Connect with us

Tech

Asus TUF Gaming RTX 3080 OC Review – TechSpot

Published

 on


Today we’re reviewing the Asus TUF Gaming RTX 3080 OC, the first third-party GeForce RTX 3080 graphics card to come our way and we’re very interested to see how it compares to Nvidia’s Founders Edition model. We’ll be checking out gaming performance which should be very similar to the FE card, but also thermal and overclocking results.

First, let’s tear the TUF Gaming RTX 3080 OC down and take a look at the cooler and PCB. Please note this was done after we collected all our thermal and performance data. Starting with LED lighting, because of course, that’s important. The TUF is minimalist which we like, there is a small light bar above the ‘TUF’ branding along with a backlit logo which faces outwards when the card is traditionally mounted.

In terms of design and appearance, the TUF Gaming looks like your typical high-end graphics card. It’s a 2.7-slot design, so it takes up 3 slots, it measures 30cm long, stands 12.7cm tall and weighs in at 1385 grams, so it’s slightly heavier than the 1355 gram Founders Edition model. It’s mostly black, so it will suit most builds and Asus has included a few ‘TUF’ theme design elements like the tire tracks on the backplate, for example.

We liked how there’s no plastic on the card, apart from the fans of course. The fan shroud has been constructed from aluminum, giving it a premium look. Asus are also using the axial-tech fans and since there are three in total, they’ve reversed the rotation of the center fan to reduce turbulence, the fans each measure 90mm in diameter. We should also note that the card includes a stop-fan feature which activates when the GPU drops below 55C.

Around at the I/O end of the card we find two HDMI 2.1 ports and three DisplayPort 1.4a outputs. That’s an extra HDMI output when compared to Nvidia’s FE version, though note it can still only support up to 4 simultaneous displays. Asus also points out that this model features a very robust stainless steel I/O bracket which they say protects against rust while providing a more durable and secure mount. Can’t say we’ve ever had an issue with the standard steel brackets, but if you have, well this will be a welcomed feature.

Pulling this thing apart to take a better look, starting with the heatsink and fans, we have a very serious looking heatsink here, in fact it looks like something you’d expect to find on an ROG Strix model and not a card with TUF branding, and it really is very large, weighing in at 820 grams.

There are two massive banks of fins which are connected using half a dozen 6mm thick nickel plated copper heatpipes, and all connect to a large nickel plated copper base plate.

Using an aluminum plate, it directly cools the VRM, so that’s good to see, but there’s a lot more to this design as there is a second heatsink which directly cools the GDDR6X memory as well as the power delivery for the memory components. It’s an impressive bit of kit that fits in snug beneath the main heatsink. Asus have also used a few high quality thermal pads to aid in heat transfer from this smaller heatsink to the primary heatsink. Oh and for those of you wondering the smaller heatsink weighs in at 70 grams.

On the back side of the card we find a rather thick aluminum back plate which weighs 138 grams and has been used to strengthen the card and reduce PCB sag, or in this case completely eliminate it. We think this is actually the thickest backplate we’ve ever seen on a graphics card. Asus has also employed a series of thermal pads to remove built up heat from the rear side of the PCB behind the VRM and GDDR6X memory chips. There’s also a few cut outs to aid in air-flow.

Now over to the PCB, we find a 24cm long by 10.6cm tall PCB, so a fairly compact board, though it is crammed full of components. Surrounding the massive GA102 die are the GDDR6X memory chips and then flanking them on either side are boatloads of inductors and power stages. In total the card packs 20 power stages, and here we’re looking at an 8 + 6 + 4 power phase design using SIC641ACD 55A powerstages, though please note two of the 8-phases drive two power stages, so a bit of an odd configuration, but this is what Asus has gone with.

If you’re wondering what the 8 and 6-phase portions power, the answer is the GPU. So 14 phases feed into the GPU, or 16 55A powerstages. The 8-phase power rail is for NVVDD and the 6-phase power rail is for MSVDD while the 4-phase power rail is for the GDDR6X memory.

For comparison, the Nvidia reference board uses an 8 + 5 + 3 power phase design featuring 50A Alpha & Omega Semiconductor power stages. In other words, Asus has beefed up the TUF Gaming well over the Nvidia reference spec.

There’s also two 8-pin PCIe power connectors feeding power into the graphics card and you’ll also find a dual BIOS switch that allows you to change from the default ‘performance’ BIOS to a ‘quiet’ BIOS. Both modes run the card at 340 watts, so we’re just looking at a change to the fan profile which will see the card run hotter in the quiet mode.

In terms of clock specifications, Asus lists a core clock frequency of 1815 MHz which is a 5% boost over the 1730 MHz default spec. The GDDR6X memory though has been left at 19 Gbps, so we’re just looking at a typical mild GPU overclock here. All that said, let’s move on to see what clock speed this model maintains when underload.

For these GeForce 30 series graphics card reviews we’ll be using Shadow of the Tomb Raider for all stress testing and will be reporting temperatures after 30 minutes of gameplay. This saw the TUF Gaming peak at just 63 degrees in a 21 degree room inside the Corsair Obsidian 500D, fully populated with fans. That’s a massive 15 degree drop in temperature when compared to Nvidia’s Founders Edition model.

To maintain this temperature the fans spin at up to 1900 RPM and while that’s a reasonably high fan speed, the card was surprisingly quiet, generating just 42 dBA of noise, making it slightly quieter than the FE version. The typical core clock speed seen during our testing was 1935 MHz and under the same conditions that’s a 5% increase over the Founders Edition model. This saw power consumption increase by 7% from 323 watts with the FE model to 344 watts with the TUF Gaming.

Overclocking

Now for overclocking, with the limits reached we again saw a peak operating temperature of 63 degrees but this time the fans spun up to 2000 RPM. The card wasn’t terribly loud at this fan speed. The overclock saw the cores operate at 2 GHz and the memory also hit 20.6 Gbps, an impressive transfer speed. Finally, when overclocked the card sucked down 357 watts, a small 4% increase from the factory OC configuration.

Let’s move into the benchmark graphs to see what overclocking gets us. We’re testing with our Ryzen 9 3950X GPU test rig with 32GB of DDR4-3200 CL14 memory. The latest drivers available at the time of testing have been used, and for this one we have just a few select games to look at.

Benchmarks

Starting with Death Stranding numbers at 1440p, the TUF Gaming was a single frame faster than Nvidia’s FE model, hitting 158 fps. That’s a pretty disappointing increase and we weren’t able to do much better with a manual overclock, boosting performance by a further 2%.

The results at 4K were just as underwhelming, here the TUF Gaming was 2% faster than the FE model and our overclock netted us an extra 2% performance.

The gains seen in Rainbow Six Siege were a little more impressive, here the TUF Gaming was 6% faster than Nvidia’s FE model, hitting 346 fps. That’s a typical factory overclock, but unfortunately through further manual tuning we were only able to extract 2 extra frames, well under a single percent gained here.

The 4K data looks much the same, again the TUF Gaming was 6% faster than Nvidia’s Founders Edition model and our manual overclock was useless.

Shadow of the Tomb Raider saw a 4% increase for the TUF Gaming over the FE model at 1440p, though quite interestingly this time our manual overclock netted us a further 3% increase in performance.

Similar margins were seen at 4K, here the TUF Gaming was 5% faster out of the box and then 8% faster with our overclock, though that was just a 3% increase over the Asus factory overclock.

Temperatures

Here’s a more in-depth look at the stock temperatures after running Shadow of the Tomb Raider for 30 minutes in a 21C room. The PCB temperatures were recorded using K-type thermocouples. For the GPU rear PCB temp the probe is attached to the backside of the PCB behind the GPU and we were expecting temperatures to be higher, so it would seem the heatsink on the frontside does a good job of extracting heat from the GPU. The GPU die temperature is the result you’ve already seen, we’re just reporting the internal sensor here.

Then the GDDR6X temperature has been reported using a thermal probe, attached to the PCB between the memory chips, so that it doesn’t interfere with the thermal pad and it’s contact with the memory chip. The same method was used to measure VRM temperatures, 4 probes were used to detect the VRM hot spot and here we’re looking at a peak of 66C which is very good. Overall, the TUF Gaming runs surprisingly cool.

What’s to Like

Asus has done an excellent job with the TUF Gaming RTX 3080 OC, it’s a great quality graphics card and better than Nvidia’s own Founders Edition in every measurable way. It’s quieter, significantly cooler, faster out of the box and best of all, it doesn’t require a silly 12-pin power adapter or specialized cable.

The only area where the Founders Edition model might be better is the physical appearance, but that’s entirely subjective. While we do like the look of Nvidia’s version, the TUF Gaming still looks great and we appreciate the use of high quality materials all around.

The fact that the TUF Gaming can be up to 6% faster while running 15C cooler seals the deal for us. There’s just no way we’d buy the Founders Edition over this Asus model for the same price.

At this point we’re still in the process of checking out other AIB RTX 3080 cards, so there might be better options, but we can’t imagine they’re going to be that much better. We’ve seen numerous 5700 XT graphics cards that run hotter than the TUF Gaming RTX 3080 OC, so there’s no denying Asus has done an excellent job. Should the price be right, then we see no reason not to buy this graphics card.

The Asus TUF Gaming RTX 3080 OC is currently listed for $700, which matches the MSRP for standard RTX 3080 boards. That said, it’s also out of stock everywhere we looked, but if you can eventually snag one for $700, then you’ve done very well.

We were surprised to find how much better this thing is than Nvidia’s intricate Founders Edition, and we like to think that our scrutiny with previous TUF Gaming cards helped motivate Asus to put together this excellent design and attention to detail.

Shopping Shortcuts:

Let’s block ads! (Why?)



Source link

Continue Reading

Tech

Apple ramps up efforts to build own search engine to rival Google, says report – CNET

Published

 on



Apple and Google

Apple is ramping up efforts to develop its own search engine, according to a Financial Times report published Wednesday, as US antitrust authorities threaten a lucrative deal that sets Google’s search engine as the default option on iPhones and Samsung phones. 

The iPhone’s latest operating system version, iOS 14, has started to show its own search results and link directly to websites when users type in search queries directly from the home screen, according to industry sources cited in the report.

This move adds to mounting evidence, according to the report, that Apple is working to build a rival to Google search, including Apple’s poaching of John Giannandrea, Google’s head of search, more than two years ago.

Earlier this month, the US Justice Department filed a long-expected antitrust lawsuit against Google over its search dominance, alleging that Google “unlawfully maintained monopolies through anticompetitive and exclusionary practices in the search and search advertising markets.” 

At the heart of the Justice Department’s case are Google’s contracts with other companies, which allow the tech giant’s search engine to be used as the default option. Google pays billions of dollars each year to maintain that default spot. 

Google has been the iPhone’s default search engine for more than a decade. It’s been widely reported that Google pays Apple between $8-12 billion each year for its search engine deal, which is a boon for both tech giants. If it turns out the deal is blocked by the DOJ, Apple will need to supply an alternative for its iPhones. 

 Apple couldn’t immediately be reached for comment. 

Let’s block ads! (Why?)



Source link

Continue Reading

Tech

Apple seems pretty interested in search

Published

 on

Apple might be building a Google competitor, Audible adds more podcasts and an ad measurement company raises $350 million. This is your Daily Crunch for October 28, 2020.

The big story: Apple seems pretty interested in search

Apple has a growing interest in search technology and might even be working on a product to compete with Google, according to The Financial Times.

The most visible change is the fact that in iOS 14, Apple is now showing its own results when you type queries in the home screen. In addition, there seems to be an increase in activity from Apple’s web crawler.

There may be more of an opportunity here as the U.S. Justice Department has sued Google over what it claims are anticompetitive behaviors around search. However, this doesn’t necessarily mean Apple and Google will soon be going head-to-head in search — it could just be a sign that Apple’s Siri voice assistant is getting more search queries.

The tech giants

Joe Rogan, Alex Jones and Spotify’s illusion of neutrality — Spotify is facing criticism after Joe Rogan brought Alex Jones of InfoWars onto his show.

Audible further expands into podcasts — Audible is adding approximately 100,000 podcasts.

Apple eyes the TikTok generation with an updated version of Clips — The update brings much-needed support for vertical videos, allowing for sharing to TikTok and the “Stories” feature in other social apps.

Startups, funding and venture capital

DoubleVerify, a specialist in brand safety, ad fraud and ad quality, raises $350M — DoubleVerify’s technology can detect fraud, viewability and brand safety.

Outrider raises $65M to bring its autonomous tech to distribution yards — The startup has built a three-part system that includes an autonomous electric yard truck, software to manage the operations and site infrastructure.

Lunchbox raises $20M to help restaurants build their own ordering experiences — CEO Nabeel Alamgir said that if restaurants can handle more online orders themselves (rather than just relying on delivery apps), they’ll make more money while also maintaining a direct relationship with their most loyal customers.

Advice and analysis from Extra Crunch

As venture capital rebounds, what’s going on with venture debt? — While venture capital is back setting new records, it appears that its lesser-known sibling won’t be able to match the past few years’ results.

Current and upcoming trends in Latin America’s mobile growth — Latin America is home to one of the fastest-growing mobile markets in the world.

Dear Sophie: Any upgrade options for E-2 visa holders interested in changing jobs? — Another edition of Sophie Alcorn’s column answering immigration questions about working at technology companies.

(Reminder: Extra Crunch is our membership program, which aims to democratize information about startups. You can sign up here.)

Everything else

Qualtrics CEO Ryan Smith is buying majority stake in the Utah Jazz for $1.6B — Smith sold Qualtrics to SAP for $8 billion in 2018.

US online holiday sales to reach $189B this year, up 33% from 2019 — That’s according to a new forecast from Adobe Analytics.

The Daily Crunch is TechCrunch’s roundup of our biggest and most important stories. If you’d like to get this delivered to your inbox every day at around 3pm Pacific, you can subscribe here.

 

 

 

 

 

 

Source:- TechCrunch

Source link

Continue Reading

Tech

AMD announces Radeon RX 6000 series gaming graphics cards – GSMArena.com news – GSMArena.com

Published

 on


AMD today announced the Radeon RX RX 6000 series of gaming graphics cards. Built on the new 7nm RDNA 2 architecture, these cards provide up to 2x improvement in performance over the previous generation AMD flagship while also including support for the new Microsoft DirectX 12 Ultimate API and will be available starting November.

The most important card in this series will likely be the $649 Radeon RX 6800 XT. It features 72 compute units and 16GB of 16Gbps GDDR6 memory on a 256-bit wide memory bus. The 6800 XT can clock up to 2015MHz under load (or Game Clock as AMD likes to call it) and can occasionally hit peaks of 2250MHz for short durations under ideal conditions (Boost Clock). AMD claims a total board power of 300W for this card.

AMD announces Radeon RX 6000 series gaming graphics cards

The 6800 XT also has 128MB of what AMD calls the Infinity Cache. This is high-density, high-speed cache based on the Zen L3 cache. It is designed to minimize DRAM bottlenecks, latency, and power consumption and is especially effective at 4K and 1440p resolutions. Combined with the VRAM, AMD claims the Infinity Cache offers 2.17x the effective bandwidth of a 384-bit wide GDDR6 memory.

AMD announces Radeon RX 6000 series gaming graphics cards

AMD also had some charts to show comparing the performance of the 6800 XT to the $699 RTX 3080. We usually take these with a grain of salt but in the absence of any reliable third-party reviews, these can be taken as a rough guidance.

Next is the $579 Radeon RX 6800. Despite being cheaper, the 6800 doesn’t lose out on much compared to the 6800 XT. The major difference is to the number of compute units, which has dropped from 72 to 60 due to one of the shader engines being disabled. The “Game Clock” has also been dropped down to 1815MHz with the Boost Clock down to 2105MHz.

Despite that, the Radeon RX 6800 will still have the full 16GB 16Gbps GDDR6 memory and 128MB Infinity Cache. And because it’s a bit slower, the board power has also dropped down to 250W instead.

AMD announces Radeon RX 6000 series gaming graphics cards

The 6800 will mostly compete with the $499 RTX 3070, even though it’s more expensive. AMD claims in its charts that the 6800 is faster than the RTX 2080 Ti, which has similar levels of performance as the RTX 3070. However, we would take this chart with an even bigger pinch of salt as it makes use of a feature called Smart Access Memory, which we will discuss shortly.

AMD announces Radeon RX 6000 series gaming graphics cards

Lastly, there is the flagship Radeon RX 6900 XT. This $999 graphics card comes with the full complement of 80 compute units along with the same game and boost clock speeds as the 6800 XT. The rest of the specifications are similar as well.

AMD announces Radeon RX 6000 series gaming graphics cards

AMD decided to take a swipe at the lofty $1499 RTX 3090 with the 6900 XT in its comparison charts. This time, however, we would suggest a small teaspoon of salt as these results are taken with Smart Access Memory and something called “Rage Mode” enabled.

So what is this comically named Rage Mode? AMD calls it a one click overclocking solution. According to Gamers Nexus, Rage Mode unlocks the power targets on these cards so they can clock higher. It doesn’t actually overclock the card itself, just removes some of the power restrictions that could prevent it from clocking higher in some instances. For manual overclockers, this is nothing new but those who are either new or uncomfortable with overclocking can just click this one button and hope it does something.

AMD announces Radeon RX 6000 series gaming graphics cards

As for the aforementioned Smart Access Memory, this one’s a bit more interesting. For this to work, you need a Radeon RX 6000 series GPU, a Ryzen RX 5000 series desktop CPU and a 500-series chipset motherboard. Once you have this trifecta and enable an option in BIOS, it allows the CPU access to the full 16GB VRAM on the graphics card, which supposedly reduces memory fragmentation on the VRAM and improves performance.

Within AMD’s test samples, they say anywhere from single digit to low double digit gains in performance with Rage Mode and Smart Access Memory enabled on the 6800 XT.

AMD announces Radeon RX 6000 series gaming graphics cards

RDNA 2 also includes support for Microsoft’s DirectX 12 Ultimate API. This enables support for features like hardware-accelerated ray tracing, variable rate shading, mesh shaders, and sampler feedback, features that were previously only available on select NVIDIA GPU.

Ray tracing will likely be the one most were looking forward to. AMD announced a handful of titles that will be available soon that support ray tracing on AMD cards, although technically nothing really stops the existing titles such as Control, Metro Exodus, Battlefield V, etc. from working on RDNA 2 hardware since they are all based on Microsoft’s DXR implementation. It will be up to AMD to enable support for them in its drivers. Eventually, all DXR based titles should work on all RTX, Radeon RX 6000, and Xbox Series X|S hardware.

AMD announces Radeon RX 6000 series gaming graphics cards

AMD is also supporting the Microsoft DirectStorage API, which aims to improve load times and improve texture quality. Along with that, the company will continue to support its own technologies, such as AMD FidelityFX, Radeon Anti-Lag, and Radeon Boost.

What’s missing from AMD’s arsenal for now is an alternative to NVIDIA’s DLSS or deep learning super sampling. This enabled AI-assisted upsampling of game assets using scans of high resolution assets, which results in close to native resolution image quality but with a lower rendering budget.

AMD has promised a technology called Super Resolution, which seems similar to DLSS but provided no further details that would let us know how well it would work. The feature is also still in development and won’t be available for some time after the launch of these cards. That should put a severe dent in AMD’s ray tracing performance compared to NVIDIA’s DLSS, which is likely why AMD hasn’t announced support for existing DXR titles as most of them have some form of DLSS.

[embedded content]

Now for availability. The RX 6800 and 6800 XT will be available starting November 18 for the aforementioned prices of $579 and $649 on AMD.com. The RX 6900 XT will be available starting December 8 for $999. Cards will also be available from the usual board partners, such as ASRock, ASUS, Gigabyte, MSI, PowerColor, SAPPHIRE, and XFX in November.

Source 1Source 2

Let’s block ads! (Why?)



Source link

Continue Reading

Trending