[Surplus By Design] ASUS TUF Gaming Nvidia GeForce RTX 4080 w/coupon $1147.5 ATL - Page 2 - RedFlagDeals.com Forums (2024)

chaosthebomb wrote: ↑6750xt is a 250w card, 4080s is rated at 320w. Since he mentioned it was a dell, it's very likely there is a junk powersupply just big enough for the current card, and has no room for any type of upgrade. If Dell put a 500w psu in that prebuilt, it might be fine for the current card, and could be just shy of what is actually needed for a 4080.

Please don't use userbenchmark, the site is unreliable and untrustworthy.

Please provide an alternative link

PirateAngel3000 wrote: ↑Please provide an alternative link

There are two problems with using user benchmark. The main problem about UB is that they have MAJOR biases in products. In a review you need to remain objective, what's good, what's bad. The owners of UB must have had their girlfriend leave them for AMD at some point based on the vendetta they have against all AMD products. To be fair, AMD has made some bad products, but they have made some great ones too, like their current CPU's, and yet every new release UB changes their scoring metric to favor Intel. They messed up their scoring metric so bad one time it said an i3 was the best gaming cpu. The other problem (not just related to UB) is that it's hard to compare a GPU down to just 1 number. They do multiple workloads, and might be better in some, and worse in others. One card might be 25% better at 1080p but on par at 4k. How do you show 1 number to describe all that detail?

Getting rid of the bias is easy. Checkout techpowerup's gpu relative performance. They have a similar score you can see how cards roughly stack up against. On this page for the 4070ti super, you can see in the relative performance section, roughly how it stacks up to the rest of the stack. 4080 is roughly 16% better. But if you were truly interested in proper comparison you should dive a little deeper and look at more specific reviews. Techspot does a good job of showing you game averages across a wide range and compiling by resolution so you can see exactly how the card will perform where you're playing. Using the techspot review here we can see at 1080p the gain from 4070ti super to 4080 is only 12.5% on average, 17.5% better at 1440p, and 20.5% at 4k. I'd still recommend checking more than one source, but TPU and Techspot are a lot more trustworthy than the crap UB spits out.

narcomonger wrote: ↑Never heard of the company.
Is it a brand new item or a refurb card?
I have an RX6750XT in current Dell but want something beefier for my first self-built. Seems like this would be a solid upgrade?

Surplus by Design has been around over a decade. They started on eBay selling overstock, returns and refurbs. I thought it was for Best Buy but now I know is Staples. Several years back they migrated to their own site. I don't buy from them anymore because they're dreaming for the prices they're listing some of the things, but I guess it appeals to many so each their own. Makes sense. They're like 2ndTurn (these guys are so old, like Future Shop old). It started out as a platform for FS and BB employees to buy overstock and returns, plus their eBay page for same thing for public, then they opened their site or changed it,( I was never an employee) where they have auctions off eBay. Sometimes you can find deals, but I haven't shopped either in years.

badOne wrote: ↑would a 750 watt psu acccomodate this card if you're oc the video card and oc a 13th gen intel i5?

As long as it’s a quality one yeah. I pull about 200w with no power limits on a 13600K at max load and even if you were pulling 350 from the 4080. You’d still have some headroom for other components. In theory you could pull more wattage but you’d be well passed the point of diminishing returns on power.

vultureca wrote: ↑I saw the 4070 super ~750 the other day. Is the price difference justified, do you get ~30% more performance?

chaosthebomb wrote: ↑There are two problems with using user benchmark. The main problem about UB is that they have MAJOR biases in products. In a review you need to remain objective, what's good, what's bad. The owners of UB must have had their girlfriend leave them for AMD at some point based on the vendetta they have against all AMD products. To be fair, AMD has made some bad products, but they have made some great ones too, like their current CPU's, and yet every new release UB changes their scoring metric to favor Intel. They messed up their scoring metric so bad one time it said an i3 was the best gaming cpu. The other problem (not just related to UB) is that it's hard to compare a GPU down to just 1 number. They do multiple workloads, and might be better in some, and worse in others. One card might be 25% better at 1080p but on par at 4k. How do you show 1 number to describe all that detail?

Getting rid of the bias is easy. Checkout techpowerup's gpu relative performance. They have a similar score you can see how cards roughly stack up against. On this page for the 4070ti super, you can see in the relative performance section, roughly how it stacks up to the rest of the stack. 4080 is roughly 16% better. But if you were truly interested in proper comparison you should dive a little deeper and look at more specific reviews. Techspot does a good job of showing you game averages across a wide range and compiling by resolution so you can see exactly how the card will perform where you're playing. Using the techspot review here we can see at 1080p the gain from 4070ti super to 4080 is only 12.5% on average, 17.5% better at 1440p, and 20.5% at 4k. I'd still recommend checking more than one source, but TPU and Techspot are a lot more trustworthy than the crap UB spits out.

Okay Vultureca, so based on this guys reply, the 4080 is probably not going to be 30% more performance for the relative 30% price increase. I would recommend the 4070 ti super at $700~ whatever it is when back in stock. Also, ive been recently encountering some interesting CPU usage being really high and recommend making sure you get a very power CPU as well. About 4 years ago, I went for a low-mid range CPU and a mid-high range gpu to save on costs. I regret that now as my gpu is totally fine but I'm finding new games are eating my CPU to 90%+ and crashing it. I will have to throttle it now.

Thanks chaosthebomb for the detail info. I do remember someone else on reddit telling me to avoid User Benchmark, and I think at the time I was using Tom's hardware benchmarks, but don't know if that's any good either.

Also, pcpartpicker is the best source for custom builds.