are best exploited.Â
The graphics card is also reportedly using 18Gbps GDDR6 memory, which offers a memory bandwidth boost of over 28.5 when compared to the RTX 2080 Ti’s 14Gbps GDDR6 memory modules. Nvidia’s RTX 3080 Ti is also rumoured to utilise a 12GB frame buffer over a 384-bit memory bus, offering an additional 32-bit memory channel over the RTX 2080 Ti’s 352-bit memory bus. This would give the RTX 3080 Ti a 40% increase in overall memory bandwidth over the RTX 2080 Ti, and an extra 1GB of VRAM. This is assuming that this information is legitimate.Â
Â
When it comes to raytracing, Nvidia’s Ampere series is said to deliver a4x increase in intersection performance per cycle over Turing, which significantly boosts the raytracing performance of Nvidia’s next-generation product offerings. This could lead to Nvidia’s RTX 20 series ageing badly in future raytracing enabled titles, though this will also partially depend on the raytracing performance of AMD’s RDNA 2 series graphics cards.Â
Nvidia’s RTX 3080 Ti is rumoured to feature a triple-fan heatsink design and dual 8-pin power inputs, support three DisplayPort outputs, a single USB Type-C (VirtualLink) output and a single HDMI 2.1 connection.Â
With the release of Ampere, Nvidia is also reportedly merging its Geforce Now applications with its Geforce Experience app. Nvidia will also be removing its mandatory login restriction for Geforce Experience users.Â
Ampere is also said to be offering gamers a 2x increase in tensor cores per SM, giving Nvidia a 2x boost in AI performance. This will be used to help push AI-based technologies like DLSS and RTX voice forward, and allow additional AI technologies to enter the gaming market.Â
Nvidia is said to be working to release Ampere as quickly as possible, as launching before AMD’s RDNA 2 architecture will help the company secure sales. With Ampere, Nvidia is also reportedly making its RTX feature set a top-to-bottom addition to Geforce graphics cards. This should help turn raytracing into a mandatory addition for future PC games.Â
Rumour has it that Nvidia will reveal Ampere’s enterprise/HPC iteration this week at Nvidia’s GTC keynote and launch its Geforce interactions in September. At this time it is unknown whether or not this will be a “paper launch”.Â
If these rumours are true, Ampere will represent a major shift for Nvidia, offering huge performance increases alongside a significant boost in performance per watt. AI and raytracing will also be secured as part of the future of PC gaming, especially if DLSS 3.0 is as good as rumoured. Â
You can join the discussion on Nvidia’s rumoured RTX 3080 Ti graphics card on the OC3D Forums.Â