Micron 24 Gbps GDDR6X memory for GeForce RTX 40 series is now in production 【搬运外媒VedioCardz报道（手工翻译）】

向晚大魔王awa

于 2022-08-09 02:24:22 发布

阅读量704

点赞数

分类专栏：外媒芯片资讯（手工英译）文章标签：电脑硬件

本文链接：https://blog.csdn.net/lolloli/article/details/126239466

版权

外媒芯片资讯（手工英译）专栏收录该内容

11 篇文章 0 订阅

订阅专栏

Micron 24 Gbps GDDR6X memory for GeForce RTX 40 series is now in production

（英伟达GeForce RTX 40系列显卡所用的镁光24Gbps 显存速率的 GDDR6X显存现已投入生产）

Micron 24Gbps GDDR6X in production

（镁光24Gbps 显存速率的 GDDR6X显存现已投入生产）

The memory manufacturer Micron, who had developed GDDR6X technology in collaboration with NVIDIA for RTX 30, is now preparing for the next-gen GPU series by offering even faster speeds.

（曾与英伟达RTX 30系列显卡合作研发的GDDR6X技术的存储器制造商镁光，现在正研发更快速率的显存为下一代显卡做准备）

在这里插入图片描述

GDDR6X Memory Reimagined
Micron’s GDDR6X memory uses innovative signaling technology to double the data rate, delivering unprecedented graphics memory
performance to feed the most data-hungry workloads.
（难以置信的GDDR6X显存
镁光的GDDR6X显存使用最新的信号技术将显存速率提升一倍，提供前所未有的显存性能去满足最严重的工作负载。）
Feed The Beast
GDDR6X feeds the most data-hungry workloads, such as gaming, professional visualization and Al inference.
（“喂饱这头野兽”
GDDR6X显存满足了最严重的工作负载，例如游戏，专业可视化以及人工智能推理。）

Company website has now been updated with new information on the new GDDR6X modules. It appears that 16Gb (2GB) modules featuring 24 Gbps speed are now in production, just 2 months ahead of the rumored launch of NVIDIA RTX 40 series.
（公司网站现已更新了关于GDDR6X模块的新信息。看起来单颗2G速率为24Gbps的显存模块已投入生产，比传闻中的英伟达RTX 40系列发布早两个月）

NVIDIA is said to use both 21 and 24 Gbps speeds for its high-end series. Should rumors be correct, NVIDIA will be using 21 Gbps GDDR6X modules for three of the high-end SKUs: RTX 4090, 4080 and 4070. At the same time, rumors on supposed flagship model, either a professional workstation GPU or a true TITAN successor, might be getting 24 Gbps modules.
（据说英伟达会在高端型号同时使用21和24Gbps速率的显存。如果传言属实，英伟达将为三款高端型号（RTX 4090，4080，4070）使用21Gpbs速率的显存，且无论是专业工作站显卡或下一代泰坦，都可能获得24Gbps的显存速率。）

在这里插入图片描述

Micron probably wouldn’t be producing such chips for no reason, but just because such memory exists, it does not guarantee that full speed will be used by next-gen GPUs. For RTX 30 series, NVIDIA only used 21 Gbps speed for its RTX 3090 Ti model released in March this year, while the original 2020 launch only went as high as 19.5 Gbps.
（镁光可能不会无缘无故地生产这样的芯片，但它确确实实存在，且并不能确保下一代显卡会使用最高速率的显存。对于RTX 30系列显卡，英伟达仅仅在今年三月发布的RTX 3090 TI上使用的21 Gbps速率的显存，而在2020年发布的显卡最高显存速率仅19.5Gbps（3090））

在这里插入图片描述
RUMORED NVIDIA GeFore RTX 40 Series Specs 传闻英伟达GeFore RTX40系列显卡规格
Architecture 架构
GPU 核心代号
Board Number 板号
SMs SM单元数量
CUDA cores CUDA核心数量
Base Clock 基础频率
Boost Clock 加速频率
Memory 显存
Memory Bus 显存位宽
Memory Speed 显存速率
BandWidth 显存带宽
TGP 整卡功耗
3DMark Time Extreme 一款跑分软件
Lunch Data 发售日期
TBC（ to be confirm ）待确认