3 ms·
I'd really question the theory that #cores will grow exponentially. Be interesting to see some evidence that's likely to happen.
by axod 18y ago
I'd really question the theory that #cores will grow exponentially.
Be interesting to see some evidence that's likely to happen.
- sanj 18y agoNow that you mention it, I'm not sure it needs to. As long as you clear ~100 cores, I think you're done.
- DabAsteroid 18y agoas you clear ~100 cores, I think you're done. Why do you think so? The 2007 book "Future Directions in Processor Design" says (p483, Ch21): http://www.springerlink.com/content/xp811205j8523537/ http://www.springerlink.com/content/xp811205j8523537/ If we think of the technology development as predicted in the ITRS roadmap [208], it is clear that more and more processors will be crammed onto a single chip. The prediction for the roadmap is that we will see hundreds or even thousands of precessors integrated within the next ten ... fifteen years, e.g., 424 precessing elements per chip in 2017. The trend can be confirmed by looking at some ambitious high-end projects in multi-core and multi-processor development. For example, Rapport Inc. is shipping a chip with 256 processing elements on board and is developing a 1,024-core processor, however these are only 8-bit elements [281]. ... My bet is that we will see more specialized processors for different specific tasks. The "one size fits all" simply cannot provide enough cost and power efficient enough solutions for the embedded sector. Thus, we will see various special-purpose off-the-shelf cores emerging.
- sanj 18y agoSorry, that was unclear. I was thinking more that few people would have >100 browsers/tabs/applications open at one time. It isn't an argument about the technology, rather about what humans will be able to manage.
- hhm 18y agoYour argument sounds a lot like "64k should be enough to anyone"; I guess we'll only need more and more processing power (and processes aren't necessarily applications, as in the case of Chrome -and you could also be using threads for that matter-, so your argument doesn't follow anyway).
- 0x44 18y agoWhile I currently only have about twelve tabs open, in the past I have gone above 100. I tend to use tabs to store things I will look at in the recent future, and also as a way to follow threaded comments, or links within articles.
- rbanffy 18y agoNiagara... cough... Larrabee... cough... If you really think Larrabee is only a GPU, think again. Each core is a full 64-bit x86 that can quite happily run anything that´s running on my Core 2 Duo.
- 13ren 18y agoMoore's Law is exponential If you believe it will continue, and that clock speeds won't increase, then you get exponentially increasing #'s of cores (or some other use of silicon estate). An alternative scenario is that there will be no mainstream demand for more computing power, and only niche markets will buy them. The mainstream will favour cheap, less powerful CPU's. The rise of eeePC and clones is suggestive of this scenario.
- anewaccountname 18y ago>If you believe it will continue, and that clock speeds won't increase, then you get exponentially increasing #'s of cores (or some other use of silicon estate). Only if you believe the wiring connecting the cores wouldn't explode with the number of cores.
- 13ren 18y agoI think the wires would just be a single bus. A single bus works fine for moving data to/from one of many memory locations. Another method of using just one bus is TCP-IP packet-style communication. Of course, there's limited bandwidth, so you want most of your computation done within a processor, not between them.
- anamax 18y agoIf a single bus works for large numbers of processors, why are data centers wired differently? For many problems, bisection bandwidth is a key constraint.
- axod 18y agoMoore's Law is more of an observation. There's nothing to suggest it can continue indefinitely forever.