Nvidia Silently Optimizes GeForce RTX 2070, 2060 Mobile GPUs for Performance

Nvidia Turing Silicon

Nvidia Turing silicon (Image credit: Nvidia)

German laptop manufacturer Schenker revealed to news outlet ComputerBase this week that Nvidia has quietly refreshed its GeForce RTX 2070 and RTX 2060 Mobile cards.

The revised GeForce RTX 2070 and RTX 2060 Mobile retain the same model names as the original versions. Therefore, the only way to differentiate the new model from the previous model is through the codename. As a matter of fact, many of the specifications from the original models are carried over in the revised models. Apparently, all Nvidia did was upgrade the memory and regulator, and that alone was enough to help squeeze some more performance out of the cards for gaming laptops.

Both the GeForce RTX 2070 and RTX 2060 Mobile utilize GDDR6 memory, and they will continue to do so. According to Schenker, the memory on the revised models operate at 1.25V as opposed to 1.35V. The lower-powered memory brings two main benefits: reduced power consumption and generated heat. As a result, the GPU can use the freed-up headroom to hit higher boost clocks and provide improved performance while still staying within the TGP (total graphics power) limit.

Specs

Swipe to scroll horizontally
Graphics CardCodenameTGP (W)
GeForce RTX 2070 Mobile (Refresh)N18E-G1R115
GeForce RTX 2070 MobileN18E-G2115
GeForce RTX 2060 Mobile (Refresh)N18E-G1-B115
GeForce RTX 2060 MobileN18E-G180

The Schenker representative noted that the refreshed versions of the GeForce RTX 2070 and RTX 2060 Mobile outperform their originals by substantial margins within the same power consumption. He suspects that a better yield in chip production somewhat plays a part, but Nvidia hasn't confirmed that.

The GeForce RTX 2070 Mobile retains the same 115W TGP rating. But since the memory is more energy efficient, the graphics card comes with a higher boost clock speed.

The GeForce RTX 2060 Mobile benefits the most from Nvidia's GDDR6 treatment. The regular version comes with a 80W TBP. Nvidia is reportedly offering its partners a vBIOS updates that increases the threshold to 115W. That's a 43.8% increase and will ultimately give the GeForce RTX 2060 Mobile a very healthy performance boost. 

But don't expect every GeForce RTX 2060 Mobile to have up to 115W of breathing room. As the Schenker spokesman noted, laptop makers are still free to use a different TGP limit and don't have to change the name of the graphics card. This is understandable, as cooling capacity varies from laptop to laptop. The problem is that unless the manufacturer strictly specifies it, there is no way of knowing which version is under the hood until you take the laptop home.

The Schenker rep said he believes we'll see public benchmarks for the refreshed GeForce RTX 2070 and RTX 2060 Mobile graphics cards on April 15.

Zhiye Liu
News Editor and Memory Reviewer

Zhiye Liu is a news editor and memory reviewer at Tom’s Hardware. Although he loves everything that’s hardware, he has a soft spot for CPUs, GPUs, and RAM.

  • watzupken
    Sounds like a performance boost due to feeding the GPU with more power in the case of the RTX 2060. 115W power requirement is going to be a problem for cooling in a laptop.

    Anyway, I suspect this sudden "free" performance from Nvidia is due to AMD's RDNA GPUs coming into the mobile space.
    Reply
  • Deicidium369
    Yeah, that tends to be how things work. Feed it more power and you get more performance - except in the case of AMD - since AMD has ultra binned ever single little slice of silicon, and offers it's pre over clocked. Along with that extra power comes extra heat - that is why it has a setting in BIOS to allow you to keep the 80W TDP or the 135W.

    Nvidia is is no way shape or form concerned about whatever marketing term AMD is using to describe whatever warmed over ancient design they are selling this week. It just isn't an issue. AMD has about as much chance taking over the discrete GPU for laptops as it does becoming the top tier full size GPU vendor. "Big Massive Totally New and Shiny Look! Squirrel Navi" is vaporware and likely to remain as such - they will undoubtedly release something with better perf - but no where near where the AMD Kiddies think it it be.

    Thing about AMD is they get things flipped - You under promise and over deliver - not over promise and under deliver - which is the history of the Ryzen line Because by now AMD should have 80% market share and Intel begging for scraps - yet that hasn't happened. Ryzen 1 (1x00 series) had the most impact, the subsequent 2 small iterative updates - not so much. Even though AMD is hitting record revenues (for AMD) it is still really nothing and R&D requires something in excess of nothing. Jim Keller's legacy at AMD (creator of Ryzen) is already in the market - they have nothing left to draw from. The changes between 1st gen Ryzen and most current in the market Ryzen is really not much at all. I know all the armchair lithography and CPU fabrication experts will say but but but "7nm"... turns out not that big of a deal. Not making money is why they spun off Global Foundries - and why they sold Ryzen designs to the Chinese - to keep the lights on while Ryzen was prepped to take over the world... a swing and a miss. I mean if Intel was in 1/10th the amount of trouble the AMD Kiddies think - then it should have been EZ Mode to fill in that gap - yet - no gap filled.

    Having a CPU with no integrated graphics (even rudimentary) with a bunch of superfluous cores (reason why Intel was at 4 cores for desktop - that's what the market wanted) is why OEMs are not selling alot of AMD systems - having to add a graphics card (another line item in the BoM) but the added load on the Support and IT depts was part of the consideration for alot of companies to stick with Intel - install the chipset drivers and you are good to go.

    The new 4x00 series APUs are great... Vast majority of businesses want basic Desktops - something to run Word, Excel, Outlook and Chrome. Don't need 8 cores and Vega graphics (still takes a sep driver)to do that.

    So no Nvidia isn't worried, Intel is not worried - The one indispensable service AMD provides to both Nvidia and Intel is to shut down ppl wanting to say they are a monopoly in their respective fields. Kinda like how a Morty is a shield for a Rick. Advanced Morty Defenses.
    Reply
  • knekker
    #Deicidium369 I own a pc with intel and Nvidia Hardware in it, and after reading this, I just couldn’t let such an incredible misinforming post left unanswered.
    Ryzen and Rome has been a massive success which is reflected in AMD’s stock skyrocketing over the year, so it would seem that there are people who disagrees with your claim, so much that they are willing to bet dollars against your claim. So when you use phrases such as “Ryzen overpromised and underdelivered” along with “AMD kiddies” One can tell that you are a moron... and a Intel kiddie. I could go into further details ripping the rest of your so called “facts” apart, but judging by your post, I will only be feeding a troll.
    It amazes me with retarded comments such as these, that there are people out there who willingly wants to be a fanboy of any company, so they ignore the facts and start wommitting crap on forums and websites such as we see above.
    Get a brain, stop getting emotionally attached to a company like intel, try start being able to think objectively. And you will find your self not expressing words and sentenses that makes other readers IQ fall after reading your posts.
    Reply
  • Deicidium369 said:
    Yeah, that tends to be how things work. Feed it more power and you get more performance - except in the case of AMD - since AMD has ultra binned ever single little slice of silicon, and offers it's pre over clocked. Along with that extra power comes extra heat - that is why it has a setting in BIOS to allow you to keep the 80W TDP or the 135W.

    Nvidia is is no way shape or form concerned about whatever marketing term AMD is using to describe whatever warmed over ancient design they are selling this week. It just isn't an issue. AMD has about as much chance taking over the discrete GPU for laptops as it does becoming the top tier full size GPU vendor. "Big Massive Totally New and Shiny Look! Squirrel Navi" is vaporware and likely to remain as such - they will undoubtedly release something with better perf - but no where near where the AMD Kiddies think it it be.

    Thing about AMD is they get things flipped - You under promise and over deliver - not over promise and under deliver - which is the history of the Ryzen line Because by now AMD should have 80% market share and Intel begging for scraps - yet that hasn't happened. Ryzen 1 (1x00 series) had the most impact, the subsequent 2 small iterative updates - not so much. Even though AMD is hitting record revenues (for AMD) it is still really nothing and R&D requires something in excess of nothing. Jim Keller's legacy at AMD (creator of Ryzen) is already in the market - they have nothing left to draw from. The changes between 1st gen Ryzen and most current in the market Ryzen is really not much at all. I know all the armchair lithography and CPU fabrication experts will say but but but "7nm"... turns out not that big of a deal. Not making money is why they spun off Global Foundries - and why they sold Ryzen designs to the Chinese - to keep the lights on while Ryzen was prepped to take over the world... a swing and a miss. I mean if Intel was in 1/10th the amount of trouble the AMD Kiddies think - then it should have been EZ Mode to fill in that gap - yet - no gap filled.

    Having a CPU with no integrated graphics (even rudimentary) with a bunch of superfluous cores (reason why Intel was at 4 cores for desktop - that's what the market wanted) is why OEMs are not selling alot of AMD systems - having to add a graphics card (another line item in the BoM) but the added load on the Support and IT depts was part of the consideration for alot of companies to stick with Intel - install the chipset drivers and you are good to go.

    The new 4x00 series APUs are great... Vast majority of businesses want basic Desktops - something to run Word, Excel, Outlook and Chrome. Don't need 8 cores and Vega graphics (still takes a sep driver)to do that.

    So no Nvidia isn't worried, Intel is not worried - The one indispensable service AMD provides to both Nvidia and Intel is to shut down ppl wanting to say they are a monopoly in their respective fields. Kinda like how a Morty is a shield for a Rick. Advanced Morty Defenses.
    You are just a salty uninformed poster. Intel fanboi
    Reply
  • mitch074
    Of course Nvidia is worried - even though they do have the performance crown, stuff like the crypto currency craze and Vulkan proved that AMD had enough resources even with a tenth the R&D budget and driver dev staff to keep within striking distance. Also AMD owns the console market, as soon as they could afford 2 GPU designs instead of 1 they would kick ass - and the half-baked RDNA design proved just that.
    RDNA2 is almost here, and since AMD actually stopped overpromising with Ryzen, if they announce 50% better performance per watt we can expect "Big Navi" to be interesting at the very least.
    I will finally be able to replace my faithful RX480 - almost 4 years old now and still good enough.
    Reply
  • TCA_ChinChin
    Mandark said:
    You are just a salty uninformed poster. Intel fanboi
    Some people are delusional. Nobody even said this was a bad thing from Nvidia (its actually quite a savvy move), but people assume whenever you don't directly praise Nvidia or Intel that you're personally insulting them and their purchasing decisions so they must justify themselves.
    Reply
  • revanthkr
    Does that mean rtx 2060 mobile which already has 80w tdp will get 115w bios update or its just the new pcs with rtx 2060 get this hike?
    Reply