New Posts  All Forums:

Posts by hajile

I know when my laptop switches from the Iris GPU to the AMD one because the fans kick up and the battery is dead in no time at all. I like having the option, but I want to be able to do moderately GPU intensive stuff while on the go.Let's say that the iGPU is reduced though. What would you do with those extra CPU cores? Bulldozer proved very clearly that consumer software can't take advantage of lots of cores. Typical users won't see too much of an increase moving from two...
Once you move away from OCN, you'll find that the vast majority of computers use only the iGPU. They make such large iGPUs because the market demands them (they could save a lot of manufacturing cost by eliminating them if they weren't needed). The vast majority of computer users aren't willing to pay for dedicated GPUs and this is way better than old chipset GPUs.You can technically run OpenCL via CPU emulation, but the cost makes it unusable (even a small GPU has far...
I'd prefer if every chip shipped with at least a small OpenCL compatible GPU to ensure that the software could always run (even if slowly). There's little reason why an iGPU can't be power-gated when a dedicated GPU exists.
I work full time in web development (and have for several years now). The coolest thing about web development (to me) is that everything a beginner needs is available for free (and as an avid collector of programming books, I assure you that the best resources are also free). If you have an internet connection and a computer, you have all the tools it takes to learn. Here's a few things I think will help you on your way. The hardest thing about learning any new occupation...
Just because Intel get's a 20% boost doesn't mean AMD gets 20% or less. POWER8 can double performance by enabling SMTx8 (actually 2.6x faster in some tests) and in SMTx2 (equivalent of Intel), gains are very often 45-65%. SMT performance is partially based on scheduling, but a huge amount has to do with how much hardware are you willing to duplicate. The more hardware you repeat, the higher your scaling can become until you switch from SMT to full-blown CMT. Bulldozer had...
5 years ago today (well, as of three minutes ago). Still miss your presence here.
I've messed with a bunch of different VR sets thanks to work (business rather than gaming applications of VR). GearVR and cardboard aren't even in the running (to be expected of mobile phones). Likewise, Hololens is AR more than VR, but it isn't super comfortable compared to the others (probably because it carries an entire computer up there) and it's low-res with a very small viewport (for the record, Google Glass also sucked the one time I played with it and it's...
This will approximately work on IE8Code:.parent {display: table;}.child { display: table-cell; vertical-align: middle;}This will approximately work on IE9Code:.parent { -webkit-transform-style: preserve-3d; -moz-transform-style: preserve-3d; transform-style: preserve-3d;}.element { position: relative; top: 50%; -webkit-transform: translateY(-50%); -ms-transform: translateY(-50%); transform: translateY(-50%);}The rest of the world can use flexbox (I don't...
This seems much bigger than GPUs. Lower frequencies are associated with lower latencies. While this isn't super important for GPUs, it is a huge interest to CPUs where latency is one of the biggest problems. A combination of closer RAM (less distance) and lower latencies could cut dozens of cycles off cache misses. For HSA, this solves the problem of either higher CPU latencies to feed the GPU (eg. PS4) or lower latencies for the CPU and choking the GPU (Xbox one). In...
Read the Rest
New Posts  All Forums: