Where are you getting that? gtx1060 3gb is beating even 4gb cards and have no problems keeping 60fps. Kepler having 30-45fps is isssue on its own, not lack of Vram. Keep in mind these benchmarks are pre-patch so nvidia is pulling the lead here.

Kepler and Tahiti don't have the color compression Maxwell and Pascal have. Also I said GK104 you know 660ti, 670, 680, 760 and 770. On the high end they were about 50% slower than the 980. GK110 would likely be around 55 for 780 and 65 for 780ti as long as they don't run out of VRAM.

Thank you Hilbert for posting this review. As for myself, Prey has been a lot of fun. The final cut scene however, really upset me - you are Morgan Yu until then. 😕

Hawaii / Grenada however still is going strong.

I missed the 8 GB 290X variant and bought 2 290x 4GB Lightnings. No regrets, the 4 GB Lightning model can OC a lot specially under water. IMHO the best AMD GPU deal of the last 5 years: 290X 8 GB https://www.msi.com/Graphics-card/R9-290X-GAMING-8G.html#hero-overview Some lucky bastards are enjoying that model! :bang: 😀

Love this game, mechanic, artworks, looks similar to System Shock 2. Dishonored 2 is a good game also, just have wrong optimisation. Idtech and cryengine in Prey was a good choice. I can play at 6K@60 lol. Just one thing. Reactor location seems stuttery, atleast for me ")

All Kepler and Tahiti GPU's would run out of VRAM at 1080. A 4GB GK104 would likely be around 30-45 fps in 1080.

They don't. I have finished the game with everything on ultra with a 3GB 7970. It doesn't stutter or do anything weird and it stays at 60fps. EDIT: Everything on the max setting except the Screen Space reflections which were sitting at half rez. It's not as if they eat a lot of vram.

Interesting review, but it seems like the conclusions drawn at the end are out of line. I appreciate the efforts that went into providing this information, but I can't help but feel the reader is being intentionally mislead. "Let me reiterate that if you'd pick up a 329 USD Ryzen 7 1700 you'd get exactly the same results, the cheapest Intel 8-core part is roughly 1100~1200 USD." - HH Except the Ryzen 7 1700 wouldn't be competing with a 5960X. It looks like the $250 i5 7600k at stock speed will beat Ryzen 7 1700 @ 4.0. The $360 7700k at stock would definitely beat Ryzen 7 1700 @ 4.0. "And yes the AMD platform was actually 1 FPS faster overall, pretty frickin nice value there eh?" - HH This one kinda burned me. It should say "YAY! the AMD platform is within the margin of error as long as both systems are GPU bound". Why would you even show the 4/8 core comparison with the 580 when you already proved 580 is GPU bound with both CPU's at all resolution - maybe so you can say Ryzen was 1 fps faster? Even the Titan is GPU bound above 1080 - the charts show the exact same numbers for the Titan+5960 @ 4.0 and 4.3. Why would you leave the numbers off the 3rd from last chart - maybe so it's not obvious that the 4 core 5960 @ 4.0 is exactly the same as the 8 core 5960 @ 4.3 and therefore is GPU bound even with Titan above 1080? You are talking about value in a game review. It may not sound new and cool, but i5 $250 i5 4 core beats $329 Ryzen 8 core - pretty frickin nice value there eh?

Interesting review, but it seems like the conclusions drawn at the end are out of line. I appreciate the efforts that went into providing this information, but I can't help but feel the reader is being intentionally mislead. "Let me reiterate that if you'd pick up a 329 USD Ryzen 7 1700 you'd get exactly the same results, the cheapest Intel 8-core part is roughly 1100~1200 USD." - HH Except the Ryzen 7 1700 wouldn't be competing with a 5960X. It looks like the $250 i5 7600k at stock speed will beat Ryzen 7 1700 @ 4.0. The $360 7700k at stock would definitely beat Ryzen 7 1700 @ 4.0. "And yes the AMD platform was actually 1 FPS faster overall, pretty frickin nice value there eh?" - HH This one kinda burned me. It should say "YAY! the AMD platform is within the margin of error as long as both systems are GPU bound". Why would you even show the 4/8 core comparison with the 580 when you already proved 580 is GPU bound with both CPU's at all resolution - maybe so you can say Ryzen was 1 fps faster? Even the Titan is GPU bound above 1080 - the charts show the exact same numbers for the Titan+5960 @ 4.0 and 4.3. Why would you leave the numbers off the 3rd from last chart - maybe so it's not obvious that the 4 core 5960 @ 4.0 is exactly the same as the 8 core 5960 @ 4.3 and therefore is GPU bound even with Titan above 1080? You are talking about value in a game review. It may not sound new and cool, but i5 $250 i5 4 core beats $329 Ryzen 8 core - pretty frickin nice value there eh?

Except one thing, that entire thing you wrote about Ryzen 7 1700 can be written word by word for that "cheapest Intel 8-core". What HH says is that if you want powerful CPU for general computing+gaming, you do not need $1000 intel chip. And since you have same old Sandy as I do. One advice: As you upgrade, take at least 6-Core CPU. Within 2 years 4-core CPUs without HT(SMT) will be dropping off the charts. i3s will be dead & forgotten.

Except one thing, that entire thing you wrote about Ryzen 7 1700 can be written word by word for that "cheapest Intel 8-core". What HH says is that if you want powerful CPU for general computing+gaming, you do not need $1000 intel chip. And since you have same old Sandy as I do. One advice: As you upgrade, take at least 6-Core CPU. Within 2 years 4-core CPUs without HT(SMT) will be dropping off the charts. i3s will be dead & forgotten.

Also in a lot of cases that good ol i5 7600k is getting beat to death by 1600x. While costing roughly the same.