Memory - Page 5 of 148

Apple WWDC 2021: iOS 15, new MacBook Pros, and what else to expect

Apple’s annual developer extravaganza, the Worldwide Developers Conference (WWDC), is coming up fast, kicking off with the keynote presentation on June 7th at 1PM ET. Like last year, WWDC will be an entirely digital and online-only event due to the COVID-19 pandemic, and for the keynote, that means we can likely expect another tightly produced video highlighting everything Apple has in store.

While we aren’t expecting any announcements on the level of Apple’s shift to custom silicon in its computers, which was WWDC 2020’s big news, Apple presumably has some notable changes in the works for iOS, iPadOS, macOS, and its other operating systems. And if the current rumors pan out, we could also see brand-new MacBook Pros with the return of some long-missed features, such as MagSafe charging.

Read on to learn everything we expect from the big show. And don’t be surprised if Apple has a few surprises in store, too.

Photo by Vjeran Pavic / The Verge

iOS 15 may bring improvements to notifications and iMessage

We haven’t heard much about what may be coming to Apple’s next version of its mobile operating system, which will presumably be called iOS 15, but we could see big changes to notifications and possibly iMessage, according to Bloomberg.

For notifications, you may be able to have different notification settings for situations like driving, working, sleeping, or even a custom category, and you’ll be able to flip those on as you need to. You might also be able to set automatic replies based on which notification setting you’re currently using, like what you can do now with Do Not Disturb while driving mode. Personally, I’m hoping iOS 15 will let me allow notifications from a select few people while silencing just about everything else.

As for iMessages, Apple is apparently working on features to make it act like “more of a social network” to compete with Facebook’s WhatsApp, Bloomberg said, but those features are still “early in development” and could be announced at a later date.

Apple also plans to add a feature that shows you apps that are silently collecting data about you, continuing the company’s trend of adding privacy-focused updates to its operating systems.

For iPadOS 15, you can apparently expect a major update to the homescreen, including the ability to put widgets anywhere you want. And with Apple just introducing the new M1-powered iPad Pros, here’s hoping we see some new upgrades to take advantage of the new chip.

In May, Apple also announced a lot of new accessibility features coming to Apple’s operating systems, such as improvements in iOS to VoiceOver, support for bidirectional hearing aids, a built-in background sounds player, and new Memoji customizations like cochlear implants. Apple said these features would arrive “later this year,” which suggests they’ll be included in iOS 15.

We don’t know much about macOS, watchOS 8, and tvOS 15 — but we could see a new “homeOS”

We haven’t heard all that much about upcoming software updates for the Mac, Apple Watch, and Apple TV, so we’ll just have to wait and see what Apple is cooking up. One tidbit: macOS could be a “more minor” update, Bloomberg says. That wouldn’t be too much of a surprise, given that the macOS operating system got a big overhaul with Big Sur last year.

However, we could see the introduction of a brand-new operating system called “homeOS,” which was recently mentioned in and later removed from an Apple job listing. While it’s unclear exactly which devices this OS is for, perhaps it will work on Apple’s home-focused products like the Apple TV and HomePod Mini.

Photo by Alexander Kramer for The Verge

New, redesigned MacBook Pros and a new Apple CPU could be announced

Apple doesn’t always introduce new hardware at WWDC, but this year, new MacBook Pros seem like a possibility. In a May 18th report, Bloomberg said that new MacBook Pros might arrive “as soon as early this summer,” which could indicate an announcement at WWDC.

These new laptops would have new Apple-designed processors that would “greatly outpace the performance and capabilities of the current M1 chips,” according to Bloomberg. The M1 is already pretty dang good, so it sounds like these new chips could be even more impressive.

Apple is apparently planning on releasing two chips for the new Pros. Both should have eight high-performance cores and two energy-efficient cores, while leaving you with the option of either 16 or 32 graphics cores. (By comparison, the M1’s CPU has four high-performance and four energy-efficient cores, while its GPU is offered with either seven or eight cores.) You’ll probably also be able to spec the laptops with as much as 64GB of memory, up from a max of 16GB on M1-equipped computers.

The new laptops should be offered with either 14-inch or 16-inch screens and those screens could have “brighter, higher contrast” displays, according to a Bloomberg report from January. The laptops may also have a new design with flat edges as in the iPhone 12, analyst Ming-Chi Kuo said in January. I’m curious to see what that design might look like in practice — I worry that the hard edges could be uncomfortable if you have the laptop on your lap.

The best rumor is that the new design may also mark the return of some of the ports and features that were taken away with the now-infamous 2016 MacBook Pro redesign, including a MagSafe charger, an HDMI port, and an SD card slot, Bloomberg said in its May report. And, according to Kuo, the OLED Touch Bar currently found on Intel-based MacBook Pros will apparently be removed in favor of physical function keys.

We could see at least one other new Mac

While it seems like MacBook Pros are the only new hardware we’ll be seeing at WWDC this year, that hasn’t stopped some other Mac rumors from swirling lately, and there’s always the chance Apple could announce more at its big event. According to Bloomberg, Apple also has “a revamped MacBook Air, a new low-end MacBook Pro, and an all-new Mac Pro workstation” in the works as well as a “higher-end Mac Mini desktop and larger iMac,” all of which would be powered by Apple’s custom silicon.

The new Mac Mini may have the same chip as the new MacBook Pros. The new Mac Pro could be a beast, with processors that are “either twice or four times as powerful as the new high-end MacBook Pro chip.”

And the redesigned “higher-end” MacBook Air could arrive as early as the end of this year. Frankly, I hope that refreshed Air arrives even later. I just bought the M1-equipped Air and it’s one of the best computers I’ve ever used, but I have a bad feeling I’ll be first in line to buy a redesigned and more capable Air anyway. (Especially if it gets the MagSafe charger that’s rumored for the new Pros.)

Apple might have dropped a hint about its AR / VR headset

Apple has long been rumored to have a mixed reality headset in the works, and recently, we’ve learned a few more potential details about it. The headset might be very expensive — approximately $3,000, according to one report — though it could be packed with 8K displays, more than a dozen cameras to track hand movements and capture footage, and might weigh less than an iPhone, too.

While the headset could be a ways out, as it’s not expected to ship until 2022 at the earliest, a few suspicious details in Apple’s WWDC promotional images may be hinting toward some kind of reveal of Apple’s upcoming headset or the software on which it runs.

Check out this image below (that I also used at the top of this post), which Apple released alongside the announcement of WWDC in March. Notice the way the app icons are reflected in the glasses — I could imagine some sort of mixed reality headset showing icons in front of your eyes in a similar way.

Those icons sure are reflected in an interesting spot.

Image: Apple

Apple continued that reflections motif with new images released in May — you can see things from the laptop screens reflected in all of the eyes of the Memojis.

Look at those reflections.

Image: Apple

Now, these reflections may just be Apple’s artists flexing their design chops. And if I had to guess, given how far out a rumored mixed reality headset is, I don’t think we’re going to see anything about it at WWDC this year.

But Apple has surprised us in the past, and maybe these images are an indication of one more thing Apple has in store for WWDC.

Google Replaces Millions of Intel’s CPUs With Its Own Homegrown Chips

(Image credit: Intel)

Google has designed its own new processors, the Argos video (trans)coding units (VCU), that have one solitary purpose: processing video. The highly efficient new chips have allowed the technology giant to replace tens of millions of Intel CPUs with its own silicon.

For many years Intel’s video decoding/encoding engines that come built into its CPUs have dominated the market both because they offered leading-edge performance and capabilities and because they were easy to use. But custom-built application-specific integrated circuits (ASICs) tend to outperform general-purpose hardware because they are designed for one workload only. As such, Google turned to developing its own specialized hardware for video processing tasks for YouTube, and to great effect.

However, Intel may have a trick up its sleeve with its latest tech that could win back Google’s specialized video processing business.

Loads of Videos Require New Hardware

Users upload more than 500 hours of video content in various formats every minute to YouTube. Google needs to quickly transcode that content to multiple resolutions (including 144p, 240p, 360p, 480p, 720p, 1080p, 1440p, 2160p, and 4320p) and data-efficient formats (e.g., H.264, VP9 or AV1), which requires formidable encoding horsepower.

Historically, Google had two options for transcoding/encoding content. The first option was Intel’s Visual Computing Accelerator (VCA) that packed three Xeon E3 CPUs with built-in Iris Pro P6300/P580 GT4e integrated graphics cores with leading-edge hardware encoders. The second option was to use software encoding and general-purpose Intel Xeon processors.

Google decided that neither option was power-efficient enough for emerging YouTube workloads – the Visual Computing Accelerator was rather power hungry itself, whereas scaling the number of Xeon CPUs essentially meant increasing the number of servers, which means additional power and datacenter footprint. As a result, Google decided to go with custom in-house hardware.

(Image credit: Google)

Google’s first-generation Argos VCU does not replace Intel’s central processors completely as the servers still need to run the OS and manage storage drives and network connectivity. To a large degree, Google’s Argos VCU resembles a GPU that always needs an accompanying CPU.

Instead of stream processors like we see in GPUs, Google’s VCU integrates ten H.264/VP9 encoder engines, several decoder cores, four LPDDR4-3200 memory channels (featuring 4×32-bit interfaces), a PCIe interface, a DMA engine, and a small general-purpose core for scheduling purposes. Most of the IP, except the in-house designed encoders/transcoders, were licensed from third parties to cut down on development costs. Each VCU is also equipped with 8GB of usable ECC LPDDR4 memory.

The main idea behind Google’s VCU is to put as many high-performance encoders/transcoders into a single piece of silicon as possible (while remaining power efficient) and then scale the number of VCUs separately from the number of servers needed. Google places two VCUs on a board and then installs 10 cards per dual-socket Intel Xeon server, greatly increasing the company’s decoding/transcoding performance per rack.

Increasing Efficiency Leads to Migration from Xeon

Google says that its VCU-based machines have seen up to 7x (H.264) and up to 33x (VP9) improvements in performance/TCO compute efficiency compared to Intel Skylake-powered server systems. This improvement accounts for the cost of the VCUs (vs. Intel’s CPUs) and three years of operational expenses, which makes VCUs an easy choice for video behemoth YouTube.

Offline Two-Pass Single Output (SOT) Throughput in CPU, GPU, and VCU-Equipped Systems

System	Throughput (MPix/s)	Throughput (MPix/s)	Performance/TCO	Performance/TCO
	H.264	VP9	H.264	VP9
2-way Skylake	714	154	1x	1x
4x Nvidia T4	2,484	–	1.5x	–
8x Google Argos VCUs	5,973	6,122	4.4x	20.8x
20x Google Argos VCUs	14,932	15,306	7x	33.3x

From performance numbers shared by Google, it is evident that a single Argos VCU is barely faster than a 2-way Intel Skylake server in H.264. However, since 20 VCUs can be installed into such a server, VCU wins from an efficiency perspective. But when it comes to the more demanding VP9 codec, Google’s VCU appears to be five times faster than Intel’s dual-socket Xeon and therefore offers impressive efficiency advantages.

Since Google has been using its Argos VCUs for several years now, it clearly replaced many of its Xeon-based YouTube servers with machines running its own silicon. It is extremely hard to estimate how many Xeon systems that Google actually replaced, but some analysts believe the technology giant could have swapped from four to 33 million Intel CPUs for its own VC. Even if the second number is an overestimate, we are still talking about millions of units.

(Image credit: Google)

Since Google needs loads of processors for its other services, it is likely that the number of CPUs that the company buys from AMD or Intel is still very high and is not going to decrease any time soon as it will be years before Google’s own datacenter-grade system-on-chips (SoCs) will be ready.

It is also noteworthy that in an attempt to use innovative encoding technologies (e.g., AV1) right now, Google needs to use general-purpose CPUs even for YouTube as the Argos does not support the codec. Furthermore, as more efficient codecs emerge (and these tend to be more demanding in terms of compute horsepower), Google will have to continue to use CPUs for initial deployments. Ironically, the advantage of dedicated hardware will only grow in the future.

Google is already working on its second-gen VCU that supports AV1, H.264, and VP9 codecs as its needs to further increase the efficiency of its encoding technologies. It is unclear when the new VCUs will be deployed, but it is clear that the company wants to use its own SoCs instead of general-purpose processors where possible.

Intel Isn’t Standing Still

Intel isn’t standing still, though. The company’s DG1 Xe-LP-based quad-chip SG1 server card can decode up to 28 4Kp60 streams as well as transcode up to 12 simultaneous streams. Essentially, Intel’s SG1 does exactly what Google’s Argos VCU does: scale video decoding and transcoding performance separately from the server count and thus reduce the number of general-purpose processors required in a data center used for video applications.

(Image credit: Intel)

With its upcoming single-tile Xe-HP GPU, Intel will offer transcoding of 10 high-quality 4Kp60 streams simultaneously. Keeping in mind that some of Xe-HP GPUs will scale to four tiles, and more than one GPU can be installed per system, Intel’s market-leading media decoding and encoding capabilities will only become even more solid.

Summary

Google has managed to build a remarkable H.264 and VP9-supporting video (trans)coding unit (VCU) that can offer significantly higher efficiency in video encoding/transcoding workloads than Intel’s existing CPUs. Furthermore, VCUs enable Google to scale its video encoding/transcoding performance independently from the number of servers.

Yet, Intel already has its Xe-LP GPUs and SG1 cards that offer some serious video decoding and encoding capabilities, too, so Intel will still be successful in datacenters with heavy video streaming workloads. Furthermore, with the emergence of Intel’s Xe-HP GPUs, the company promises to solidify its position in this market.

RTX 3070 Ti Benchmark Results Leaked for Ashes of The Singularity and Geekbench

(Image credit: Nvidia)

We have our first glimpse at the performance of Nvidia’s recently announced GeForce RTX 3070 TI, thanks to two leaked benchmark runs from Ashes of the Singularity and Geekbench 5. They were brought to our attention by @leakbench on Twitter. However, the benchmark results are about as vague as it gets, giving us a little indication as to the real performance of the 3070 Ti.

Unfortunately for us, the Ashes of the Singularity score in particular tells us almost nothing about the performance of Nvidia’s new mid-range part, due to the system configuration and game quality settings.

The RTX 3070 Ti was paired with a Ryzen 9 3900X and 32GB of system memory, and the resolution used was 1080P with a high-quality preset (not the crazy preset). The 3070 Ti scores an average frame rate of 105.5 frames per second and a CPU frame rate of 105.9 fps.

(Image credit: Twitter)

While the frame rate looks good, this is probably one of the worst Ashes benchmarks to date, as the resolution is locked to 1080P and the teseter didn’t even use the maximum quality preset. On top of that, the Ryzen 9 3900X is a Zen 2 part, and at 1080P there should be a noticeable amount of bottlenecking, enough to skew performance results when comparing graphics cards specifically.

What we really needed from the leaker is a better testbed with a more powerful CPU like a Ryzen 5000 series or a highly overclocked Intel 10th gen or 11th gen part. But most importantly, we would like to see resolutions of 1440P or 4K and have the game running at the crazy preset.

As it stands now, there aren’t enough benchmarks on the high preset, with the same CPU and memory configuration to judge performance. For instance, the top result for the 1080P high preset benchmark results belongs to a system running on an i7-11700KF running 16GB of RAM, with an RTX 3080, which has exactly 1-2 FPS less on its score than the leaked RTX 3070 Ti test.

Geekbench

(Image credit: Twitter)

Hopefully, Geekbench 5’s result can give us some better guesses as to the performance of the RTX 3070 Ti. Paired with a Core i9-11900K, the RTX 3070 Ti scored 155763 points in the OpenCL test.

For comparison, the RTX 3080 on Geekbench’s browser earned a score of 183,452 and the RTX 3070 scores a 135,886. So the RTX 3070 Ti is sandwiched right in-between the vanilla 3080 and 3070, being 17% slower than the RTX 3080 and 14% faster than the RTX 3070, which is to be expected.

Take this result with a pinch of salt, until you see full reviews of the RTX 3070 Ti come online. This is especially true of Geekbench 5, which focuses on the raw compute performance of GPUs, which rarely applies to the actual gaming performance of Nvidia and AMD’s graphics cards.

NZXT N7 Z590 Motherboard Review: Unique Style, and Capable

Our Verdict

NZXT’s N7 Z590 is a full-featured Z590 motherboard that includes two M.2 sockets, Intel-based Wi-Fi 6E, capable power delivery, premium audio, and more. It’s a well-rounded mid-ranger for Intel’s Z590 platform, though for a similar price, there are other options with more and better parts.

For

+ Wi-Fi 6E/2.5 GbE Networking
+ Improved and capable power delivery

Against

– Last-gen audio codec
– Only two M.2 sockets and four SATA ports

Go to page:

Features and Specifications

NZXT has joined the Intel Rocket Lake motherboard party with the N7 Z590, bringing PCIe 4.0 capability (with a Rocket Lake CPU), capable power delivery, Intel Wi-Fi 6E, and a unique design aesthetic that easily matches most themes. With an MSRP of $279.99, the N7 positions itself as a strong competitor in the mid-range Z590 space.

NZXT worked with ASRock to bring you this board which appears to be based on the ASRock Z590 Steel Legend. We reviewed that board a couple of months back, and overall liked what ASRock offered at the price point. The N7 Z590, like previous NZXT motherboards, employs unique-looking shrouds/heatsinks that cover most of the board, giving it that signature NZXT appearance.

Performance-wise, the N7 Z590 did well overall. Its results traded punches with the other boards in most tests. Like the Steel Legend it’s modeled from, this board follows Intel specifications, and in a couple of tests (Handbrake, Cinebench/POV-Ray single thread), the times/scores were lower than the others. To bypass that, simply adjust the power limits up, as the other boards do from the factory. The N7 board set our DDR4 3600 sticks at 1:1 with the memory controller, and we saw solid results in our memory bandwidth and latency tests. Overclocking was a breeze too, as we set our CPU to 5.1 GHz while running the memory at DDR4 4000 (with a few tweaks for stability).

Although you can’t tell the difference by looking at the board, the Z590 version of NZXT’s board strives to improve upon the last generation and does so with aplomb. The N7 Z590 includes PCIe 4.0 M.2 socket (two total sockets, the other is PCIe 3.0) and a PCIe 4.0 x16 slot, additional USB ports including a USB 3.2 Gen2x2 Type-C on the rear panel, improved power delivery, and more. We’ll cover these features in detail and more below. But first, here are the full specs from NZXT.

Specifications – NZXT N7 Z590

Socket	LGA 1200
Chipset	Z590
Form Factor	ATX
Voltage Regulator	12 Phase (12+2, 50A MOSFETs)
Video Ports	(1) HDMI (v2.0)
USB Ports	(1) USB 3.2 Gen 2, Type-C (10 Gbps)
	(3) USB 3.2 Gen 2, Type-A (10 Gbps)
	(4) USB 3.2 Gen 1, Type-A (10 Gbps)
	(2) USB 2.0 (480 Mbps)
Network Jacks	(1) 2.5 GbE
Audio Jacks	(5) Analog + SPDIF
Legacy Ports/Jacks	✗
Other Ports/Jack	✗
PCIe x16	(1) v4.0 x16
	(1) v3.0 x4
PCIe x8	✗
PCIe x4	✗
PCIe x1	(2) v3.0 x1
CrossFire/SLI	AMD CrossFireX
DIMM slots	(4) DDR4 4600(OC), 128GB Capacity
M.2 slots	(1) PCIe 4.0 x4 / PCIe (up to 80mm)
	(1) PCIe 3.0 x4 / PCIe and SATA (up to 80mm)
U.2 Ports	✗
SATA Ports	(4) SATA3 6 Gbps
	Supports RAID 0, 1, 5 and 10
USB Headers	(1) USB v3.2 Gen 2 (Front Panel Type-C)
	(2) USB v3.2 Gen 1
	(2) USB v2.0
Fan/Pump Headers	(7) 4-Pin
RGB Headers	(2) NZXT RGB
	(1) ARGB (3-pin)
	(1) RGB (4-pin)
Legacy Interfaces	✗
Other Interfaces	FP-Audio, TPM
Diagnostics Panel	Q-LED
Internal Button/Switch	Power and Reset buttons
SATA Controllers	✗
Ethernet Controller(s)	(1) RealtekRTL8125BG (2.5 GbE)
Wi-Fi / Bluetooth	(1) Intel AX210 (WiFi 6, 2×2 160 Hz/6 GHz, MU-MIMO, OFDMA, BT 5.2)
USB Controllers	ASMedia ASM1074
HD Audio Codec	Realtek ALC1220
DDL/DTS Connect	✗ / ✗
Warranty	3 Years

Inside the box included with the motherboard are a few accessories. Our review board didn’t even have a driver disk (perhaps because this is a pre-launch sample), but what’s inside should allow you to get going without an extra trip to the store. For drivers, if a disk isn’t included in the retail packaging, you can get them from the NZXT website. Below is a complete list of the included parts.

Motherboard Guide
Wi-Fi Antenna
(4) SATA cables
(2) Screw package for M.2 sockets

Image 1 of 3

(Image credit: NZXT)

Image 2 of 3

(Image credit: NZXT)

Image 3 of 3

(Image credit: NZXT)

After taking the board out of the box, we get a chance to see the nearly completely covered 6-layer matte-black PCB. Just about the only design cues are the punched-out circles above the chipset and left VRM bank heatsinks. Additionally, some NZXT branding resides on the top M.2 socket cover and over the rear IO area.

If you’re looking for integrated RGB lighting, you’ll have to look somewhere else, as the N7 Z590 doesn’t include any. That said, the board has two RGB headers, a 3-pin ARGB and 4-pin RGB, along with two NZXT RGB LED connectors. The NZXT CAM software handles all RGB lighting control. Overall, I like how this board looks. The mostly covered PCB gives the board a premium look and feel that matches most any build theme. If a stealthy all-black motherboard isn’t your thing, the N7 Z590 also comes in white.

(Image credit: NZXT)

Looking at the top half of the board, there isn’t much to see outside of shrouds. On the left, the IO cover reaches out over the rear IO chips and touches the left VRM bank. The heatsink doesn’t have a lot of surface area. but does an excellent job keeping the power bits cool, even while overclocked as we’ll see later.

Above the heatsink on the top edge, we find the 8-pin EPS (required) and a 4-pin EPS connector (optional) to feed power to the CPU. To the right are two (of seven) 4-pin fan headers. Each fan/pump header on the board is capable of 2A/24W output, as well as auto-detecting what type of fan is connected (PWM or DC). Continuing right are two 4-pin NZXT RGB headers and two more fan headers.

To the right of the socket area are four unreinforced DRAM slots that support up to 128GB of RAM. NZXT lists supported speeds up to DDR4 4600(OC). This is on the lower side compared to other boards, but most users will run RAM well under that speed. Our DDR4 4000 kit worked with a minor voltage change to the VccIO Memory voltage (0.10+), so we know it’s good to at least that point. Much beyond that and you pay a steep premium for little in the way of gains anyway.

On the right edge of the board, the only thing visible from the top is the 24-pin ATX for board power and a USB 3.2 Gen 2 Type-C front-panel port.

(Image credit: Tom’s Hardware)

NZXT lists the N7 Z590 as a 12-phase Dr.MOS VRM, which breaks down to a 12+2 configuration for the Vcore and SOC. A Richtek RT3609BE (X+Y=8) 8-channel controller handles the CPU while a Renesas RAA229001 controls the SOC. The eight-channel controller feeds 12 Vishay Sic654 50A MOSFETs for CPU Vcore in a teamed/parallel configuration. In other words, NZXT does not use phase doublers on this board. This configuration is plenty for both 10th and 11th generation CPUs on this platform.

(Image credit: NZXT)

Focusing in on the bottom half of the board, we’re again greeted primarily by shrouds. Hidden underneath on the left-hand side is a Realtek ALC1220 codec, along with five audio caps and a Texas Instruments NE5532 OpAmp. While the ALC1220 codec is a solid audio solution (it was the flagship of Z490) most users are happy with, I would like to have seen the latest and greatest ALC4080/4082 here instead.

In the middle are a few PCIe slots and two M.2 sockets. Starting with PCIe, there are two full-length slots and three x1 size slots. The top slot is PCIe 4.0-capable with a Rocket Lake-based CPU, while the second full-length slot runs PCIe 3.0 x4 max and is fed from the chipset. The documentation doesn’t mention multi-GPU support. But by lane count, it should be able to run 2-Way CrossfireX. The three x1 slots are PCIe 3.0 and fed from the chipset as well. I like the x1 slot location, so you can easily insert an AIC without worrying about the location covering a full-length slot.

The top M.2 socket is located to the right of the top x1 slot. When using a Rocket Lake-based CPU, this socket runs at PCIe 4.0 x4 speeds and supports up to 80mm modules. The bottom M.2 socket runs at PCIe 3.0 and accepts SATA-based modules up to 80mm. The manual does not mention RAID functionality. Worth a mention is that although the M.2 covers are metal, they do not contact the M.2 modules to help with cooling. If you have hot-running M.2 modules, you may want to keep an eye out on temperatures.

As we move further to the right, we pass over the chipset heatsink on the way to the right edge. Here, hidden under the shroud with horizontally oriented connectors, we spy a USB 3.2 Gen 1 header and the four SATA ports. Although most users would be happy with four SATA ports and two M.2 sockets, this is less than most other boards at this price point. Many have three M.2 sockets and six SATA ports, all of which can be active (though not always) in specific configurations.

Across the board’s bottom are several headers, including more USB ports, fan headers and more. Below is the complete list, from left to right:

Front-panel audio
UART header
RGB and ARGB headers
(3) USB 2.0 header
(3) System Fan headers
Q-Code LEDs
Clear CMOS jumper
USB 3.2 Gen 1 connector
Power/Reset buttons
Front panel header

(Image credit: NZXT)

Flipping the board around to the rear IO area, we see the black pre-installed IO plate, which matches the colors and design of the rest of the board, along with the NZXT branding in white. In total, there are 10 USB ports: You get four USB 3.2 Gen 2 ports (3x Type-A, 2x Type-C), four USB 3.2 Gen 1 ports, and two USB 2.0 ports. A single HDMI port handles video output when using the integrated graphics on the CPU. You’ll also find the Wi-Fi antenna headers, a Clear CMOS button, the Realtek 2.5 GbE and the 5-plug plus SPDIF audio stack.

MORE: Best Motherboards

MORE: How To Choose A Motherboard

MORE: All Motherboard Content

Today’s Newegg Shuffle Is All RTX 3080 Tis

(Image credit: Newegg)

Today’s Newegg shuffle is actually worth paying attention to, since it’s all RTX 3080 Ti cards, three of which match the Founders Edition’s $1,199 MSRP. In a sea of sold out signs and gray “add to cart” buttons, even the minimal chance of winning a Newegg Shuffle actually starts to look enticing.

All of today’s picks are from third party manufacturers, ranging from Gigabyte to EVGA to Asus to Zotac. That’s not necessarily a bad thing, though. We’ve only reviewed Nvidia’s own GeForce RTX 3080 Ti so far, but in our review, we said “Anyone still hoping to pick up a 3080 Ti should also opt for a third party card with higher factory clocks and a beefier cooler.” That’s because, in our testing, the RTX 3080 Ti Founders Edition tended to run a bit hot with higher fan speeds than previous 30-series Founders Edition cards — no surprise considering it’s the same cooler as the 3080, but now with a higher TDP and two more GDDR6X modules.

Luckily, all of today’s cards have three fans each, including the ones that match the FE’s MSRP. Specifically, here’s what you can get today:

Gigabyte Eagle GeForce RTX 3080 Ti for $1,199
EVGA XC3 GeForce RTX 3080 Ti for $1,199
Asus TUF Gaming GeForce RTX 3080 Ti for $1,199
EVGA FTW3 GeForce RTX 3080 Ti for $1,399
Gigabyte Gaming GeForce RTX 3080 Ti for $1,479
Gigabyte Vision GeForce RTX 3080 Ti for $1,529
Zotac AMP Holo GeForce RTX 3080 Ti for $1,799

For most people, even $1,199 will be too expensive, but that’s just the starting price. In a world of perpetually sold out GPUs, prices have skyrocketed. There’s a $200 gap between the three MSRP-level cards and the next costom card, the EVGA FTW3. Zotac’s AMP Holo meanwhile kicks reason to the curb and sports a $1,799 price tag. At that point, you should be getting a custom RTX 3090 with twice the memory.

Today’s Newegg Shuffle is active for just over 2 more hours. If you’re not aware of how the Newegg Shuffle works, it uses a lottery format. Just select the component(s) you’d like to potentially buy. Then Newegg will hold a drawing later today, after which the ‘winners’ will be notified by email with the chance to purchase an item (only one) within a several-hour period.

For other ways to get hard-to-find graphics cards, check out our RTX 3080 stock tracker and our feature on where to buy RTX 30-series cards. And for more Newegg savings, visit out page of Newegg promo codes.

Samsung Unveils ZNS SSD: QLC V-NAND for Datacenter-Grade Endurance & Performance

(Image credit: Samsung)

Samsung has introduced its first Zones Namespaces (ZNS) solid-state drives that combine high performance, long endurance, relatively low price enabled by the company’s QLC V-NAND, and improved quality of service (QoS) for datacenters. To use Samsung’s new PM1731a ZNS SSDs, datacenters will have to deploy new storage systems and software.

ZNS SSDs work differently than conventional block-based drives and have a number of advantages. ZNS SSDs write data sequentially into large zones and have better control over write amplification, which reduces over-provisioning requirements by an order of magnitude. For example, some enterprise drives rated for 3 DWPD (drive writes per day) reserve about a third of their raw capacity for over-provisioning, but for ZNS SSDs about 10% of that is enough. In addition, since ZNS uses large zones instead of many 4KB blocks, garbage collection is not needed as often as traditional SSDs, it also improves real-world read and write performance.

Samsung’s 2.5-inch PM1731a ZNS SSDs are based on the company’s proprietary dual-port controller as well as 6th Generation QLC V-NAND memory. The drives will be available in 2TB and 4TB capacities.

(Image credit: Samsung)

Samsung says that its new PM1731a ZNS drives will last up to four times longer than conventional NVMe SSDs, which will reduce their total cost of ownership (TCO) and will simplify server infrastructure.

But enhanced endurance and performance come at a cost. ZNS ecosystem requires new software infrastructure that is not yet widely available. In a bid to make ZNS more widespread, Samsung is participating in a number of open-source ZNS projects. The manufacturer also plans to make its ZNS technology available to xNVMe and participate in the Storage Performance Development Kit (SPDK) community to enable NVMe and SPDK users to implement ZNS more easily.

Samsung will start mass production of its PM1731a ZNS SSDs in the second half of the year. The company is the second major maker of SSDs, the first being Western Digital, to unveil a ZNS SSD.

“Samsung’s ZNS SSD reflects our commitment to introducing differentiated storage solutions that can substantially enhance the reliability and lifetime of server SSDs,” said Sangyeun Cho, senior vice president of the Memory Software Development Team at Samsung Electronics. “We plan to leverage quad-level cell (QLC) NAND technology in our next-generation ZNS drives to enable higher thresholds for storage performance and capacity in the enterprise systems of tomorrow.”

Nvidia goes back to 4K for more than $1K

Last year’s Nvidia RTX 3080 was the first GPU to make 4K gaming finally feasible. It was a card that delivered impressive performance at 4K, especially for its retail price of $699 — far less than the 2080 Ti cost a generation earlier. That was before the reality of a global chip shortage drove the prices of modern GPUs well above $1,000. Now that the street prices of RTX 3080s have stayed above $2,000 for months, Nvidia is launching its RTX 3080 Ti flagship priced at $1,199.

It’s a card that aims to deliver near identical levels of performance to the $1,499 RTX 3090, but in a smaller package and with just 12GB of VRAM — half what’s found on the RTX 3090. Nvidia is effectively competing with itself here, and now offering three cards at the top end. That’s if you can even manage to buy any of them in the first place.

I’ve spent the past week testing the RTX 3080 Ti at both 4K and 1440p resolutions. 4K gaming might have arrived originally with the RTX 2080 Ti, but the RTX 3080 Ti refines it and offers more headroom in the latest games. Unfortunately, it does so with a $1,199 price tag that I think will be beyond most people’s budgets even before you factor in the inevitable street price markup it will see during the current GPU shortage.

Hardware

If you put the RTX 3080 Ti and the RTX 3080 side by side, it would be difficult to tell the difference between them. They look identical, with the same ports and fan setup. I’m actually surprised this card isn’t a three-slot like the RTX 3090, or just bigger generally. The RTX 3080 Ti has one fan on either side of the card, with a push-pull system in place. The bottom fan pulls cool air into the card, which then exhausts on the opposite side that’s closest to your CPU cooler and rear case fan. A traditional blower cooler also exhausts the hot air out of the PCIe slot at the back.

This helped create a quieter card on the original RTX 3080, and I’m happy to report it’s the same with the RTX 3080 Ti. The RTX 3080 Ti runs at or close to its max fan RPM under heavy loads, but the hum of the fans isn’t too distracting. I personally own an RTX 3090, and while the fans rarely kick in at full speed, they’re certainly a lot more noticeable than the RTX 3080 Ti’s.

Nvidia has used the same RTX 3080 design for the 3080 Ti Model.

That quiet performance might have a downside, though. During my week of testing with the RTX 3080 Ti, I noticed that the card seems to run rather hot. I recorded temperatures regularly around 80 degrees Celsius, compared to the 70 degrees Celsius temperatures on the larger RTX 3090. The fans also maxed out a lot during demanding 4K games on the RTX 3080 Ti in order to keep the card cool. I don’t have the necessary equipment to fully measure the heat output here, but when I went to swap the RTX 3080 Ti for another card after hours of testing, it was too hot to touch, and stayed hotter for longer than I’d noticed with either the RTX 3080 or RTX 3090. I’m not sure if this will result in problems in the long term, as we saw with the initial batch of 2080 Ti units having memory overheating issues, but most people will put this in a case and never touch it again. Still, I’m surprised at how long it stayed hot enough for me to not want to touch it.

As this is a Founders Edition card, Nvidia is using its latest 12-pin single power connector. There’s an ugly and awkward adapter in the box that lets you connect two eight-pin PCIe power connectors to it, but I’d highly recommend getting a single new cable from your PSU supplier to connect directly to this card. It’s less cabling, and a more elegant solution if you have a case window or you’re addicted to tidy cable management (hello, that’s me).

The RTX 3080 Ti inside my case.

I love the look of the RTX 3080 Ti and the pennant-shaped board that Nvidia uses here. Just like the RTX 3080, there are no visible screws, and the regulatory notices are all on the output part of the card so there are no ugly stickers or FCC logos. It’s a really clean card, and I’m sorry to bring this up, but Nvidia has even fixed the way the number 8 is displayed. It was a minor mistake on the RTX 3080, but I’m glad the 8 has the correct proportions on the RTX 3080 Ti.

At the back of the card there’s a single HDMI 2.1 port and three DisplayPort 1.4a ports. Just like the RTX 3080, there are also LEDs that glow around the top part of the fan, and the GeForce RTX branding lights up, too. You can even customize the colors of the glowing part around the fan if you’re really into RGB lighting.

Just like the RTX 3080, this new RTX 3080 Ti needs a 750W power supply. The RTX 3080 Ti even draws more power, too, at up to 350 watts under load compared to 320 watts on the RTX 3080. That’s the same amount of power draw as the larger RTX 3090, which is understandable given the performance improvements, but it’s worth being aware of how this might impact your energy bills (and the cost of your PC build to run it).

1440p testing

I’ve been testing the RTX 3080 Ti with Intel’s latest Core i9 processor. For 1440p tests, I’ve also paired the GPU with a 32-inch Samsung Odyssey G7 monitor. This monitor supports refresh rates up to 240Hz, as well as Nvidia’s G-Sync technology.

I compared the RTX 3080 Ti against both the RTX 3080 and RTX 3090 to really understand where it fits into Nvidia’s new lineup. I tested a variety of AAA titles, including Fortnite, Control, Death Stranding, Metro Exodus, Call of Duty: Warzone, Microsoft Flight Simulator, and many more. You can also find the same games tested at 4K resolution below.

All games were tested at max or ultra settings on the RTX 3080 Ti, and most exceeded an average of 100fps at 1440p. On paper, the RTX 3080 Ti is very close to an RTX 3090, and my testing showed that plays out in most games at 1440p. Games like Microsoft Flight Simulator, Assassin’s Creed: Valhalla, and Watch Dogs: Legion all have near-identical performance across the RTX 3080 Ti and RTX 3090 at 1440p.

Even Call of Duty: Warzone is the same without Nvidia’s Deep Learning Super Sampling (DLSS) technology enabled, and it’s only really games like Control and Death Stranding where there’s a noteworthy, but small, gap in performance.

However, the jump in performance from the RTX 3080 to the RTX 3080 Ti is noticeable across nearly every game, with the exception of Death Stranding and Fortnite, which both perform really well on the base RTX 3080.

RTX 3080 Ti (1440p)

Benchmark	RTX 3080 Founders Edition	RTX 3080 Ti Founders Edition	RTX 3090 Founders Edition
Benchmark	RTX 3080 Founders Edition	RTX 3080 Ti Founders Edition	RTX 3090 Founders Edition
Microsoft Flight Simulator	46fps	45fps	45fps
Shadow of the Tomb Raider	147fps	156fps	160fps
Shadow of the Tomb Raider (DLSS)	154fps	162fps	167fps
CoD: Warzone	124fps	140fps	140fps
CoD: Warzone (DLSS+RT)	133fps	144fps	155fps
Fortnite	160fps	167fps	188fps
Fortnite (DLSS)	181fps	173fps	205fps
Gears 5	87fps	98fps	103fps
Death Stranding	163fps	164fps	172fps
Death Stranding (DLSS quality)	197fps	165fps	179fps
Control	124fps	134fps	142fps
Control (DLSS quality + RT)	126fps	134fps	144fps
Metro Exodus	56fps	64fps	65fps
Metro Exodus (DLSS+RT)	67fps	75fps	77fps
Assassin’s Creed: Valhalla	73fps	84fps	85fps
Watch Dogs: Legion	79fps	86fps	89fps
Watch Dogs: Legion (DLSS+RT)	67fps	72fps	74fps
Watch Dogs: Legion (RT)	49fps	55fps	56fps

Assassin’s Creed: Valhalla performs 15 percent better on the RTX 3080 Ti over the regular RTX 3080, and Metro Exodus also shows a 14 percent improvement. The range of performance increases ranges from around 4 percent all the way up to 15 percent, so the performance gap is very game dependent.

Even when using games with ray tracing, the RTX 3080 Ti still managed high frame rates when paired with DLSS. DLSS uses neural networks and AI supercomputers to analyze games and sharpen or clean up images at lower resolutions. In simple terms, it allows a game to render at a lower resolution and use Nvidia’s image reconstruction technique to upscale the image and make it look as good as native 4K.

Whenever I see the DLSS option in games, I immediately turn it on now to get as much performance as possible. It’s still very much required for ray tracing games, particularly as titles like Watch Dogs: Legion only manage to hit 55fps with ultra ray tracing enabled. If you enable DLSS, this jumps to 72fps and it’s difficult to notice a hit in image quality.

The RTX 3080 Ti stays quiet during heavy loads.

4K testing

For my 4K testing, I paired the RTX 3080 Ti with Acer’s 27-inch Nitro XV273K, a 4K monitor that offers up to 144Hz refresh rates and supports G-Sync. I wasn’t able to get any of the games I tested on both the RTX 3080 Ti and RTX 3090 to hit the frame rates necessary to really take advantage of this 144Hz panel, but some came close thanks to DLSS.

Metro Exodus manages a 14 percent improvement over the RTX 3080, and Microsoft Flight Simulator also sees a 13 percent jump. Elsewhere, other games see between a 4 and 9 percent improvement. These are solid gains for the RTX 3080 Ti, providing more headroom for 4K gaming over the original RTX 3080.

The RTX 3080 Ti comes close to matching the RTX 3090 performance at 4K in games like Watch Dogs: Legion, Assassin’s Creed: Valhalla, Gears 5, and Death Stranding. Neither the RTX 3080 Ti nor RTX 3090 is strong enough to handle Watch Dogs: Legion with ray tracing, though. Both cards manage around 30fps on average, and even DLSS only bumps this up to below 50fps averages.

RTX 3080 Ti (4K)

Benchmark	RTX 3080 Founders Edition	RTX 3080 Ti Founders Edition	RTX 3090 Founders Edition
Benchmark	RTX 3080 Founders Edition	RTX 3080 Ti Founders Edition	RTX 3090 Founders Edition
Microsoft Flight Simulator	30fps	34fps	37fps
Shadow of the Tomb Raider	84fps	88fps	92fps
Shadow of the Tomb Raider (DLSS)	102fps	107fps	111fps
CoD: Warzone	89fps	95fps	102fps
CoD: Warzone (DLSS+RT)	119fps	119fps	129fps
Fortnite	84fps	92fps	94fps
Fortnite (DLSS)	124fps	134fps	141fps
Gears 5	64fps	72fps	73fps
Death Stranding	98fps	106fps	109fps
Death Stranding (DLSS quality)	131fps	132fps	138fps
Control	65fps	70fps	72fps
Control (DLSS quality + RT)	72fps	78fps	80fps
Metro Exodus	34fps	39fps	39fps
Metro Exodus (DLSS+RT)	50fps	53fps	55fps
Assassin’s Creed: Valhalla	64fps	70fps	70fps
Watch Dogs: Legion	52fps	55fps	57fps
Watch Dogs: Legion (DLSS+RT)	40fps	47fps	49fps
Watch Dogs: Legion (RT)	21fps	29fps	32fps

Most games manage to comfortably rise above 60fps in 4K at ultra settings, with Microsoft Flight Simulator and Metro Exodus as the only exceptions. Not even the RTX 3090 could reliably push beyond 144fps at 4K without assistance from DLSS or a drop in visual settings. I think we’re going to be waiting on whatever Nvidia does next to really push 4K at these types of frame rates.

When you start to add ray tracing and ultra 4K settings, it’s clear that both the RTX 3080 Ti and RTX 3090 need to have DLSS enabled to play at reasonable frame rates across the most demanding ray-traced titles. Without DLSS, Watch Dogs: Legion manages an average of 29fps (at max settings), with dips below that making the game unplayable.

DLSS really is the key here across both 1440p and 4K. It was merely a promise when the 2080 Ti debuted nearly three years ago, but Nvidia has now managed to get DLSS into more than 50 popular games. Red Dead Redemption 2 and Rainbow Six Siege are getting DLSS support soon, too.

DLSS also sets Nvidia apart from AMD’s cards. While AMD’s RX 6800 XT is fairly competitive at basic rasterization at 1440p, it falls behind the RTX 3080 in the most demanding games at 4K — particularly when ray tracing is enabled. Even the $1,000 Radeon RX 6900 XT doesn’t fare much better at 4K. AMD’s answer to DLSS is coming later this month, but until it arrives we still don’t know exactly how it will compensate for ray tracing performance on AMD’s GPUs. AMD has also struggled to supply retailers with stock of its cards.

That’s left Nvidia in a position to launch the RTX 3080 Ti at a price point that really means it’s competing with itself, positioned between the RTX 3080 and RTX 3090. If the RTX 3090 wasn’t a thing, the RTX 3080 Ti would make a lot more sense.

The RTX 3080 Ti comes close to RTX 3090 performance.

Nvidia is also competing with the reality of the market right now, as demand has been outpacing supply for more than six months. Nvidia has introduced a hash rate limiter for Ethereum cryptocurrency mining on new versions of the RTX 3080, RTX 3070, and now this RTX 3080 Ti. It could help deter some scalpers, but we’ll need months of data on street prices to really understand if it’s driven pricing down to normal levels.

Demand for 30-series cards has skyrocketed as many rush to replace their aging GTX 1080 and GTX 1080 Ti cards. Coupled with Nvidia’s NVENC and professional tooling support, it’s also made the RTX 30-series a great option for creators looking to stream games, edit videos, or build games.

In a normal market, I would only recommend the RTX 3080 Ti if you’re really willing to spend an extra $500 to get some extra gains in 1440p and 4K performance. But it’s a big price premium when the RTX 3090 exists at this niche end of the market and offers more performance and double the VRAM if you’re really willing to pay this much for a graphics card.

At $999 or even $1,099, the RTX 3080 Ti would tempt me a lot more, but $1,199 feels a little too pricey. For most people, an RTX 3080 makes a lot more sense if it were actually available at its standard retail price. Nvidia also has a $599 RTX 3070 Ti on the way next week, which could offer some performance gains to rival the RTX 3080.

Either way, the best GPU is the one you can buy right now, and let’s hope that Nvidia and AMD manage to make that a reality soon.

Photography by Tom Warren / The Verge

Memory Kit	Part Number	Capacity	Data Rate	Primary Timings	Voltage	Warranty
Crucial Ballistix Max	BLM2K8G51C19U4B	2 x 8GB	DDR4-5100 (XMP)	19-26-26-48 (2T)	1.50	Lifetime
Adata XPG Spectrix D50 Xtreme	AX4U500038G19M-DGM50X	2 x 8GB	DDR4-5000 (XMP)	19-28-28-46 (2T)	1.60	Lifetime
Thermaltake ToughRAM RGB	R009D408GX2-4600C19A	2 x 8GB	DDR4-4600 (XMP)	19-26-26-45 (2T)	1.50	Lifetime
Predator Apollo RGB	BL.9BWWR.255	2 x 8GB	DDR4-4500 (XMP)	19-19-19-39 (2T)	1.45	Lifetime
Patriot Viper 4 Blackout	PVB416G440C8K	2 x 8GB	DDR4-4400 (XMP)	18-26-26-46 (2T)	1.45	Lifetime
TeamGroup T-Force Dark Z FPS	TDZFD416G4000HC16CDC01	2 x 8GB	DDR4-4000 (XMP)	16-18-18-38 (2T)	1.45	Lifetime
TeamGroup T-Force Xtreem ARGB	TF10D416G3600HC14CDC01	2 x 8GB	DDR4-3600 (XMP)	14-15-15-35 (2T)	1.45	Lifetime

Memory Kit	DDR4-4400 (1.45V)	DDR4-4500 (1.50V)	DDR4-4600 (1.55V)	DDR4-4666 (1.56V)	DDR4-4866 (1.60V)	DDR4-5100 (1.60V)
Crucial Ballistix Max DDR4-5100 C19	N/A	N/A	N/A	N/A	N/A	17-25-25-48 (2T)
Adata XPG Spectrix D50 Xtreme DDR4-5000 CL19	N/A	N/A	N/A	N/A	19-27-27-46 (2T)	N/A
Thermaltake ToughRAM RGB DDR4-4600 C19	N/A	N/A	18-24-24-44 (2T)	20-26-26-45 (2T)	N/A	N/A
Patriot Viper 4 Blackout DDR4-4400 C18	17-25-25-45 (2T)	21-26-26-46 (2T)	N/A	N/A	N/A	N/A

	Price	Cores	ROPs	Core Clock	Boost Clock	Memory Clock	GPU	Transistors	Memory
RTX 3060 Ti	$1300	4864	80	1410 MHz	1665 MHz	1750 MHz	GA104	17400M	8 GB, GDDR6, 256-bit
RX 6700 XT	$1000	2560	64	2424 MHz	2581 MHz	2000 MHz	Navi 22	17200M	12 GB, GDDR6, 192-bit
RTX 2080 Ti	$1400	4352	88	1350 MHz	1545 MHz	1750 MHz	TU102	18600M	11 GB, GDDR6, 352-bit
RTX 3070	$1300	5888	96	1500 MHz	1725 MHz	1750 MHz	GA104	17400M	8 GB, GDDR6, 256-bit
RX 6800	$1400	3840	96	1815 MHz	2105 MHz	2000 MHz	Navi 21	26800M	16 GB, GDDR6, 256-bit
RX 6800 XT	$1700	4608	128	2015 MHz	2250 MHz	2000 MHz	Navi 21	26800M	16 GB, GDDR6, 256-bit
RTX 3080	$1500	8704	96	1440 MHz	1710 MHz	1188 MHz	GA102	28000M	10 GB, GDDR6X, 320-bit
RTX 3080 Ti	$2000 MSRP: $1200	10240	112	1365 MHz	1665 MHz	1188 MHz	GA102	28000M	12 GB, GDDR6X, 384-bit
RX 6900 XT	$2100	5120	128	2015 MHz	2250 MHz	2000 MHz	Navi 21	26800M	16 GB, GDDR6, 256-bit
RTX 3090	$2900	10496	112	1395 MHz	1695 MHz	1219 MHz	GA102	28000M	24 GB, GDDR6X, 384-bit

	Price	Cores	ROPs	Core Clock	Boost Clock	Memory Clock	GPU	Transistors	Memory
RTX 3060 Ti	$1300	4864	80	1410 MHz	1665 MHz	1750 MHz	GA104	17400M	8 GB, GDDR6, 256-bit
RX 6700 XT	$1000	2560	64	2424 MHz	2581 MHz	2000 MHz	Navi 22	17200M	12 GB, GDDR6, 192-bit
RTX 2080 Ti	$1400	4352	88	1350 MHz	1545 MHz	1750 MHz	TU102	18600M	11 GB, GDDR6, 352-bit
RTX 3070	$1300	5888	96	1500 MHz	1725 MHz	1750 MHz	GA104	17400M	8 GB, GDDR6, 256-bit
RX 6800	$1400	3840	96	1815 MHz	2105 MHz	2000 MHz	Navi 21	26800M	16 GB, GDDR6, 256-bit
RX 6800 XT	$1700	4608	128	2015 MHz	2250 MHz	2000 MHz	Navi 21	26800M	16 GB, GDDR6, 256-bit
RTX 3080	$1500	8704	96	1440 MHz	1710 MHz	1188 MHz	GA102	28000M	10 GB, GDDR6X, 320-bit
RTX 3080 Ti	$2000 MSRP: $1200	10240	112	1365 MHz	1665 MHz	1188 MHz	GA102	28000M	12 GB, GDDR6X, 384-bit
MSI RTX 3080 Ti Suprim X	$2100 MSRP: N/A	10240	112	1365 MHz	1830 MHz	1188 MHz	GA102	28000M	12 GB, GDDR6X, 384-bit
RX 6900 XT	$2100	5120	128	2015 MHz	2250 MHz	2000 MHz	Navi 21	26800M	16 GB, GDDR6, 256-bit
RTX 3090	$2900	10496	112	1395 MHz	1695 MHz	1219 MHz	GA102	28000M	24 GB, GDDR6X, 384-bit

	Price	Cores	ROPs	Core Clock	Boost Clock	Memory Clock	GPU	Transistors	Memory
RTX 3060 Ti	$1300	4864	80	1410 MHz	1665 MHz	1750 MHz	GA104	17400M	8 GB, GDDR6, 256-bit
RX 6700 XT	$1000	2560	64	2424 MHz	2581 MHz	2000 MHz	Navi 22	17200M	12 GB, GDDR6, 192-bit
RTX 2080 Ti	$1400	4352	88	1350 MHz	1545 MHz	1750 MHz	TU102	18600M	11 GB, GDDR6, 352-bit
RTX 3070	$1300	5888	96	1500 MHz	1725 MHz	1750 MHz	GA104	17400M	8 GB, GDDR6, 256-bit
RX 6800	$1400	3840	96	1815 MHz	2105 MHz	2000 MHz	Navi 21	26800M	16 GB, GDDR6, 256-bit
RX 6800 XT	$1700	4608	128	2015 MHz	2250 MHz	2000 MHz	Navi 21	26800M	16 GB, GDDR6, 256-bit
RTX 3080	$1500	8704	96	1440 MHz	1710 MHz	1188 MHz	GA102	28000M	10 GB, GDDR6X, 320-bit
RTX 3080 Ti	$2000 MSRP: $1200	10240	112	1365 MHz	1665 MHz	1188 MHz	GA102	28000M	12 GB, GDDR6X, 384-bit
ASUS RTX 3080 Ti STRIX LC	$2200 MSRP: $2100	10240	112	1365 MHz	1830 MHz	1188 MHz	GA102	28000M	12 GB, GDDR6X, 384-bit
RX 6900 XT	$2100	5120	128	2015 MHz	2250 MHz	2000 MHz	Navi 21	26800M	16 GB, GDDR6, 256-bit
RTX 3090	$2900	10496	112	1395 MHz	1695 MHz	1219 MHz	GA102	28000M	24 GB, GDDR6X, 384-bit

iOS 15 may bring improvements to notifications and iMessage

We don’t know much about macOS, watchOS 8, and tvOS 15 — but we could see a new “homeOS”

New, redesigned MacBook Pros and a new Apple CPU could be announced

We could see at least one other new Mac

Apple might have dropped a hint about its AR / VR headset

Loads of Videos Require New Hardware

Increasing Efficiency Leads to Migration from Xeon

Offline Two-Pass Single Output (SOT) Throughput in CPU, GPU, and VCU-Equipped Systems

Intel Isn’t Standing Still

Summary

Geekbench

Our Verdict

For

Against

Features and Specifications

Specifications – NZXT N7 Z590

Hardware

1440p testing

RTX 3080 Ti (1440p)

4K testing

RTX 3080 Ti (4K)

Our Verdict

For

Against

Comparison Hardware

AMD Performance

Overclocking and Latency Tuning

Lowest Stable Timings

Bottom Line

Introduction

Introduction

Introduction

Introduction