| Horizon68 <horizon@horizon.com>: May 06 03:12PM -0700 Hello.. I think X86 CPUs are also the future, here is why: The final ISA showdown: Is ARM, x86, or MIPS intrinsically more power efficient? The RISC vs. CISC argument should've passed into history a long time ago. It may still have some relevance in the microcontroller realm, but has nothing useful to contribute to the modern era. An x86 chip can be more power efficient than an ARM processor, or vice versa, but it'll be the result of other factors — not whether it's x86 or ARM. Read more here: https://www.extremetech.com/extreme/188396-the-final-isa-showdown-is-arm-x86-or-mips-intrinsically-more-power-efficient Thank you, Amine Moulay Ramdane. |
| Horizon68 <horizon@horizon.com>: May 06 12:17PM -0700 Hello.. Intel demos first Lakefield chip design using its 3D stacking architecture The Foveros era at Intel is approaching Lakefield will specialize in lower-power devices with unique hardware constraints, including foldable phones and tablets, drones, smart home devices, and other gadgets that require a tiny, all-in-one chip. Read more here: https://www.theverge.com/2019/1/7/18173001/intel-lakefield-foveros-3d-chip-stacking-soc-design-ces-2019 Thank you, Amine Moulay Ramdane. |
| Horizon68 <horizon@horizon.com>: May 06 11:48AM -0700 Hello.. Intel's Plan to Defeat Qualcomm in Laptops Read more here: https://www.fool.com/investing/2019/01/23/intels-plan-to-defeat-qualcomm-in-laptops.aspx Thank you, Amine Moulay Ramdane. |
| Horizon68 <horizon@horizon.com>: May 06 08:28AM -0700 Hello... About scalability of my efficient Threadpool engine with priorities that scales very well.. I have to explain something important about my efficient Threadpool engine with priorities that scales very well: In NUMA enabled systems all memory operations require snooping across all CPU sockets to keep cache data coherent. In other words, before data is retrieved from RAM snooping operation will check cache content of local and remote CPUs to find the copy of this data. In Haswell CPUs there is the following more powerful snooping mode: Cluster-on-Die (COD) snooping mode is ideal for highly NUMA optimized workloads. Compared to two previous modes where snoops are simply broadcasted the COD will first snoop the directory cache and then the home agent. With Cluster on Die snooping mode each Memory controller now serves only half of the memory access requests thus increasing memory bandwidth and reducing the memory access latency. And obviously two memory controllers can serve twice as much memory operations. So you have to understand that my efficient Threadpool engine with priorities that scales very well is only sharing some variables across cores on the producers side, so i think it will scale well with Cluster-on-Die (COD) snooping mode. You can download my efficient Threadpool engine with priorities that scales very well from: https://sites.google.com/site/scalable68/an-efficient-threadpool-engine-with-priorities-that-scales-very-well I have also implemented a powerful scalable ParallelFor() in it. Thank you, Amine Moulay Ramdane. |
| You received this digest because you're subscribed to updates for this group. You can change your settings on the group membership page. To unsubscribe from this group and stop receiving emails from it send an email to comp.programming.threads+unsubscribe@googlegroups.com. |
No comments:
Post a Comment