Search the Community

Just over a month ago, the purported specifications of Nvidia's RTX 4060 laptop GPU had leaked. The reports suggested that the performance gap in favor of the 4060 against its RTX 3060 predecessor could be slim. Nvidia has since released the RTX 4060 laptop GPU in various devices. The card features 8GB of VRAM and it looks like the desktop counterpart is also going to be the same. Spotted by VideoCardz, Gigabyte appears to have accidentally leaked the specifications of the desktop RTX 4060 via its Gigabyte Control Center (GCC) tool. The release notes for the latest GCC version 23.03.02.01 lists support for colours of the RTX 4060 Gaming OC 8GB model. As a bonus, Gigabyte has also confirmed the memory capcity of the RTX 4070 as well. Its Aero OC RTX 4070 apparently has 12GB VRAM, just like the RTX 4070 Ti. The last gen RTX 3060 had 12GB and 3070/Ti featured 8GB. Here is a screenshot of the release notes showing the 4060 with 8GB and the 4070 with a 12GB memory buffer. In terms of the core specifications themselves, a report last year suggested that the RTX 4060 will have 36 steaming multiprocessors (SMs) or 4,608 CUDA cores. However, this is still unconfirmed and is unlikely. The laptop variant comes with 3,072 shaders which means the desktop counterpart might pack around 3,840. As for the RTX 4070, it purportedly features 5,888 CUDA cores per previous rumors (via RedGamingTech YouTube channel). Source: Gigabyte GCC release notes Gigabyte confirms Nvidia RTX 4060 has less VRAM than 3060, RTX 4070 also leaks out

The reviews for Nvidia GeForce RTX 4070 GPU show performance similar to RTX 3080 10GB and AMD Radeon RX 6800 XT GPU while being very power efficient. After months of waiting, Nvidia GeForce RTX 4070 is going to release to the public. Just before its release, the restriction on their reviews has been lifted. But that lifting of the review embargo isn’t without its restrictions. though. Nvidia has two sets of cards in the RTX 4070 series. One that run at default clocks and are priced at MSRP and another which are sold by graphics cards makers with overclocked specs and at higher MSRP. Today, only the default clocked reviews are out. With the reviews of aftermarket overclocked graphics cards allowed only a day later. The interesting thing about RTX 4070 is people think it’s mid-range. While it’s being officially priced at $599. We never understood the definition of a mid-range card costing a $600, that too at the base price. Whatever it is. The reviews are out and worth looking at. Tom’s Hardware Nvidia GeForce RTX 4070 Average FPS At 1440p. Credit: Tom’s Hardware. As always, Tom’s Hardware has done a review for the RTX 4070 Founders Edition by Nvidia. In terms of performance, they decided to include Ray Tracing performance in the default overall list. Something we disagree slightly. We don’t believe Ray Tracing is used by everyone and showing in as main numbers is unfair to AMD too. That’s why we have used just the non-RTX performance image above. At 1080p (page 5), Nvidia RTX 4070 performs as fast as RTX 3080 Ti and is slightly behind AMD Radeon RX 6800 XT. But no one buys such an expensive graphics card to game at 1080p. At 1440p (page 4), for which this graphics card is made for, the performance numbers change. Here, both RTX 3080 10GB and AMD Radeon RX 6800 XT performs similar to the Nvidia RTX 4070. These numbers are disappointing. The only thing here is that it leaves AMD far behind in Ray Tracing tests. At 4K (page 6), RTX 4070 performs nicely. While the performance similarity with AMD 6800 XT is comparable to 1440p above, Nvidia RTX 3080 10GB is found beating it. Which again is a disappointment. What impresses us is the overall performance at 4K, it still manages to run at a good average of 55 FPS, which is a good number for a “mid-range” graphics card depending on the game. Here too, the Ray Tracing numbers are way too good. Tom’s Hardware also did upscaling tests with DLSS2, DLSS3 and FSR2. The numbers for DLSS3, which is exclusive to the RTX 40 series, are really good. However, they complain that it adds latency, which doesn’t help things. The power usage numbers weren’t available on time, hence they have not been published by them. igor’sLAB Nvidia GeForce RTX 4070 Average FPS At 1440p. Credit: igor’sLAB. igor’sLAB tested the Nvidia GeForce RTX 4070 Founder’s Edition at base 200W and also at 220W. At 1080p (page 4), RTX 4070 200W performs the same as RTX 3080 10GB and is slightly faster than the AMD RX 6800 XT. However, when overclocked to 220W, it performs faster than AMD RX 6900 XT. The average 1% lows it’s slightly faster than RTX 3080 10GB though. At 1440p (page 5), RTX 4070 200W’s performs 3.5% slower than RTX 3080 10GB. The performance is similar to AMD RX 6800 XT too. That gap is reduced when overclocked to 220W, though. In 1% lows, AMD RX 6800 XT is slightly faster than Nvidia RTX 4070. At 4K (page 6), RTX 4070 is 1% faster than RTX 3080 10GB. This is because of the extra 2GB RAM which RTX 4070 offers. Here too, the performance is similar to that of AMD RX 6800 XT. In 1% lows however, AMD RX 6800 XT is 3% ahead, but when RTX 4070 is overclocked to 220W, that difference becomes negligible. Nvidia GeForce RTX 4070 Average Power Consumption In Watts At 4K. Credit: igor’sLAB. What’s important is the power consumption. Nvidia RTX 4070 uses average 200W in gaming at 4K. Whereas AMD RX 6800 XT requires 299W and Nvidia RTX 3080 10GB requires 312W for the same. TechPowerUp Nvidia GeForce RTX 4070 Average FPS At 1440p. Credit: TechPowerUp. Open the review link for full images. TechPowerUp has made the most comprehensive review for the Nvidia GeForce RTX 4070 Founders Edition. Additionally, they have written a review for the Gainward, PNY, MSI Ventus 3X and ASUS Dual versions too. However, the performance on all of them are similar. At 1080, Nvidia RTX 4070 is 2% faster than RTX 3080 10GB and is just 1% faster than AMD RX 6800 XT. That difference is reduced at 1440p, where the performance between all the three cards is virtually the same. However, while it performs the same as AMD RX 6800 XT, the Nvidia RTX 4070 a huge 5% slower than the RTX 3080 10GB at 4K. TechPowerUp however was very impressed by the DLSS3 performance and the temps of the FE card (page 37). Another thing is the power consumption of the card overall (page 39). It uses 200W max, whereas AMD RX 6800 XT used 298W and Nvidia RTX 3080 10GB 335W in gaming. TechSpot Nvidia GeForce RTX 4070 Average FPS At 1440p. Credit: TechSpot. TechSpot too reviewed the Nvidia GeForce RTX 4070 Founders Edition. At 1080p, the RTX 4070 was faster than both RTX 3080 10GB and AMD RX 6800 XT. However, at 1440p, that difference is slightly reduced. At 4K, all three perform virtually the same. In total PC power consumption, it doesn’t go above 320W, whereas an AMD RX 6800 XT PC used 409W and an Nvidia RTX 3080 10GB PC used 460W for the same. Conclusion Nvidia GeForce RTX 4070 overall performs similar to the Nvidia GeForce RTX 3080 and AMD Radeon RX 6800 XT. At 1080p is it is faster than the other two, at 144p, it is the same and at 4K, it lags behind a bit. Why, because it comes with just 192-bit memory bus, which is a major cut-down from previous-gen offerings from both Nvidia and AMD. It’s a massively power efficient card and the Ray Tracing performance is excellent too. But its problem is something else. RTX 4070 costs $599, which is the same as the original cost of RTX 3070 Ti. Whereas it performs similar to RTX 3080. This is not the generational upliftment expected. For example, generation after generation Nvidia has offered around 50% upliftment in performance at the same price. This time around, it’s just 30% increase at a higher price. Ideally it should have been costed $499 instead for it to have a RTX 3080 level performance. Even then it wouldn’t be justified. To take things further, a tweet by a well known info sharer @harukaze5719 shows, that the $599 price is just for the base model. The overclocked models by one of the card maker starts from $659 and goes till $829. Remember, we are talking about a mid-range graphics card here. Not to forget, the price in Europe and Asia is only going to be higher. Additionally. If Ray Tracing and DLSS3 performances are not up for consideration, then AMD Radeon RX 6800 XT can be found at a similar or even cheaper price, with 16GB VRAM, which is 4GB more than what Nvidia RTX 4070 offers. In all. Nvidia GeForce RTX 4070 is a good card, which should have been better and it costs too much. Nvidia RTX 4070 Reviews Show RTX 3080 Level Performance

But it's the cheapest way (so far) to add DLSS 3 support to your gaming PC. Nvidia's GeForce RTX 4070 Founders Edition is dwarfed by the 4090 and 4080 design, though third-party GPU makers can make bigger cards if they want. Andrew Cunningham Its specs have been thoroughly leaked for anyone paying close attention, but today Nvidia is making it official: It's launching its GeForce RTX 4070 graphics card for a starting price of $599. This GPU will be the cheapest way to buy into Nvidia's Ada Lovelace GPU family, which, in addition to better performance and power efficiency, gets you access to Nvidia's DLSS 3 upscaling and frame-generation features and hardware acceleration for AV1 video encoding. Cards will be available from Nvidia and other partners starting on April 13th. Nvidia's Founders Edition card is significantly smaller than its RTX 4090 and 4080 GPUs, with a relatively svelte dual-slot cooler that will easily fit into most PC cases. GPUs made by Nvidia's partners can (and will) be more expensive and physically larger, and some of them will also come with 8-pin PCI Express power connectors rather than the newer but somewhat controversial 12VHPWR power connector that Nvidia uses. The RTX 4070 is $100 more expensive than the $499 starting price of the RTX 3070 and RTX 2070, which were $120 more expensive than the $379 starting price of the GTX 1070. Even accounting for inflation since 2016 ($379 in 2016 is roughly $475 now), it continues a tradition of bumping up the price for GPUs with the same relative position in Nvidia's lineup, something Nvidia partially (and not untruthfully) justifies by pointing out how much faster its modern graphics cards are. What does $599 get you? RTX 4090 RTX 4080 RTX 4070 Ti RTX 4070 RTX 3080 Ti RTX 3080 10GB RTX 3070 Ti RTX 3070 RTX 3060 CUDA Cores 16,384 9,728 7,680 5,888 10,240 8,704 6,144 5,888 3,584 Boost Clock 2,520 MHz 2,505 MHz 2,610 MHz 2,475 MHz 1,665 MHz 1,710 MHz 1,765 MHz 1,725 MHz 1,777 MHz Memory Bus Width 384-bit 256-bit 192-bit 192-bit 384-bit 320-bit 256-bit 256-bit 192-bit Memory Clock 1,313 MHz 1,400 MHz 1,313 MHz 1,313 MHz 1,188 MHz 1,188 MHz 1,188 MHz 1,750 MHz 1,875 MHz Memory size 24GB GDDR6X 16GB GDDR6X 12GB GDDR6X 12GB GDDR6X 12GB GDDR6X 10GB GDDR6X 8GB GDDR6X 8GB GDDR6 12GB GDDR6 TGP 450 W 320 W 285 W 200 W 350 W 320 W 290 W 220 W 170 W The RTX 4070 shares a lot in common with the 4070 Ti, the GPU originally announced as the "RTX 4080 12GB" before being "unlaunched" and relaunched. They share the same AD104 GPU die, a smaller chip than the AD102 flagship used in the RTX 4090 or the AD103 used in the 4080 series. They also both use 12GB of GDDR6X memory on a 192-bit interface. The 4070 has fewer CUDA cores (5,888, down from 7,680), but the 4070 and 4070 Ti are a whole lot more similar than the 16GB and 12GB RTX 4080 cards would have been. That 192-bit memory interface is narrower than the 256-bit interface used by the 3070 and 3070 Ti, which Nvidia has compensated for by adding 32MB of extra L2 cache to the 4070 (for a total of 36MB). That cache, plus much faster base and boost clock speeds, are the 4070's biggest improvements over the last-gen cards. There's a lot to recommend the 4070 on paper. It's the first Ada Lovelace graphics card to launch for closer to $500 than $1,000, it's the cheapest way to buy into RTX 4000-series features, and it's at least possible to buy versions that will physically fit inside smaller computer cases. It also promises to be significantly more power-efficient than the 3070 and 3080-series cards it's replacing (and competing with, given how widely available new 3000-series GPUs still are). But in terms of raw rasterized game performance, it doesn't look like quite as big a slam dunk as the 3070 did when it was released. That GPU could soundly beat the more-expensive previous-generation RTX 2080 and 2080 Super for less money while trading blows with the RTX 2080 Ti. Nvidia's performance numbers for the 4070 show it beating a 3070 Ti and either narrowly beating or narrowly losing to the 10GB version of the RTX 3080. That's not bad, especially if you can actually buy a 4070 for $599—the 10GB 3080 launched for $699, and it still sells for at least that much unless you're buying used or refurbished—but previous midrange GeForce cards have done a slightly better job of bringing last-generation flagship performance down to more reasonably priced PCs. When combining DLSS upscaling with DLSS frame generation, Nvidia can (under the right conditions) produce a more detailed, higher framerate image without requiring the raw GPU performance it would normally take to render that image. Nvidia Nvidia is leaning on DLSS 3 and its AI-accelerated frame generation to make those numbers more impressive. This feature, you might recall from our RTX 4090 review, "doubles" your frame rate by creating one AI-interpolated frame for every frame that the GPU renders. DLSS FG works in concert with the typical "super resolution" upscaling feature that adds pixels to the image your GPU is rendering to make it look more detailed. The result can be a clean-looking, high-resolution, high-frame rate image rendered using just a fraction of the GPU performance that would be needed to render the same scene natively. Nvidia claims the RTX 4070 can be 1.4 times as fast as an RTX 3080 when DLSS FG is enabled. But as we found in our RTX 4090 review, it tends to work best when your GPU is already pumping out reasonable frame rates, making imperfections in the AI-generated frames harder to perceive. In other words, it's a lot better at making a 60 fps image look like a 120 fps image than it is at making a 15 fps image look like a 30 fps image. DLSS FG can also add latency, making it harder to recommend for twitchy competitive first-person shooters, and it's compatible with a relatively narrow selection of games (though Nvidia's list of "50-plus titles" includes big names like Microsoft Flight Simulator, Diablo IV, and Cyberpunk 2077). On AMD's end, we're still waiting for the RX 7000 series and the RDNA3 architecture to trickle down to cheaper GPUs, so AMD is fighting this battle with last-gen cards. A recent email from AMD, likely sent in anticipation of the 4070's release, played up the $579 and $649 MSRPs of the Radeon RX 6800 XT and 6950 XT. If you can't find a 4070 in stock for $599—and if you don't mind the lack of DLSS support and inferior ray-tracing performance—either is a plausible alternative to the 4070 for high-end 1440p to low-end 4K gaming. We'll take a closer look at the RTX 4070 in a full review soon. Nvidia’s $599 GeForce RTX 4070 is a more reasonably priced (and sized) Ada GPU

It's been the worst-kept secret in the graphics chip world for a while, but today NVIDIA has finally revealed the new GeForce RTX 4070 and 4070 TI GPUs. These new graphics chips are designed to offer more affordable performance in high-end PC gaming compared to the earlier, and more expensive, GeForce RTX 4080 and 4090 GPUs. In fact, you may be able to get a new 4070 graphics card for just $599. The GeForce RTX 4070 has been designed to give gamers a way to play their high-end PC games at frame rates of 100 fps at 1440p resolution, while still also supporting features like DLSS 3 and ray tracing. NVIDIA states: Compared to the RTX 2070 SUPER, the GeForce RTX 4070 is 2.6x faster with DLSS 3 on average, and the GeForce RTX 4070 is 1.4x faster than the GeForce RTX 3080 with DLSS 3 on average. DLSS 3 provides Ada GPUs with a tremendous performance boost, but the GeForce RTX 4070 also excels in traditional games that don’t include more advanced features such as ray tracing and DLSS. In these rasterized games, the GeForce RTX 4070 is on par with the GeForce RTX 3080 while running at nearly half the power, and offering an additional 2GB of memory. The NVIDIA GeForce RTX 4070 Ti GPU is available in limited quantities right now. The NVIDIA GeForce RTX 4070 will launch tomorrow, April 13. NVIDIA will offer its Founders Edition version of the graphics card, while several third-party companies will release their own GeForce RTX 4070 cards. From now until May 8, people who buy a GeForce RTX Series 40 card, or get one inside their new gaming PCs, will be able to redeem a code for an Overwatch 2 Ultimate Battle Pass Bundle, with a lot of in-game currency and items for Blizzard's multiplayer shooter. NVIDIA finally reveals the GeForce RTX 4070 and 4070 TI GPUs

Nvidia recently "unlaunched" its RTX 4080 12GB SKU as a result of the backlash from the community over the confusing naming scheme. The company, as usual, thought it was a perfectly fine idea to announce two very differently specced GPUs under the same model name. Despite the SKU names suggesting that the RTX 4080 16GB and the RTX 4080 12GB only had a difference in memory capacity, the latter, in reality, sports 20% fewer CUDA cores as well, since it comes with 7,680 CUDA cores compared to the 4080 16GB's 9,728. Reports suggested that Nvidia was rebranding the 4080 12GB model to an RTX 4070 Ti, and the report has since been confirmed by AIB partner colourful, which listed the specs on its website by mistake before taking it down. Today, Twitter user and leakster has kopite7kimi has shared the alleged details of the RTX 4070 non Ti. Apparently the car will share the same memory subsystem of the 4070 Ti model but will have fewer CUDA cores. This means it will also pack 12GB GDDR6X running at 21Gbps for a bandwidth of 504.2GB/s. The CUDA core count gets down to 5,888 in the 4070, up from 7,680 in the 4070 Ti. As such, the TGP(total graphics power) also gets reduced by 35W down to 250W. Some updates of RTX 4070: PG141-SKU336/337 AD104-250-A1 5888FP32 12G 21Gbps GDDR6X 36M L2 250W Compared to the Ampere-based RTX 3070 and 3070 Ti, the 4070 Ti should be bigger jump in relation to the alleged RTX 4070. This is because the 4070 is said to have the same core count as last gen's 3070 but the 4070 Ti is seeing a big increase in cores up from 6 ,144 in the 3070 Ti to 7,680. There is no info on pricing of the RTX 4070 series GPUs but Nvidia may stick the original MSRP of $899 of the 4080 12GB on the 4070 Ti. But this will entirely depend on the competence of the AMD RX 7900 series and how well they do. Speaking of the performance of the RX 7900 XTX and 7900 XT, we now have leaked a few leaked benchmarks of those cards, and they don't paint the prettiest of pictures all the time. Source: kopite7kimi (Twitter) Alleged Nvidia RTX 4070 and 4070 Ti specification details are now out

New leaks suggest that Nvidia could be making a cut-down GDDR6 VRAM version of GeForce RTX 4070 graphics card instead of currently used GDDR6X. Nvidia launched the GeForce RTX 4070 graphics card more than a year ago. At launch, the reviews found it to be performing similarly to the RTX 3080 and AMD Radeon RX 6080 XT graphics cards. It was later rumored that RTX 4070 wasn’t selling that much after the launch and Nvidia had cut its GPU’s supply to the board partners. Later, Nvidia was working on the RTX 4070 Super version, along with Super versions of other graphics cards. These Super versions have improved performance over their non-super variants. They also have better value and seem to selling better. Interestingly, Nvidia used GDDR6X memory chips in all the graphics cards from RTX 4070 and above in the RTX 4000 series. But it looks like that’s likely to change. RTX 4070 Cut To GDDR6 Well known Twitter (now X) based leaker MEGAsizeGPU has revealed (via VideoCardz and Tom’s Hardware) that Nvidia is working on another variant of the RTX 4070 with GDDR6 VRAM instead of GDDR6X one. In the tweet, MEGAsizeGPU has revealed that the GDDR6 VRAM model will be running at 20Gbps, without any memory overclocking. Which is very similar to the GDDR6X model, which runs at 21Gbps. The other specs like the CUDA core count and GPU seems to be the same. The overall memory bandwidth will likely decrease from 504 GB/s to 480 GB/s. Which is just 5% decrease. Though, we feel the power usage might go down further. Why is Nvidia doing this, though. The likely explanation seems to be the shortage of GDDR6X memory. Several days ago, Tom’s Hardware reported that GDDR6X memory is in short supply. But that issue is just temporary. So it’s unlikely that Nvidia would be making a new model just for that temporary shortage. Another possibility is that Nvidia is trying to save some costs. The GDDR6X is a VRAM that Micron specially designed in collaboration with Nvidia for its graphics cards. At the time of its launch, it was faster than the GDDR6 VRAM and still likely costs more than it. The pricing of this model isn’t available yet though. Interestingly, MEGAsizeGPU had predicted way back in October 2023 that Nvidia would be making a GDDR6 version of RTX 4070. That tweet had also predicted the launch of the RTX 4070 Super variant and that turned out to be true. Whatever the reason. If Nvidia does launch the GDDR6 variant of RTX 4070, we are hoping that Nvidia labels it correctly, so that the buyers are not misguided by it. Source Hope you enjoyed this news post. Thank you for appreciating my time and effort posting news every single day for many years. 2023: Over 5,800 news posts | 2024 (till end of July): 3,313 news posts

The sales of Nvidia GeForce RTX 4070 are so bad that Nvidia has temporarily stopped supplying the GPUs to its board partners. More than a month ago, Nvidia released the GeForce RTX 4070 graphics card to the public. The reviews for the card weren’t that great. It costs as much as RTX 3070 Ti while performing only as good as RTX 3080. The generational uplift in performance is just 30%. Which is far lesser than 50% performance uplift which Nvidia usually provides. While the power efficiency and ray tracing performance of RTX 4070 is good, overall the card has been a big disappointment. Also, though the card is having good sales as per the data available for the German retailer Mindfactory, that data is limited just to the region. In fact, outside , the sales of the RTX 4070 are so bad that Nvidia has temporarily halted the supply of the GPU to its board partners. RTX 4070 Supply Halted MoneyDJ reports (translated) that Nvidia had stopped the supply of the RTX 4070 GPU to its board partners in April itself due to poor sales. It says that while the supply was expected to resume in May. it hasn’t. In fact, it could be that this halting of GPU supply might continue well into the month of June, too. The report mentions that while initial supply could have been good, overall it’s not selling much due to inflation and other reasons such as hike in interest rates. Because of that, only the loyal Nvidia fan-base is expected to buy the card, which is not enough to increase the overall sales. Meanwhile, the Nvidia board partners are eagerly waiting for Nvidia to release RTX 4060 series, so that sales of graphics card board partners increase again. The report also points out that similarly, even RTX 4080 wasn’t selling much. At least as compared to RTX 4090 which was released earlier. Hence, to prevent over-inventory of not one, but two GPUs, Nvidia seems to have stopped the supply of RTX 4070. Reasons The report gives out some reason why sales of RTX 4070 are not great. Like Nvidia concentrating more on servers. Then Nvidia controlling the inventory to make sure RTX 30 series graphics cards are completely sold out first. The third reason given is to control the price. Nvidia probably doesn’t want prices to get too cheap. The biggest reason we personally feel is that RTX 4070 is just not good enough. It costs a lot and it competes only with RTX 3080 and AMD’s RX 6800 XT. So nothing ground breaking. Additionally, it comes with 12GB VRAM, which would have been fine, but games these days are really demanding so 16GB VRAM is becoming a must for a card like RTX 4070. Overall sales of such cards are expected to disappoint. We feel the only way Nvidia can fix it is not to control it’s supply, but to make them cheaper for everyone. Because it doesn’t cost that much to make them. RTX 4070 GPU Hardly Selling, Nvidia Has Halted Its Supply