Samsung has introduced all-new GDDR6 reminiscence modules which can be anticipated to be included in high-end gaming and workstation GPUs to be launched subsequent 12 months. The reminiscence chips include a switch price of 16Gbps and are capable of attain bandwidths of as much as 64 GB/s (per chip) at a mere 1.35v. To place issues into perspective, current-generation GDDR5 reminiscence discovered on a Radeon RX 480 or a GTX 1070 is just across the 8Mbps mark at finest, which suggests the promise of double the bandwidth and better clock speeds at an analogous or decrease voltage is definitely one thing to look ahead to. Density help can also be anticipated to rise to 32GB max from the 16GB commonplace in GDDR5 and GDDR5X. Speeds, in the meantime, are anticipated to leap from 12 Gbps in GDDR5X to 16 Gbps in GDDR6.
Samsung isn’t the one firm, nevertheless, that’s on monitor with its GDDR6 manufacturing schedule. SK Hynix and Micron are additionally stated to be engaged on bringing their very own GDDR6 modules to the market subsequent 12 months, with the previous really asserting its personal GDDR6 modules earlier this 12 months at Nvidia’s GTC 2017. The famous graphics chip-maker is anticipated to utilize the most recent reminiscence chips on its next-gen Volta-based graphics playing cards in 2018, changing the intermediate GDDR5X chips they’ve been utilizing on their playing cards as a stop-gap resolution.
Whereas Nvidia is confirmed to be utilizing GDDR6 for its next-gen graphics chips, there’s some confusion about AMD’s plans. The corporate has been utilizing HBM2 on its present merchandise, however the firm has been struggling due to HBM2’s manufacturing difficulties. It’s anticipated, nevertheless, that the corporate will proceed to persevere with HBM2 on their high-end Vega playing cards, though, extra pocket-friendly playing cards might nonetheless find yourself with GDDR6 reminiscence interface, which will likely be sooner in efficiency phrases and cheaper to provide. Unconfirmed on-line speculations counsel that, it may very well be one of many causes for the delay within the rollout of Vega 11.
