New Posts  All Forums:

Posts by TranquilTempest

Maybe if Zen is competitive Intel won't be able to get away with crippling their products so much. I'm not holding my breath though. As for overclocking mobile parts, that's a really stupid idea if the cooling and power delivery systems aren't designed for it, so the decision to enable/disable overclocking on those systems should probably be left to the OEM.
Obviously perf/w scales into overall perf for GPUs, because those workloads are inherently parallel(you can always add more die area to get more performance). It's much harder to extract instruction level parallelism, so it doesn't really translate to single threaded performance. Hell, it doesn't translate to multithreaded performance either until Intel decides to start increasing core counts on their mainstream parts.Also, "Internet of things" is pretty much a garbage...
I think it's more likely he'll be designing a custom vision ASIC than CPU cores, sure there will probably be an arm core included in there somewhere, but I don't see why they'd need an architecture license over one of the off the shelf cores.
He's competitive and disruptive, and it's driving technical innovation in entire industries. The reactions from his competitors are just as important as the direct impact of his companies.
Google "skylake errata"
AA can have a big impact on performance, and whether I enable or disable it is pretty game specific, If you have the GPU headroom, go ahead and enable it, though take care not to put yourself in a GPU bottlenecked situation. AF doesn't impact performance much I'll normally leave it at 8x or 16x.
You need to be more specific. "Turbo mode" is a name that gets slapped on to just about every feature a marketing department hears about but can't understand. So, exactly which product are you enabling "turbo mode" on, and what does the manual say that feature does?Sometimes it's just a mild automatic overclock that you'd normally override with your own settings.
What is the thinnest layer of paste you can reliably apply without trapping air? Multiply that thickness by thermal conductivity for a more useful metric.There are a few factors that impact this, like viscosity, contact angle, and application method.
There are a few problems with the test, like measuring system power consumption instead of video card power consumption, and cherry picking the title to give the best result. Also, the 950 isn't Nvidia's most power efficient card right now, that would probably be the 980.Basically, wait for TPU to measure power consumption before you trust claims about how efficient it is.
I would set it at 150, for consistency. You don't want big spikes in latency during heavy combat, do you?
New Posts  All Forums: