Re: Exclusive core for a process, is it reasonable?

2018-04-10 Thread John Hening
Thanks for your responses W dniu niedziela, 8 kwietnia 2018 14:51:52 UTC+2 użytkownik John Hening napisał: > > Hello, > > I've read about thread affinity and I see that it is popular in > high-performance-libraries (for example > https://github.com/OpenHFT/Java-Thread-Affinity). Ok, jugglery

Re: Exclusive core for a process, is it reasonable?

2018-04-09 Thread Wojciech Kudla
John, What I'm referring to is basically reducing topological distance between relevant nodes (ie. co-location, alternative network routes), or changing means/medium (copper vs fiber vs microwave). For instance microwave has gained a lot of popularity recently due to the speed of propagation of

Re: Exclusive core for a process, is it reasonable?

2018-04-09 Thread Greg Young
Yep. Most the ASIC stuff I have seen has been trivial algorithms where speed was the most important thing. On Mon, Apr 9, 2018 at 5:03 PM, Wojciech Kudla wrote: > Some of the stuff I had a chance to work on managed to handle market data > in single digit micros and

Re: Exclusive core for a process, is it reasonable?

2018-04-09 Thread Wojciech Kudla
Some of the stuff I had a chance to work on managed to handle market data in single digit micros and trading in low tens. That's Java/c++. With modern day hardware it would be extremely hard (and costly) to push it much further. I can easily imagine how going for ASIC and staying under 1

Re: Exclusive core for a process, is it reasonable?

2018-04-09 Thread Avi Kivity
Seriously, people are trading on ASICs? The amount of effort going into this is astounding. I can't help thinking that it won't end well. On 04/09/2018 10:55 AM, Greg Young wrote: To be fair many of the fpga based things have also moved to asics. You know you are in for fun when a fpga is

Re: Exclusive core for a process, is it reasonable?

2018-04-09 Thread Jean-Philippe BEMPEL
Hi John, As always, it depends :) In my previous job, We had a system that processing orders in 100us. Here thread affinity and core isolation were mandatory to achieve the SLA. The difference measured was from 2x to 4x. But not all systems need this setting/tuning. if you are accessing a lot of

Re: Exclusive core for a process, is it reasonable?

2018-04-08 Thread Gil Tene
“Reasonable people adapt themselves to the world. Unreasonable people attempt to adapt the world to themselves. All progress, therefore, depends on unreasonable people.”― George Bernard Shaw To your question tho: there are plenty