mardon85

Well picked up.. Compare just the GPU score and my point still stands.

On its release the 2080 Super got a real panning from the youtube community saying it was pointless etc. I managed to get my hands on one at a great price (£480) and set about overclocking it.

The GDDR6 Samsung Memory on these things is unreal. I was able to achieve a stable overclock of 9000Mhz (18Ghz double data rate) on the RAM which will close the gap on the larger memory bus of the 2080Ti.

The card I had wasn't a great overclocker and only achieved 2050Mhz locked using a Kraken G12 AIO with temps around 50C. Performance was fantastic and I was very happy with the card.

Unfortunately due to an error all of my own creation the card is now dead. I replaced the 2080 Super with a 2080Ti which arrived yesterday. Unfortunately the Ebay listing was wrong and I got a Gigabyte Windforce none A chip vs the Gaming OC A Chip I had paid for so it's going back today. However this didn't stop me running a few Time Spy Extreme benchmarks. The overclocked 2080 Super is within 502 graphics score points of a stock 2080 Ti. Obviously overclocked the difference will be much bigger. However an impressive result nevertheless for a significantly cheaper product. 

Compare link: https://www.3dmark.com/compare/spy/10597956/spy/9838792

... port royal on the other hand is a bit of a blood bath for the 2080 Super 

https://www.3dmark.com/compare/pr/217933/pr/192895

So.. i've been bitten by the bug recently and have owned the following over the last 2 months:

Sold my 1070 for £240 on ebay and bought a 2070 Super for £480 including COD/MW (which I was going to purchase anyway) so call it £450. The 2070 was a good match to my 3440x1440 100hz monitor. However I was playing Metro Exodus at the time and didn't rate the RTX performance or the impact DLSS had on image quality.  It also was good but not great at 4k on my living room OLED but not able to hold 60fps at the settings i'd be happy with. Wasn't a great chip ether so was never truly happy with it.

A 2080 Super appeared on Facebook marketplace for £480 so I jumped at that and sold the 2070 Super for £440 so a £40 upgrade. Now the 2080 Super overclocks very well on the memory. At this point I also fitted a Kraken G12 and was running 2050mhz on the core. This was killing everything I played at 4k/60 bar metro but I was really happy with it.. Until one day the fan adaptor from GPU fan header to standard fan header somehome broke and caused one of my fans to spin up to 100%. Stupidly during my investigations I unplugged the fan from the header and then reattached it.. while the PC was turned on.. Idiot.. fried my 2080 Super...  Gutted.

I had to claim on the house insurance and they've just paid out £630. I've just won a Gigabyte 2080 Ti Gaming OC for £710 which is arriving tomorrow so £80 upgrade.

I'll report back how it performs in 4K because TBH the 2080 Super overclocked was ample. I just thought i'd make the most of a bad situation.

I've probably not answered your question there but despite the bad youtube videos if you can source a 2080 Super for a good price and overclock it you'll be very happy at 4k/60

4 minutes ago, BigDaddy0078 said:

3DMark is purely synthetic, the best way to test is through actual gaming, if your fps goes up, then it's fine.

 

There's also a few things you could try,

1) Setting your pc to "Performance" This tends to fix

2) Maybe OC the GPU?? 

3) Look to see if your drivers are installed (I know this is said all the time but it's basic)

4) Try and reinstall windows and see if that works (I know it's tedious but it's probably the best thing you could do)

1) I'll give that a go

2) I've already OC'd the GPU a fair bit was really impressed with the 2080 Super's samsung memory 9000mhz vs 7800mhz on the 2070 Super. I know the 2080 has been panned in the reviews but most the reviews i've seen have been stock. For the same price as the 2070 Super and overclocked its a great card.

3) Fresh drivers using DDU software

4) Fresh Windows 10 UEFI Install on my new M.2 drive and the 2070 results which were the same were on my old legacy SSD install.

Its gone from 0.9 Shader rate 60fps at 3440x1440 on Metro Exodus at the settings I run (inc RTX) to being able to run at 1.0 Sharder rate at over 60FPS. I can finally enjoy Metro the way I wanted!

I'm currently in the process of building a new PC. I've swapped various parts in and out as i've seen deals on FB or Ebay, maybe I have a tech addiction I don't know..

Anyways I've gone from a not great silicone lottery 8700k @ 5.0Ghz to an 'ok' 9700k @ 5.1Ghz

I've also gone from a 2070 Super which I got new for £480 with a game I was going to purchase anyway to a new (without box) 2080 Super for £480 which I thought was a pretty good deal.

In addition i'm also running 3200 LPX Vengeance at 3750Mhz OC and a Mini-ITX Z370 MSI pro Carbon Mobo + 1TB Saberant  M.2 SSD.

I'm seeing some strange 3Dmark results mainly with my combined score varying massibly. Clocks and temps have remained pretty much the same. Please see below for some examples:

9700k - 2080 Super:

Compare - https://www.3dmark.com/compare/fs/21209161/fs/21209358

59 vs 53 FPS combined same clocks on card and CPU

9700 - 2070 Super:

Compare - https://www.3dmark.com/compare/fs/21097497/fs/21097901#

58 vs 49 FPS combined

What could be causing the large fluctuation in FPS on the combined score? This is usually the test which throws the numbers out.