汉字News

← 汉字NewsChinese News for Learners
technology
kē jì
科技
·HSK 4

OpenAI (AI company)
OpenAI
OpenAI
and / together with
yǔ
与
Broadcom (semiconductor company)
Bótōng
博通
jointly / in partnership
liánhé
联合
to release / to announce
fābù
发布
the first / the debut model
shǒu kuǎn
首款
self-developed / proprietary
zì yán
自研
AI (artificial intelligence)
AI
AI
inference / reasoning
tuīlǐ
推理
chip / semiconductor
xīnpiàn
芯片
Jalapeño (chip name)
Jalapeño
Jalapeño
。

OpenAI (AI company)
OpenAI
OpenAI
and
hé
和
Broadcom (semiconductor company)
Bótōng
博通
on / at (formal)
yú
于
June 24th
6 yuè 24 rì
6月24日
to release / to announce
fābù
发布
(completion particle)
le
了
the first / debut model
shǒu kuǎn
首款
self-developed / proprietary
zì yán
自研
AI (artificial intelligence)
AI
AI
inference / reasoning
tuīlǐ
推理
chip / semiconductor
xīnpiàn
芯片
，
named / called
míng wéi
名为
Jalapeño (chip code name)
Jalapeño
Jalapeño
。
this / the said
gāi
该
chip / semiconductor
xīnpiàn
芯片
specifically designed for
zhuān wéi
专为
large language model (LLM)
dà yǔyán móxíng
大语言模型
inference / reasoning
tuīlǐ
推理
to optimize / optimized
yōuhuà
优化
，
compared with / and
yǔ
与
mainstream / dominant
zhǔliú
主流
GPU (graphics processing unit)
GPU
GPU
in comparison / compared to
xiāngbǐ
相比
cost / expense
chéngběn
成本
to reduce / to lower
jiàngdī
降低
approximately / about
yuē
约
50 percent
50%
50%
。

June 25, 2026 at 2:00 AM

OpenAI (AI company)
OpenAI
OpenAI
and / together with
yǔ
与
Broadcom (semiconductor company)
Bótōng
博通
on / at (formal)
yú
于
June 24th
6 yuè 24 rì
6月24日
jointly / together
liánhé
联合
to release / to announce
fābù
发布
(completion particle)
le
了
the first / debut model
shǒu kuǎn
首款
self-developed / proprietary
zì yán
自研
AI (artificial intelligence)
AI
AI
inference / reasoning
tuīlǐ
推理
chip / semiconductor
xīnpiàn
芯片
，
code name / designation
dàihào
代号
Jalapeño (chip code name)
Jalapeño
Jalapeño
。
this / the said
gāi
该
chip / semiconductor
xīnpiàn
芯片
specifically / exclusively
zhuānmén
专门
targeting / aimed at
zhēnduì
针对
large language model (LLM)
dà yǔyán móxíng
大语言模型
(possessive/modifier particle)
de
的
inference / reasoning
tuīlǐ
推理
task / mission
rènwù
任务
to carry out / to perform
jìnxíng
进行
to optimize / optimized
yōuhuà
优化
。
Broadcom (semiconductor company)
Bótōng
博通
CEO / Chief Executive Officer
shǒuxí zhíxíng guān
首席执行官
Hock Tan (Broadcom CEO)
Huòkè Tán
霍克·谭
to state / to indicate
biǎoshì
表示
，
Jalapeño (chip name)
Jalapeño
Jalapeño
compared with / and
yǔ
与
mainstream / dominant
zhǔliú
主流
GPU (graphics processing unit)
GPU
GPU
in comparison / compared to
xiāngbǐ
相比
cost / expense
chéngběn
成本
can / able to
kě
可
to reduce / to lower
jiàngdī
降低
approximately / about
yuē
约
50 percent
50%
50%
。
in terms of / regarding
zài
在
energy efficiency / performance per watt
néng xiào
能效
aspect / dimension
fāngmiàn
方面
，
early-stage / initial
zǎoqī
早期
testing / evaluation
cèshì
测试
to show / to indicate
xiǎnshì
显示
，
Jalapeño (chip name)
Jalapeño
Jalapeño
(possessive/modifier particle)
de
的
per watt
měi wǎ
每瓦
performance / capability
xìngnéng
性能
significantly / by a large margin
dàfú
大幅
to surpass / better than
yōuyú
优于
current / present-day
dāngqián
当前
state-of-the-art / most advanced
zuì xiānjìn
最先进
(possessive/modifier particle)
de
的
of the same kind / comparable
tónglèi
同类
product / device
chǎnpǐn
产品
。
this time / on this occasion
cǐ cì
此次
collaboration / partnership
hézuò
合作
from
cóng
从
initial / at the outset
zuìchū
最初
design / blueprint
shèjì
设计
to / until
dào
到
tape-out / chip fabrication sign-off
liú piàn
流片
only / merely
jǐn
仅
to take (time) / duration of
lìshí
历时
nine months
jiǔ gè yuè
九个月
，
to set (a record)
chuàng xià
创下
high-end / premium
gāoduān
高端
semiconductor
bàndǎotǐ
半导体
industry / sector
hángyè
行业
fastest / most rapid
zuì kuài
最快
research and development / R&D
yánfā
研发
record
jìlù
纪录
one of
zhī yī
之一
。
OpenAI (AI company)
OpenAI
OpenAI
to state / to indicate
biǎoshì
表示
，
under its umbrella / affiliated
qíxià
旗下
(possessive/modifier particle)
de
的
ChatGPT (AI chatbot)
ChatGPT
ChatGPT
、
Codex (AI coding tool)
Codex
Codex
and / as well as
jí
及
API (application programming interface)
API
API
and so on / etc.
děng
等
product / service
chǎnpǐn
产品
(possessive/modifier particle)
de
的
daily / routine
rìcháng
日常
to run / to operate
yùnxíng
运行
demand / requirement
xūqiú
需求
all / each
jūn
均
served as / constituted
wéi
为
this / the said
gāi
该
chip / semiconductor
xīnpiàn
芯片
(possessive/modifier particle)
de
的
design / blueprint
shèjì
设计
to provide / to supply
tígōng
提供
(completion particle)
le
了
basis / grounds
yījù
依据
。
Jalapeño (chip name)
Jalapeño
Jalapeño
is / serves as
shì
是
the two companies
liǎng jiā
两家
currently / in the process of
zhèngzài
正在
jointly / together
gòngtóng
共同
to build / to develop
gòujiàn
构建
(possessive/modifier particle)
de
的
multi-generation / across generations
duō dài
多代
computing / calculation
jìsuàn
计算
platform / infrastructure
píngtái
平台
(possessive/modifier particle)
de
的
the first step
dì yī bù
第一步
，
expected / anticipated
yùjì
预计
in / at (formal)
yú
于
end of 2026
2026 nián dǐ
2026年底
before / prior to
qián
前
to put into / to deploy
tóurù
投入
deployment / rollout
bùshǔ
部署
。

News in English

Science & Technology

OpenAI and Broadcom Unveil First Proprietary AI Inference Chip Jalapeno

OpenAI and Broadcom unveiled Jalapeño, their first jointly developed AI inference chip, on June 24. The chip is optimized for large language model inference and promises roughly 50 percent cost savings compared with mainstream GPUs.

More in Science & Technology

United States
měiguó
美国
Congress
guóhuì
国会
approves
pīzhǔn
批准
artificial intelligence
réngōng zhìnéng
人工智能
regulation
jiānguǎn
监管
framework
kuàngjià
框架
。

US Congress Approves Comprehensive AI Regulation Framework

OpenAI (AI company)
OpenAI
OpenAI
to launch
tuīchū
推出
GPT-6 (AI model)
GPT-6
GPT-6
and
yǔ
与
intelligent
zhìnéng
智能
interface
jièmiàn
界面
feature
gōngnéng
功能
。

OpenAI Launches GPT-6 With Intelligent Interface Features

Microsoft
Wēiruǎn
微软
to launch / to release
tuīchū
推出
Work IQ (Microsoft enterprise AI tool)
Work IQ
Work IQ
，
to let / to allow
ràng
让
Copilot (Microsoft AI assistant)
Copilot
Copilot
to connect to
liánjiē
连接
enterprise data
qǐyè shùjù
企业数据
。

Microsoft Launches Work IQ, Letting Copilot Connect to Enterprise Data

Nous Research (AI research company)
Nous Research
Nous Research
to complete
wánchéng
完成
$90 million
jiǔ qiān wàn měiyuán
九千万美元
to raise funding
róngzī
融资
，
Hermes (open-source AI model)
Hermes
Hermes
to advance into / to enter
jìnjūn
进军
enterprise market
qǐyè shìchǎng
企业市场
。

Nous Research Raises $90 Million, Hermes Enters Enterprise Market