The same Si cost problems apply to Intel as well. Both broadly as a result of exploding fab costs across the industry, but in this case in an even more specific way because Intel is using TSMC for their GPUs.Intel should be in a position to take the budget gpu segment because both AMD and Nvidia missed.
I thought it was valid of GN to make the points about the launch confusion, it gave a feel of just how disjointed and out of step AMD is right now with their products and the community. Steve likes to base his reviews on value and by dropping the price they messed with his whole style of reviewing which I can get is irritating to him, but you have a valid point that it's a bit overblown and the intro and extro can correct any thoughts on price/performance without affecting the numbers.
HU was brutal on both launches, which I think is the way to be for both launches since they both are major disapointments.
Computerbase gets 42 FPS on a 6800 1440p, AMDUnboxed has 65FPS:HU tested the heaviest section of TLOU in the woods. Daniel Owen tested the prologue so it could be why.
I'm not saying that, it's just changing it on reviewers at last minute makes their jobs a lot harder and it shouldn't have to be. I was just meaning to opine that I think there was a lot of confusion within AMD about this launch and they didn't handle it very well.
And drivers. Imagine running 15 benchmarks multiple times in various resolutions only to have to rerun them a few days before because "they gave out the wrong drivers". That would SUCK, if it's even possible to pull off. AMD has long had a branch of PR that deals with media relations, and for now youtubers make up probably the largest segment of that for them. They have people dedicated to working with reviewers and back in my day they were pretty freaking awesome about trying to not dick over the press. (I'm looking at you Chris Hook, and man I miss you there! )From my perspective, if these things can happen, then film all of the discussion about the performance data leaving out mention of price to performance and have all of the price to performance discussion in one segment of the video so you can swap it out without having to re-edit the whole thing. It's a highly competitive industry, that kind of stuff is going to happen, especially from the underdog players that have to be very reactive to the market leaders. There are basically two things that can change at the last second: release date, price. They should be helping themselves by anticipating change changes at the last second and building that into their workflow. That's basic professional stuff in any job. AMD can't cater their business decisions specifically to techtubers. Youtube is a crunch industry if you targetted day one release videos.
Why do you think there was confusion? It’s perfectly normal for a company to constantly monitor changing variables and model various pricing strategies before launch.I think there was a lot of confusion within AMD about this launch and they didn't handle it very well.
They just test a different area.I don't think a 6800 should be dropping down to 42fps under any circumstances at 1440p. Unless they reviewed the launch version or something.
If ray tracing is involved don't underestimate AMD's ability to choke hard.
There is no RT in TLOU Part I. The PS5 runs it at High settings 1440p and generally 60-80fps. A 6800 shouldn't drop to 42fps at 1440p. HU's results seem more in line with what I've seen from the people here, DF, and my own tests.They just test a different area.
Look at Spider-Man and a more heavy scene with raytracing - 45% more costs get you 70% more performance.That is even a fair comparision, in 1440p this "better" product is useless.
That is a $150 GPU at best and at $269 much worse than a 4060TI... (that is actually an archivement).
My mistake, thanks for correcting me and letting me know the facts.There is no RT in TLOU Part I. The PS5 runs it at High settings 1440p and generally 60-80fps. A 6800 shouldn't drop to 42fps at 1440p. HU's results seem more in line with what I've seen from the people here, DF, and my own tests.
I was talking about GN's Steve (don't really watch HUB's Steve anymore, too much BS to sort through). He's spent like 15 minutes ranting about how messy the launch was behind the scenes including the sudden lowering of the MSRP - as if that never happened previously - and only once (I think?) during this period he said that lower price is in fact a good thing.I'm not saying that, it's just changing it on reviewers at last minute makes their jobs a lot harder and it shouldn't have to be. I was just meaning to opine that I think there was a lot of confusion within AMD about this launch and they didn't handle it very well.
Microsoft has provided a path in DirectML for vendors like AMD to enable optimizations called ‘metacommands’. In the case of Stable Diffusion with the Olive pipeline, AMD is building driver support for a metacommand implementation intended to improve performance and reduce the time it takes to generate output from the model. This feature will be validated on AMD RDNA™ 3 devices including AMD Radeon™ RX 7900 Series graphics cards and AMD Ryzen™ 7040 Series Mobile processors with Radeon™ graphics.
I doubt anyone is buying the 6800 XT for AI performance.Microsoft talks a little bit about WinML right now. AMD will support WinML, Metacommands and better efficiency and performance only on RDNA3 cards:
AMD now has driver support for a metacommand implementation intended to improve performance and reduce the time it takes to generate output from the model.gpuopen.com
Another great example what happens when you buy a "recommended" GPU like a 6800XT(6900XT) instead of a RTX4070.