It’s been another insanely busy two weeks for us here at GN, sandwiched between various product launches. Of course, this week has been anchored by the arrival of AMD’s RX 6000-series RDNA 2 GPUs. As ever, you can find our RX 6800XT and RX 6800 reviews, as well as our usual teardowns, on our YouTube channel. We’ll briefly recap them below, but you’ll need to watch the reviews for the full scope and context.
Outside of consumer GPUs, there’s also new GPUs from both AMD and Nvidia this week aimed at HPC and supercomputing. There’s also news of a new security co-processor from Microsoft, developed in collaboration with partners such as AMD, Intel, and Qualcomm. There’s also an interesting prospect of using supercomputers to help fuel progress in silicon manufacturing.
As usual, article and video embed follow below.
Hardware news this week has been busy, once again, slotting right in between silicon product releases. Our AMD Ryzen 5000 coverage is mostly done, but we're now ramping into RX 6000 GPU coverage. While preparing work for the RX 6800 XT (and subsequent) GPU launches, we opened a dialogue with NVIDIA to ask about a potential PCIe resizable BAR implementation as a counter to AMD's SAM. That's our leading story for this one, followed-up by some coverage of the Zen 3 delidding work done recently, Intel's add-in GPU for servers, and more.
This past week was slammed for us. We posted reviews of the AMD Ryzen 5 5600X, Ryzen 7 5800X, Ryzen 9 5900X, and Ryzen 9 5900X, all available on our YouTube channel. In the time since, we’ve been working on Ryzen memory benchmarks, including an upcoming piece featuring Wendell of Level1Techs and Buildzoid of AHOC. That piece will focus on ranks, channels, and Zen 3’s newly exaggerated behaviors with regard to interleaving and scaling. More on that soon. For hardware news this week, the big story is a GN-exclusive about NVIDIA MSRP targets and BOM cost suggestions for an RTX 2060-style replacement.
This hardware news episode was filmed prior to the AMD news update, which we covered in full depth on the YouTube channel over here: AMD RX 6900 XT, 6800 XT, & 6800 Specs. We won't be going back over that announcement in this content, so you can check out the full RX 6000 details in that fully dedicated piece with additional AMD Q&A. In this one, we'll be talking RTX models being canceled or pushed back, AMD supercomputer pushes, Intel selling off part of its business, and more.
It’s been something of a busy week in hardware and enthusiast related news this past week, even with Apple’s iPhone event and Amazon’s glutenous Amazon Prime Day sales seemingly dictating part of the news cycle. Still, we’ve got a few stories worth talking about, and as ever, we’ve been busy with other coverage here at GN.
It looks like EVGA is among the first vendors to attempt to address the power limit issue for overclockers looking to push the RTX 3080 ever further, as the company has released a new beta VBIOS that raises the power target. Additionally, there are the usual rumblings in the memory market about price drops, a China-designed 7-nm class chip, an interesting vector supercomputer headed for Japan, and Gundam parts branded by ASUS.
At GN, we’ve been continuing our RTX 30-series coverage with a review and teardown of the ASUS RTX 3080 TUF OC, and we looked at Nvidia’s Reflex suite, including its Latency Analyzer and Reflex Low Latency Mode.
As usual, find the news recap and video embed below.
Hardware news this past week has been busy, with the main coverage being AMD's Zen 3 CPUs (5000 series, like the 5950X, 5900X, 5600X, and 5800X), which we covered in a news piece previously. Following that, for this news recap, we've been updated on the MSI "scalping" story, RTX 3080 and 3090 inventory numbers for a European retailer, and how EVGA is still getting through day-one orders for the RTX 3080. Additional stories include Intel's quasi-announcement of Rocket Lake's timelines, NVIDIA's A6000 and A40 GPU specs, and Razer's cringe-worthy credit card.
At GN, we’re slowly emerging from our RTX 30-series coma, where we’ve pushed our testing and coverage perhaps as far as we ever have. We’re getting ready to slow down for a week or so to revamp and improve processes internally and get ready to do it all again with Zen 3, RDNA2, and the RTX 3070.
As ever, there’s plenty to cover outside of our reviews and testing. This week, we have news regarding NVIDIA delaying the RTX 3070 launch window to the end of October in an effort to avoid the previous RTX 3080 and 3090 catastrophe. There’s also a credible rumor suggesting that Zen 3 will come in under the Ryzen 5000-series banner, which would probably be for the best, given how convoluted CPU naming is getting.
Elsewhere, we discuss Intel’s Omni-Path being resurrected under the new Cornelis Networks, leaked Windows XP source code, an interesting new HPE-Cray built supercomputer, and more. Check out the article and video embed below.
Even with Nvidia's RTX 30-series launches the past couple of weeks, it has still been busy on fronts outside of new GPUs. We’ve been exhaustively testing the various aspects of the RTX 3080 and 3090 and have been diving into enthusiast overclocking.
Outside of that, we’ve got plenty of hardware and industry news to cover, offering some additional reporting and analysis on Nvidia’s botched launches and Microsoft's titanic ZeniMax/Bethesda buyout. We also have some news regarding Nvidia’s use of Micron’s GDDR6X memory and why it opted to keep clock rates modest -- for now.
We also have some information on AMD’s new Ryzen and Athlon 3000 C-series CPUs, Intel still shipping secret products to Huawei (with a license), Western Digital forming new business units internally, and the arrival of Amazon’s new Luna game streaming service. News article and video embed follow below, as usual.
The past week has been filled with RTX 3000-series coverage on the YouTube channel, including a world record-setting livestream with liquid nitrogen. Now, though, it's time to go through the hardware news over the past 10 days or so. We'll be talking about the recent Bethesda/Microsoft acquisition, NVIDIA attempting to buy ARM (and, unrelatedly, change SLI support), and 5nm capacity at TSMC. Show notes to follow, along with the embedded video.
We’re reviewing the NVIDIA RTX 3080 FE today, where new testing methods for pressure analysis, acoustics, and game benchmarks are all in place. We also have a separate upcoming piece involving Schlieren photography of the RTX 3080 FE card, but we can show a short clip of that here as a preview. More on that soon. These cards are so complex -- especially thermally -- that the hardest part was figuring out how to segment the content in a way that’s usable and also possible to complete. Today, our focus is on rasterized games, hybrid rendered games, and path-traced games, alongside basic thermals, acoustics, pressure, power, and coldplate flatness. We have a separate piece going up today for a tear-down of the RTX 3080 (but a quick note that we finished all testing prior to the tear-down, as always). We’ll have more on PCIe generation results, but rest assured that our benchmarks use the best-performing bench, and more thermals.
Note: This is a transcript from our video review of the NVIDIA RTX 3080 Founders Edition card. You can watch that here (or embedded below). We also have a tear-down video coming up.
If you would like to learn about our new GPU testing methodology, we have a video on the channel here.
We moderate comments on a ~24~48 hour cycle. There will be some delay after submitting a comment.