Tag Archives: Nvidia

300 Nvidia GPUs Seized After High Speed Boat Chase

This site may earn affiliate commissions from the links on this page. Terms of use.

As GPU prices have increased, cryptocurrency miners and gamers have resorted to once-unusual tactics for buying video cards, like paying monthly fees to bot subscription services or paying thousands of dollars on eBay.

It’s always fun to watch the cost of something explode until you can play games calculating whether the street price of various illegal substances is more or less per ounce than your average GPU. The illegal substances are generally still winning, but the skyrocketing price of cryptocurrency has had a similar impact on the cost of video cards. Driving up the cost of video cards has made them rarer and more-desired commodities, which is how we arrive at the point in this story where Hong Kong customs officials are chasing smugglers in speed boats to crack down on illicit GPU sales.

There’s a certain dark hilarity in imagining drug dealers across the world offering their clientele multiple ounces of weed or an RTX 3060, but in this case, the haul consisted of low-end 30HX CMP cards. Nvidia offers a range of CMP cards, with performance ranging from 26MH/s to 86MH/s.

The 30HX and 40HX are believed to be based on Turing silicon — the GTX 1660 Super and RTX 2070, respectively. The 50X and 90HX are harder to pin down. The 50HX is a touch faster than the known mining performance of the RTX 2080 Ti, while the 90HX is about 10 percent slower than the known mining performance of an RTX 3080. If the 50HX is based on the RTX 2080 Ti, it’s fielding a smaller amount of VRAM; the RTX 2080 Ti offered 11GB, while the 50HX has just 10GB.

We’re hoping they just took the cards *out* of the boxes rather than thinking the entire lot was shipped this way. Seawater exposure is not known to enhance overclocking performance.

We’ve been tracking the peanut-butter-and-clusterf**** sandwich that is the modern GPU (and CMP) market for several months now. There are three reasons why the GPU market looks like it does: cryptocurrency mining, Samsung yields, and pandemic-related semiconductor demand.

At the moment there’s mixed messaging on when semiconductor shortages could ease, depending on which parts you care about. CPU shortages might ease by the back half of the year, but we’re still hearing that GPU shortages could drag on into 2022. Whether they do is going to depend on which factors are starving the market today and whether the companies responsible for them can meaningfully improve the situation in the next few months. If you’re looking for a card, we’ve written a guide to the most affordable GPU options available today.

Now Read:

Let’s block ads! (Why?)

ExtremeTechGaming – ExtremeTech

Report: Nintendo ‘Switch Pro’ Will Feature DLSS, New Nvidia Silicon

This site may earn affiliate commissions from the links on this page. Terms of use.

A few weeks ago, we wrote about the possibility the rumored Switch refresh coming from Nvidia and Nintendo later this year would use DLSS. Rumors have claimed that the new Switch (sometimes referred to as the Switch Pro) would target 4K rendering when plugged into a dock. Modern gaming GPUs aren’t efficient enough to drive AAA titles at native 4K on handheld power budgets, so it seemed likely that any 4K-capable hybrid solution would use either DLSS or an approach like checkerboard rendering to hit that resolution target.

Today, agreeably, Bloomberg ran a story claiming that the new Switch will indeed use DLSS to reach 4K when played in docked mode. Because the feature will require additional code, it may only be supported on new titles going forward, and the report reiterates claims that the upcoming Switch will have a larger display (7 inches, up from the current 6.2 inches). This doesn’t necessarily require a form factor change. Eliminating the bezels from the current Switch would allow Nintendo to increase the screen size up to 7.5 inches, without changing the physical device at all.

A hypothetical Switch with no bezels. Image posted to Reddit by user agiantbluewhale

DLSS is Nvidia’s deep learning technology for enhancing low-resolution images into high-resolution renders using AI. The technique requires tensor cores, which means a Volta or later GPU, but it offers the potential for much higher resolutions than a device like the Switch can render natively.

Making DLSS a built-in capability of newer titles also allows Nvidia to keep offering support for the mainstream Switch. Both flavors of the platform will target 720p by default in handheld mode, so the only difference between the Switch and Switch Pro would be the latter’s ability to hold a 720p image where the Switch drops below that target to keep its frame rate up. When docked, the standard Switch would raise its clock somewhat and continue to target somewhere between 720p and 1080p, depending on the title. The Switch Pro would have the option to engage DLSS and aim for 4K. In theory, Nintendo and Nvidia could work to enable other resolution targets besides 4K, or to boost the frame rate up to 60fps instead of holding 30fps.

Sony and Microsoft allow game developers to target a mixture of experiences, and gamers can sometimes choose between playing a game at 1080p60 or 4K30, for example. Nintendo hasn’t previously offered this degree of flexibility, but the Switch has been something of a game-changer for the Japanese company.

Nintendo Finally Has a Reason to Chase Cutting-Edge Silicon

Nintendo has never particularly chased silicon specifications. There have been times in the console cycle when the company’s platforms led its competition — the SNES was a better console than the Genesis, despite a lower CPU clock — but the last time Nintendo attempted to position its own platform (mostly) head to head against Sony and Microsoft was 20 years ago, with the GameCube.

The Switch followed Nintendo’s conservative design philosophy by tapping a commodity chip and a known silicon design, but that’s not how mobile device manufacturers maximize performance or battery life. It’s still possible Nintendo would tap a newer bulk Nvidia SoC, but the more I look at the situation, the more likely it seems Nintendo might have paid Nvidia for some custom design work. Company executives have made statements implying Nintendo may have changed its approach to system design:

If Nintendo stuck with an existing Nvidia solution, Xavier NX would be the likely chip of choice. Xavier is huge, though, with a die size of 350mm sq compared with just 118mm sq. for the original Tegra X1 SoC. Its CPU cores are only clocked at 1.2GHz in quad-core mode (when operated in a 10W power envelope) and it’s built on a 12nm FinFET process. Either 7nm or 5nm (7nm being more likely) would be better for power consumption. The GPU is larger (512 cores), but based on the older Volta architecture, not Ampere or Turing.

Paying Nvidia for a custom SoC would give Nintendo the freedom to target the features it wanted for Switch Pro without paying (in $ or die space) for hardware blocks it doesn’t need. Nvidia doesn’t currently manufacture a quad-core Cortex-A78 SoC with 256-512 Ampere GPU cores. Nvidia’s Ampere-equipped Orin SoC won’t hit the market until 2022, and it’s a 750W TDP part intended for vehicles. We’re guessing the Switch Pro probably doesn’t represent a truly radical experiment in die recovery.

Bloomberg reports that the Switch Pro could be priced between $ 349 and $ 399 and a price increase makes sense if Nintendo is planning to pay for custom silicon. This could be seen as somewhat in tension with reports Nvidia has ceased production of the current Switch SoC. Nintendo has introduced larger versions of its handheld devices before, but it’s never forced its customers to adopt a more-expensive version of a handheld by phasing out an older model.

The company’s last set of earnings slides state:

“When adapting a new technology, we go to great lengths to ensure we offer new and innovative content that complements it… when we do decide to use an existing technology, we actively pursue collaborations with partner companies that specialize in that technology.”

DLSS could be a truly game-changing technology for the Switch Pro. The current Switch is based on a 7-year-old GPU and an eight-year-old CPU. The Cortex-A57 was never a particularly power-efficient CPU core — it’s the same CPU that powered the Snapdragon 810 — so stepping up to a modern design based on the A76 or A78 would be advantageous for Nintendo. Xavier NX offers 51.2GB/s of memory bandwidth and we can expect any custom silicon to at least match this level, or at least 2x what’s available on the Switch today. If Nintendo had pushed the envelope back in 2017, the gap between the Switch and Switch Pro would be smaller, but Nintendo relied on commodity hardware when it built the handheld in the first place, which means a jump to either the leading edge or n-1 would deliver a proportionally larger boost.

A custom, Ampere-based Switch Pro ought to be able to offer a rock-solid 30 or 60fps in 720p mode, while a DLSS-powered 4K mode worth playing seems plausible, given the combined gains in memory bandwidth efficiency and raw memory bandwidth. Nintendo would almost certainly follow in the footsteps of Sony and Microsoft in requiring that Switch games work well on both Switch and Switch Pro, to avoid alienating customers.

Nintendo doesn’t try to compete head-to-head with Sony and Microsoft the way it used to, but it still pays close attention to how its products are positioned in-market against the other firms. Adding 4K capabilities to the Switch allows Nintendo’s handheld to claim the same render target as the far larger consoles, delivered in a fraction of the power. Nintendo cannot build a Switch Pro that competes with the PS5 or Xbox Series S|X on raw power, but the combination of cutting-edge manufacturing, updated hardware architectures, and AI could shrink the gap between the current Switch and other consoles significantly. This time around, Nintendo may feel the gains are worth the cost.

Now Read:

Let’s block ads! (Why?)

ExtremeTechGaming – ExtremeTech

ET Deals: Save $550 On Dell XPS 8940 Intel Core i5 Nvidia GTX 1660 Ti Gaming Desktop, Roborock S6 Robot Vacuum and Mop for $401

This site may earn affiliate commissions from the links on this page. Terms of use.

Today you can save $ 550 on a gaming desktop from Dell that comes equipped with an Intel Core i5 processor and an Nvidia GeForce GTX 1660 Ti graphics card. This makes the system well suited for gaming at 1080p resolutions. There’s also an excellent discount on a Roborock S6 robot vacuum that’s marked down to just $ 401.99.

  • Dell XPS 8940 Intel Core i5-10400 Gaming Desktop w/ Nvidia GeForce GTX 1660 Ti GPU, 16GB DDR4 RAM, 256GB NVMe SSD and 1TB HDD for $ 729.99 from Dell with promo code DTXPSAFF323 (List price $ 1,279.99)
  • Roborock S6 Robot Vacuum and Mop for $ 401.99 from Amazon with promo code ROBOROCKS6 (List price $ 649.99)
  • Dell Vostro 5000 Intel Core i7-10700 Desktop w/ Nvidia GeForce GT 730, 8GB DDR4 RAM, 256GB M.2 NVMe SSD and 1TB HDD for $ 779.00 from Dell (List price $ 1,427.14)
  • Logitech M330 Silent Plus Wireless Mouse for $ 12.99 from Amazon (Regularly $ 29.99)
  • Dell Vostro 15 7500 Intel Core i7-10750H 15.6-Inch 1080p Laptop w/ Nvidia GeForce GTX 1650 GPU, 8GB DDR4 RAM and 256GB NVMe SSD for $ 949.00 from Dell (List price $ 1,712.86)
  • Dell UltraSharp U2520D 25-Inch 2K USB-C Monitor + $ 100 Gift Card for $ 389.99 from Dell (List price $ 519.99)

Dell XPS 8940 Intel Core i5-10400 Gaming Desktop w/ Nvidia GeForce GTX 1660 Ti GPU, 16GB DDR4 RAM, 256GB NVMe SSD and 1TB HDD ($ 729.99)

Dell’s new XPS 8940 features an updated design and it comes loaded with strong processing hardware that’s able to tackle just about any task you throw at it. The mid-range Intel Core i5-10400 with its six CPU cores is well suited for running numerous applications at the same time. As the system also has an Nvidia GeForce GTX 1660 Ti graphics card, the system’s is able to run games with high settings fairly well, making it a fitting machine for gaming and work. Currently you can get one of these systems from Dell marked down from $ 1,279.99 to just $ 729.99 with promo code DTXPSAFF323.

Roborock S6 Robot Vacuum and Mop ($ 401.99)

This high-powered robot vacuum has 2,000Pa of suction power and it has a built-in mop function, which makes it a versatile cleaning tool for your home. The Roborock S6 was also built to be fairly quiet with an average cleaning volume of just 56dB. Currently, these robot vacs are selling on Amazon marked down from $ 649.99 to just $ 401.99 with promo code ROBOROCKS6.

Dell Vostro 5000 Intel Core i7-10700 Desktop w/ Nvidia GeForce GT 730, 8GB DDR4 RAM, 256GB M.2 NVMe SSD and 1TB HDD ($ 779.00)

Dell’s Vostro computers were designed as office and business solutions, and this Vostro 5000 is no different. It’s equipped with an Intel Core i7-10700 processor and 8GB of RAM, which gives the system solid performance that’s perfect for a wide range of office and work tasks. Dell is offering these systems for a limited time marked down from $ 1,427.14 to $ 779.00.

Logitech M330 Silent Plus Wireless Mouse ($ 12.99)

The M330 mouse from Logitech was built to be an affordable wireless mouse with solid performance. The mouse reportedly produces 90 percent less noise when clicked than a standard mouse and it can also last for up to 2 years on a fully charged battery. Amazon is offering these mice at the moment marked down from $ 29.99 to just $ 12.99.

Note: Terms and conditions apply. See the relevant retail sites for more information. For more great deals, go to our partners at TechBargains.com.

Now read:

Let’s block ads! (Why?)

ExtremeTechExtreme – ExtremeTech

Nvidia Doubles GeForce Now Subscription Price to $10 Per Month

This site may earn affiliate commissions from the links on this page. Terms of use.

Nvidia has been trying to make cloud gaming workable longer than any of the other major players. Its current GeForce Now platform has gone through a few incarnations, and the current one is changing in a way subscribers won’t like much. Nvidia is increasing the paid tier from $ 5 per month to $ 10. This brings it more in-line with other streaming platforms, but the price was Nvidia’s big advantage until now. 

GeForce Now started its life in 2013 as Nvidia Grid, which was available only on Nvidia’s Shield Android handheld. Nvidia eventually moved to a storefront model like Google Stadia uses today, but it was unable to convince gamers to buy games and pay for a monthly service. Today’s GeForce Now is bring-your-own-games, but you’ll need a monthly subscription to have a good experience. 

Today’s changes don’t eliminate the free tier of GeForce Now — it’s just as frustrating as it was before. If you don’t pay for priority access, it can take time to get space on a server to run your game, and sessions are limited to an hour. With Priority (previously called Founders), you can get playing immediately, and sessions are valid for up to six hours. A subscription also enables ray tracing in supported games. 

The $ 5 per month pricing was a great deal, particularly if you already have a large library of games. GeForce Now connects to Steam, Epic, and GOG to mirror your games in the cloud, and it’s added support for a good number of titles. There are hundreds of compatible GeForce Now games, whereas Stadia only has about 100. Microsoft’s xCloud does a little better by the numbers, but it’s a pure subscription model that drops games on occasion. 

Nvidia is only charging the new $ 10 monthly fee to new subscribers. Unfortunately, if you’re reading this and have not subscribed, it’s too late to get in on the grandfathered price. Founders who cancel will only be able to re-subscribe at the higher $ 10 rate. 

In addition to the pricing change, Nvidia has announced it’s bringing new data centers online in Phoenix, Arizona and Montreal, Canada. This should help keep latency low for anyone in these general regions. The service is also now available in Turkey, and it will expand to Saudi Arabia and Australia soon.

Now read:

Let’s block ads! (Why?)

ExtremeTechGaming – ExtremeTech

Nvidia Accidentally Releases RTX 3060 Driver With No Mining Limiter

This site may earn affiliate commissions from the links on this page. Terms of use.

Up until now, claims that miners had found a way around Nvidia’s Ethereum mining limitations on the RTX 3060 have been fake. That’s now changed after Nvidia inadvertently provided a beta driver capable of mining at full speed on the RTX 3060.

The driver in question was beta 470.05 and it’s already been yanked. With previous drivers, mining performance on the RTX 3060 was hitting about 28MH/s. With the 470.05 driver, the RTX 3060 is hitting 44-48MH/s. Whoops.

This news raises questions about the accuracy of the following tweet from Bryan Del Rizzo, Nvidia’s communications director:

These events don’t mean there isn’t some kind of secure handshake between the BIOS and the driver under normal circumstances, but clearly, that handshake can itself be modified or omitted by Nvidia in driver code without any kind of failure or problem. Hopefully, miners won’t be able to figure out how this driver bypasses the code lockout.

A specific BIOS version is rumored to be required to use the driver, but other reports have indicated this isn’t necessary. Nvidia has yanked access to the driver from its own portals, but it’s likely to pop up again in other places. Miners aren’t going to let this one go. Nvidia’s rate-limiting should still be of some value here, since not literally every miner will find out about this driver, but this release effectively undoes some of the work the company had previously done to limit mining on its GPUs.

Here’s the good news. If Nvidia releases a hypothetical RTX 3080 Ti with the same mining limits as the RTX 3060 (limits not shared by the vanilla RTX 3080), said card won’t be supported in the GeForce 470.05 beta driver. If Nvidia was going to make a mistake like this, it made it at the right time — namely, before it launches any refreshed cards with mining limits.

Limiting cryptocurrency mining on GPUs may or may not improve overall availability, but it’s one of the only options for manufacturers to plausibly deploy. Nvidia has the right idea with the RTX 3060, in our opinion, this mistake notwithstanding.

GPU Availability Is Not Improving

In other news, DigiTimes is reporting that Nvidia GPU availability is “unlikely to ease by the third quarter of this year.” This could still change as the year progresses, but it’s likely to be due to a change in demand rather than supply. TSMC and Samsung should have full visibility into their own yields and wafer starts for the various chips AMD and Nvidia need to ship. AMD isn’t mentioned in the DigiTimes story, but Team Red has been facing its own severe GPU shortages.

If yields and shipments can’t grow, demand will have to drop. The two most likely causes for such an event are the slow end to the pandemic and a loosening of restrictions as more people are vaccinated and the end of the current crypto-mining craze. We don’t know how demand for electronics will shift as people can go outside again. We do know that the crypto market will probably cool off eventually, but not when. If demand stays high and the market is undersupplied through the end of the year, it could be 2022 before pricing stabilizes.

Now Read:

Let’s block ads! (Why?)

ExtremeTechGaming – ExtremeTech

Nvidia Reportedly Ending Production of Nintendo Switch Processor

This site may earn affiliate commissions from the links on this page. Terms of use.

Nvidia has been interested in making its own mobile chips for years, going all the way back to 2008 and the original Tegra chip that powered the Microsoft Zune HD. Tegra has been at the heart of more successful devices in recent years, most notably the Nintendo Switch. Things might be changing in Nintendo’s lineup if a new report is to be believed. According to “a person familiar with the matter,” Nvidia is ending production of the Tegra chip that powers the Switch, PCMag reports. 

The Tegra X1 system-on-a-chip (SoC) is designed around technology that’s a few years old, but it can still pull its weight thanks to the powerful GPU. The original Switch used the stock Tegra X1 (codename Erista), but that chip had a vulnerability that modders used to modify the Switch’s system software. Nintendo updated the console in 2019 to use a new Tegra X1+ (codenamed Mariko). Now that chip is apparently finished. 

There are a few ways this could affect Nintendo, including not at all. It’s possible Nintendo intends to use the same Mariko chip for all its future Switch hardware, and Nvidia ending production won’t change that. Nintendo could be working to stockpile all the chips they’ll need for the rest of the Switch’s life cycle, in which case, this news won’t affect you at all. 

It’s also plausible that Nintendo plans to upgrade the Switch with better hardware. There are rumors of a high-end Switch revamp, featuring a larger OLED display and support for 4K docked gaming. The current Tegra X1+ would struggle with 4K, so many have speculated it could use AI-powered DLSS to upscale graphics. However, a more powerful SoC could just do 4K natively. 

TegraK1-Full

Nvidia, which is trying to acquire UK-based Arm Holdings, has several newer Tegra chip designs, but they’ve only appeared in developer-focused products and a few cars. The newest is Orin, which was announced in 2019, but we know almost nothing about it other than it would be much faster than the Mariko chip. 

There may also be an outside chance that Nintendo is going to struggle to get enough parts to keep building the Switch. Although, I have to think Nvidia wouldn’t end production if Nintendo was still buying millions of current-gen Tegras. If this report proves accurate, I think it will serve to legitimize the rumors of a revamped Switch console. When we’ll see that is anyone’s guess, though.

Now read:

Let’s block ads! (Why?)

ExtremeTechGaming – ExtremeTech

Nvidia Wins Billion-Dollar Crypto Mining Lawsuit

This site may earn affiliate commissions from the links on this page. Terms of use.

Nvidia has won a lawsuit brought against it by multiple investors who claimed the company had deliberately and recklessly misrepresented the provenance of over a billion dollars in crypto-mining sales during the last boom in 2017-2018.

To briefly recap: A few years ago, when cryptocurrency-related demand for GPUs was through the roof, Nvidia told investors that the bulk of the demand it was seeing was actually for gaming GPUs. The previous cryptocurrency boom of 2013-2014 had targeted AMD cards almost exclusively, so the idea that miners were still preferring GCN over Pascal had a certain credibility to it. Nvidia’s CEO Jen-Hsun Huang dismissed the idea that cryptocurrency was having a significant impact on Nvidia sales at multiple points. This part of the record is uncontested.

The reason Judge Haywood Gilliam has dismissed the case is straightforward: None of the evidence presented by the plaintiffs was sufficiently detailed to support the weight of the accusations brought against Nvidia. While the lawsuit identified resources it claimed guided Nvidia’s actions with respect to the cryptocurrency market, it didn’t produce enough direct evidence to satisfy the judge. Having already been granted leave to amend their case once, the judge denied them leave to do so a second time.

Judge Gilliam may have cleared Nvidia of legal wrongdoing, but the question of what the hell the PC channel is supposed to do about the ongoing GPU disaster hasn’t gotten the attention it needs. Even if the market improves more quickly than expected — which is to say, it improves during 2021 — there’s nothing stopping this from happening again. Every time the price of cryptocurrencies spikes, GPU availability dies. Things have gotten bad enough that miners are targeting laptops, too. Nvidia’s plan to limit hashing may not work if cards are still profitable at reduced hash rates.

The RTX 3060 has mining restrictions on it. The rumored RTX 3080 Ti may as well.

The first time this happened, people thought the rise of FPGAs and ASICs would end the disaster. But people like the ease and ubiquity of mining on consumer GPUs, and cryptocurrency coins specifically designed to mine poorly on FPGAs and ASICs soon appeared. So long as cryptocurrencies remain deflationary, this problem is going to keep happening.

Gaming, as a hobby, relies on the idea that upgrades and replacement parts are available over time. If people can’t buy the upgrades and replacement parts they need, they’ll find other hobbies. In the past, customers who wanted a GPU could either buy Nvidia (2012-2013) or buy a pre-built OEM system (2017-2018). This time around, higher tariffs and worldwide shortages have driven OEM prices upwards, too.

Ampere is now six months old. Nvidia tends to launch new GPU architectures every other year these days. If it takes another six months for these GPU shortages to clear — and based on what we’ve heard from the industry, six months seems pretty reasonable — it means we’ll be halfway to Ampere’s projected replacement before you can stroll out and buy the current card at whim as opposed to hovering over the keyboard, hitting refresh like a jockey in a race.

The industry needs a real, long-term solution to this problem. I’m really not looking forward to the day, a few months from now, when the GPU market will have been overheated for more than half the time since Pascal launched in 2016. We won’t cross that bridge until the July / August time frame, so there’s still a chance to fix this problem first. Demand for GPUs isn’t good for gaming if the cards aren’t actually being used to game.

Now Read:

Let’s block ads! (Why?)

ExtremeTechGaming – ExtremeTech

Nvidia Hints at More GPU Mining Restrictions

This site may earn affiliate commissions from the links on this page. Terms of use.

Getting a new GPU is almost impossible right now, with prices skyrocketing as supply continues to lag demand. Nvidia has already announced some steps to separate cryptocurrency applications into a new product category, and it’s even got a new gaming GPU that blocks high-speed crypto mining. During the company’s recent earnings call, Nvidia CFO Colette Kress hinted that more gaming GPU restrictions could be coming

In the unfortunate event your GPU dies, you’ll be dismayed to learn the only way to get a halfway decent replacement is to pay a reseller at least 50 percent over MSRP. The demand for GPUs is particularly high right now thanks to the COVID pandemic that’s kept everyone inside for a year, as well as the surging price of cryptocurrency. The global semiconductor shortage isn’t helping, either. 

Nvidia’s first real stab at addressing the out-of-control pricing was to nerf its latest RTX 3060 GPU. This card has an artificially decreased Etherium hash rate, making it less interesting to crypto miners. Nvidia hopes this will increase supply for gamers who just want to frag some noobs. It’s also planning to release a line of cryptomining processors (CMPs) that are only good for mining (they don’t even have video outputs).

Large mining rigs can contain dozens of GPUs, and not one of them is being used for fragging noobs.

In discussing the shortage, Kress noted that Nvidia has limited hash rates on GPUs “starting with the 3060.” That strongly implies future GPUs will come with similar restrictions, and the company did go out of its way recently to stress that it would not nerf any existing video cards. Nvidia’s desire to keep the gaming and mining markets separate supports the idea future cards will have mining limits, as stressed by CEO Jensen Huang. “I think proof of work is going to be around for a bit,” he said, referring to the workloads that miners run to unlock new digital money. “We developed CMP for this very reason.”

It plans to ship approximately 15 million CMP cards for industrial miners in the first quarter of the year. That might help get GPUs back on store shelves for gamers, but the trend of splitting mining and gaming might not appeal to everyone. Hopefully, Nvidia leaves at least some product lines alone — if you’re spending big on a flagship GPU, it’s reasonable to expect it to be unencumbered by artificial restrictions of any kind.

Now read:

Let’s block ads! (Why?)

ExtremeTechGaming – ExtremeTech

Nvidia Says It Won’t Nerf Crypto Mining on Existing GPUs

This site may earn affiliate commissions from the links on this page. Terms of use.

We’re living in a perfect storm for GPU price inflation —  cryptocurrency mining is on the rise, as is interest in gaming during the pandemic. There’s also a global semiconductor shortage at the moment. The result: even if you can find a high-end GPU, it’ll probably cost at least twice what it should. Nvidia hopes to combat this with the upcoming RTX 3060, which will have its crypto mining capabilities nerfed. The company has now clarified its plans, saying it won’t make crypto changes to any existing GPUs. 

Nvidia raised eyebrows when it announced the restrictions for its next RTX video card. Using driver-level limits, the RTX 3060 will be limited to 50 percent of the usual Etherium hash rate. So, why just Etherium? Nvidia says Etherium is the only popular cryptocurrency that benefits heavily from GPU-based mining. The hope is that by intentionally handicapping the cards for this one, very specific use case, that the supply for gamers will increase. 

In place of RTX cards, Nvidia wants miners to buy one of its upcoming CMP (Crypto-Mining Processor) cards. These parts don’t have video output capabilities, but most of the internal components should be identical to RTX video cards. Nvidia has said RTX supply won’t be impacted by CMP production because these chips “could not meet the specifications of GeForce.” That vague wording suggests Nvidia is binning chips, a common practice across the industry that involves selling chips with minor manufacturing errors as lower-tiered models. 

This GPU should only cost about $ 550.

Knowing that Nvidia could just slash the hash rate of GPUs at the driver level has naturally led some to wonder if changes are coming to other RTX cards. After all, if Nvidia wants miners to buy CMP cards, nerfing RTX cards would certainly drive adoption. However, Nvidia says that’s not going to happen — no existing GPUs will see forced hash rate changes. 

The fate of future GPU models will probably depend on how the CMP launch goes. The first mining-specific Nvidia cards will launch in the first quarter of the year, with more to follow later. Still, the RTX 3060 is a mid-range model. People buying the top-of-the-line hardware might balk at any mining limits. Still, it’s possible future high-end video cards will have limits on Etherium mining if CMP takes off. That might not be all bad if it means gamers can buy cards at (or at least within spitting distance of) MSRP.

Now read:

Let’s block ads! (Why?)

ExtremeTechGaming – ExtremeTech

Nvidia Limits RTX 3060 Crypto Speeds, Tells Miners to Buy New CMP Cards

This site may earn affiliate commissions from the links on this page. Terms of use.

Nvidia has announced that it will take steps to ensure RTX 3060 GPUs wind up in the hands of gamers, not crypto miners. On Thursday, the company said it would launch a new lineup of CMP (Crypto-Mining Processor) cards, while also restricting the maximum performance of the RTX 3060 when mining Ethereum.

RTX 3060 GPUs will only mine Ethereum at 50 percent of their potential hash rate due to driver-level restrictions. According to Nvidia, its new CMP cards will run at full speed. These cards lack video outputs and they offer both lower peak voltage and lower maximum core clocks, in order to improve mining efficiency.

Nvidia claims CMPs shouldn’t be referred to as video cards or GPUs because they lack the fundamental hardware to fulfill that role. This is true, as far as it goes, but it’s not because the cards actually lack support for “RTX real-time ray-tracing, DLSS AI-accelerated image upscaling technology, [and] Reflex super-fast response rendering for the best system latency.” All of that capability is still baked into the CMP — you just can’t make practical use of it.

This initiative is meant to increase the supply of RTX 3060 cards in-market.

Nvidia’s push to rebrand mining GPUs as “CMPs” makes some objective sense, since you literally can’t use these cards for gaming or video output, but an actual custom-designed CMP wouldn’t just be missing some video outputs.

A custom processor designed for mining Ethereum would be considered a specific type of ASIC. A GPU design repurposed for mining Ethereum without any additional customization work would at least include additional processing cores rather than dedicating the physical silicon to useless gaming features. Nvidia calling this iteration of the RTX 3060 a CMP makes sense from a marketing perspective, but from a technical one you’d expect the chip to be a little more specialized than these cards actually are.

Will This Actually Help Anything?

We’re glad to see Nvidia trying to fix this situation, but it’s not clear how much this is going to help. We should see some improvement in RTX 3060 availability, but GPUs are hard to find right now across the product spectrum, and that means there’s additional upward price pressure on any available GPU, even if we ignore cryptocurrency mining demand.

If TSMC and Samsung are currently capacity-constrained and shipping every chip they can manufacture, then splitting the pile of RTX 3060 GPUs into “CMP” and “GPU” buckets just means there are less chips available to fill demand in both segments. Unmet demand for CMP cards will waterfall into the GPU space, even if GPUs are disfavored for these workloads. Specialized cards might blunt the impact, but it’s doubtful they’ll completely offset it.

If the entire market wasn’t supply constrained, Nvidia would have more options. Well-priced CMP cards could be counted on to divert buyers from the GPU market, and there would be enough inventory to meet demand. With GPUs hard to come by at every price point and both TSMC and Samsung running at maximum production, this looks a bit more like subdividing the same pie to carve out another slice. We’d like to be wrong on this, because the semiconductor market needs all the help it can get. But it’s not clear that limiting the RTX 3060 alone in this fashion will meaningfully improve GPU supplies in and of itself. We’re hopeful this will help at the margins, but only ramping up

Now Read:

Let’s block ads! (Why?)

ExtremeTechGaming – ExtremeTech