Sir Beregond

Welcome!

Thanks ChatGPT!

I agree that that would be a great feature for some of the longer threads out there.

Yes I agree its the faster raster card, but its definitely a worse card for RT w/ FSR. It's Navi31 sure, but I wouldn't call it "full fat". That's the 7900XTX. Its 80 compute units for the GRE vs 96 on the XTX, not to mention a reduction in memory bandwidth from 384-bit to 256-bit. The architecture is basically that each compute unit has a ray accelerator, so it is 80 vs 96. That's about a 17% cut-down on the GRE. Sure we can all say "its not that bad" but its still for all intents and purposes a generation behind which is why I think trying to slot in price wise to Nvidia's current gen really isn't all that compelling to me. Yeah I just feel like that "generation behind" on RT / FSR like features makes these launch MSRPs for AMD cards really lackluster - they are after all pricing within Nvidia's current generation. I can at least say the price usually course corrects, unlike Nvidia.

It's a solid enough card as a slot in between the 7900XT and the 7800XT. But looking at the performance vs the 4070 and 4070 Super I think the price is kind of a hard sell for what you give up in RT. It's not terrible, but it's definitely a worse card if RT matters for you and I think the 16GB vs 12GB argument at this price segment is less important. So asking for the same money as a 4070...I don't know. Its also considerably more power consumption than the 4070 and 4070 Super under load, so I don't see a win there either. I guess if 12GB vs 16GB is make or break for your use case, otherwise I don't know, not excited about this card at $550.

I agree. The only difference is I don't think they'll try to launch a 5060 8GB. I think they'll try to hold out for the higher density memory chips into 2025 before launching those cards. If they can get 3GB modules, then even a 128-bit card can have 12GB.

That would be nice, but sounded like no 24Gb chips till next year, so if the 5090 launches this year, its stuck with 2GB modules so 24GB or 48GB if a 384-bit card. I personally don't buy the 512-bit rumors.

My guess is 5090 is either another 24GB card or 48GB if they decide to clamshell 3090 style. And the 5080 if it's 256-bit again, is another 16GB card then, unless it comes out when the 24Gb chips arrive in which case it becomes possible for 24GB as well. I could see 48GB, then 24GB. And then if the 70 and 60 cards also use higher density memory into 2025, enables options for some odd sizes like 18GB for 192-bit cards and 12GB for 128-bit. Or they could just go lame and pull another 40-series lineup VRAM wise. Who knows.

@Avacadohit the nail on the head. +1 to avoiding EK. They are really overpriced yet have massive quality control issues with most of their products. You never know what you are going to get whether its faulty DD blocks, dirty misfitting tubing, to really crappy coolants. Instead I agree with Avocado's list that the likes of Aquacomputer, Heatkiller (Watercool), Optimus, Bitspower, Phanteks, Alphacool (generally) are all generally considered to be quality and should cost considerably less than brands like EK for similar or better performance. I'm using a clear premix right now since I had it and figured I should just use them, but I agree that distilled water with Mayhems Inhibitor+ and Hades+ drops is all you need. Having had issues with glycol based coolants collecting little bits of congealed glycol in the Optimus block, I'll probably either look at other premixes or more likely just switch back to distilled + those additives. You don't have to get all your parts from one vendor. It might make it easier as a first timer, but I generally mix and match. I like Aquacomputer or Heatkiller reservoirs for the material quality, but I might like blocks from various companies depending on who has the best performer for a given block type, while choosing Alphacool or Hardware Labs radiators, with fittings from Bitspower, Monsoon, Barrow, or whoever is making what I need that's quality. For pumps I am also more of a D5 guy. For a single component loop, one should serve your needs just fine. D5's utilize the coolant/water of your loop for cooling, and in my experience are generally quiet (though not silent). Once my fans are ramping up a bit I can't hear my 100% speed D5 at all. DDC's on the other hand use external heatsinks, but I have also heard they tend to be louder. I've never actually used a DDC in any of my loops so just speaking from what I've heard and not from experience. The other difference is that D5's tend to provide better flow rates while DDC's tend to be better head pressure. Head pressure is good for high restriction loops to maintain flow, but a simple loop a single D5 will do just fine. So you can really go either way here. Tubing - I wouldn't go into hard tubing a first go around. Hell I'm not sure I'll ever do it again either. Stick to soft tubing and again Avacado gave good recommendations for which tubing to get, and I agree with those recs. Pull vs push does not matter and performs the same. If the question is should you have fresh air intaking into the radiator vs exhausting used up case air through it, I imagine its not a big deal either way as long as you have good airflow.

Just feeling some nostalgia for some Dreamcast games I forgot about playing. Forgot how good this particular soundtrack was.

But I wanted it to cook breakfast.. All seriousness though, good review and I agree, I've yet to really see a point in even upgrading my Gen 3 drives and when I do get some more, will probably stick to Gen 4.

Yeah I don't get why companies do that. Good review and always appreciate the accompanying video reviews/walkthroughs!

Agreed, good review! I once again have to compliment your builds and the themes. Love the Half Life themed PC.

Not just custom loop, probably direct die too.

Unless they are going for a Titan branded product again that really is top of the line in both spec and price, I seriously doubt you'll see anything more than a 384-bit bus still. Density on GDDR7 modules is supposedly still going to be the same as what GDDR6X is now (16Gb which is 2GB per module) starting out, so unless they clamshell the memory modules to get 48GB in 3090 style, then it's probably another 24GB card. There will still be an increase in bandwidth at least going from G6X to G7 as the 4090 is 21Gbps and the G7 is rumored to be 28-32Gbps. GeForce 200-series (circa 2008-2009) was the last time they did any consumer cards with larger than 384-bit for traditional GDDRx based cards. Color me extremely skeptical of 512-bit rumors.