Their bias is so absurdly clear you can see it from a mile away.
They've given much worse ratings to much better products with significantly less to criticize.
Their bias is so absurdly clear you can see it from a mile away.Anyone else think Tom's was a bit... generous giving this thing 4/5/5 stars?
And risk Jensen's wrath? Nah.Anyone else think Tom's was a bit... generous giving this thing 4/5/5 stars?
Probably due to more physical cores. Raytracing must really punch all eight cores hard and create too much cache pressure for SMT to be of much benefit. I would like to see SMT off results to see if the situation improves with the 9800X3D in raytracing.In this case, "faster" is relative. Is a 13900k faster than a 9800X3D? Thanks to crappy NV drivers, apparently so . . . but only for ray tracing.
Anyone else think Tom's was a bit... generous giving this thing 4/5/5 stars?
Now that you mention it, I wanna see #"14900KS Extreme Performance profile + Overclocked 5090" trendingThe new meta will be building rapture lake combo with 5090 that lacks a hotspot sensor. What could possible go wrong? ☠️
It's been a while that I last used a Windows computer with Nvidia card (and last time I played video games was on C= Amiga anyway) — but isn't the actual Nvidia device driver single-threaded (on single-GPU systems)? That is, other threads coming from the game engines, not from the device driver?I would like to see SMT off results to see if the situation improves with the 9800X3D in raytracing.
Nvidia drivers could also be using the E-cores properly and offloading less important tasks to them, relieving the pressure on the P-cores.
They seem to have gone multi-threaded way back: https://techgage.com/news/nvidia_graphics_drivers_to_go_multithreaded/but isn't the actual Nvidia device driver single-threaded (on single-GPU systems)?
Then we should see improvement with the 7950X3D and 9950X3D? It isn't like the E cores are world beaters for gaming.So having extra cores does help with driver overhead, at last on high-end NV cards. There was speculation that higher core counts would also help with Intel A580 driver overhead as well, and while those charts aren't direct evidence, it does seem to hint that they would be useful.
Moral of the story: fix your stupid drivers NV and Intel.
US is 120v 15-20A so 1800-2400w our 240v is 2 120v dual phase and has 3600w+ of power. Old houses maybe but newer houses, 70s and later, should be fine.Would that even run in America with your 110V mains power? I thought things got a bit sketchy at 1.5KW for a normal outlet.
In the UK we can pull 3KW from a standard outlet and it is pretty cold here so it would double as a pretty good room heater as well.
In Europe there's a lot of strange things in some places. 230V is either phase+neutral or dual phases or a subset of triple phases (3x230V) or subset of 400V (3x phases + neutral).US is 120v 15-20A so 1800-2400w our 240v is 2 120v dual phase and has 3600w+ of power. Old houses maybe but newer houses, 70s and later, should be fine.
+1I'd like to see the 7950X3D with the 5090 to test the hypothesis that cores and threads are why raptor slightly outperforms the 9800X3D with max RT/PT. Until that is determined, it is possible some of the cause is driver related.
This is a relatively minor quibble especially in this context, but US 240V residential is still technically single-phase, it's just that the neutral is center-tapped out of the 240V secondary winding, which gives 120V from neutral to either of the two hot wires.US is 120v 15-20A so 1800-2400w our 240v is 2 120v dual phase and has 3600w+ of power. Old houses maybe but newer houses, 70s and later, should be fine.