Showing posts with label hardware platforms. Show all posts
Google’s Making Its Own Chips Now. Time for Intel to Freak Out
The
Internet’s most powerful company sent a few shock waves through the tech world
yesterday when it revealed that a new custom-designed chip helps run what is
surely the future of its vast online empire: artificial intelligence.
In
building its own chip, Google has taken yet another step along a path that has
already remade the tech industry in enormous ways. Over the past decade, the
company has designed all sorts of new hardware for the massive data centers
that underpin its myriad online services, including computer servers,
networking gear, and more. As it created services of unprecedented scope and
size, it needed a more efficient breed of hardware to run these services. Over
the years, so many other Internet giants have followed suit, forcing a seismic
shift in the worldwide hardware market.
With
its new chip, Google’s aim is the same: unprecedented efficiency. To take AI to
new heights, it needs a chip that can do more in less time while consuming less
power. But the effect of this chip extends well beyond the Google empire. It
threatens the future of commercial chip makers like Intel and
nVidia—particularly when you consider Google’s vision for the future. According
to Urs Hölzle, the man most responsible for the global data center network that
underpins the Google empire, this new custom chip is just the first of many.
No,
Google will not sell its chips to other companies. It won’t directly compete
with Intel or nVidia. But with its massive data centers, Google is by far the
largest potential customer for both of those companies. At the same time, as
more and more businesses adopt the cloud computing services offered by Google,
they’ll be buying fewer and fewer servers (and thus chips) of their own, eating
even further into the chip market.
Indeed,
Google revealed its new chip as a way of promoting the cloud services that let
businesses and coders tap into its AI engines and build them into their own
applications. As Google tries to sell other companies on the power of its AI,
it’s claiming—in rather loud ways—that it boasts the best hardware for running
this AI, hardware that no other company has.
Google’s
Need for Speed
Google’s
new chip is called the Tensor Processing Unit, or TPU. That’s because it helps
run TensorFlow, the software engine that drives the Google’s deep neural
networks, networks of hardware and software that can learn particular tasks by
analyzing vast amounts of data. Other tech giants typically run their deep
neural nets with graphics processing units, or GPUs—chips that were originally
designed to render images for games and other graphics-heavy applications.
These are well-suited to running the types of calculations that drive deep
neural networks. But Google says it has built a chip that’s even more
efficient.
According
to Google, it tailored the TPU specifically to machine learning so that it
needs fewer transistors to run each operation. That means it can squeeze more
operations into the chip with each passing second.
For
now, Google is using both TPUs and GPUs to run its neural nets. Hölzle declined
to go into specifics on how exactly Google was using its TPUs, except to say
that they handle “part of the computation” needed to drive voice recognition on
Android phones. But he said that Google would be releasing a paper describing
the benefits of its chip and that Google will continue to design new chips that
handle machine learning in other ways. Eventually, it seems, this will push
GPUs out of the equation. “They’re already going away a little,” Hölzle says.
“The GPU is too general for machine learning. It wasn’t actually built for
that.”
That’s
not something nVidia wants to hear. As the world’s primary seller of GPUs,
nVidia is now pushing to expand its own business into the AI realm. As Hölzle
points out, the latest nVidia GPU offers a mode specifically for machine
learning. But clearly, Google wants the change to happen faster. Much faster.
The
Smartest Chip
In
the meantime, other companies, most notably Microsoft, are exploring another
breed of chip. The field-programmable gate array, or FPGA, is a chip you can
re-program to perform specific tasks. Microsoft has tested FPGAs with machine
learning, and Intel, seeing where this market was going, recently acquired a
company that sells FPGAs.
Some
analysts think that’s the smarter way to go. An FPGA provides far more
flexibility, says Patrick Moorhead, the president and principal analyst at Moor
Insights and Strategy, a firm that closely follows the chip business. Moorhead
wonders if the new Google TPU is “overkill,” pointing out that such a chip
takes at least six months to build—a long time in the incredibly competitive
marketplace in which the biggest Internet companies compete.
But
Google doesn’t want that flexibility. More than anything, it wants speed. Asked
why Google built its chip from scratch rather than using an FPGA, Hölzle said:
“It’s just much faster.”
Core
Business
Hölzle
also points out that Google’s chip doesn’t replace CPUs, the central processing
units at the heart of every computer server. The search giant still needs these
chips to run the tens of thousands of machines in its data centers, and CPUs
are Intel’s main business. Still, if Google is willing to build its own chips
just for AI, you have to wonder if it would go so far as to design its own CPUs
as well.
Hölzle
plays down the possibility. “You want to solve problems that are not solved,”
he says. In other words, CPUs are a mature technology that pretty much works as
it should. But he also said that Google wants healthy competition in the chip
market. In other words, it wants to buy from many sellers—not just, say, Intel.
After all, more competition means lower prices for Google. As Hölzle explains,
expanding its options is why Google is working with the OpenPower Foundation,
which seeks to offer chip designs that anyone can use and modify.
That’s
a powerful idea, and a potentially powerful threat to the world’s biggest chip
makers. According to Shane Rau, an analyst with research firm IDC, Google buys
about 5 percent of all server CPUs sold on Earth. Over a recent year-long
period, he says, Google bought about 1.2 million chips. And most of those
likely came from Intel. (In 2012, Intel exec Diane Bryant told WIRED that
Google bought more server chips from Intel than all but five other
companies—and those were all companies that sell servers.)
Whatever
its plans for the CPU, Google will continue to explore chips specifically
suited to machine learning. It will be several years before we really know what
works and what doesn’t. After all, neural networks are constantly evolving as
well. “We’re learning all the time,” he says. “It’s not clear to me what the
final answer is.” And as it learns, you can bet that the world’s chip makers
will be watching.
When
the movie The Terminator was released in 1984, the notion of computers
becoming self-aware seemed so futuristic that it was almost difficult to
fathom. But just 22 years later, computers are rapidly gaining the
ability to autonomously learn, predict, and adapt through the analysis
of massive datasets. And luckily for us, the result is not a nuclear
holocaust as the movie predicted, but new levels of data-driven
innovation and opportunities for competitive advantage for a variety of
enterprises and industries.
Artificial
intelligence (AI) continues to play an expanding role in the future of
high-performance computing (HPC). As machines increasingly become able to learn
and even reason in ways similar to humans, we’re getting closer to solving the
tremendously complex social problems that have always been beyond the realm of
compute. Deep learning, a branch of machine learning, uses multi-layer
artificial neural networks and data-intensive training techniques to refine
algorithms as they are exposed to more data. This process emulates the
decision-making abilities of the human brain, which until recently was the only
network that could learn and adapt based on prior experiences.
Deep
learning networks have grown so sophisticated they’ve begun to deliver even
better performance than traditional machine learning approaches. One advantage
of deep learning is that there is little need to "train" the system
and define features that might be useful for modeling and prediction. With only
basic labeling, machines can now learn these features independently as more
data is introduced to the model. Deep learning has even begun to surpass the
capabilities and speed of the human brain in many areas, including image,
speech, or text classification, natural language processing, and pattern
recognition.
The
core technologies required for deep learning are very similar to those
necessary for data-intensive computing and HPC applications. Here are a few
technologies that are well-positioned to support deep learning networks.
Multi-core
processors:
Deep
learning applications require substantial amounts of processing power, and a
critical element to the success and usability of deep learning comes with the
ability to reduce execution times. Multi-core processor architectures currently
dominate the TOP500 list of the most powerful supercomputers available today,
with 91% based on Intel processors. Multiple cores can run numerous
instructions at the same time, increasing the overall processing speed for
compute-intensive programs like deep learning, while reducing power
requirements, increasing performance, and allowing for fault tolerance.
The
Intel® Xeon Phi™ Processor, which features a whopping 72 cores, is geared
specifically for high-level HPC and deep learning. These many-core processors
can help data scientists significantly reduce training times and run a wider
variety of workloads, something that is critical to the computing requirements
of deep neural networks.
Software
frameworks and toolkits:
There
are various frameworks, libraries, and tools available today to help software
developers train and deploy deep learning networks, such as Caffe, Theano,
Torch, and the HPE Cognitive Computing Toolkit. Many of these tools are built
as resources for those new to deep learning systems, and aim to make deep
neural networks available to those that might be outside of the machine
learning community. These tools can help data scientists significantly reduce
model training times and accelerate time to value for their new deep learning
applications.
Deep
learning hardware platforms:
Not
every server can efficiently handle the compute-intensive nature of deep
learning environments. Hardware platforms that are purpose-built to handle
these requirements will offer the highest levels of performance and efficiency.
New HPE Apollo systems contain a high ratio of GPUs to CPUs in a dense 4U form
factor, which enables scientists to run deep learning algorithms faster and
more efficiently while controlling costs.
Enabling
technologies for deep learning is ushering in a new era of cognitive computing
that promises to help us solve the world’s greatest challenges with more
efficiency and speed than ever before. As these technologies become faster,
more available, and easier to implement, deep learning technologies will secure
their place in real-world applications – not in science fiction.
