汉字News

← 汉字News
Chinese News for Learners
technology
kē jì
科技
·HSK 2

Google
Gǔgē
谷歌
to release / to launch
fābù
发布
eighth generation
dì bā dài
第八代
artificial intelligence
AI
AI
chip / microchip
xīnpiàn
芯片
,
to challenge / to compete with
tiǎozhàn
挑战
Nvidia (semiconductor company)
Yīngwěidá
英伟达
。

Google
Gǔgē
谷歌
at / in
zài
在
cloud computing
yún jìsuàn
云计算
conference / summit
dàhuì
大会
at / on (an event)
shàng
上
released / announced
fābù le
发布了
eighth generation
dì bā dài
第八代
artificial intelligence
AI
AI
chip / microchip
xīnpiàn
芯片
。
new
xīn
新
chip / microchip
xīnpiàn
芯片
divided into / split into
fēn wéi
分为
training / model training
xùnliàn
训练
and
hé
和
inference / reasoning
tuīlǐ
推理
two types / two kinds
liǎng zhǒng
两种
,
performance / capability
xìngnéng
性能
significantly / by a large margin
dàfú
大幅
improved / enhanced
tíshēng
提升
。

April 23, 2026 at 10:30 AM

Google
Gǔgē
谷歌
at / in
zài
在
annual / yearly
niándù
年度
cloud computing
yún jìsuàn
云计算
conference / summit
dàhuì
大会
at (an event)
shàng
上
released / announced
fābù le
发布了
eighth generation
dì bā dài
第八代
artificial intelligence
AI
AI
chip / microchip
xīnpiàn
芯片
。
new
xīn
新
chip / microchip
xīnpiàn
芯片
divided into / split into
fēn wéi
分为
two types / two kinds
liǎng zhǒng
两种
:
TPU 8t (AI training chip)
TPU 8t
TPU 8t
used for / designed for
yòng yú
用于
model / AI model
móxíng
模型
training / to train
xùnliàn
训练
,
TPU 8i (AI inference chip)
TPU 8i
TPU 8i
used for / designed for
yòng yú
用于
inference / reasoning
tuīlǐ
推理
。
TPU 8t (AI training chip)
TPU 8t
TPU 8t
structural particle
de
的
performance / capability
xìngnéng
性能
compared to / than
bǐ
比
previous generation
shàng yī dài
上一代
improved / increased by
tígāo le
提高了
nearly / almost
jìn
近
three times / threefold
sān bèi
三倍
。
TPU 8i (AI inference chip)
TPU 8i
TPU 8i
structural particle
de
的
inference speed
tuīlǐ sùdù
推理速度
compared to / than
bǐ
比
previous generation
shàng yī dài
上一代
faster by
kuài le
快了
80 percent / eighty percent
bǎifēn zhī bāshí
百分之八十
。
Google
Gǔgē
谷歌
to state / to indicate
biǎoshì
表示
,
these / this
zhè
这
two models / two products
liǎng kuǎn
两款
chip / microchip
xīnpiàn
芯片
will / shall (future marker)
jiāng
将
at / in (formal)
yú
于
this year
jīnnián
今年
later this year / at a later time
wǎn xiē shíhòu
晚些时候
to / for
xiàng
向
users / customers
yònghù
用户
to open / to make available
kāifàng
开放
。
this
zhè
这
is / are
shì
是
Google
Gǔgē
谷歌
in / within
zài
在
artificial intelligence
AI
AI
chip / microchip
xīnpiàn
芯片
market
shìchǎng
市场
in (a field / domain)
shàng
上
to challenge / to compete with
tiǎozhàn
挑战
Nvidia (semiconductor company)
Yīngwěidá
英伟达
structural particle
de
的
important / significant
zhòngyào
重要
a step / a move
yī bù
一步
。

New Words

Gǔgē
谷歌
Google
xīnpiàn
芯片
chip
xùnliàn
训练
training
tuīlǐ
推理
inference
xìngnéng
性能
performance
bèi
倍
times / fold
tuīlǐ sùdù
推理速度
inference speed
bǎifēnbǐ
百分比
percentage
yònghù
用户
users
tiǎozhàn
挑战
to challenge

News in English

Science & Technology

Google Releases Eighth Generation AI Chips to Challenge Nvidia

Google unveiled its eighth-generation AI chips at Google Cloud Next, split into a training chip and an inference chip. The new lineup delivers up to three times faster AI model training and 80 percent better inference performance.

Google released the eighth generation AI chip at its annual cloud computing conference.

The new chip is divided into two types: TPU 8t for model training and TPU 8i for inference.

The performance of TPU 8t improved nearly threefold compared to the previous generation.

The inference speed of TPU 8i is 80 percent faster than the previous generation.

Google indicated that these two chips will be made available to users later this year.

This is an important step for Google to challenge Nvidia in the AI chip market.

More in Science & Technology

OpenAI (AI company)
OpenAI
OpenAI
to announce / to declare
xuānbù
宣布
to postpone / to delay
tuīchí
推迟
IPO / to go public
shàngshì
上市
plan
jìhuà
计划
,
CEO / Chief Executive Officer
shǒuxí zhíxíng guān
首席执行官
stated / said
chēng
称
now / at this time
xiànzài
现在
timing / moment
shíjī
时机
inappropriate / ill-advised
bùdàng
不当
。

OpenAI Delays IPO Plans as CEO Sam Altman Cites AI Safety Concerns

Apple Inc.
píngguǒ gōngsī
苹果公司
officially / formally
zhèngshì
正式
to release / to launch
fābù
发布
iPhone 18 Pro (Apple flagship smartphone)
iPhone 18 Pro
iPhone 18 Pro
and / simultaneously
bìng
并
to roll out / to introduce
tuīchū
推出
iOS 27 (Apple mobile operating system)
iOS 27
iOS 27
operating system
xìtǒng
系统
update / upgrade
gēngxīn
更新
。

Apple Officially Launches iPhone 18 Pro and Rolls Out iOS 27 Update

Beijing
běijīng
北京
respond
huíyìng
回应
Anthropic
Anthropic
Anthropic
CEO
CEO
CEO
artificial intelligence
AI
AI
development
fāzhǎn
发展
。

Beijing Responds to Anthropic CEO Remarks on China AI Development

United States
měiguó
美国
investigation
diàochá
调查
indicate
zhǐ
指
China
zhōngguó
中国
government
zhèngfǔ
政府
participate
cānyù
参与
artificial intelligence
AI
AI
technology
jìshù
技术
theft
dàoqiè
盗窃
。

US Investigators Link China to AI Technology Theft