[Intel’s Arc GPUs](https://www.techradar.com/news/intel-arc) just got a lot faster with a new driver update – and apparently the graphics cards may have received a small performance bump for clock speeds (at least in one case – more on that later).

First off, let’s look at the driver improvements Intel has ushered in, which as [Tweak Town](https://www.tweaktown.com/news/92907/intel-arc-driver-update-brings-massive-20-performance-improvement-to-directx-11-gaming/index.html) spotted were built around boosting DX11 performance.

Driver version 31.0.101.4644 means an average 19% boost to a clutch of DX11 games with the Arc A750 (at 1080p resolution), and that includes a 33% gain for [Overwatch 2](https://www.techradar.com/reviews/overwatch-2), and a 27% boost for GTA V.

Dota 2 is 20% faster, Valorant is up 18%, Apex Legends gets a 17% increase in frame rates, and there are some slightly more modest boosts like 5% for Destiny 2.

As [Neowin](https://www.neowin.net/news/alongside-dx-11-performance-boost-intel-firmware-update-quietly-bumps-up-arc-specs/) further points out, Intel has done some work with the interface of the driver installer, so it now shows when the graphics card firmware is being updated (and warns you not to turn off or reboot your PC, or otherwise interrupt the upgrade, as that can be seriously bad news if it happens).

Another interesting nugget here is that a Neowin forum member called ‘Eternal Tempest’ observed that their Arc A380 got an actual speed boost in terms of the default clock being upped by 150MHz.

* * *

Analysis: Go heavy with the seasoning
-------------------------------------

There's no denying that 150MHz is a tidy boost for a graphics card, but the catch is this is just a single report. We did some searching but couldn’t find any other reports of this driver giving an Arc GPU a boost to clocks, so we’ll have to take this with a couple of huge fistfuls of salt for the time being.

Even if it did happen for this one A380 owner, maybe it was for a different reason, or there’s an anomaly with the driver somehow. (There are reports of the new graphics driver also adjusting fan speeds – and bugs occurring with that – on [Reddit](https://www.reddit.com/r/IntelArc/comments/15s9el4/new_whql_intel_arc_driver_3101014644/), so that could maybe tie in here somehow).

Whatever the case with the purported speed boost, the driver does provide a really impressive uplift for some popular games as outlined above. And what’s even better is that Intel has an established and ongoing track record of continually delivering improvements – [sometimes huge ones](https://www.techradar.com/news/intel-arc-graphics-cards-just-got-another-major-performance-upgrade) – for a variety of top titles.

While it should be noted that many of the massive frame rate gains are for older games, as we see here, even contemporary games are having their fps cranked up considerably.

Another theme for Arc GPUs is these graphics cards getting more affordable, with some serious discounts making them great candidates for the [best cheap graphics card](https://www.techradar.com/news/best-cheap-graphics-cards-2020-the-top-graphics-cards-on-a-budget) list. As [VideoCardz](https://videocardz.com/newz/asrock-arc-a770-gpu-with-16gb-vram-drops-to-299-8gb-model-at-249) spotted, this is happening again, with ASRock having knocked down the price of its A770 and A750 models.

You can now get the ASRock A750 8GB Challenger from [Newegg at $200](https://www.newegg.com/asrock-arc-a750-a750-cld-8go/p/14-930-078), and the A770 8GB Phantom Gaming has been [reduced to $250](https://www.newegg.com/asrock-arc-a770-a770-pgd-8go/p/N82E16814930077). For those looking for the 16GB variant of the latter card, [that’s now $300](https://www.newegg.com/asrock-arc-a770-a770-pg-16go/p/N82E16814930102?Item=N82E16814930102), which is about as cheap as you’ll get for a board sporting that quantity of VRAM.

New Intel Arc GPU driver gives a big free frame rate boost with some popular games

Intel’s Raptor Lake Refresh desktop processors have been spotted in another leak, and this time round we have some more benchmarks to feast our eyes on.

In this case, the benchmarks are drawn from CrossMark (a suite produced by BAPCo to provide a gauge of app performance and responsiveness), and we have results of both Intel’s incoming Core i9-14900K and the Core i7-14700K.

As spotted by [Tom’s Hardware](https://www.tomshardware.com/news/core-i9-14900k-benchmarked), the Core i9-14900K recorded an overall score of 2,265 (add seasoning, as ever), with the Core i7-14700K hitting 1,980. Thus we can conclude the 14900K is around 14% faster than the 14700K, at least going by this one result.

The 14900K is slightly faster still in some of the tests that make up the overall benchmark – notably, it’s 20% quicker in the responsiveness test.

What this leak also shows is the purported core counts of these Raptor Lake Refresh CPUs, and it backs up what has previously been aired on the rumor mill – namely that the 14700K is going to have more cores, 20 of them. That’ll break down to 8 performance cores and 12 efficiency cores, with the latter being 4 more than the 13700K.

The 14900K will keep the same core configuration as its [Raptor Lake](https://www.techradar.com/news/intel-raptor-lake-rumors-and-what-we-want-to-see) predecessor, which occupies the number three spot in our list of the [best processors](https://www.techradar.com/news/best-processors) (the 13700K is number one, incidentally, to give you some idea of how important these sequels are).

Unfortunately, the spec details provided by this leak don’t cover clock speeds, but Raptor Lake Refresh is only expected to juice things up to a relatively minor extent compared to Intel’s current-gen CPUs. One theory we’ve heard is that 14th-gen processors will be 200MHz faster across most of the range.

* * *

Analysis: Why we shouldn’t focus on the scores here
---------------------------------------------------

Why aren’t we comparing these 14th-gen processors to the results of their 13th-gen counterparts here? Well, because the latter are actually faster in the CrossMark results database.

While that might seem worrying, or indeed even ridiculous, remember that these Raptor Lake Refresh processors are still engineering samples – or that’s what we presume, anyway – and clearly not running at their full performance levels. After all, there’s no way Intel would release a new generation of silicon that is slower than the previous range of CPUs, for obvious reasons.

For the same reason, there’s no point in drawing comparisons between these leaked results and AMD Ryzen CPUs.

Don’t worry about the raw performance side of the equation here. The value of this leak is in showing that the Core i7-14700K does indeed up the efficiency core count – or at least, this is yet another indication that this is the case – and that it’s not too far off the performance of the 14900K. Actually, we would expect the gap to be a bit closer – based on current-gen Raptor Lake results in CrossMark – but there could be something else going on behind the scenes here.

As ever with spilled benchmarks, we must be careful about drawing too much in the way of conclusions.

The other interesting thing about this leak is its very existence, and the fact that we’re now seeing multiple Raptor Lake Refresh benchmarks popping up on the grapevine (there was a Cinebench leak for the 14700K last month). As the release of any hardware nears, more leaks appear, and so this is a hint that Intel’s launch date for 14th-gen CPUs is nearing. October – the [rumored month for these chips to hit the shelves](https://www.techradar.com/computing/cpu/intels-next-gen-cpus-could-be-coming-in-october-but-i-wont-be-buying-them) – looks a good bet at this point.

After all, we just heard that [Intel’s Innovation event in September will see Team Blue reveal Meteor Lake processors](https://www.techradar.com/computing/cpu/intel-to-reveal-meteor-lake-cpus-in-september-and-to-talk-about-a-bold-vision-for-ai), which are the laptop chips to run alongside Raptor Lake Refresh on the desktop – so surely the latter will get a mention, too? We shall see.

Leaked benchmarks for Intel’s next-gen Raptor Lake Refresh CPUs look shaky – but don’t panic

Building a DIY custom PC can be arduous, even for seasoned PC builders like myself. But iCUE Link, a new product line from component manufacturer Corsair, could make the whole process a lot more streamlined.

I remember spending hours on my very first PC build years ago, struggling to fit everything together inside my case - and the worst part of it all, something that still plagues me even now, was cable management. Good cable management is something of a holy grail among PC builders, with all too many of us settling for an opaque side panel that conceals our cable clutter.

iCUE Link, which Corsair describes as a ‘smart component ecosystem’, uses patent-pending single-cable connections to link (see what they did there?) all your components together. For example, an LED fan won’t need both a power cable and an RGB control cable routed through the case; it’ll just have a single cable.

Cleaning up your case
---------------------

![Multiple PC components connected together via the Corsair iCUE Link system.](https://cdn.mos.cms.futurecdn.net/RukqJyFRyX2NjroPNL2cTR.png)

Daisy-chaining components is great, but even better is how the iCUE Link fans can connect to each other simply by slotting together on a radiator or fan mount. (Image credit: Corsair)

The real draw here isn’t just that each component only needs a single cable, though, it’s that you can ‘daisy-chain’ any compatible parts together, with only one ‘end’ cable connecting to the Corsair control hub.

Fans mounted side-by-side on the same radiator can be linked together directly without any cables at all. The image above illustrates how this will work. Previously, every single fan would need a separate connection to the hub in the bottom left.

The picture below shows what your case could look like with iCUE Link at work - sure, it probably won’t be _quite_ as neat and tidy as this, but even my own current [gaming PC](https://www.techradar.com/news/the-best-gaming-pc) (which is somewhere in the triple digits of ‘computers I’ve built’) doesn’t look this good behind the side panel.

![The backplate of a PC showing neat cable management with the Corsair iCUE Link hub.](https://cdn.mos.cms.futurecdn.net/zQ7W8yDaw7P2XcwuWpemFR.png)

Look at how clean that cable management is. This is bordering on pornographic for PC-builders like me. (Image credit: Corsair)

Building your own PC is still a serious undertaking - and if you prefer the easy life, all the [best computers](https://www.techradar.com/news/computing/pc/10-of-the-best-desktop-pcs-of-2015-1304391) are ‘no assembly required’ - but Corsair claims that using Link products can speed up the overall process by as much as 50%. I’m not entirely sure where that figure comes from, but I can attest that cable management can be a time-consuming aspect of putting together a high-performance PC. The only downside, of course, is that you’ll need to stick with compatible Corsair components to get the best experience - and while Corsair’s products are typically very high quality, there are cheaper options out there.

Corsair offers the full details in a [blog post](https://www.corsair.com/newsroom/press-release/corsair-revolutionizes-diy-pc-building-with-the-new-icue-link-smart-component-ecosystem), confirming that the first wave of Link-compatible products will be landing very soon in June, with more products coming later in 2023. Personally, I think this is a great step forward in simplifying the PC-building process; I just wish Corsair made more stuff, so this wouldn’t be limited to just cooling components. Corsair [graphics cards](https://www.techradar.com/news/computing-components/graphics-cards/best-graphics-cards-1291458), please?

Corsair introduces iCUE Link to make building your next PC easier

Oops – it looks like Gigabyte might have just spoiled a big surprise from Nvidia, with the third party GPU manufacturer accidentally leaking two new graphics cards: the Gigabyte RTX 4070 AERO OC 12GB and Gigabyte RTX 4060 GAMING OC 8GB.

The timing isn’t great (not that leaks are ever well timed), as Nvidia’s GTC (GPU Technology Conference) is currently taking place. While Nvidia would rather we concentrated on news coming out of the conference regarding its AI advancements and enterprise-grade technology, the latest hints that we could finally be getting more affordable consumer cards could overshadow much of what Team Green shows.

As [Videocardz reports](https://videocardz.com/newz/gigabyte-confirms-geforce-rtx-4070-with-12gb-memory-and-rtx-4060-featuring-8gb), references to “GV-N4070AERO OC-12G” and “GV-N4060GAMING OC-8GD” products have been spotted in the latest version of the Gigabyte Control Center software, which is used for things like monitoring and overclocking GPUs and configuring RGB lighting.

Unlike some manufacturers, Gigabyte’s internal product naming conventions means its relatively easy to decipher what GPUs they could be referring to: an [RTX 4070](https://www.techradar.com/news/nvidia-rtx-4070) with 12GB of memory and an [RTX 4060](https://www.techradar.com/news/nvidia-rtx-4060) with 8GB.

> Bad news for Nvidia, good news for gamers

![Happy PC gamer shows heart sign with hands](https://cdn.mos.cms.futurecdn.net/dwEqCuZGMkgkFwruzCeHtn.jpg)

(Image credit: Parilov / Shutterstock)

This leak looks pretty conclusive, and it’s a slip up that could annoy Nvidia, which has yet to officially announce the RTX 4070 or RTX 4060. While I wasn’t expecting Nvidia to announce either GPU at GTC, due to the conference’s focus on AI and enterprise-level computing, leaks like this threaten to undermine any launch plans Nvidia might be cooking up.

Videocardz claims it has information that the RTX 4070 will be announced on April 12, but doesn’t have similar info on the RTX 4060.

While this leak may be bad news for Nvidia, as it may have wanted to keep these GPUs as a big surprise, it’s pretty good news for gamers.

So far, Nvidia has only released expensive, high-end, desktop GPUs in its latest 40 series ‘[Ada Lovelace](https://www.techradar.com/news/nvidia-lovelace-rtx-40-series-everything-we-know-so-far)’ lineup: the [RTX 4090](https://www.techradar.com/reviews/nvidia-geforce-rtx-4090), [RTX 4080](https://www.techradar.com/reviews/nvidia-geforce-rtx-4080), and [RTX 4070 Ti](https://www.techradar.com/reviews/nvidia-geforce-rtx-4070-ti-review-the-next-gen-nvidia-card-for-the-rest-of-us). While these are some of the [best graphics cards](https://www.techradar.com/news/computing-components/graphics-cards/best-graphics-cards-1291458) ever made, not everyone can afford (or needs) such powerful GPUs.

For many PC gamers on a budget, it began to feel like Nvidia had forgotten about them. Neglecting a big part of the market is never a good idea, no matter how solid your position as market leader is, and rivals AMD and Intel were more than happy to swoop in with [affordable graphics cards](https://www.techradar.com/news/best-cheap-graphics-cards-2020-the-top-graphics-cards-on-a-budget).

While Nvidia has released more mainstream and affordable [RTX 40 series mobile GPUs for laptops](https://www.techradar.com/news/live/nvidia-ces-2023-live-will-we-see-the-rumored-rtx-4070-ti), that wasn’t much use for desktop gamers looking to upgrade their rigs.

So, it’s great to see that Nvidia may be planning to finally release affordable GPUs again – although, without an official announcement, we currently don’t know what prices they’ll launch up. I’d have once said that Nvidia isn’t foolish enough to launch these GPUs at much higher prices than their RTX 30 series counterparts, but Team Green _has_ been making some [weird decisions recently](https://www.techradar.com/news/nvidias-embarrassing-rtx-4080-unlaunch-could-cost-it-dearly), so I won’t get too excited until I hear official prices.

Oops! Leaked RTX 4070 and 4060 GPUs spoil Nvidia’s party

US storage giant Western Digital already covers [SSD](https://www.techradar.com/news/best-solid-state-drives-ssds), [memory cards](https://www.techradar.com/news/best-sd-and-microsd-memory-cards), [RAM](https://www.techradar.com/news/best-ram) and [hard disk drives](https://www.techradar.com/news/best-external-desktop-and-portable-hard-disk-drives) and has dipped its toes in [DNA data storage](https://www.techradar.com/news/dna-storage-could-be-the-savior-weve-been-waiting-for) (it is a founding member of the DNA Data Storage Alliance). 

Now, newly-unearthed patents appear to show that the company may want to (re) add tape to complete its current media portfolio after the sun-setting of its Arkeia product range a few years ago. 

The company was recently assigned a number of patents mentioning “_tape embedded drive”_ in recent years:

*   11393498 ([PDF](https://image-ppubs.uspto.gov/dirsearch-public/print/downloadPdf/11393498)) (head assembly with suspension system for a tape embedded drive) 
*   20200258544 ([PDF](https://image-ppubs.uspto.gov/dirsearch-public/print/downloadPdf/20200258544)) (Tape embedded drive) 
*   11081132 ([PDF](https://image-ppubs.uspto.gov/dirsearch-public/print/downloadPdf/11081132)) Tape embedded drive with HDD components
*   And a few more

What this alludes to is the intriguing possibility of getting the basic components of a tape drive merged with the actual tape media in a bid to reduce the inherent environmental and technological complexity of tape libraries as well as improving the access time by at least one order of magnitude.

Having the read and write heads closer to the media in an enclosed form factor is nothing new. That’s what hard disk drives do and what others, most notably Iomega with its Zip Drive, has tried to do in the past. Western Digital’s patent suggests adopting a standardized form factor, 2.5-inch or 3.5-inch, for easier adoption by data centers and hyperscalers.

The cost factor
---------------

An embedded tape would still be more expensive than a normal one (LTO-9 tapes retail for about $130 a pop) because of the extra electronics but you don’t need a tape drive to get started. As long as it sits somewhere between tapes ($4 per TB) and enterprise hard drives ($20 per TB), there will be a significant market for it.

A standard LTO (Linear Tape-Open) tape is 102 x 105 x 21mm while your average 3.5-inch hard drive is about 147 x 101 x 26mm while weighing about a quarter of the weight. LTO-9 has a compressed capacity of 45TB (18TB uncompressed) with the next generation - likely to come in the second half of this decade - doubling in capacity (obviously, there might be some adjustments as it was the case going from Gen 8 to Gen 9).

A sealed LTO-based tape drive would likely be lighter, cheaper, consume/dissipate less power but also have more onboard compute capabilities than your standard hard disk. Thicker and wider tape reels would also allow far bigger capacities (LTO-9 uses 1Km of tape with a 5.2µm thickness and 12.65mm width). 

Western Digital is uniquely positioned to convert this into a workable reality especially as it can use its expertise in hard drive components. That new tape could, all things considered, use a similar PCB and interface as an enterprise hard disk drive; it doesn’t need to have the traditional tape look.

Perhaps the biggest challenge is to take this novel approach to tape to the LTO consortium, an organization that oversees the development of LTO and comprises IBM, HPE and Quantum, all of whom may have different commercial strategies that require having an expensive drive and cheap tapes.

I can’t believe this huge storage company wants to mix tape and hard drive technology

Nvidia’s Lovelace laptop GPUs are inbound and we already knew portables carrying these RTX 4000 [graphics cards](https://www.techradar.com/news/computing-components/graphics-cards/best-graphics-cards-1291458) would debut in February, but now we’ve heard exactly when the various cards should be emerging next month.

This is according to Chinese tech site [IT Home](https://www.ithome.com/0/667/855.htm) (flagged up by [Wccftech](https://wccftech.com/nvidia-geforce-rtx-4090-rtx-4080-laptop-pre-orders-commence-on-february-1st/), via [VideoCardz](https://videocardz.com/newz/nvidia-geforce-rtx-4090-4080-laptop-preorders-start-on-february-1st)), and as ever, we must be skeptical around rumors, but the contention is that Nvidia will first launch RTX 4090 and 4080 models on February 8.

That’s when these laptops will start shipping, but you’ll actually be able to order the machines on February 1, a week in advance, the report asserts.

[Laptops](https://www.techradar.com/news/mobile-computing/laptops/best-laptops-1304361) which run with the RTX 4070, 4060 and 4050 GPUs will be available later in the month, going on sale February 22.

Pricing for the lower-tier RTX 40 laptops released later in the month starts from $999 (around £820, AU$1,430), with price tags for the RTX 4080 and 4090 portables starting from $1,999 (around £1,640, AU$2,620).

We’ve already seen that the notebooks carrying the Lovelace flagship will be seriously pricey, with the cost of an [upgrade to the RTX 4090 being shown by XMG](https://www.techradar.com/news/nvidia-rtx-4090-laptop-gpu-price-revealed-and-its-seriously-expensive). It adds an eye-watering €1,687 (around $1,830, £1,500, AU$2,630) to the price tag of XMG’s Neo 16 gaming laptop (and that’s just the _additional_ cost for upgrading the GPU, not the total price).

* * *

Analysis: A swift launch for the Lovelace laptop range
------------------------------------------------------

If this proves correct, in just a touch over two weeks’ time, you’ll be able to pre-order a laptop with an RTX 4090 or 4080. And before the end of February, all Lovelace mobile graphics cards will be out there in notebooks on the market – so Nvidia isn’t hanging around deploying the entire range for laptops.

This feels especially pertinent when you consider that in the desktop sphere, only the RTX 4090, 4080, and 4070 Ti have been released so far – and those releases were staggered over quite some time (the [4070 Ti has only just got here](https://www.techradar.com/news/live/where-to-buy-the-nvidia-rtx-4070-ti)).

Whether this push with deploying laptop GPUs swiftly means we might see the RTX 4070 or [RTX 4060](https://www.techradar.com/news/nvidia-rtx-4060) (or even 4050) for desktop PCs anytime in the near future, well, we wouldn’t bank on it – certainly not in the latter two cases anyway. All recent rumors have suggested the desktop [RTX 4060 is still quite some way off](https://www.techradar.com/news/nvidia-rtx-4060-gpu-could-arrive-disappointingly-late-in-2023), and its release may not even be considered by Nvidia until enough lower-tier RTX 3000 stock – which is still hanging around – has been sold off.

As we mentioned above, the cost of the laptop RTX 4090 is not for the faint-hearted. Indeed, it’s actually every bit as pricey as the [desktop RTX 4090](https://www.techradar.com/reviews/nvidia-geforce-rtx-4090), while delivering performance in line with the desktop 4080. (Remember that the mobile 4090 is built on the AD103 chip and has a considerably chopped CUDA Core count of 9,728 compared to AD102 on the desktop 4090, which offers 16,384).

Leak Reveals When You Can Preorder A Gaming Laptop With A GeForce RTX 4090 Or 4080

Laptops with Nvidia RTX 4090 through to 4050 GPUs are about to go on sale

Samsung quietly revealed that it’s working on the next generation of [SSDs](https://www.techradar.com/news/cheap-ssd-deals) (solid state drives) with the upcoming PM9C1a. The company intends for the drive to “provide elevated computing and gaming performance in PCs and laptops” for everyday scenarios.

With that design philosophy in mind, the [PM9C1a](https://news.samsung.com/global/samsung-electronics-unveils-high-performance-pc-ssd-that-raises-everyday-computing-and-gaming-to-a-new-level) will sport the tried and true PCIe 4.0 (Peripheral Component Interconnect Express) expansion bus for a more consistent performance (PCIe 5.0 drives are still in the works). Boosting the older standard, the upcoming drive will be “integrated with a new controller based on Samsung’s… 5-nanometer process \[and\] seventh-generation V-NAND technology”. Translated into plain English, that means the PM9C1a will have better performance and energy efficiency compared to previous offerings.

Speed and efficiency
--------------------

Samsung claims the future drive will be able to reach a sequential read speed of 6,000 MB/S (megabytes per second) and a sequential write speed of 5,600 MB/s, making it almost twice as fast as the PM9B1 SSD. Read speed refers to how fast the drive can access stored files while write speed refers to how fast it can “write” or save a file. For the sake of comparison, the [Samsung 990 Pro](https://www.techradar.com/reviews/samsung-990-pro-review-a-pro-ssd-champ-to-fill-that-next-gen-gap), which TechRadar rated very highly, has a sequential read and write speed of 7,465.79 MB/s and 6,887.68.19 MB/s, respectively. So, yes, the PM9C1a will be slower, but it does get pretty close, which is particularly impressive for a drive that’s meant to be for general, everyday usage. It's not some top-of-the-line model.

Energy efficiency, as alluded to earlier, has been improved compared to the PM9B1. It’s now “70 percent more \[efficient\] per watt than its predecessor.” Samsung gives the example that when a laptop with this SSD goes into “standby mode,” the drive will consume “approximately 10 percent less power.” 

Security is also being boosted as the PM9C1a will support the Device Identifier Composition Engine security standard or DICE for short. This will give the drive the ability to “securely \[generate\] cryptographic keys” as a way to provide device authentication. The idea is it’ll help defend the computer against supply chain attacks and firmware that's been tampered with by bad actors.

Availability pending
--------------------

Since the PM9C1a is still in development, it’ll probably be a while until it launches. The announcement doesn’t say anything about a release date or price tag; although we did reach out to Samsung for clarification. However, we do know the SSDs will come in three different models according to storage capacity and size. The 256GB drive will have a form factor of 22mm by 30mm, the 512GB model will be 22mm by 42mm, and the largest is the 1TB option at 22mm by 80mm.

While it may be a while before we see these SSDs in computers, you can assume that they'll eventually end up in some of [the best ultrabook laptops](https://www.techradar.com/news/mobile-computing/laptops/best-ultrabook-18-top-thin-and-lights-1054355) on the market.

Samsung starts building SSDs that could make laptops faster, more power efficient

[Nvidia’s RTX 4070](https://www.techradar.com/news/nvidia-rtx-4070) might be inbound sooner than we thought, if a clue in the form of a regulatory filing that has just popped up is anything to go by.

[VideoCardz](https://videocardz.com/newz/inno3d-confirms-geforce-rtx-4070-ti-and-rtx-4070-in-eec-filing) highlighted a tweet from a regular hardware leaker, @harukaze5719, which pointed out an EEC (Eurasian Economic Commission) filing from Inno3D for its RTX 4070 Ti and RTX 4070 graphics cards.

> NVIDIA GeForce RTX 4070 TiNVIDIA GeForce RTX 4070https://t.co/dup2WQ9RPD pic.twitter.com/8d75sIVELo[December 21, 2022](https://twitter.com/harukaze5719/status/1605515816002281474)

The listed name, Ask Technology Group Limited, is actually the firm behind Inno3D, in case you were wondering, and filings like this with the EEC often appear quite soon before a product is launched.

Now, we’re pretty sure that the [RTX 4070 Ti](https://www.techradar.com/news/nvidia-geforce-rtx-4070-ti-everything-we-know-about-the-rumored-graphics-card) is about to be revealed at CES in just a couple of weeks’ time, [and should go on sale just after](https://www.techradar.com/news/nvidia-rtx-4070-ti-could-arrive-soon-but-team-green-needs-a-christmas-miracle) (January 5 in theory). But with the RTX 4070 being filed alongside the GPU at the EEC, does that suggest it’s ready to be unveiled alongside the 4070 Ti, perhaps? That’s certainly the question some folks are asking.

* * *

Analysis: A hopeful sign for the RTX 4070
-----------------------------------------

As noted, this clearly underlines that the RTX 4070 Ti is imminent – there’s a heap of evidence and spillage for that graphics card now, [including leaked photos](https://www.techradar.com/news/nvidia-rtx-4070-ti-gpu-leaked-photos-give-us-hope-for-our-energy-bills) most recently – but what about the possibility of the RTX 4070 coming with it?

The simple truth is that there hasn’t been nearly as much leakage about the RTX 4070, which would indicate that it’s not all that likely to be about to appear at CES 2023. If it was, we’d surely have heard more on the grapevine about the GPU, wouldn’t we?

Well, most likely, but on the other hand, we have been hearing _some_ rumors about the 4070. (By the way, they claim the card may be considerably cut-down compared to the 4070 Ti, at least in terms of its core count – although that may not be a fair reflection of its [true performance levels as we discuss in-depth here](https://www.techradar.com/news/nvidia-rtx-4070-leak-is-an-eye-opener-but-lets-not-dismiss-this-gpu-yet)).

So, it’s not beyond imagining that the RTX 4070 could launch alongside its faster sibling, or at least follow quite quickly after the 4070 Ti. With previous rumors pointing to a March debut, maybe this regulatory filing is a solid hint that we could see the RTX 4070 turn up in February, perhaps; or even late January?

In short, the RTX 4070 could arrive sooner than previously anticipated, but don’t count on it being at CES 2023, by any means. The RTX 4070 Ti is very much expected at CES, but even strong and plentiful rumors are no guarantee as Nvidia might change its mind at the last minute.

Nvidia RTX 4070 GPU could arrive soon – but don’t get your hopes too high

Nvidia has released a fix for problems with its new [RTX 4090](https://www.techradar.com/reviews/nvidia-geforce-rtx-4090) and RTX 4080 [graphics cards](https://www.techradar.com/news/computing-components/graphics-cards/best-graphics-cards-1291458) experiencing a black (or blank) screen on bootup in some unfortunate cases.

This isn’t quite as bad as it first sounds, as the operating system will still load okay and produce the desktop on your display as normal, eventually – you just get the blank screen on boot. However, that means you can’t access the BIOS for example, or see anything at all on the screen, until you reach the OS environment. And not being able to get in the BIOS could be a big problem in itself.

As [Tom’s Hardware](https://www.tomshardware.com/news/uefi-firmware-update-rtx-4090-4080-fixes-blank-screens) reports, though, Nvidia has released a GPU firmware update for the RTX 4090 and [RTX 4080](https://www.techradar.com/reviews/nvidia-geforce-rtx-4080) to cure these blank screen blues.

Note that the problem will only happen with certain [motherboards](https://www.techradar.com/news/computing-components/motherboards/best-motherboard-14-reviewed-and-rated-904229) which are in UEFI mode.

Nvidia says: “The Nvidia GPU Firmware Update Tool will detect whether the firmware update is needed, and if needed, will give the user the option to update it.”

For those experiencing a blank screen on boot with their new Lovelace GPU, you can grab the [relevant update here](https://nvidia.custhelp.com/app/answers/detail/a_id/5411/?=&linkId=100000163736896), where Nvidia also provides instructions on how to get going with the update.

* * *

Analysis: Another Lovelace GPU glitch for Nvidia – but the solution has arrived quickly this time
-------------------------------------------------------------------------------------------------

By all accounts, this problem is a rare one, so that’s some good news. Still, if you’re affected, it’ll obviously be frustrating not to be able to reach the BIOS (though at least your PC will still boot fine to the desktop).

As Nvidia notes, don’t apply the update unless you’re actually experiencing the blank screen – there’s no need to risk a firmware update if you don’t have a tangible problem to resolve, of course.

Note that this gremlin in the works has nothing to do with the other well-publicized issue that has hit the RTX 4090 graphics card, namely the melting cable controversy. (We’ve just heard an update from Nvidia on that problem, by the way, and it seems the culprit, as thought, is the [power adapter not being plugged in correctly](https://www.techradar.com/news/nvidia-update-on-rtx-4090-gpu-melting-adapter-reveals-likely-cause)).

Nvidia RTX 4080 and 4090 GPUs get a fix for nasty blank screen issue on boot

The [Nvidia RTX 4080 16GB](https://www.techradar.com/news/nvidia-geforce-rtx-4080) will go on sale on November 16, and it has every chance of being one of the [best graphics cards](https://www.techradar.com/news/computing-components/graphics-cards/best-graphics-cards-1291458) money can buy. Good luck getting one, though: according to YouTuber and hardware leaker [Moore’s Law Is Dead](https://www.youtube.com/c/MooresLawIsDead), the stock situation for the new GPU looks pretty dire.

The flagship [RTX 4090](https://www.techradar.com/reviews/nvidia-geforce-rtx-4090#:~:text=Without%20a%20doubt%2C%20the%20Nvidia,it's%20unquestionably%20worth%20that%20enthusiasm.) sold out almost immediately when it hit retailers back in October, and a [new video](https://www.youtube.com/watch?v=_PRXJZHWqMI) from MLID has claimed that anonymous retailers have confirmed that RTX 4080 stock will be between 20-40% less than that of the RTX 4090 on day one.

It’s purported that there will still be a fairly large number of cards for a non-flagship GPU launch, but there’s another kicker in the video: MLID adds that retailers are expecting fewer restocks than usual in the final quarter of 2022, less than for the 4090.

The obvious reason for the lower RTX 4080 stocks is that Nvidia wants to push consumers towards buying older Ampere-series RTX 3000 GPUs. These seem to still be readily available (after the months of scalping and scamming that followed the Ampere launch), and Nvidia no doubt wants to clear out its old stock before committing fully to the new Lovelace-series cards.

Analysis: Nvidia is being crafty with its supply of GPUs
--------------------------------------------------------

This isn’t an entirely surprising turn of events, and it could actually benefit some gamers: with [Black Friday](https://www.techradar.com/black-friday/black-friday-graphics-card-deals-2022-our-expert-predictions) swiftly approaching, many will be looking to upgrade their GPU, and we could see some great deals on graphics cards in the coming weeks.

The Ampere cards (especially the excellent midrange [RTX 3060 Ti](https://www.techradar.com/reviews/nvidia-geforce-rtx-3060-ti)) are still solid choices if you’re rocking an older GTX GPU and want a bit more graphical oomph in your PC – so if you’re after a [cheap graphics card](https://www.techradar.com/news/best-cheap-graphics-cards-2020-the-top-graphics-cards-on-a-budget), Black Friday could be the time to snap one up.

Of course, another reason why the RTX 4080 is allegedly languishing with lower launch stocks is likely Nvidia’s [sudden cancelation](https://www.techradar.com/news/nvidia-caught-with-its-pants-down-will-unlaunch-the-12gb-rtx-4080) of the 4080 12GB, which it admitted was “not named right”. We were [actually relieved](https://www.techradar.com/opinion/nvidia-made-the-right-call-to-unlaunch-the-12gb-rtx-4080-bring-on-the-4070) to see Team Green have a change of heart over the less powerful 12GB model, but it's likely caused fewer 4080 cards to be available for the day one release.

Nvidia is still in hot water over claims that the high-end RTX 4090 is actually a [fire hazard](https://www.techradar.com/opinion/nvidias-melting-rtx-4090-cables-are-symptomatic-of-a-bigger-problem) due to issues with the power adapter, and the GPU giant has yet to release a statement regarding the problem. According to a [dedicated Reddit megathread](https://www.reddit.com/r/nvidia/comments/ydh1mh/16_pins_adapter_megathread/), 20 cases have been reported in the past two weeks. We very much hope the RTX 4080 doesn’t share these thermal issues.

Nvidia RTX 4080 set to sell out on day one, with even less stock than RTX 4090

It’s been only a few days since [AMD rolled out its Ryzen 7000 processors](https://www.techradar.com/reviews/amd-ryzen-9-7950x) with up to 16 Zen 4 cores and there are already rumblings of its next big CPU announcement.

What appears to be the follow-up to the world’s fastest x86 processor - the [Ryzen Threadripper Pro 5995WX](https://www.techradar.com/news/amds-new-64-core-threadripper-pro-cpu-is-now-on-sale-but-you-may-need-to-take-out-a-loan-to-afford-it) -  has appeared in a popular scientific computing project, [einstein@home](https://einstein11.aei.uni-hannover.de/host/12981157), as reported by prolific CPU chaser [Benchleaks](https://twitter.com/BenchLeaks/status/1575218825880276993) ([via Tom'sHardware](https://www.tomshardware.com/news/ryzen-threadripper-7000-storm-peak-cpu-surfaces-with-64-zen-4-cores)).

The processor is listed as an “AuthenticAMD AMD Eng Sample: 100-000000454-20\_Y \[Family 25 Model 24 Stepping 1\]” which, with 128 threads (64 cores, two threads each) and an unusual identifier, doesn’t correspond to any current AMD processor.

Following in the footsteps of its predecessor, this impending Storm Peak product is likely to be called named 7995WX and have four dies stuck together (probably underclocked 65W AMD Ryzen 9 7950X), reaching a TDP of around 260W with some overheads (I/O etc.), which will nicely match the 280W TDP of the Zen 3-powered 5995WX. The latter had a base clock of 2.7GHz with a maximum boost clock up to 4.5GHz, 292MB combined cache (L1 + L2 + L3).

Workstation galore
------------------

The new processors will find their way into [desktop workstation PCs](https://www.techradar.com/best/best-workstations), where high processor/thread counts help tremendously - more than just higher frequencies - with a wide variety of tasks: video editing, rendering, financial analysis, GIS, AI, ML etc.

[Lenovo](https://www.techradar.com/news/new-amd-ryzen-threadripper-pro-cpu), [Dell](https://www.techradar.com/news/dell-just-released-a-formidable-mac-pro-workstation-rival-but-you-cant-buy-it-yet) and boutique providers ([Maingear](https://www.techradar.com/news/the-maingear-pro-workstation-is-guaranteed-to-turn-heads-and-empty-wallets), [Velocity Micro](https://www.techradar.com/news/the-worlds-fastest-pc-will-cost-you-less-than-you-think-but-youll-have-to-be-patient), Puget Systems) are likely to launch products based around the latest Threadripper, probably in June 2023.

Intel will also be watching closely as it seeks to ramp up the number of cores, mirroring what we have seen in the consumer market. For the first time, Core i7 processors ([the 13th generation launched just a few days ago](https://www.techradar.com/news/intel-13th-gen-cpus-launch-with-big-boosts-should-amd-be-worried)) now have 16 cores, matching AMD’s Ryzen 7 family.

Will Intel up the ante and boost its Xeon core count (currently peaking at 40 cores)? We wouldn't bet against it, although that may require a big price drop to maintain some sense of competitiveness; the 40-core 8380 costs around 50% more than the 64-core 5995WX.

It will also be interesting to see whether AMD launches a new range of Threadripper processors aimed at [mobile workstations](https://www.techradar.com/news/best-mobile-workstations), a small but still very lucrative niche currently dominated by Intel.  

64-core Ryzen 7000 Threadripper monster caught in the wild

Kioxia, the company formerly known as Toshiba, has started to ship a new storage device based on a new form factor, XFMEXPRESS, which aims to bring the best of both [SSD](https://www.techradar.com/news/best-solid-state-drives-ssds) and [microSD](https://www.techradar.com/news/best-sd-and-microsd-memory-cards) worlds.

The XT2 is a removable PCIe attached NVMe storage device which, according to Kioxia’s marketing literature, will “enhance next-generation mobile and embedded applications”.

The technology - which is three years in the making - is removable like a microSD card, with dimensions (18 x 14 x 1.4mm) slightly bigger than the latter (14 x 11 x 1mm). Its connector is similar to what smartphones use for nano SIM cards, requiring the cover to be slid aside to access the card.

A spokesperson for the company told _TechRadar Pro_ that the XT2 will be priced competitively with other SSD form factors of similar performance and will be offered in capacities ranging from 128GB to 1TB. As for speeds, given that it comes with PCIe Gen 4 x 2 lanes, sequential reads and writes should exceed 2500Mbps and 650Mbps respectively.

A rival to microSD?
-------------------

Don’t expect XFMExpress to replace microSD any time soon, even if demand for the tiny cards has taken a downward trajectory as mainstream smartphones abandon removable storage altogether.

There’s also the conundrum associated with rival format SD Express, which is also a PCIe NVMe standard, which can reach up to 4GBps, offers advanced security features and is backed by the SD association.

Which brings us to the question of whether the market needs another new form factor, when the existing incumbent benefits from bigger market support, better performance, minimal retooling and enhanced security. We’ve put that question to Kioxia and will update the article when we hear back.

As for use cases that require high performance, high capacity storage, there’s plenty (e.g. prosumer cameras, IoT/edge [servers](https://www.techradar.com/news/best-small-business-servers) etc.), but they are likely to be niche and unlikely to sustain a market where rival formats exist.

SSD performance in a MicroSD format? We’re almost there, but a battle is brewing

More AMD and Nvidia [graphics cards](https://www.techradar.com/uk/news/computing-components/graphics-cards/best-graphics-cards-1291458) are now at their recommended price – or have dipped below that level – than GPUs which exceed their MSRP, according to a new report; but it’s Team Red which offers the most competitive pricing.

This comes from [Tom’s Hardware](https://www.tomshardware.com/uk/news/graphics-card-prices-dropped-15-percent-in-may-2022) which crunched a load of data on sales to discover that for current-gen RTX 3000 and RX 6000 graphics cards in May, prices dropped by 15% on average compared to what retailers were asking in April.

As noted, just over half of current-gen GPUs are now at or below recommended pricing, although it’s AMD which is leading the way here, in that of its RDNA 2 offerings, only the RX 6800 and 6800 XT are now priced above their MSRP.

Unfortunately, most Nvidia RTX 3000 graphics cards are still over the MSRP, with Tom’s observing that the RTX 3050 through to [RTX 3080](https://www.techradar.com/uk/news/nvidia-geforce-rtx-3080-ti-vs-rtx-3080) 10GB models have their price bumped up by about 20% on average. That said, those cards are still seeing price drops compared to the previous month, and everything is moving in the right direction.

* * *

Analysis: Nvidia GPUs in great demand
-------------------------------------

It’s obviously good news to see prices dropping, and by quite a chunk in May, even if Nvidia’s Ampere offerings still haven’t normalized and returned to their MSRP. Of course, this is a measure of demand in many ways, because as we’ve seen in recent times, [Team Green’s GPUs are selling strongly](https://www.techradar.com/uk/news/nvidia-is-now-selling-a-lot-more-rtx-3080-gpus) compared to AMD.

[Recall that as we reported a month ago](https://www.techradar.com/uk/news/nvidia-is-blowing-away-amd-for-gpu-sales-right-now), at big US retailer Newegg, Nvidia models were all but one of the top 20 best-selling graphics cards, and indeed looking now, the situation is even worse for AMD – the company hasn’t got a single GPU in that top 20 chart (the highest entry is an RX 6600 model at number 24, in fact).

So greater demand seems like a plausible enough reason why prices aren’t going down quite so quickly for Nvidia GPUs. What might change things is the possibility that [Nvidia could soon announce its next-gen Lovelace graphics cards](https://www.techradar.com/uk/news/nvidias-plans-to-hype-up-its-next-gen-rtx-4000-gpus-could-backfire), and those [RTX 4000 GPUs](https://www.techradar.com/uk/news/nvidia-lovelace-rtx-40-series-everything-we-know-so-far) could go on sale sooner than expected – which would doubtless persuade some (or indeed many) folks to consider playing a waiting game, forcing RTX 3000 prices to drop to seem more tempting.

Tom’s also did an interesting bit of comparison in terms of working out a rough performance per dollar metric for the top current-gen (and past) GPUs, and the interesting news is that AMD is winning here, perhaps unsurprisingly as Team Red’s prices are dropping faster as observed.

So, which is the best value proposition among the contemporary graphics cards out there right now? That would be [AMD’s RX 6600](https://www.techradar.com/uk/reviews/amd-radeon-rx-6600) which represents an excellent choice for those gaming at 1080p, with a performance per dollar rating of 0.23 (lower is better) – although if you need something with more oomph for 1440p duties, then the top value pick is the RX 6700 XT rated at 0.201. For Nvidia, the value king for 1080p is the RTX 3060 (at 0.175), and for 1440p it’s the 3060 Ti (also ranked at 0.175, which isn’t _that_ far behind the 6700 XT to be fair).

For those interested in some historical perspective, Tom’s ranked GPUs going back to 2015, and the best performer relative to its asking price turned out to be the GTX 780 (on 0.245, so the RX 6600 isn’t far off that anyway – and obviously offers way better overall performance, of course).

Majority of current-gen GPUs now selling at or below recommended price

Nvidia is totally ruling the [GPU](https://www.techradar.com/uk/news/computing-components/graphics-cards/best-graphics-cards-1291458) roost over AMD right now, at least going by Newegg’s ranking of graphics card sales, with Team Green dominating its rival to a seemingly unprecedented extent.

[Tom’s Hardware](https://www.tomshardware.com/uk/news/newegg-best-sellers-dominated-by-nvidia) pointed out that the top 10 best-selling graphics card listed by the US retailer are GeForce models at the time of writing.

In fact, the [entire top 20 models](https://www.newegg.com/d/Best-Sellers/Desktop-Graphics-Cards/s/ID-48?) are all Nvidia products, with one exception – an AMD entry at number 15, which is the MSI Mech Radeon RX 6600. To frame that another way, 95% of the top 20 graphics cards are made by Nvidia, and those are all Ampere (RTX 3000) GPUs, except for the RTX 2060 at number 17.

What’s also interesting to note is that the top dog sellers are quite expensive Nvidia GPUs, namely the [RTX 3070 Ti](https://www.techradar.com/uk/news/nvidia-rtx-3070-ti-vs-rtx-3070), a trio of models of which hold the top three positions, and they’re followed by a 3060 Ti, then an RTX 3050 – an affordable option, of course – but at number six there’s an RTX 3080 Ti.

There’s a bit of an oddity here given that the RTX 3060 Ti at number 4 is actually a tad more expensive than the RTX 3070 Ti in second place (both are Gigabyte cards), but the latter has had a [big price cut to $700](https://www.newegg.com/gigabyte-geforce-rtx-3070-ti-gv-n307tgaming-oc-8g/p/N82E16814932443), and the former is only being offered by a third-party seller now, so presumably Newegg was selling it a good chunk cheaper until very recently when it ran out of stock.

* * *

Analysis: A real head-scratcher, make no mistake
------------------------------------------------

This is certainly a surprising situation. Nvidia’s dominance in desktop GPUs is well-established if you look at any report, mind – it’s [generally at around the 80% market share level](https://www.techradar.com/uk/news/graphics-card-sales-for-2021-might-shock-you-but-theyre-good-news-for-amd), or high 70s – but Team Green’s ownership is taken to new levels at Newegg currently.

So what’s going on here? Of course, [Nvidia graphics card prices have been dropping recently](https://www.techradar.com/uk/news/gpu-prices-are-crashing-will-the-nvidia-rtx-3080-finally-be-affordable), and there are some compelling looking deals right now as a result – but AMD’s price tags have also been heading downwards, and equally there are some [tempting price reductions for Team Red](https://www.techradar.com/uk/news/gpu-prices-fall-again-with-one-amd-graphics-card-now-35-below-its-launch-price), too.

That RX 6600 – which is the only AMD GPU to make the cut for the top 20 – is not far off its MSRP now, and is just over 10% cheaper than the most affordable RTX 3060 in Newegg’s rankings at the time of writing – with both of those cards being fairly even in overall performance terms, [as we observed in our review](https://www.techradar.com/uk/reviews/amd-radeon-rx-6600). Indeed, that RX 6600 doesn’t cost much more than the RTX 3050 much higher up in the list (you’re looking at 10% more for the AMD GPU).

We’re really not sure why AMD’s RX 6000 products are being kept out of the top 20 so convincingly right now. Tom’s suggests this could be down to Nvidia mindshare, but as noted already, Team Green has always had that advantage – and there are no real reasons why anything would have changed in that regard all that much in recent times.

Perhaps there has been some serious pent-up demand for beefier Nvidia RTX 3000 models, and with pricing finally starting to normalize, those buyers are making a move. Although that still doesn’t explain the extent of Nvidia’s domination here, even further down the pricing tiers, and all-in-all, this is a bit of a head-scratcher.

Naturally, we can’t put too much stock in a single source – this is just a snapshot of the market from one retailer, after all – but one thing it might point to, perhaps, is more compelling price cuts from AMD to get back into the game a bit more. Particularly as [new RDNA 2 refreshes are rumored to be due imminently](https://www.techradar.com/uk/news/amds-rumored-new-gpus-may-not-be-as-fast-as-we-hoped-but-pricing-could-be-tempting), including the RX 6950 XT, which should knock down the pricing of the existing graphics cards that these revamped models will be a step up from (namely the RX 6900, 6700 and that 6600).

All of this should be good news for gamers looking for GPU prices to come down further still...

Nvidia is blowing away AMD for GPU sales right now

At Qualcomm’s recent earnings call, it was revealed that its powerful Nuvia chips are finally coming, but not until late 2023.

While this still feels like a long wait, it could be well worth it, as we may finally see [Windows 11](https://www.techradar.com/news/windows-11-home-and-pro) laptops powered by ARM chips that can seriously rival Apple’s latest batch of [Macs and MacBooks](https://www.techradar.com/news/computing/apple/mac-buyer-s-yRqpp9BVva9YiRhf3Lr3NQe-2015-1295725), which run on the company’s ARM-based [M1 chips](https://www.techradar.com/news/apple-m1-chip).

Unlike the processors from Intel and AMD that are often found in [laptops](https://www.techradar.com/news/mobile-computing/laptops/best-laptops-1304361) and PCs, ARM chips were more commonly found in mobile devices such as smartphones and tablets, but are increasingly found in laptops as well, bringing benefits such as long battery lives and almost instant wake-up times that you’d usually find in a smartphone.

However, while Apple’s M1-powered MacBooks have really impressed us, ARM-based Windows laptops have been less impressive, mainly due to how expensive and underpowered they are.

Qualcomm’s Nuvia chips could change that, however, with the company’s President and CEO, Christian Amon, stating that this new chip will be “going after the performance tier,” which makes us think that it’ll specifically aim to take on Apple’s M1 chip, as well as the more powerful [M1 Pro and M1 Max](https://www.techradar.com/news/m1-pro-vs-m1-max) versions.