Hardware news this week talks Intel Arc GPUs, RTX 40 card coolers as leaked by wxnod, PCI-SIG discussion on melting 12VHPWR connectors, and more.
Intel Refutes Claims of AXG Cancellation, Says ARC Launch is Soon
There's not really much to this one -- we'll leave it mostly to the video above. Basically, rumors of Intel Arc being canceled were directly refuted by contacts GN has at Intel, in addition to public tweets by Raja Koduri at Intel.
Melting 12VHPWR Cables
PCI-SIG, the group that governs PCIe standards, has issued an email to all of its members and their suppliers regarding melting 12VHPWR cables. That email was technically private, but so many groups were on the email that it inevitably got passed around and ended up in our hands too.
As a reminder, 12VHPWR is the official name for the new PCIe 5.0 12+4-pin cable used on some current and upcoming high-power graphics cards and is capable of handling 600W sustained on a single cable.
You may have already seen this story covered on some rumor sites citing that the problem is with adapters converting to 8-pin PCIe connectors, but that’s incorrect and may have been based on incomplete information. Here’s what the email had to say:
“Please be advised that PCI-SIG has become aware that some implementations of the 12VHPWR connectors and assemblies have demonstrated thermal variance, which could result in safety issues under certain conditions. Although PCI-SIG specifications provide necessary information for interoperability, they do not attempt to encompass all aspects of proper design, relying on numerous industry best-known methods and standard design practices. As the PCI-SIG workgroups include many knowledgeable experts in the field of connector and system design, they will be looking at the information available about this industry issue and assisting in any resolution to whatever extent is appropriate.”
“As more details emerge, PCI-SIG may provide further updates. In the meantime, we recommend members work closely with their connector vendors and exercise due diligence in using high-power connections, particularly where safety concerns may exist.”
The PDF explains that Nvidia has been testing 12VHPWR connectors to validate that prototype power supplies and cables can meet the specification of 55A continuous. During this testing, Nvidia found certain cable routing conditions led to excess heat and, in some cases, melting.
The conditions required for the excess heat were either subjecting the cables to severe bending or a high number of mating cycles (about 40). Cables tested in these scenarios exhibited hot spots at roughly 2 and a half hours, and melting at 10 to 30 hours. Connectors from multiple suppliers have failed.
This is with a continuous 55A of current (or 660W at 12V), which would not be a typical load condition, especially not in gaming. Nvidia did not observe any failures on connectors with low mating cycles and without any bend.
Photos provided in the PDF show some pretty gruesome melting, and it’s not in the same area each time. The failures occurred on different pins depending on the direction the cable was bent. This could be seriously dangerous.
The document includes per-pin measurements taken during the testing. As the cable was bent in various directions, severe current imbalance resulted from huge swings in resistance. We’ll use the last set of data as an example as it’s the most severe. The resistance in pins 3 and 4 measured high, especially pin 3, resulting in a measured 36.4A on a single pin, or 436W on a single pin, leading to a hotspot temperature of 180 degrees celsius. By the way, the current rating of stranded 16AWG is only between 5A to 7A, so we’re talking about 5 to 7 times the rated current.
The PDF goes on to hypothesize that the bending and side-loading cause the plug to improperly seat in the receptacle, perhaps deforming it. The testing conditions might seem extreme, with the cable being bent around at full load for hours on end, but this kind of thing is done to ensure a margin of safety as products get used and age in various circumstances.
Group members are encouraged to do independent testing and share the results with Nvidia, who also has volunteered to work with the manufacturers of the connectors to fix this issue. Nvidia and the PCI-SIG are trying to get ahead of a potential problem before it is allowed to become widespread.
Our opinion is that while the situation is serious and should be taken seriously, this isn’t likely to be a problem that you’ll encounter in your gaming PCs. The test conditions are intentionally extreme in a way that you likely won’t have in your own system, especially if you take care not to put too much strain on any of your cables and connectors. It’s never a good practice to shove, bend, or cram any cables.
Leak: More 4090 Pictures, Insane Coolers
Hardware leaker wxnod has been on a tear digging up leaked images relating to the upcoming Nvidia 40-series graphics cards. This week wxnod gathered photos of an insane cooler, a complete Lenovo card, and a Gigabyte RTX 4090 box.
Starting with the cooler, we see a dense fin stack with cutouts for board components and a large vapor chamber coldplate. Vapor chambers in essence act like one large heatpipe, spreading the concentrated heat from the die to the coldplate, which evaporates liquid inside the camber, spreading out heat quickly across the entire surface area. This vapor chamber looks large enough to contact memory and possibly some VRM components near the die area as well.
The next picture shows us what’s under the vapor chamber – 13 heatpipes. We’ve never seen a cooler with this many before. These heatpipes will be bonded to the vapor chamber, taking the heat that the chamber is already spreading and moving it to the fin stacks. This many pipes wouldn’t be able to be properly leveraged without the vapor chamber in between. We don’t know what card this belongs to, but it’s a seriously complex design.
Moving on to the next leak, wxnod shows a Lenovo Legion pre-built PC with a massive 4-slot, 3-fan video card. The text on the card is pixelated, but just enough was left unobscured to assume this is a 4090. Additionally, we can plainly see the 12VHPWR connector about two-thirds of the way down the card. We like that Lenovo left a cutout for the fins to exhaust the hot air, avoiding a trap we used to see far too often.
Lastly, wxnod shared two images of a Gigabyte RTX 4090 Gaming OC box originally posted on a Baidu forum. We can only gather a few details here, namely it being a 24GB card with 3 fans, utilizing a vapor chamber cooler design with a flow-through area at the end of the card.
Like the Zotac card we saw in the last Hardware News, the consistent trend throughout all of these leaks is massive size and cooling potential. Rumors have the 4090 pulling at least 450W, so this bulk will probably be necessary.
Intel has been on tour hyping up and slowly releasing information for its new 13th Gen. Raptor Lake CPUs, and now we know why they exist. Dr. Ian Cuttress confirmed with Intel that Raptor Lake only came into the pipeline due to Meteor Lake not being ready in time, and that development for Raptor Lake started only two years ago, which is shorter than most architectures. Additionally, Dr. Cuttress confirmed that the IGP and I/O haven’t changed since Alder Lake, and that Intel claims a single thread performance increase of 15%.
By extension, we also take this to mean that socket LGA1700 was originally planned for use with Alder Lake only, since Meteor Lake will be on LGA1800 and require new motherboards.
One other point of interest from Intel’s event is this slide claiming that Raptor Lake can hit 6GHz without an overclock, which would be a first for the x86 space, and 8GHz overclocked. Intel’s slide claims 8GHz would be an OC world record, but we think The Stilt’s 8.7GHz FX-8370 would like a word here. We don’t know if the 6GHz number will be broadly attainable by Raptor Lake CPUs, or if it will be limited to a top SKU, like a 13900KS.
In a recent Hardware News, we covered the existence of ASRock’s Intel ARC A380 graphics card, and now we have confirmation of the existence of an MSI offering.
This card was part of a prebuilt that was reviewed by Japanese website ITmedia (as reported by Videocardz), and is the first time we’ve seen MSI’s A380 in the wild. ITmedia provided one photo of the Intel GPU, showing it to be a half-height 2-slot design. The card has two small fans on the cooler and lacks any visible power connector. The primary inclusion for an A380 in a workstation style PC like this is likely for Arc’s excellent AV1 encoding capabilities, which we understand to be good.
This low-profile MSI A380 isn’t available anywhere as a standalone card, and only seems to be available as part of prebuilts, like the one tested by ITmedia in the original Japanese source. Our main takeaway from this is that MSI has produced some ARC cards despite rumors to the contrary, but that doesn’t necessarily mean we’ll see higher-end ARC cards from MSI.
TechPowerUp has launched GPU-Z version 2.48.0, which among other things adds Nvidia-requested features to clamp down on pre-release products being added to the database.
There are two important changes, the first of which is that when an Nvidia engineering sample GPU is installed, GPU-Z will run entirely offline. As we said, this was requested by Nvidia itself. The second is that GPU-Z will now only send traffic to www.gpu-z.com, instead of www.techpowerup.com. This makes it easier for network administrators to block only data coming from GPU-Z, and not the entire TechPowerUp site. Previous en dpoints on the techpowerup site will be disabled entirely soon.
It’s interesting that this has become enough of a problem for Nvidia that it chose to make this request, but is in line with other secretive practices, such as tightly guarding pricing until the CEO is on stage to present. The change doesn’t stop good old fashioned screenshots from getting out, but those are generally intentional leaks.
We’ve seen a lot of dual chamber cases coming to market in the wake of the Lian Li O11 Dynamic and its predecessors like the Corsair Air 740, but they all lack a certain angle that the Mars Gaming MCB brings to the table.
We mean that literally, because the MCB has the main chamber tilted at an angle, creating a showcase look. The main chamber holds the motherboard, graphics card, up to 10 fans, and a few radiators. It does look like you’d have to be careful with planning an open loop, as the thickness of fans and radiators above and below the motherboard may block access to some parts of the board or straight up interfere and not fit.
The rear chamber is where the PSU and storage are shunned to hide in shame, to the chagrin of RGB PSU manufacturers. We would like to point out that if you were to do a build in this case to avoid having the PSU in this orientation shown on the MCB’s product page. As far as we can tell, that PSU’s fan is right up against a solid steel panel, which last we checked is not very tolerant of allowing air to pass through. Flip it around and let the thing breathe.
On the same topic, if you populate the fan positions on the wall dividing the two chambers, you’d want to push air into the rear chamber so as not to fight the PSU’s own fan, which will be trying to pull air from the rear chamber and exhaust it out the rear of the case. In a way, this limits the effectiveness of the case’s front panel fans. One way around this situation would be to use a passively cooled PSU, which wouldn’t have a fan of its own.
Despite the unorthodox layout, the construction is relatively simple with the entire case being made out of stamped steel components, aside from the tempered glass main chamber window. There are no exterior front or top panels or dust filters, with only basic fan grilles built into the fan mounts on the walls of the case. If you want dust filtration, you’re on your own to figure it out.
The MCB comes with no fans and will be available in either white or black, but we don’t know the price at this time. Mars Gaming appears to serve mostly Southern Europe and Mexico, so its products might not be easily attainable in the US, but we wanted to mention it for being unique.
Backblaze has published a mid-year review of the long term reliability of SSDs in its fleet, and the short version is that SSDs are doing well. Backblaze is an offsite data backup company which publishes data on the failure rates of its drives, giving insights into the general reliability of different models.
Starting with just SSD data, Backblaze published both quarterly and annualized reports, but we’ll focus on the latter. The SSDs in question were used as boot drives for Backblaze’s storage servers. The important part to look at here is the Annualized Failure Rate (AFR) on models with a low Confidence Interval, and lower is better. As Backblaze points out, a high Confidence Interval isn’t bad, it just means that Backblaze doesn't have enough data gathered on that model yet to call it conclusive.
Standout SSDs from this dataset are the Dell drive listed here, and the Seagate drive here. The Dell drive is an enterprise M.2 drive that retails for $469, which Backblaze points out is a lot more than consumers are willing to pay. The Seagate drive, however, is a consumer SATA SSD available for $45 on Amazon when it’s in stock.
The report goes on to compare the lifetime AFR of its SSDs to its HDD boot drives, which the company is phasing out. The comparison was done between drives of the same age to keep it fair. Backblaze’s data through Q2 2022 shows us a few things.
In the first few years of life, SSDs and HDDs follow roughly the same AFR curve, with SSDs posting slightly fewer failures throughout years 1 to 4. In year 5, the relative parity breaks, with HDDs rising to an AFR of 3.55%, and SSDs actually dropping slightly to 0.92%. So for now, SSDs are trending more reliable than HDDs.
Backblaze points out that it’s possible SSDs could hit a “wall,” whe re they start to rapidly fail due to media wearout limits. Backblaze will continue to gather this data and present more reports in the future. This is just another example of SSDs supplanting HDDs in all areas save for price per gigabyte.
The price of most large studio video games has been $60 for a long time, starting with Call of Duty 2 in 2005. Now however, it looks to be slowly creeping upward to $70 with more AAA games hitting that price tag on new consoles.
Ubisoft recently stated that its new AAA games will cost $70, starting with Skull and Bones due to be released later this year. Though Ubisoft is the latest to make a statement like this, it isn't the first. In recent years we’ve seen a handful of $70 games such as NBA 2K21, Demon’s Souls, and Call of Duty: Modern Warfare 2. As more publishers tiptoe into the $70 slot and gradually normalize the price increase, it becomes easier for the next one to step up and follow suit.
It’s not hard to see why these game companies want to start charging more, with rising development costs and the effort required to deliver a AAA game rising gradually over time. You’d be hard pressed to find many other things that have remained the same price since 2005 or 2006.
This could benefit indie developers over time, as their costs don’t scale in the same way as AAA firms and their prices can remain disproportionately competitive. From the consumer’s point of view, it’s easy to understand the desire for the price to stay static over time in a world where everything else has jumped up significantly in the past 15 years. Adding to the mindset, there have been tons of good indie games priced anywhere between $5 to $40, so it sets a sort of precedent.
That said, it’s not like these AAA game companies are barely getting by; they have found ways to get the money out of you. DLC, micro-transactions, and season passes have all handily filled the void and the bank accounts. We think that it’s reasonable to pay $70 if the game is complete and isn’t aggressively monetized post-sale, but our level of faith in AAA developers and publishers is almost zero when it comes to that.
Aside from the cost of the PC or console, gaming is still a relatively cheap hobby. We’d be curious to know what your opinion is here. Would you pay $70 or even $80 if you were guaranteed no micro-transactions or season passes? Let us know in the comments below.
Valve’s Steam Deck continues to be a hit in the market, with more users getting the device delivered every week. Now, Valve is offering first-party repair for the Steam Deck for both in and out of warranty issues. Until now, any out of warranty repair situations had to be DIY or third-party utilizing parts and guides from iFixit.
The new repair centers will diagnose any issues with your Steam Deck, then let you know if the required repairs are in or out of warranty. In the event of the latter, Valve will quote you for the repair to be done in house, but offer to just send it back to you if you want to handle it yourself.
Making parts and guides publicly available is huge for repairability, but not everyone is comfortable with working on their own expensive electronics. That’s not within everyone’s capability, and that’s OK. It’s the option that’s important. Valve is taking responsibility for the product in more stages of the life cycle, and we like this move.
As a last piece of incredibly important information this week, AMD has just launched its most anticipated product of 2022 … socks. They’re so new, in fact, they don’t even have a real product image.
The socks pay tribute to the classic AMD logo and color, green, and are available for $20 on the AMD Fan Store. AMD is late to the game though, as Intel already has two “Heritage” sock SKUs available. We checked, but it appears Nvidia has yet to enter this competitive market. Be sure to make your foot warming choice appropriately.