Micron
Micron has share some public presentation figures for its next - gen graphic memory , and if these numbers turn out to be true , it could intend great thing for upcominggraphics cards .
The company claims that its GDDR7 VRAM will offer up to 30 % execution improvement in gaming scenario , and this reportedly applies both to games that rely heavily on ray tracing and on pure rasterization . Will Nvidia’sRTX 50 - series , which is said to be using GDDR7 retentivity , wrench out to be a bigger upgrade than await ?
Micron
In the data , which was first partake byharukaze5719 on X(formerly Twitter ) , Micron pits its new GDDR7 32Gb / s modules against its GDDR6 VRAM that max out at 20GB / s. For comparison , GDDR6X memory can go up to 24GB / s per stick , so it ’s a grown upgrade no matter what , but it ’s worth mention that GDDR7 retentivity will begin at 28Gb / s.
Micron also lists a few of the fundamental benefits of its GDDR7 tech . For starters , the bandwidth is increase by up to 60 % if you compare the top GDDR6 and GDDR7 modules , with GDDR7 being able-bodied to accomplish over 1.5TB / s in system bandwidth .
Despite these telling gains , Micron also reports that GDDR7 is up to 50 % more efficient in term of power consumption , and its unexampled sleep modes can massively disregard back on the amount of power consumed while on standby — up to 70 % . Micron also managed to reduce response time by up to 20 % , which will be helpful for reproductive AI workload .
micron , 30 % FPS improve
28Gbps : https://t.co / hj9RtqyraN32Gbps : https://t.co / sWJAe1oDnGpic.twitter.com / lsToW4nwr9
& mdash ; 포시포시 ( @harukaze5719)June 29 , 2024
With first - political party benchmark like this , it ’s important to always take them with a footling bit of incredulity . However , some numbers are laborious to challenge . While we ca n’t be indisputable how universal that 30 % improvement in frames per 2nd ( fps ) is live on to be , we do know for a fact that the firm GDDR7 remembering will volunteer a massive growth in retentivity bandwidth .
Micron ’s comparability match up a setup with 12 GDDR7 integrated circuits ( ICs ) and a 384 - bit computer memory interface against an exact friction match from the previous generation , and that bring the bandwidth over the 1.5TB / s line . Meanwhile , the RTX 4090 , equipped with faster GDDR6X modules that polish off up to 21Gbps , maxes out at a little over 1TB / s. The gains are emphatically there .
This sounds like good intelligence for Nvidia . late hearsay about thespecs for the RTX 50 - series leave behind a lot to be desired , with a like huge improvement in the RTX 5090 and significantly smaller boost across the lower end of the product quite a little . But with this much faster memory , we might see GPUs that share similar CUDA core counts still shoot ahead in Federal Protective Service just because of the faster VRAM .
Of course , that 30 % flesh is n’t go to apply to every GPU . A minute memory bus topology can stifle even the best VRAM , as we ’ve seen in GPUs like theRTX 4060 Ti 16 GB . In the RTX 40 - series , Nvidia kept the amount of VRAM moderately buttoned-down for most of the graphics cards , and the memory port was the same way . With that same approach in the RTX 50 - serial publication , GDDR7 memory will still be an improvement , but that 30 % may be earmark for GPUs like the RTX 5090 .
Micron ’s new memory module sound pretty amazing , and with any chance , we ’ll be seeing them this way in Nvidia ’s upcoming graphics card . AMD is said to be vex to GDDR6 inRDNA 4 . This could terminate up giving Nvidia an edge over its competitor in more ways than one .
