Slashdot Mirror
IBM Discovery May Lead To Exascale Supercomputers
alphadogg writes "IBM researchers have made a breakthrough in using pulses of light to accelerate data transfer between chips, something they say could boost the performance of supercomputers by more than a thousand times. The new technology, called CMOS Integrated Silicon Nanophotonics, integrates electrical and optical modules on a single piece of silicon, allowing electrical signals created at the transistor level to be converted into pulses of light that allow chips to communicate at faster speeds, said Will Green, silicon photonics research scientist at IBM. The technology could lead to massive advances in the power of supercomputers, according to IBM."
...that the metal connections between individual components would not be fast enough.
I only wonder how long before this sort of technology makes its way to the consumer market, if only for show. Of course I can't see a use for an exascale databus on the mobo anytime soon.
GPUs are indeed an inexpensive way to boost speed in some cases. But they have been rather oversold; while some specific types of problems benefit a lot from them, many problems do not. If you need to frequently share data with other computing nodes (neural network simulations come to mind), then the communications latency between card and main node eats up much of the speed increase. And as much of the software you run on this kind of system is customized or one-off stuff, the added development time in using GPUs is a real factor in determining the relative value. If you gain two weeks of simulation time but spend an extra month on the programming, you're losing time, not gaining it.
Think about it this way: GPU's are really the same thing as specialized vector processors, long used in supercomputing. And they have fallen in and out of favour over the years depending on the kind of problem you try to solve, the relative speed boost, cost and difficulty in using them. The GPU resource at the computing center is used much less than the general clusters themselves, indicating most users do not find it worth the extra time and trouble to use.
It is a good idea, but it's not some magic bullet.
Trust the Computer. The Computer is your friend.
We have reached an informational threshold which can only be crossed by harnessing the speed of light directly. The quickest computations require the fastest possible particles moving along the shortest paths. Since the capability now exists to take our information directly from photons travelling molecular distances, the final act of the information revolution will soon be upon us.
-- Academician Prokhor Zakharov, "For I Have Tasted The Fruit"
Now I just need room temperature superconductors to build my gatling laser speeders.
Exascale is not a word
A whole dictionary full of perfectly good words and they have to make one up to mean “very large”...
Exascale is a perfectly cromulent word.
IBM may be patent-happy, but it's only reasonable to protect their "inventions". There's a huge difference between a patent troll who buys patents solely for litigation purposes, and IBM, who has been among the leading tech innovators for decades, defending their investments using the legal system. We may not love the current state of affairs for patents, but it's important to distinguish between bottom feeders out for a dirty buck and successful entities making use of their R&D department.
I live in constant fear of the Coming of the Red Spiders.
He's sped up links between chips from something like one-third c to c.
Architecturally that reduces inter-chip latency by 66%, which does indeed open up a new overall speed range for applications that are bandwidth-limited by interconnects. But in no sense does it imply a 1000-fold increase in overall performance. It's only a 3X improvement in bandwidth of the physical layer of the interconnect to which the speedup applies.
It may allow architectures that pack in more computing units, since light beams don't interfere physically or electrically the way wires do. And light can carry multiple channels in the same beam if multiple frequency or phase or polarization accesses can be added. Those will further improve bandwidth and possibly allow a further increase in the number of computing units, which could help get to the 1000X number.
BTW, didn't Intel have an announcement on optical interconnects just a while ago? Yes. They did.
Of course it's a weapon! Have you ever been hit over the head with a Power Mac G4?
I've abandoned my search for truth; now I'm just looking for some useful delusions.
Well, electricity does travel slightly slower than light (physical electrons, which have mass, do move, although not from one end of the wire to the other). However, I suspect what they're after is improved switching speed. High frequency photons can switch on & off more sharply (i.e. in less time) than electrons in a typical electrical flow.
I've had a PDP-11/40 drop on my foot -- does that count?