C*5
2 楼
腾讯最近从百度挖了不少人,看来是要撸起袖子大干一场了。
m*a
3 楼
能不能举例一个公司的AI计画需要多少张GPU?
c*a
5 楼
If 10,000 Tesla P100, say on a discount, 5000 each. That will be 50,000,000
m*n
9 楼
搬一段网文过来,说不定你已经看到过了。
Many companies, perhaps most, opt to access GPUs in the cloud instead of
buying and deploying the hardware directly. AI startups alone are a big
market; over 2300 investors have now funded over 1700 startups, according to
data compiled by Angel List, and the vast majority of these cash-conscious
firms use cloud based NVIDIA GPUs to develop their innovative products and
services. The exception is the world’s largest datacenters, aka the “Super
Seven” (Alibaba, Amazon, Baidu, Google, Facebook, Microsoft, and Tencent)
whose server farms probably crunch proprietary machine learning algorithms
with thousands, or even tens of thousands, of speedy GPUs, 100% of which
bear the bright green NVIDIA logo.
【在 m*******a 的大作中提到】
: 能不能举例一个公司的AI计画需要多少张GPU?
Many companies, perhaps most, opt to access GPUs in the cloud instead of
buying and deploying the hardware directly. AI startups alone are a big
market; over 2300 investors have now funded over 1700 startups, according to
data compiled by Angel List, and the vast majority of these cash-conscious
firms use cloud based NVIDIA GPUs to develop their innovative products and
services. The exception is the world’s largest datacenters, aka the “Super
Seven” (Alibaba, Amazon, Baidu, Google, Facebook, Microsoft, and Tencent)
whose server farms probably crunch proprietary machine learning algorithms
with thousands, or even tens of thousands, of speedy GPUs, 100% of which
bear the bright green NVIDIA logo.
【在 m*******a 的大作中提到】
: 能不能举例一个公司的AI计画需要多少张GPU?
l*8
11 楼
extremely overestimated.
lz的思维要是和他的骂人一样强大就好了。
很难想象受过良好教育的人,能说出,
“谁操了你妈你就认谁是爹,太猴急了”。
每次看到您的帖子,我就拿这句话讽刺您一下,
希望古板不再整天脏话。这里不都是美国phd么。
lz的思维要是和他的骂人一样强大就好了。
很难想象受过良好教育的人,能说出,
“谁操了你妈你就认谁是爹,太猴急了”。
每次看到您的帖子,我就拿这句话讽刺您一下,
希望古板不再整天脏话。这里不都是美国phd么。
P*a
13 楼
amd 完全占不了角么?
to
conscious
Super
)
【在 m*****n 的大作中提到】
: 搬一段网文过来,说不定你已经看到过了。
: Many companies, perhaps most, opt to access GPUs in the cloud instead of
: buying and deploying the hardware directly. AI startups alone are a big
: market; over 2300 investors have now funded over 1700 startups, according to
: data compiled by Angel List, and the vast majority of these cash-conscious
: firms use cloud based NVIDIA GPUs to develop their innovative products and
: services. The exception is the world’s largest datacenters, aka the “Super
: Seven” (Alibaba, Amazon, Baidu, Google, Facebook, Microsoft, and Tencent)
: whose server farms probably crunch proprietary machine learning algorithms
: with thousands, or even tens of thousands, of speedy GPUs, 100% of which
to
conscious
Super
)
【在 m*****n 的大作中提到】
: 搬一段网文过来,说不定你已经看到过了。
: Many companies, perhaps most, opt to access GPUs in the cloud instead of
: buying and deploying the hardware directly. AI startups alone are a big
: market; over 2300 investors have now funded over 1700 startups, according to
: data compiled by Angel List, and the vast majority of these cash-conscious
: firms use cloud based NVIDIA GPUs to develop their innovative products and
: services. The exception is the world’s largest datacenters, aka the “Super
: Seven” (Alibaba, Amazon, Baidu, Google, Facebook, Microsoft, and Tencent)
: whose server farms probably crunch proprietary machine learning algorithms
: with thousands, or even tens of thousands, of speedy GPUs, 100% of which
I*S
18 楼
感觉NVDA太火了,
w*e
19 楼
These datacenters use GPU for DL for themselves and customers. However
threat is looming. INTL and Google are developing DL custom chips (FPGA).
Current publications show they have performance advantage at much lower
power. So NVDA/AMD could be left out in the cold. AMD Vega will enter market
late 2017 and already signed with Google and Baba. so GPU alone can not
secure NVDA lead position. NVDA is also working on its own self-drive
solution and find several partners. In this market, it will compete with
Google,INTL.
threat is looming. INTL and Google are developing DL custom chips (FPGA).
Current publications show they have performance advantage at much lower
power. So NVDA/AMD could be left out in the cold. AMD Vega will enter market
late 2017 and already signed with Google and Baba. so GPU alone can not
secure NVDA lead position. NVDA is also working on its own self-drive
solution and find several partners. In this market, it will compete with
Google,INTL.
i*s
20 楼
我用24片 nvidia jetson tx1 做了个服务器
C*5
25 楼
老马不可能是CS PhD。
莱布妮子应该是PhD。
Guvest有可能只不过是个硕士。但是Guvest毕竟工作经验长,水平还可以(只不过没有
他自我感觉那么好)。数学功底尚可,但是对DL刚入门。假以时日也许能有小成。但是
如果继续炖鸡汤的话也困难。
老马不但不可能是CS PhD,学识在股版只能算中等偏下。
莱布妮子应该是PhD。
Guvest有可能只不过是个硕士。但是Guvest毕竟工作经验长,水平还可以(只不过没有
他自我感觉那么好)。数学功底尚可,但是对DL刚入门。假以时日也许能有小成。但是
如果继续炖鸡汤的话也困难。
老马不但不可能是CS PhD,学识在股版只能算中等偏下。
相关阅读