hopefully some of you do realize that those flashy heatsink and flashing led are meant to detract you away form the fact that you are actually getting slower, binned, ball to wall overclock ram.
vs
non overclock server grade 32gb ecc ram that does not need any nonsense heatsink nor led.
As far that I know, ECC by itself isn't slower than non-ECC. Buffered IS slower than Unbuffered, since the buffer was effectively like an extra CAS cycle, if memory serves me correctly (That info is what I recall back from DDR1 days, not sure if it remains the same). Buffered vs Unbuffered performance was like... 2 or 3% performance on a worst case scenario. That is pretty much nothing. Benchmarks were done back when the A64 FX-53 was relaunched in Socket 939, since previously there were the Socket 940 FX-51 and FX-53 using the Opterons platform, and they required Buffered DDR to POST.Server grade RAM is often slower than "gaming" memory, just a lot more reliable. In fact, clock for clock, ECC/Registered memory is going to be slower than desktop RAM. It has other advantages like higher density per stick and error correction, but pure speed isn't one of them.
DDR2 FBDIMMs DID require heatsinks. But they were extremely power hungry. Fully Buffered DIMMs tech wasn't reused for DDR3/DDR4.Heatsinks on RAM are not really needed most of the time, it'd just make the build look cooler I guess.
Seems to me he has some proper technical knowledge. Sounds like you're just hating.
No, this is impressive because it can respond to the questionStuff like $250 8800GTs or $183 C2Ds OCed to 3GHz+ in their heydays are impressive. This can't be any further than that to the point it's comical.
He seems a bit "strange" to me. I definitely couldn't watch him with regularity. But this vid was pretty cool.
He seems a bit "strange" to me. I definitely couldn't watch him with regularity. But this vid was pretty cool.
Linus is a bit nutty and that's kinda always been his approach but lately he's embraced and extended this. It's working out good for him too, just look how many views he has compared to other similar channels. Now that he's found success he's doing a lot more interesting and entertaining ones like this. I say more power to him. There's enough room on YouTube for other more professional reviews if that's what you want.
How do you define efficient? As I have seen many developers state, rasterization is a compromise between speed and accuracy. Ray/Path tracing is easy to implement with only the cost of raw computational power. Maybe something like the wizard or some kind of fixed function raytracing hardware might make the computational requirements moot.
rasterization is a compromise between speed and accuracy.
As far that I know, ECC by itself isn't slower than non-ECC. Buffered IS slower than Unbuffered, since the buffer was effectively like an extra CAS cycle, if memory serves me correctly (That info is what I recall back from DDR1 days, not sure if it remains the same). Buffered vs Unbuffered performance was like... 2 or 3% performance on a worst case scenario. That is pretty much nothing. Benchmarks were done back when the A64 FX-53 was relaunched in Socket 939, since previously there were the Socket 940 FX-51 and FX-53 using the Opterons platform, and they required Buffered DDR to POST.
Also, since Buffered put less electrical load on the Memory Controller, you can run more modules/ranks maintaining the same Frequency. With Unbuffered, you usually have to lower their speed if you fill everything (Say, dropping from 1600 to 1333 MHz after X amount of modules/ranks). At that point, Buffered becomes faster because of the Frequency difference, no questions asked.
Buffered+ECC RAM can still overclock, just that very few Server platforms will allow you to do that (That is a Supermicro Sandy Bridge-E era Motherboard with overclocking capabilities, which SuperMicro calls HyperSpeed on validated Servers). Based on what usually happens when enthusiast tries to overclock Server parts like S939 Opterons or Westmere based Xeons, chances are that they are higher quality bins than consumer gaming/enthusiast parts and had a lot of headroom.
DDR2 FBDIMMs DID require heatsinks. But they were extremely power hungry. Fully Buffered DIMMs tech wasn't reused for DDR3/DDR4.
Modern GPUs already struggle with modern games. Software is designed to push the hardware.
And there's generally no reason to raytrace instead of using current render techniques. Image quality is more than fine if implemented correctly.
If you're thinking Pixar-quality visuals in realtime, it isn't going to happen. Raytracing just isn't efficient to do on a computer, period.
According pcper podcast the machine ultimately did not work at CES but was only displayed. I wonder what went wrong.
Imagination demos PowerVR real-time raytracing tech
Wouldn't it be interesting to have a 3rd player (or 4th if you want to include Intel in there. and why not?) in the GPU business?