New WD Black SN750 SSD Slashes Prices, Adds 2TB Model

Western Digital has launched its new SN750 SSD with a few updated features and capabilities, including a tweaked firmware intended to boost performance. The new drives (full name: Western Digital WD Black SN750) will also be offered in a separate retail package with an included custom heatsink designed by EKWB. Whether said heatsink is necessary or an improvement compared with the standard M.2 heatsinks that are built into many motherboards these days is something reviewers will have to test, but if you like the black anodized aluminum cover look, it’s something to consider.

The biggest change to the SN750 family from last year’s WD Black NVMe SSD 2018 is the pricing. The SN750SEEAMAZON_ET_135 See Amazon ET commerce will be available in 250GB, 500GB, 1TB, and 2TB capacities, at prices of $80, $130, $250, and $500, respectively. As Anandtech points out, the 2TB model is $50 more than the 1TB model cost just last year. Prices of 25 cents per GB for premium NVMe storage is excellent. Cost, however, also explains some of the design decisions behind the SN750 itself. It’s a very minor update to last year’s model as far as performance is concerned and it uses the same type of 64-layer 3D NAND. Western Digital could have begun transitioning over to 96-layer memory already, but both it and Toshiba are reportedly concerned about oversupply issues in the NAND market as it is. This dovetails with previous reports that manufacturers might slow their technology rollouts to limit the impact of overcapacity.

According to Anandtech, the gamer-oriented features of the SSD amount to disabling system features that allow the drive to slip into its lowest power states. This will slightly improve performance, provided that your system wasn’t already configured for this in the first place.

Western Digital SN750, sans heatsink.

It’s harder to pick out a single benchmark from Anandtech’s stack that actually captures the drive’s overall performance in a single test, so I’ll just say this. While the rankings change depending on which specific workloads you examine, the SN750 is often at or near the top.

For all practical purposes, the new WD Black SN750 can be regarded as more or less identical to its predecessor, except that the SN750 is launching at far better prices, and will soon be adding a 2TB model.

By 2018 standards, the new WD Black is still a very competitive high-end NVMe SSD, and probably the best overall SSD using Toshiba/SanDisk 3D NAND.

The only place you can really ding the WD Black SN750 for missing anything is in its lack of support for PCIe 4.0 and its use of 64-layer 3D NAND instead of upcoming 96-layer cells. Then again, there’s a very real question as to how much consumers will practically notice these impacts. We’re not going to claim that SSDs have stopped getting faster — they haven’t — but the visible performance impact of going from an HDD to an SSDSEEAMAZON_ET_135 See Amazon ET commerce is still much larger than what you’d expect to see when shifting from a SATA SSD to an NVMe drive in all but the most storage-limited performance scenarios. We’ve reached the point where SSD performance can generally be described as “excellent” for the standard use cases that consumers will encounter.

The Self-Driving Industry Is Finally Becoming More Realistic

Over the last decade, most of the buzz around self-driving has focused on the vision originally launched by Google (now Waymo) and later Tesla and others, of a Level 5 (fully, really, truly, autonomous no matter what) personal automobile that you could afford and blissfully relax in the back seat of from then on. You could even send it to fetch your children from school or tell it to wander around waiting for you or maybe even find itself a parking space. More recently, it has become clear that Level 5 personal autos are going to take much longer and be much harder to create, than the initial rosy projections. At CES 2019 last week, this was reinforced by the way vendors moved from pie-in-the-sky promises at previous CES shows to a much more step-by-step, practical approach this year.

Level 5 Is Getting Further Away, Not Closer

Last year, in particular, was chock full of warning signs and pullbacks for autonomous vehicle efforts. Most famously, a number of high-profile accidents (aka crashes) on the part of Uber and Tesla cars have called into question whether some of the companies involved actually know what they’re doing. Along with that, Waymo, the undisputed leader in progress towards Level 5, met its self-imposed deadline of a commercial robo-taxi launch in 2018 only with the very softest possible launch to a small number of testers in restricted environments with limited routes. Tesla pulled the “Full Self-Driving” option from its order page, citing that it was causing confusion (if you mean confusion that something you ordered years ago may never actually be available for your specific vehicle, I guess).

However, on the bright side, there have been some positive developments in the autonomous vehicle space that point to likely near-term use cases and winners. Three of these stood out to me.

Aptiv: There’s a Lot of Money to Be Made In Levels 2, 3, and 4

One problem for nearly everyone aiming to produce a Level 5 vehicle is that it’s a total money sink. Unless you are Google and essentially print profits, or Tesla and Uber, who seem to be able to raise whatever money they need so far, that makes it a very expensive proposition with no near-term payoff. But for industry-leading automotive supplier Aptiv (formerly Delphi) the march from Level 2 through Level 5 is simply the natural progression of its automotive electronics business. They have laid out an advanced architecture that they believe will pave the way for car companies to get started with the process by integrating various existing driver assistance systems and then progressively improving them over time.

Over the last decade, most of the buzz around self-driving has focused on the vision originally launched by Google (now Waymo) and later Tesla and others, of a Level 5 (fully, really, truly, autonomous no matter what) personal automobile that you could afford and blissfully relax in the back seat of from then on. You could even send it to fetch your children from school or tell it to wander around waiting for you or maybe even find itself a parking space. More recently, it has become clear that Level 5 personal autos are going to take much longer and be much harder to create, than the initial rosy projections. At CES 2019 last week, this was reinforced by the way vendors moved from pie-in-the-sky promises at previous CES shows to a much more step-by-step, practical approach this year.

Level 5 Is Getting Further Away, Not Closer

Last year, in particular, was chock full of warning signs and pullbacks for autonomous vehicle efforts. Most famously, a number of high-profile accidents (aka crashes) on the part of Uber and Tesla cars have called into question whether some of the companies involved actually know what they’re doing. Along with that, Waymo, the undisputed leader in progress towards Level 5, met its self-imposed deadline of a commercial robo-taxi launch in 2018 only with the very softest possible launch to a small number of testers in restricted environments with limited routes. Tesla pulled the “Full Self-Driving” option from its order page, citing that it was causing confusion (if you mean confusion that something you ordered years ago may never actually be available for your specific vehicle, I guess).

However, on the bright side, there have been some positive developments in the autonomous vehicle space that point to likely near-term use cases and winners. Three of these stood out to me.

Aptiv: There’s a Lot of Money to Be Made In Levels 2, 3, and 4

One problem for nearly everyone aiming to produce a Level 5 vehicle is that it’s a total money sink. Unless you are Google and essentially print profits, or Tesla and Uber, who seem to be able to raise whatever money they need so far, that makes it a very expensive proposition with no near-term payoff. But for industry-leading automotive supplier Aptiv (formerly Delphi) the march from Level 2 through Level 5 is simply the natural progression of its automotive electronics business. They have laid out an advanced architecture that they believe will pave the way for car companies to get started with the process by integrating various existing driver assistance systems and then progressively improving them over time.

Aptiv is well positioned to evolve the electronics architecture for vehicles to support future innovations

As a compelling demonstration of how this is working, I got a test ride in one of the cars Aptiv has provided to the Lyft fleet in Las Vegas. The car itself was similar to the one they were using last year, but the system had been upgraded to provide improved functionality. For example, implementing RTK (Real-Time Kinematic GPS augmentation) has allowed the cars to locate themselves within 2.5 cm (instead of 10 cm). That makes the difference between not knowing and knowing whether a pedestrian is standing on the edge of the curb or in the crosswalk. Impressively, but not surprisingly, in the several mile test drive I took, no operator intervention was required. That included dodging buses, pedestrians, and making unprotected U-turns on a busy 6-lane street.

Nvidia also made a major move in this direction. Previously it had primarily touted how its high-end Pegasus computer was perfect for Level 5 projects. Perhaps realizing that any type of volume sales for Level 5 or even Level 4 are likely to be slow in coming, Nvidia launched a Level 2+ solution, DRIVE AutoPilot, based on the less-expensive Xavier SoC — aiming to capture the large and growing volume of vehicles that have various driver assist and automated safety systems.

Aptiv is well positioned to evolve the electronics architecture for vehicles to support future innovations

As a compelling demonstration of how this is working, I got a test ride in one of the cars Aptiv has provided to the Lyft fleet in Las Vegas. The car itself was similar to the one they were using last year, but the system had been upgraded to provide improved functionality. For example, implementing RTK (Real-Time Kinematic GPS augmentation) has allowed the cars to locate themselves within 2.5 cm (instead of 10 cm). That makes the difference between not knowing and knowing whether a pedestrian is standing on the edge of the curb or in the crosswalk. Impressively, but not surprisingly, in the several mile test drive I took, no operator intervention was required. That included dodging buses, pedestrians, and making unprotected U-turns on a busy 6-lane street.

Nvidia also made a major move in this direction. Previously it had primarily touted how its high-end Pegasus computer was perfect for Level 5 projects. Perhaps realizing that any type of volume sales for Level 5 or even Level 4 are likely to be slow in coming, Nvidia launched a Level 2+ solution, DRIVE AutoPilot, based on the less-expensive Xavier SoC — aiming to capture the large and growing volume of vehicles that have various driver assist and automated safety systems.

Intel Optane Memory H10: Cache, NAND Flash on Single M.2 Device

Intel has been busy at CES this week, discussing a plethora of new initiatives, projects, and products that included a significant number of announcements related to its 10nm technology. Optane hasn’t been quite as prominent at the show, but the company has announced a new memory product that could prove quite popular. Intel’s Optane Memory H10 combines an onboard Optane cache with NAND flash on the same M.2 drive.

This is, presumably, an alternative method of improving QLC (quad-level cell) NAND performance compared with what SSD manufacturers typically deploy. Normally, SSDsSEEAMAZON_ET_135 See Amazon ET commerce that rely on slower, higher-density NAND like QLC or TLC (triple-level cell) also include a small cache of very fast NAND. Some systems deploy this cache in a static configuration while others expand or contract the cache pool depending on how much free drive space is actually available.

In either case, however, the end goal is the same: Use some of the drive’s free capacity in a low-density configuration to improve the overall responsiveness and performance of the product in general. This approach actually works fairly well, but it does leave the drives in question subject to performance declines as they fill (again, the exact details are drive-and-OEM-dependent).

 

Integrating an Optane cache has several benefits for Intel. First, QLC NAND has definite trade-offs as far as density versus performance and longevity. Optane’s performance relative to NAND depends on the product SKU you’re comparing, but it’s generally equivalent even in the worst-case, with the potential to exceed NAND SSD performance in low queue depth scenarios. It’s caching performance relative to NAND is generally equivalent or better. Right now, the product is only intended for OEM systems, but it could make an appearance as boxed retail hardware as well, Intel representatives told PCWorld.

Optane Memory H10 will be available in three configurations:

  • 16GB Optane / 256GB NAND
  • 32GB Optane / 512GB NAND
  • 32GB Optane / 1TB NAND

Up until now, we’ve mostly seen SSDs cache for spinning disks, but using a high-speed Optane cache to buffer QLC NAND should work quite well. Putting the solution on the same M.2 silicon allows Intel to slip these drives into even the thinnest thin-and-light systems, while the density improvements of QLC NAND should help keep overall costs low.

This seems like the kind of consumer application where Optane could have a meaningful impact on system configurations, improving the amount of SSD storage available in laptops while simultaneously buffering the performance impact of adopting slower NAND as primary storage capacity.

Western Digital Plans First 16TB MAMR HDDs in 2019

Late last year, Seagate announced that it would bring 16TB HDDs to market in 2019, assisted by its development of new technology to boost drive areal density. Now, Western Digital has declared that it’s also sampling its new 16TB drives, only they don’t use the HAMR technology Seagate has developed. Instead, WD is relying on a different approach, dubbed MAMR (Microwave Assisted Magnetic Recording).

With HAMR (Head Assisted Magnetic Recording), the drive media is heated before writing. This heating makes it easier to write to the media, but also requires some material changes. The media has to be capable of heating and cooling very quickly and must withstand this cycle thousands of times. MAMR doesn’t require the same thermal transitions or material changes — it embeds a spin torque oscillator in the drive head. This lowers the surface resistance of the magnetic material and allows the drive to be written to more easily. Both technologies were developed to allow HDD areal density to continue to scale. Drive density has been improving much more slowly than it once did, allowing SSDs to close the gap with HDDs in terms of capacities more quickly than they might otherwise. Hard drives still have an absolute advantage as far as affordability per GB, but the gap has been shrinking for years.

 

The first 16TB drives will use eight platters, with 18TB drives also forecast to arrive this year. Like Seagate, Western Digital intends to adopt multi-actuator technology as well. Using multiple actuators will improve latencies and overall performance, bringing HDDs more into line with SSDs, at least in certain use cases. Hard drive seek times, for example, are intrinsically incapable of matching an SSD — there’s no substitute for solid-state storageSEEAMAZON_ET_135 See Amazon ET commerce seek timing when compared with physically needing to reposition a drive head and spinning magnetic discs.

But while Western Digital is reportedly confident enough in MAMR to forecast availability in 2019, it isn’t clear what kind of long-term plans are being made around the technology. On the one hand, WD is forecasting a years-long roadmap for its newest drive technology and capacities of 20TB+ drives by 2020. On the other, it’s also publicly saying that it continues to invest in some HAMR research. This is a touch unusual, because Western Digital’s entire public messaging has revolved around the idea that Seagate’s HAMR approach is too expensive and unwieldy to be effectively adopted. Either Western Digital’s MAMR approach isn’t as baked as the company wants to imply or it may have a tougher roadmap to reaching new density improvements. Even if WD has MAMR working at current drive capacity targets, it could be concerned about the long-term scaling of the technology. Anandtech notes that Seagate has claimed HAMR offers superior long-term density scaling, which could imply a lower ceiling for MAMR drives.

Image by Anandtech

Then again, it’s not clear how aggressively either Seagate or WD will actually be when it comes to rolling out new tech, or if these drive technologies will come to the consumer space at all. At least initially, we’re likely to see them reserved for data center deployments, surveillance systems, and cloud enterprises.

AMD Vaults Onto the NASDAQ 100 Stock Index

AMD has announced its imminent addition to the NASDAQ 100, a list composed of the 100 largest companies on the NASDAQ stock market based on total market capitalization. AMD’s stock price has increased enormously over the past few years and doubled since early January 2017. While it’s fallen from its high earlier this year, that drop has occurred in the context of a great deal of stock market uncertainty and concerns about the overall impact of the crypto market’s decline. In retrospect, AMD was hit less-hard by the fall-off in crypto sales than Nvidia, which has publicly declared that it would ship virtually no midrange GPUs this quarter in order to reduce its overall inventory levels.

Equally important, AMD has improved its margins and long-term margin guidance. The company expects its Q4 margin to be 41 percent, up from 34 percent in Q4 2017, “driven by the ramp of Ryzen, EPYC and datacenter GPU processors.” During the same call, CEO Lisa Su identified the 41 percent target as “the low end of our long-term guidance.”

 

Su has also stated that while AMD wants to gain unit share, it isn’t willing to sacrifice margins to do it, saying: “As it relates to unit share, I think unit share is certainly important. We look at revenue growth overall as important for that business, but I believe we’ll be able to do that at good margins and to continue on our margin path.” Keeping its margins up is critical to keeping Wall Street happy, and AMD has pledged to continue bringing its margins up throughout 2019 and into 2020 as its product mix improves and the last of its Carrizo and Piledriver CPUs shuffle off the mortal coil.

Looking ahead to 2019, AMD’s major challenge will be demonstrating that its 7nm ramps for CPU and GPU have both come off without a hitch. Navi, Epyc (Rome) and updated Ryzen desktop chips are all expected in-market in 2019, but despite rumors to the contrary, nothing we’re hearing suggests a near-term launch date for any consumer 7nm hardware. Our understanding is that Epyc will debut on 7nm before Ryzen does, and AMD has said that volume ramp on Epyc will happen after Q2 2019. This implies Ryzen updates coming in the fall rather than the March-April timeframe. Launch dates for Navi are still unknown, but we’re not hearing anything to suggest a near-term time frame for that part, either. AMD will undoubtedly talk about its roadmaps and plans for 2019 at CES, but we don’t expect to see 7nm hardware launching at the event.

NASA’s InSight Lander Places the First-Ever Seismic Sensor on Mars

Landing InSight on the surface of Mars was an incredible feat all by itself, but the robot has just successfully completed its first major mission milestone. After carefully surveying the nearby terrain, NASA pinpointed a location to deploy the lander’s Seismic Experiment for Interior Structure (SEIS). InSight successfully placed the sensor on the surface as instructed.

InSight set down on Mars in late November, but it couldn’t just play a game of claw machine to set down the dome-shaped SEIS package. Mars is too far away to control the lander’s robot arm in real time, so NASA engineers had to work out exactly where the instrument would go and map out exactly how InSight would deploy it with minimal risk.

NASA had to ensure the sensor made good contact with the ground. Otherwise, it wouldn’t be able to track the seismic activity beneath the surface. Luckily, InSight’s landing zone is very flat and free of debris. Still, the team constructed a model of the landing zone to confirm placement before sending instructions to the lander on December 18th. This is ahead of schedule based on the vague timeline NASA offered previously.

InSight placed the instrument about 5.3 feet (1.6 meters) away — about as far as the arm can reach. In the coming days, NASA will perform tiny adjustments to the SEIS package to get it perfectly level with the surface. It’s currently sitting on ground that it tilted by two or three degrees.

NASA’s InSight landing zone model.

NASA will start collecting data as soon as it’s in the right position, but there’s another step before SEIS is at full functionality. Engineers will check and possibly adjust the cable that tethers the seismometer to the rover to ensure there is no interference. In January, InSight will drop a wind shield over top of the sensor to make sure atmospheric conditions don’t affect the data.

Later in January, the team expects to deploy the Heat Flow and Physical Properties Probe (HP3). This instrument will drill into the surface of Mars about the same distance from the lander as SEIS but in a different direction. Once in position, HP3 will relay temperature data from the planet’s interior.

BMW, Porsche Demo Super-Fast Electric Car Charger

Electric vehicles may soon eliminate one of the most often cited pain points for owners: It takes a long time to recharge them. BMW and Porsche have all demoed technologies that could juice up electric vehicles considerably faster than current charging methods. These test vehicles showed that it’s possible to get most of a charge from a high-power cable in just a few minutes, but it’ll require new vehicles and infrastructure.

Both companies are part of the FastCharge consortium, a collection of companies looking at ways to faster-charging vehicles and the infrastructure to support them. BMW and Porsche demoed similar systems relying on a 450kW charger, but Porsche pushed the power higher. The Porsche electric prototype vehicle set a record charging at 400kW. It picked up 100 km (62 miles) of range in just three minutes. The experimental BMW i3 took it a little slower at 350kW. That was still enough to push the battery from 10 to 80 percent in 15 minutes.

The system works at up to 900 volts and 500 amps—multiply those, and you get 450,000 watts or 450kW. That’s about 25,000 times faster than your average smartphone fast charger. Porsche was able to get closer to the charger’s maximum speed thanks to a cooling rig that kept its 90kWh battery stable.

The charger BMW and Porsche used isn’t some bespoke piece of hardware that could never work in the real world. The FastCharge consortium is building technology compatible with Europe’s standard Combined Charging System (CCS). Tesla has also pledged to support that system.

 

The time it takes to charge an electric vehicle is one of the primary reasons for slow adoption. Even when a consumer can afford a higher-priced EV, they often pass because of the compromises they have to make to keep it charged. Current vehicles charge at a fraction of the speed demoed by BMW and Porsche. Tesla has done the most to make fast vehicle charging viable, but its cars only charge at a maximum of 120kW right now. Audi’s less popular e-Tron vehicles can output 150kW. Tesla plans to begin rolling out enhanced super chargers next year, but those will only boost the speed to double what it is right now.

BMW, Porsche Demo Super-Fast Electric Car Charger

Electric vehicles may soon eliminate one of the most often cited pain points for owners: It takes a long time to recharge them. BMW and Porsche have all demoed technologies that could juice up electric vehicles considerably faster than current charging methods. These test vehicles showed that it’s possible to get most of a charge from a high-power cable in just a few minutes, but it’ll require new vehicles and infrastructure.

Both companies are part of the FastCharge consortium, a collection of companies looking at ways to faster-charging vehicles and the infrastructure to support them. BMW and Porsche demoed similar systems relying on a 450kW charger, but Porsche pushed the power higher. The Porsche electric prototype vehicle set a record charging at 400kW. It picked up 100 km (62 miles) of range in just three minutes. The experimental BMW i3 took it a little slower at 350kW. That was still enough to push the battery from 10 to 80 percent in 15 minutes.

The system works at up to 900 volts and 500 amps—multiply those, and you get 450,000 watts or 450kW. That’s about 25,000 times faster than your average smartphone fast charger. Porsche was able to get closer to the charger’s maximum speed thanks to a cooling rig that kept its 90kWh battery stable.

The charger BMW and Porsche used isn’t some bespoke piece of hardware that could never work in the real world. The FastCharge consortium is building technology compatible with Europe’s standard Combined Charging System (CCS). Tesla has also pledged to support that system.

 

The time it takes to charge an electric vehicle is one of the primary reasons for slow adoption. Even when a consumer can afford a higher-priced EV, they often pass because of the compromises they have to make to keep it charged. Current vehicles charge at a fraction of the speed demoed by BMW and Porsche. Tesla has done the most to make fast vehicle charging viable, but its cars only charge at a maximum of 120kW right now. Audi’s less popular e-Tron vehicles can output 150kW. Tesla plans to begin rolling out enhanced super chargers next year, but those will only boost the speed to double what it is right now.

InSight Lander Records the Sound of Martian Wind

Several days ago, a gust of wind swept across Elysium Planitia on the surface of Mars. This has happened uncountable times over the eons, but never before was a robot on hand to record the sound of that wind. Now, NASA’s InSight lander is sitting on Elysium Planitia with a suite of instruments designed to study the geology of Mars. It turns out, they can also sample the sound of Martian wind for the first time ever.

The new recording comes from two of InSight’s sensors, neither one of which was intended for this particular use. The lander’s seismometer is currently stowed while NASA plans the delicate process of setting it on the ground and shielding it from the wind. In the meantime, NASA managed to use the seismometer to measure the vibrations from the nearby solar panel array. This sound is at the very bottom range of human hearing, but you’ll be able to hear it if you’ve got a good set of speaker or headphones. To help you make it out, the video below includes the original recording and a version shifted up two octaves.

Elsewhere on InSight, there’s the air pressure sensor. NASA used this instrument to record vibrations in the air directly, which is closer to what we on Earth would think of as sound. However, it’s nowhere close to the gusty noises you’ve heard on Earth. It’s eerie and, well, alien.

NASA had to make some tweaks to the original air pressure recording to bring it into the realm of human hearing. Mars’ atmosphere is mainly carbon dioxide, and it’s much thinner than Earth’s atmosphere. The blowing wind on Mars doesn’t have enough energy to produce eddies on the millimeter scale like on Earth. Currents on Mars tend to fall apart around a centimeter, which results in a low rumble that’s difficult to hear. Thus, the pitch of the gusting was far too low to hear unaided despite the wind blowing at 10 to 15 miles per hour. The NASA team sped up the recording by 100 times to make it audible. The final 29-second recording came from about 48 minutes of audio. That doesn’t mean you’d never be able to hear the wind while standing on Mars. Once it hit about 30 miles per hour, you’d be able to hear a low hum. Of course, you’d have bigger problems if you were in a position to hear the Martian wind blow.

Qualcomm’s ‘Extreme’ Snapdragon 8cx CPU Could Power Your Next Laptop

Intel and AMD could be looking at some stiff competition in the processor game. Fresh off announcing its new Snapdragon 855 mobile chip, the company has announced the Snapdragon 8cx. It’s for laptops instead of smartphones and is by far the most powerful processor the company has ever made. How can you tell? The “X” in the name stands for “extreme.”

While the Snapdragon 8cx is not the company’s first PC chip (that honor goes to the quickly forgotten Snapdragon 850), it’s the first one that could make Intel take note. Like the 855, the Snapdragon 8cx uses a 7nm manufacturing process. It has the same octa-core design with four high-performance cores based on the Cortex A76 and four low-power cores based on the A55. That’s really the end of the similarities, though.

Qualcomm has cranked the clock speed of all of its “Kryo 495” cores way up in the Snapdragon 8cx, but it won’t say exactly how high. The chip has 10MB of cache between L2 and L3 — the 855 only has 3MB. That makes the Snapdragon 8cx better at running heavy apps, and there’s support for up to 16GB of system memory. You can also check the boxes for NVMe and UFS3.0 storage.

The 7nm process, which no other chip maker has yet managed, gives the Snapdragon 8cx excellent performance for the power draw. At a sustained 7W of power, Qualcomm says the 8cx is twice as fast as a comparable Intel U-series CPU and four times as fast as the fanless Y-series chip. The ARM architecture doesn’t need to throttle as quickly due to heat buildup, so Qualcomm isn’t talking about peak performance. It’s likely Intel chips still clobber Qualcomm when it comes to single-threaded maximum throughput.

Intel and AMD could be looking at some stiff competition in the processor game. Fresh off announcing its new Snapdragon 855 mobile chip, the company has announced the Snapdragon 8cx. It’s for laptops instead of smartphones and is by far the most powerful processor the company has ever made. How can you tell? The “X” in the name stands for “extreme.”

While the Snapdragon 8cx is not the company’s first PC chip (that honor goes to the quickly forgotten Snapdragon 850), it’s the first one that could make Intel take note. Like the 855, the Snapdragon 8cx uses a 7nm manufacturing process. It has the same octa-core design with four high-performance cores based on the Cortex A76 and four low-power cores based on the A55. That’s really the end of the similarities, though.

Qualcomm has cranked the clock speed of all of its “Kryo 495” cores way up in the Snapdragon 8cx, but it won’t say exactly how high. The chip has 10MB of cache between L2 and L3 — the 855 only has 3MB. That makes the Snapdragon 8cx better at running heavy apps, and there’s support for up to 16GB of system memory. You can also check the boxes for NVMe and UFS3.0 storage.

The 7nm process, which no other chip maker has yet managed, gives the Snapdragon 8cx excellent performance for the power draw. At a sustained 7W of power, Qualcomm says the 8cx is twice as fast as a comparable Intel U-series CPU and four times as fast as the fanless Y-series chip. The ARM architecture doesn’t need to throttle as quickly due to heat buildup, so Qualcomm isn’t talking about peak performance. It’s likely Intel chips still clobber Qualcomm when it comes to single-threaded maximum throughput.

 

There’s also a new Adreno 680 Extreme GPU in the Snapdragon 8cx, but Qualcomm keeps a lot of its Adreno specs secret. What we do know is it’s fast. The 680 Extreme is twice as fast as the new GPU in the Snapdragon 855 thanks to twice the transistors and twice the memory bandwidth. You can connect two 4K monitors to a Snapdragon 8cx system. Like the CPU cores, the GPU can also remain cranked up for longer than comparable solutions in current laptops.

Qualcomm hasn’t offered battery life specifics, but we expect plenty of longevity. The 850 promised 25 hours of usage, and the Snapdragon 8cx is even more efficient. The SoC also includes an integrated LTE modem, so laptops running on the 8cx should have always-on connections as an option.

There’s also a new Adreno 680 Extreme GPU in the Snapdragon 8cx, but Qualcomm keeps a lot of its Adreno specs secret. What we do know is it’s fast. The 680 Extreme is twice as fast as the new GPU in the Snapdragon 855 thanks to twice the transistors and twice the memory bandwidth. You can connect two 4K monitors to a Snapdragon 8cx system. Like the CPU cores, the GPU can also remain cranked up for longer than comparable solutions in current laptops.

Qualcomm hasn’t offered battery life specifics, but we expect plenty of longevity. The 850 promised 25 hours of usage, and the Snapdragon 8cx is even more efficient. The SoC also includes an integrated LTE modem, so laptops running on the 8cx should have always-on connections as an option.