Tag Archives: GPUs

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

Steam Data Shows Ampere GPUs Barely Trickling Into Market

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

As the GPU shortage continues, what constitutes “success” is being rapidly recast. Several publications have recently run stories claiming that an uptick in Ampere GPU deployments according to the Steam Hardware Survey constitutes evidence that these cards are making their way to gamers and that miners aren’t soaking up all the demand. This is factually true, inasmuch as cryptocurrency mining isn’t literally consuming every single GPU. When evaluated in a historical context, however, the current SHS doesn’t support an optimistic narrative about Ampere availability.

The RTX 3070 gained 0.17 percent market share from February 2021 to March 2021. That’s the most market share any GPU gained last month. But according to past Steam data, a single GPU topping out at 0.17 percent adoption isn’t very good at all.

I’ve surveyed several multiple data points in the SHS over the past two years. In November 2019, no fewer than nine GPUs gained more than 0.17 percent market share. The RTX 2060 picked up 0.42 percent that month, for example. In February 2020, before the pandemic hit, the GTX 1660 Ti and GTX 1650 gained 0.34 percent and 0.51 percent share, respectively, with other cards above 0.17 percent. Even in March 2020, with the pandemic gearing up, cards like the RTX 2060 (0.51 percent), RTX 2070 (0.31 percent), and RTX 2070 Super (0.28 percent) saw stronger growth than what’s being reported for Ampere today.

Steam has only included data on the RTX 3070 for two months, but the RTX 3080 has been included for longer. The trend is not encouraging:

Market share data above is for the period November 2020 – March 2021. Look at what happened to the RTX 3080’s adoption rate after December. We see gaming market share more than double in a single month. Thereafter, the growth rate falls off a cliff. It took the RTX 3080 a single month to grow by 2.08x, then a further three months to grow by 1.77x. In an ordinary year, this might reflect nothing more than seasonality, but this isn’t an ordinary year. There are still a lot of would-be Ampere gamers waiting for prices to fall.

This is reflected in how the numbers for all the other cards stop moving. Anyone with a GTX 1070 Ti, RTX 2080 Super, or RTX 2080 is a potential RTX 3070 customer (the 2080 Ti is a bit too high to really see the RTX 3070 as a replacement card). In November and December, the percent of users with each of these cards bounces around. From January forward, the percentages have been nearly static. RTX 2080 Ti customers aren’t upgrading to RTX 3090’s. RTX 2080 and 2080 Super owners clearly aren’t upgrading to Ampere. The Pascal gamers that Nvidia said it was explicitly targeting with this launch remain largely wedded to their hardware. The GTX 1060 has dropped 1.14 percent since November and the 1050 Ti has dropped about 0.5 percent. The RTX 1070 is 0.36 percent less common now than in November 2020. The GTX 1080 has dropped even less.

One reason why the RTX 3070 looks good is Steam didn’t add the GPU to its tracking until it had hit 1.12 percent. If we actually had the month-by-month report, however, I’m betting we’d see exactly the same thing as with the RTX 3080 — an initial jolt, followed by slow growth for such an attractively positioned high-end part. The RTX 3060 Ti entered the SHS at 0.27 percent in January and has risen to just 0.38 percent three months later.

In aggregate, the RTX 3080 and RTX 3070’s market share is growing on par with how the RTX 2080 and 2070 performed back in 2019. At MSRP, Ampere is everything Turing wasn’t. It offers ray tracing performance we feel more comfortable recommending and much stronger AI and gaming performance. It’s also much less expensive (theoretically) than the RTX 2070 and RTX 2080 were at launch back in 2018. In this context, a 0.17 percent rise in the number of RTX 3070 GPUs in-market isn’t a ray of hope. It’s a demonstration of how bad the market continues to be.

As of this writing, anyone who needs a replacement GPU should consider AMD’s R9 290 and R9 290X the best options for a reasonably priced card. We continue to keep an eye on this situation and it continues to offer the best price/performance ratio outside of getting lucky. RDNA2 GPUs are not contemplated in the story above because Steam has not yet added these cards to its database. It can take Valve months to update the database with new cards, however, and it does not add them at a consistent market share level. By all accounts, however, AMD availability is poor.

It is not clear if these shortages are being driven mostly by cryptocurrency-related demand, by low yields, or by a mixture of both. We may get some better data on that point when Nvidia eventually gives Q1 results, but that won’t happen for another couple of months. Several GPU manufacturers have implied they can’t get sufficient GPU inventory, but cryptocurrency demand is also a known impact right now.

We don’t typically refer to the SHS for hardware information because of doubts about its accuracy. But to the extent we can rely on this data to show us anything, what it shows is not positive. Four to five cards with a >0.17 percent rise might constitute some positive sign that we’re headed back towards normal. A single GPU just illustrates how far we’ve got to go.

Now Read:

Let’s block ads! (Why?)

ExtremeTechGaming – ExtremeTech

PS5 Availability Is Improving, but GPUs Prices Are the Worst We’ve Ever Tracked

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

For the past six months, every CPU, GPU, and console gamers have wanted to buy has been between difficult and impossible to find, depending on your luck, skill, and willingness to pay scalpers a monthly fee for a chance to buy hardware. We’ve kept an eye on the situation as it has developed, and there’s even the ghost of some good news to report if you’re a console buyer. PC customers, unfortunately, shouldn’t expect relief any time soon.

First, the good news: A new analysis by The Verge shows prices for the PlayStation 5 coming down. More and more systems are selling on eBay for lower prices or aren’t selling at all at sky-high targets. Sean Hollister writes, “Over a seven-day period, eBay moved 5,284 PS5 consoles, and yet plenty of PS5s that were listed didn’t sell. PS5 scalping is becoming less profitable, eBay’s getting flooded, and things are slowing down.”

“Slowing down” is still a careful qualifier, considering that Sony has told people it doesn’t expect big improvements until the back half of the year, but at least the analysis demonstrates positive motion. The Verge’s findings on GPUs were not so rosy.

GPU-Pricing-LateMarch

Image by The Verge

Now, it just so happens that I performed a similar analysis almost exactly three years ago. Let’s take a look at MSRP data from today against how hot things were running back three years ago. Verge’s data is above, here’s our 2018 data below:

Nvidia-Chart-GPU-Price

AMD-GPU-Pricing

Back in 2018, we had five GPUs out of 16 running more than 2x above MSRP. Today, no fewer than 7 GPUs out of 9 are running above 2x, while two are running above 3x. This is the worst GPU prices have been in my two-decade career. It’s not even close. The cheapest GT 1030 from an OEM brand I recognize (Gigabyte) is $ 133. The cheapest 1050 Ti is $ 262. A new RX 580 is currently selling for $ 649 for the 8GB flavor. The GTX 1080 is no longer on sale at a $ 500 price point, but the RTX 3070 is supposed to be. The current price of $ 1,239 as measured by the Verge makes the $ 776 inflation from several years ago seem nostalgic.

This tweet summarizes the situation quite well:

GPU manufacturers reportedly expect Nvidia availability to remain scarce into Q3 2021, and we can expect holiday demand to surge for Q4 like always. I’ve tried to remain positive in our various update articles on the GPU pricing debacle, but it increasingly looks like the market will be tight for the entire year. I’m not going to say this means nobody will get a card for MSRP — GPUs continue to ship, and clearly, some customers are getting them at retail — but it doesn’t look like you can depend on getting a GPU at MSRP this year, at any price point. Hopefully, as the year progresses, TSMC will be able to allocate more capacity for GPU manufacturing and finally get ahead of mining demand.

Now Read:

Let’s block ads! (Why?)

ExtremeTechGaming – ExtremeTech

AMD Will Not Limit Cryptocurrency Mining on RDNA2 GPUs

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

AMD and Nvidia often look for ways to differentiate themselves from the competition and one of those ways going forward, apparently, will be cryptocurrency mining. Unlike Nvidia, which opted to limit RTX 3060 mining performance before releasing a driver that obviated much of the goal, AMD told us it explicitly won’t be limiting any workload at all.

“The short answer is no,” said Nish Neelalojanan, a gaming product manager at AMD. “We will not be blocking any workload, not just mining.”

Nish went on to explain some specific factors weighing into AMD’s decision, including the fact that the company believes RDNA2 and its Infinity Cache are specifically tuned towards gaming rather than mining or compute workloads. There’s some evidence to support this, inasmuch as the high-end RDNA2 GPUs are not as good at mining as the RTX 3080 or RTX 3090.

But let’s be real about this: Right now, GPU prices are so broken, you can get an extra $ 50 for a houseplant so long as it supports HDMI. Anything that can mine cryptocurrency is being used to mine cryptocurrency, and if AMD’s GPUs can crank out profitable hash rates, no miner is going to care if AMD optimized for the workload. It’s also not clear how much leeway AMD has to block cryptocurrency workloads on Linux, given that AMD’s Linux GPU driver is entirely open source. Nvidia probably has more leeway here.

AMD’s new RTX 6700 XT. Open for business, if you can find one.

I’ve stated in several articles that I think Nvidia is doing the right thing by blocking cryptocurrency workloads. It’s not a position I like taking. The entire point of a PC, as opposed to a console, tablet, or smartphone, is the freedom the end-user has to create, modify, and run applications. It’s extremely difficult to create a programmable graphics processor only to turn around and try to prevent it from running very specific programs. It’s not clear that AMD could prevent mining the way Nvidia tried to do, even if it wanted to.

But if that’s true, it only speaks to how much the cryptocurrency industry is warping computing. When AMD decided to fully open-source its Linux driver, it wasn’t considering the impact that might have on GPU availability down the road. There wasn’t a reason to believe one topic had anything to do with the other.

Even if we assume the cryptocurrency market eventually cools off, this is cold comfort to anyone who wants a GPU. Long-term, we’re looking at a situation in which each new microarchitectural launch has a chance to spark a 9-18 month GPU shortage every 24-36 months. That’s not going to be good for the long-term survival of our hobby. Best-case is a massive buying cycle realignment, but that’s still going to burn everyone whose GPU dies during the 9-18 month period and who can’t find a card at MSRP.

These impacts aren’t theoretical. I’ve heard from a number of readers who are variously delaying purchases, saving additional funds to buy a complete system, or who wound up paying a lot more money than they intended. If they continue long enough, people who can’t afford to game on PC will game on something else or find a different hobby altogether.

I don’t know if there’s a way for Intel, AMD, and Nvidia to build a GPU that continues the industry’s historical trend of improving both game performance and performance per watt while absolutely sucking as a cryptocurrency miner, but the future of the gaming market may belong to the firm that figures out the best answer to this puzzle. At this point, the most optimistic predictions suggest the GPU semiconductor shortage might begin to ease by the end of Q2, with Q3 and Q4 floated as alternatives. No one is suggesting a widespread outage that drags on into Q1 2022 — yet.

If AMD’s reason for not limiting cryptographic mining is philosophical, not technical, I have to say I disagree with the decision under the current circumstances. Six months after Ampere debuted, GPUs of every stripe remain very difficult to find.

Now Read:

Let’s block ads! (Why?)

ExtremeTechGaming – ExtremeTech

GPUs Used For Crypto Mining Might Lose Game Performance, Long-Term

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

One question that’s come up from time to time in gaming is whether older GPUs get slower over time. ExtremeTech has examined this question before with respect to software updates, meaning we checked whether older GPUs lost performance as a result of later driver updates that were not as well optimized for older GPU architectures. While our driver tests found no evidence of software-driven slowdowns, we didn’t check the impact of whether aging GPU hardware could impact performance.

A new investigation of an 18 month-old RTX 2080 Ti claims to have uncovered evidence that an old GPU will run more slowly than a newer card, based on a comparison of two (we think) MSI GeForce RTX 2080 Ti Gaming X Trio GPUs. Unfortunately, based on available information, there’s no way to confirm that conclusion. At best, what YouTube channel Testing Games has established is that long-term mining might slow down a GPU.

At first glance, the findings seem equivocal. Testing Games runs a suite of games, including Assassin’s Creed Valhalla, Battlefield V, Cyberpunk 2077, Forza Horizon 4, Horzion: Zero Dawn, Kingdom Come Deliverance, Mafia Definitive Edition, and Red Dead Redemption. The MSI Gaming X Trio GPU card used for crypto mining for the past 18 months is typically about 10 percent slower than the new MSI Gaming X Trio GPU that hasn’t been used for mining. Spot checks of various games show that the used RTX 2080 Ti runs 15-20 degrees hotter than the new card, uses somewhat less power, and hits a lower maximum clock speed. This would seem to be an open-and-shut demonstration of the fact that mining can wear out a GPU, but there are some problems with this analysis.

First, the authors don’t appear to have re-pasted or dusted the used GPU. Dust is an absolutely magnificent insulator and enough of it will easily destabilize a gaming rig. This alone could account for the higher temperatures and lower clocks on the used card, no explanation needed.

Second, it is not clear if this represents the same GPU being tested, or two different versions of the card purchased at two different times. The former would be more useful. The wide use of Turbo clocks in GPUs and CPUs today allow for variations in binning that can impact the final result. It could be that the newer card fielded a better core, allowing for higher base performance, and effectively invalidating our ability to derive any useful information from this comparison. The official boost clock on the MSI Gaming Trio X is 1755MHz, which means both GPUs are shown running above this specification. It is possible that some of the variance between the two GPUs reflects SoC quality.

If these are two different GPUs, we also don’t know if they use an identical VBIOS version or if they use exactly the same brand of RAM. Micro-timings and VBIOS updates can introduce their own performance changes. The newer GPU is also often faster than the older GPU than the difference in its clock speed would indicate. The clock gap as measured in-game is on the order of 3-5 percent (it varies depending on where you are in the run), while the performance variation varies by 8-12 percent. The RAM clock is supposedly locked to an effective 7GHz (14Gbps) across both cards.

There’s another point I want to bring up: These numbers are a little odd as far as the implied relationship between clock variation and actual observed performance.

From Testing Games. The clock speed gap, current FPS difference, and the average FPS score across the benchmark run do not agree with each other. A nearly 20 percent momentary performance difference is shown to correspond to a 6.6 percent clock variation, with the older card running nearly 10 percent slower overall. These gaps are present in almost every title, at any given moment of measurement.

GPU clocks and performance results do not typically move in lockstep. Increase the GPU core and memory clocks by 10 percent, and a game’s performance may only improve by 6-8 percent. This is expected because there’s always the chance that a benchmark is slightly limited by some other aspect of the system. The expected result from a linear clock speed increase is a linear-to-sublinear improvement in performance. It therefore follows that the expected impact of reducing clock is a linear-to-sublinear reduction in performance.

The results in this video show the opposite, in almost every case. Apart from Kingdom Come: Deliverance, the gap between the reported GPU clocks is about half the size of the performance improvement. 4-6 percent clock speed differences are associated with 8-15 percent performance shifts.

This could be a result of polling errors in the utilities being used to gather this information. Alternately, it suggests some other variable in play that hasn’t been accounted for in the YouTube video above. The used GPU could be hitting thermal limits and throttling itself back, but doing so more quickly than monitoring utility can detect. Most polling utilities only poll once per second, while GPUs are capable of adjusting their clocks in a matter of milliseconds. It’s possible that the used GPU’s clock looks more stable than it would if we had finer-grained reporting tools.

Testing Games has not released any follow-up information on their testing protocols or whether this comparison was performed on the same GPU at two different points in time or on two different GPUs purchased at different times. It also hasn’t released any discussion of why these results point to greater-than-linear performance improvements despite linear increases in GPU clock and no changes to memory clock.

Until these questions are answered, the idea that heavily-mined cards lose gaming performance can only be considered a theory. We’re not saying the theory is wrong, but it hasn’t been properly tested yet. More data is needed, either from Testing Games or from other sources, to illustrate the accuracy of this claim.

Now Read:

Let’s block ads! (Why?)

ExtremeTechGaming – ExtremeTech

GPUs Used For Crypto Mining Might Lose Game Performance, Long-Term

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

One question that’s come up from time to time in gaming is whether older GPUs get slower over time. ExtremeTech has examined this question before with respect to software updates, meaning we checked whether older GPUs lost performance as a result of later driver updates that were not as well optimized for older GPU architectures. While our driver tests found no evidence of software-driven slowdowns, we didn’t check the impact of whether aging GPU hardware could impact performance.

A new investigation of an 18 month-old RTX 2080 Ti claims to have uncovered evidence that an old GPU will run more slowly than a newer card, based on a comparison of two (we think) MSI GeForce RTX 2080 Ti Gaming X Trio GPUs. Unfortunately, based on available information, there’s no way to confirm that conclusion. At best, what YouTube channel Testing Games has established is that long-term mining might slow down a GPU.

At first glance, the findings seem equivocal. Testing Games runs a suite of games, including Assassin’s Creed Valhalla, Battlefield V, Cyberpunk 2077, Forza Horizon 4, Horzion: Zero Dawn, Kingdom Come Deliverance, Mafia Definitive Edition, and Red Dead Redemption. The MSI Gaming X Trio GPU card used for crypto mining for the past 18 months is typically about 10 percent slower than the new MSI Gaming X Trio GPU that hasn’t been used for mining. Spot checks of various games show that the used RTX 2080 Ti runs 15-20 degrees hotter than the new card, uses somewhat less power, and hits a lower maximum clock speed. This would seem to be an open-and-shut demonstration of the fact that mining can wear out a GPU, but there are some problems with this analysis.

First, the authors don’t appear to have re-pasted or dusted the used GPU. Dust is an absolutely magnificent insulator and enough of it will easily destabilize a gaming rig. This alone could account for the higher temperatures and lower clocks on the used card, no explanation needed.

Second, it is not clear if this represents the same GPU being tested, or two different versions of the card purchased at two different times. The former would be more useful. The wide use of Turbo clocks in GPUs and CPUs today allow for variations in binning that can impact the final result. It could be that the newer card fielded a better core, allowing for higher base performance, and effectively invalidating our ability to derive any useful information from this comparison. The official boost clock on the MSI Gaming Trio X is 1755MHz, which means both GPUs are shown running above this specification. It is possible that some of the variance between the two GPUs reflects SoC quality.

If these are two different GPUs, we also don’t know if they use an identical VBIOS version or if they use exactly the same brand of RAM. Micro-timings and VBIOS updates can introduce their own performance changes. The newer GPU is also often faster than the older GPU than the difference in its clock speed would indicate. The clock gap as measured in-game is on the order of 3-5 percent (it varies depending on where you are in the run), while the performance variation varies by 8-12 percent. The RAM clock is supposedly locked to an effective 7GHz (14Gbps) across both cards.

There’s another point I want to bring up: These numbers are a little odd as far as the implied relationship between clock variation and actual observed performance.

From Testing Games. The clock speed gap, current FPS difference, and the average FPS score across the benchmark run do not agree with each other. A nearly 20 percent momentary performance difference is shown to correspond to a 6.6 percent clock variation, with the older card running nearly 10 percent slower overall. These gaps are present in almost every title, at any given moment of measurement.

GPU clocks and performance results do not typically move in lockstep. Increase the GPU core and memory clocks by 10 percent, and a game’s performance may only improve by 6-8 percent. This is expected because there’s always the chance that a benchmark is slightly limited by some other aspect of the system. The expected result from a linear clock speed increase is a linear-to-sublinear improvement in performance. It therefore follows that the expected impact of reducing clock is a linear-to-sublinear reduction in performance.

The results in this video show the opposite, in almost every case. Apart from Kingdom Come: Deliverance, the gap between the reported GPU clocks is about half the size of the performance improvement. 4-6 percent clock speed differences are associated with 8-15 percent performance shifts.

This could be a result of polling errors in the utilities being used to gather this information. Alternately, it suggests some other variable in play that hasn’t been accounted for in the YouTube video above. The used GPU could be hitting thermal limits and throttling itself back, but doing so more quickly than monitoring utility can detect. Most polling utilities only poll once per second, while GPUs are capable of adjusting their clocks in a matter of milliseconds. It’s possible that the used GPU’s clock looks more stable than it would if we had finer-grained reporting tools.

Testing Games has not released any follow-up information on their testing protocols or whether this comparison was performed on the same GPU at two different points in time or on two different GPUs purchased at different times. It also hasn’t released any discussion of why these results point to greater-than-linear performance improvements despite linear increases in GPU clock and no changes to memory clock.

Until these questions are answered, the idea that heavily-mined cards lose gaming performance can only be considered a theory. We’re not saying the theory is wrong, but it hasn’t been properly tested yet. More data is needed, either from Testing Games or from other sources, to illustrate the accuracy of this claim.

Now Read:

Let’s block ads! (Why?)

ExtremeTechExtreme – 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

Intel Launches Iris Xe, Its First Desktop GPUs in More Than 20 Years

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

Intel is finally back in the desktop graphics business, at least if you squint. The company has announced a partnership with certain PC OEMs to bring DG1 silicon to specific pre-built systems.

This isn’t exactly a full-on desktop graphics launch — that’ll come later in 2021 with the launch of DB2 — but Intel is still indisputably shipping at least a handful of discrete GPUs in the low end of the desktop market, for the first time in more than 20 years. The company announced it had partnered with “two ecosystem partners, including Asus” in its initial PR, but LegitReviews thinks the GPU featured in the image above is manufactured by Colorful. The other DG1 card identifies itself, ships without a fan, and is clearly an Asus-branded product.

Unfortunately for anyone hoping to play around with Intel’s latest desktop card and/or hoping to find a low-end GPU at a reasonable price, this GPU is OEM-only. Sometimes, OEM-only products will surface on secondary markets like eBay, but that’s not going to happen in this case. According to Intel, these GPUs will only work on very specific systems. LegitReviews inquired on this point and was told:

The Iris Xe discrete add-in card will be paired with 9th gen (Coffee Lake-S) and 10th gen (Comet Lake-S) Intel® Core™ desktop processors and Intel(R) B460, H410, B365, and H310C chipset-based motherboards and sold as part of pre-built systems. These motherboards require a special BIOS that supports Intel Iris Xe, so the cards won’t be compatible with other systems.

So, that’s that, then. It’s not clear why a motherboard would require a special UEFI to use a new GPU. Presumably DG2, when it arrives, will not have this problem.

It’s interesting to see Intel back in the graphics market because it’s been so long since we had an actual three-way fight. Once 3dfx died, the only company to offer any kind of competition to the ATI/Nvidia duopoly was PowerVR with the Kyro and Kyro II. While these GPUs were an interesting alternative to the Radeon and GeForce product lines, they did not find mainstream success and faded from the market.

Intel has not previously covered itself in glory where GPUs are concerned. The company’s first attempt at a discrete GPU, the Intel i740, was custom-designed to showcase the capabilities of Intel’s new AGP bus. It didn’t compare well against GPUs with onboard RAM, and Intel didn’t stay in the market very long. Intel’s Larrabee was based on a modified Pentium architecture with 512-bit vector processing units. Overall interest in Larrabee was high, but Intel canceled the product and used Larrabee as the basis for the first generation of Xeon Phi processors.

It’s not unfair to be skeptical of Intel’s ability to launch a competitive GPU. We don’t even necessarily expect Intel’s first-generation cards to be all that great, objectively speaking. So long as they’re good enough to get a little traction somewhere in the market, Intel has an opportunity to iterate and improve the design. With a third player on the field, both AMD and Nvidia are presented with new challenges — but also, potentially, with new opportunities depending on how Intel’s presence impacts user GPU purchases.

All of this hinges on Intel building competitive products and being willing to stay the course over the long term. Incoming CEO Pat Gelsinger may have his own ideas about where to take the company. These DG1-equipped systems aren’t going to make a huge splash in the wider market, but we should know how effectively DG2 will compare against AMD and Nvidia before the end of the year. Given how hard AMD and Nvidia cards have both been to source, Intel could spin a modest GPU into a smash hit just by shipping it on-time, at MSRP.

Now Read:

Let’s block ads! (Why?)

ExtremeTechGaming – ExtremeTech

Quake II RTX Now Runs on AMD GPUs Thanks to Vulkan Ray Tracing

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

Nvidia’s decision to rebuild Quake II as a ray tracing title was, in my opinion, a really clever way to show off what the feature was capable of, while simultaneously giving people a reason to return to the well-regarded classic title. Up until now, however, the only gamers who could enjoy it were folks who owned an Nvidia RTX GPU. While some games will run ray tracing on Pascal, Quake II RTX isn’t one of them.

Now, however, AMD fans have an opportunity to enjoy Quake II RTX as well, apparently with Nvidia’s explicit help. Any GPU that supports the Vulkan Ray Tracing extensions maintained and developed by the Khronos Group can also run Quake II RTX. On AMD’s side of the fence, that support appears to be limited to the RX 6000 series, at least for now.

Nvidia’s support is much wider. In addition to Ampere and Turing, Vulkan Ray Tracing is also supported on the GTX 1660 family of GPUs, the Volta-based Nvidia Titan V and the Quadro GV100, and on Pascal GPUs with at least 6GB of RAM. If you haven’t seen the Quake II RTX launch trailer, we’ve embedded it below.

Nvidia actually did some of the heavy lifting to bring ray tracing to Vulkan in the first place. There are already existing tools in place to assist with translating DX12 calls and HLSL (High Level Shader Language) code into Vulkan and SPIR-V, respectively. Nvidia’s specific contribution to the project was to add ray tracing support to Microsoft’s open source DirectXCompiler, which is commonly used to port HLSL code to Vulkan.

In other words, Nvidia’s open source work is a key part of why AMD GPUs can now run Quake II RTX. This kind of ‘coopetition’, if you will, is a key part of ensuring standards are widely supported and making certain gamers can expect certain features on a wide range of systems. In theory, developers that already have a Vulkan and a DX12 path could keep Nvidia RTX support for one API and support both Nvidia and AMD in Vulkan. So far, we haven’t heard much about whether or not RTX-enabled games will receive an update to allow AMD to use ray tracing, or how much additional optimization is required to make use of the feature on RDNA2 GPUs as opposed to Turing / Ampere.

For now, it doesn’t look as though AMD has any plans to enable ray tracing on the 5700 or 5700 XT. One thing we know is that the performance impact of enabling ray tracing is heavy, especially on GPUs that weren’t designed to support it. On paper, Nvidia supports a huge range of cards. In practice, most of the Pascal cards other than the 1080 Ti and maybe the GTX 1080 have often proven too slow to run ray tracing effectively. Even if you turn down other detail levels enough to get the feature running, you may not feel ray tracing makes up for the losses in other places.

It would be nice to see AMD offer Vulkan support for RX 5700 and 5700 XT cards, but keep in mind, the 6800 XT and 6800 lag Nvidia’s ray tracing performance already. It’s entirely possible that the reason the company isn’t enabling ray tracing support in its lower end products is that it can’t guarantee a high-quality experience.

Either way, if you want to check out Quake II ray traced, AMD’s beta drivers with Vulkan RT support can be downloaded here, while Nvidia’s beta driver with the same can be downloaded here. For now, the only title using Vulkan Ray Tracing is Quake II RTX, as far as I know, but Khronos has designed the standard for use across computers and mobile devices and is hoping for robust uptake, long term.

Now Read:

Let’s block ads! (Why?)

ExtremeTechGaming – ExtremeTech

$340K Worth of MSI RTX 3090 GPUs Stolen in Factory Heist

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

MSI has announced today that $ 340,000 worth of Nvidia RTX 3090 GPUs have been stolen from its factory. The theft, which is theorized to have involved as many as 10 GPUs given current prices, may have been an inside job.

More seriously, MSI states that 40 “boxes” of RTX 3090 GPUs were affected with a value of $ 336,500. MSRP on the RTX 3090 is $ 1,500, so we’re talking about 170 to 220 GPUs depending on selling prices. The company’s shipping area is covered by factory surveillance and it inspects trucks coming and going from the facility, which is part of why the firm suspects it may have a thief on the inside.

This is the second issue for MSI in recent months that involved inventory allocation ending up somewhere it wasn’t supposed to be. Back in early October, the company took heat for the fact that one of its subsidiaries, Starlit Partner, was selling MSI RTX 3080 and RTX 3090 cards for well above MSRP, even as these GPUs were sold out at Amazon and Newegg.

We have no information on whether the two incidents could be related in any fashion. In one case, an OEM partner somehow gained access to GPUs it should not have been allowed to resell and then sold them far above MSRP, but this could be explained by anything from an inventory system error to an inside decision to “accidentally” funnel some cards to a company that shouldn’t have been allowed to sell them. That’s potentially a very different type of crime from arranging for a bunch of MSI GPUs to get loaded on the wrong truck without anyone noticing.

Anyone with knowledge of the theft can contact MSI for a 100,000 Yaun / $ 15,000 reward. Any participant in the theft is offered clemency, provided they help to locate the stolen items.

The theft is another symptom of how bad the allocation of pretty much everything is right now, and what’s that doing to prices. There are RTX 3090 GPUs on eBay listing as having sold for $ 2,500 or more, which shows why someone would take the challenge on in the first place. Two hundred RTX 3090 GPUs at $ 2,500 each is a cool $ 500K.

Unfortunately, these prices are also why there’s no way to get your hands on much of the latest technology from Intel, AMD, Nvidia, Sony, or Microsoft. Low yields, high demand, and COVID-damaged supply lines have collectively made shipping products difficult. Hopefully, MSI is able to recover its property and find whichever individual at the company is responsible for the thefts.

Now Read:

Let’s block ads! (Why?)

ExtremeTechGaming – ExtremeTech