NVIDIA GeForce GTX 1660 Ti Linux Gaming Benchmarks

[Weasel]

It's crazy that people are still gaming on 1080p

It's crazy that people are gaming on anything above 1080p when in reality it makes no difference unless you stay unhealthily close to your monitor.

[cj.wijtmans]

It's crazy that people are still gaming on 1080p (unless those are PCs hooked up to a TV in a living room). I got my first 1440p monitor almost 10 years ago, and even before that I got my 24" 1200p screen more than 15 years ago.

I wouldn't even know where to find a non-TV 1080p screen nowadays except maybe on a smartphone or some cheap 13" notebook.

where do you live ? the market is full of cheap 1080p monitors while 1440p is expensive and hard to find a good one.

[cj.wijtmans]

This is the first time I hear of overclocking as a means to reduce power consumption

its better for CPUs to be at full clock speed so they can idle faster, it saves more power than being underclocked and taking a long time to finish their task so they cant idle.

[JanW]

its better for CPUs to be at full clock speed so they can idle faster, it saves more power than being underclocked and taking a long time to finish their task so they cant idle.

That might be true under two conditions:

• The workload is fixed, so the CPU or GPU can race-to-idle.

Not true here as the benchmarks record FPS over a fixed time. The GPU is never idle.

• The operating point on the power-performance curve you are starting from is favorable

This is why there is likely a difference between overclocking (what you initially claimed, and which is not usually beneficial as the standard boost clocks are usually fairly high up on the power-performance curve) and underclocking (which you are talking about here, and which might be detrimental). See an example here of a power-performance curve (would be happy to link to Phoronix data on this, but haven't found any).

[cj.wijtmans]

It's crazy that people are gaming on anything above 1080p when in reality it makes no difference unless you stay unhealthily close to your monitor.

my laptop has high pixel density 4k. my computer monitor is still 1080p. Its night and day difference. and no i am not that close. Although i do hate the high pixel density which is why i would look a good 1440p monitor for gaming. But they cost as much as mid range PC. so not gonna buy.

[creative]

Holy hogging three slots batman, that is one hunch back of a card. I have never seen a card that deep good lord! Im guessing this is not targeting itx builders.

[creative]

It's crazy that people are gaming on anything above 1080p when in reality it makes no difference unless you stay unhealthily close to your monitor.

Well I think 1440p most likely a better place to be if you have a display that is 27inch and veritable frequency display up to something like 120 or 144hz. I am on 1080p@60hz myself with hardware that can push that resolution solidly at my displays frequency. Freesync is getting cheaper and cheaper at 27inches 144hz.

[creative]

55 less watts and and around 384 less cores than the gtx 1070

little less power draw and a little less performance considering game title

[Jabberwocky]

55 less watts and and around 384 less cores than the gtx 1070

little less power draw and a little less performance considering game title

+1 I am really impressed by this card's power draw and price! If this card had good open source drivers I would have bought it already.

[Buntolo]

Is it just me or Rise of the Tomb Raider runs horribly on Linux compared to Windows?

Although i do hate the high pixel density which is why i would look a good 1440p monitor for gaming. But they cost as much as mid range PC. so not gonna buy.

They've dropped quite a bit in prices, years ago they costed like 700€, now a 1440p 144Hz can be found for ~300€.

+1 I am really impressed by this card's power draw and price! If this card had good open source drivers I would have bought it already.

Vega 56 will be cut at sub 300$, so you might consider it. It will consume 80W more, but I think it won't be a big issue.