Power, Temperature, and Noise

As always, we'll take a look at power, temperature, and noise of the GTX 1660, though after having seen the GTX 1660 Ti in a similar if not identical design, we aren't expecting anything out of the ordinary. As mentioned earlier, we've seen the XC Black board with the GTX 1660 Ti not too long ago.

Using the same TU116 GPU as the GTX 1660 Ti, the voltages are unsurprisingly the same.

NVIDIA GeForce Video Card Voltages
Model Boost Idle
GeForce GTX 1660 1.037V 0.656V
GeForce GTX 1660 Ti 1.037V 0.656V
GeForce RTX 2060 1.025v 0.725v
GeForce GTX 1060 6GB 1.043v 0.625v

As for clockspeeds, the same broad points from the GTX 1660 Ti review apply. Clocks at +10W TDP and at reference 120W TDP are only slightly altered, and the trend of NVIDIA's conservative boost estimates continues.

GeForce Video Card Average Clockspeeds
Game GTX 1660 EVGA
GTX 1660 XC
GTX 1660 Ti GTX 1060 6GB
Max Boost Clock
2160MHz
2160MHz
2160MHz
1898MHz
Boost Clock 1830MHz 1830MHz 1770MHz 1708MHz
Battlefield 1 1880MHz 1885MHz 1888MHz 1855MHz
Far Cry 5 1889MHz 1897MHz 1903MHz 1855MHz
Ashes: Escalation 1874MHz 1872MHz 1871MHz 1837MHz
Wolfenstein II 1832MHz 1861MHz 1825MHz 1835MHz
Final Fantasy XV 1865MHz 1869MHz 1855MHz 1850MHz
GTA V 1894MHz 1898MHz 1901MHz 1872MHz
Shadow of War 1879MHz 1882MHz 1860MHz 1861MHz
F1 2018 1880MHz 1886MHz 1877MHz 1865MHz
Total War: Warhammer II 1890MHz 1893MHz 1908MHz 1875MHz

Compared to the official average boost clock of the GTX 1660 Ti, the differences are also minor.

Power Consumption

Idle Power Consumption

Load Power Consumption - Battlefield 1

Load Power Consumption - FurMark

Meanwhile when it comes to idle power consumption, the GTX 1660 falls in line with everything else at 83W. With contemporary desktop cards, idle power has reached the point where nothing short of low-level testing can expose what these cards are drawing.

All told, NVIDIA has very good and very consistent power control here. and it remains one of their key advantages over AMD, and key strengths in keeping their OEM customers happy.

Temperature

Idle GPU Temperature

Load GPU Temperature - Battlefield 1

Load GPU Temperature - FurMark

Noise

Idle Noise Levels

Load Noise Levels - Battlefield 1

Load Noise Levels - FurMark

Turning again to EVGA's card, despite being a custom open air design, the GTX 1660 XC Black doesn't come with 0db idle capabilties and features a single smaller but higher-RPM fan. The default fan curve puts the minimum at 33%, which is indicative that EVGA has tuned the card for cooling over acoustics. But the curve is a little more forgiving at higher temperatures, and doesn't ramp up as much, reducing their noise levels significantly from the Ti XC Black.

Compute & Synthetics Overclocking
POST A COMMENT

69 Comments

View All Comments

  • oleguy682 - Thursday, March 14, 2019 - link

    Not really sure this works as a 970 replacement. I'll have to peruse Bench to see what the difference between the 960 and 970 were to see if it's worth the money. Might need to wait another generation.... or refresh. Reply
  • Ryan Smith - Thursday, March 14, 2019 - link

    On which note I've gone ahead and unlocked the GTX 1660 cards in bench.

    https://www.anandtech.com/bench/product/2456?vs=21...

    The GTX 1660 is a good bit faster than the 970, but it may not be enough to satisfy your needs. Then again, this is a $219 card versus a $329 card.
    Reply
  • nathanddrews - Thursday, March 14, 2019 - link

    Going by current Newegg prices - the GTX 1070 and Vega56 can be had for just under $329 (original 970 pricing) and are both literally 2X faster in many Bench scenarios than the 970. Reply
  • 0ldman79 - Thursday, March 14, 2019 - link

    Considering the 970 is $100 on ebay right now...

    Roughly 85% of the performance.
    Reply
  • nathanddrews - Thursday, March 14, 2019 - link

    You can justify anything using used pricing. In a freak bit of luck, I got two used 1080Ti GPUs for $400 each last October, virtually eliminating any possible price/performance comparison. Assuming the OP bought the 970 new, it's a fair comparison to look at current $329 cards. Reply
  • Gigaplex - Thursday, March 14, 2019 - link

    It's irrelevant what they paid for the 970. They already have it, so it's effective cost on whether to upgrade is what they can get for it when they sell it, or if they don't sell it, $0. Reply
  • Old Dog - Friday, March 15, 2019 - link

    Of course it's not. He's not talking about going out and buying a 970 today at that price. The real question is if the increase in performance is worth the $150 or so it'll cost him to upgrade (assuming he nets ~$80 by selling the 970 for $100 on eBay. (I'm afraid that price is about to take a hit though.) This is the same question we're all asking ourselves. Original price of our current cards is irrelevant. It's all about current value. Reply
  • Flunk - Thursday, March 14, 2019 - link

    And about 4 years of time. That's not a very good deal. Reply
  • flyingpants265 - Thursday, March 14, 2019 - link

    While you're at it, can you make Bench compare multiple cards at the same time? This site seems oddly trapped in 2007 in some ways. Reply
  • catavalon21 - Friday, March 15, 2019 - link

    Or, it ties or bests the GTX 980 in 43 of 44 benchmarks. Not bad for a $219 card with a 3 year warranty (compared to whatever life one will get out of a 980 after years of mining...) Reply

Log in

Don't have an account? Sign up now