vacancies advertise contact news tip The Vault
facebook rss twitter

Snapdragon SC8280 might be Qualcomm's next PC chip

by Mark Tyson on 18 January 2021, 13:11

Tags: Qualcomm (NASDAQ:QCOM), Microsoft (NASDAQ:MSFT)

Quick Link: HEXUS.net/qaep3g

Add to My Vault: x

Reports are emerging of Qualcomm's next high-performance SoC being readied for Windows on Arm machines, Always Connected PCs, Cellular PCs, Snapdragon PCs or whatever they might be called later this year. Qualcomm's latest collaborations with Microsoft and system making partners started back in 2017 with Windows 10 laptops based on the Snapdragon 835, it has since created PC-specific 'extreme' processors such as the Snapdragon 8cx, and the Snapdragon 8cx Gen 2 5G.

The new Qualcomm processor for Windows laptops is dubbed, at least internally, the Snapdragon SC8280. WinFuture has shared the first details of this SoC, and expects it to reach Windows 10 laptops sometime this year. It has unearthed the following significant indications with regard to the SC8280:

  • There will be a basic and advanced (more powerful) variant
  • Test platforms have been spotted with 8GB of LPDDR5, and 32GB of LPDDR4X, showing flexibility in this area
  • Variants have been spotted in systems with an accompanying Snapdragon X55 modem on board
  • Reference designs have screens up to 14-inches in diagonal
  • The SoC packages seen are "significantly larger" than known Qualcomm mobile processors

The above details aren't of the depth we are accustomed to when bullet pointing the tech specs of a platform, but a lot of the specs are probably not finalised at this early-ish stage of development. We don't even know how many cores, or rough clock speeds to expect from the SC8280. Perhaps benchmark leaks will help provide an early sketch of the SoC and its performance but as usual the proof of the pudding will be in third party testing of shipping PCs based upon these processors. With Apple's M1 now being a known quantity, let us hope Qualcomm can lift its competitiveness to rival AMD and Intel – at least in the thin and light segment.

Reminder: Qualcomm bought out high performance chip startup Nuvia last week.



HEXUS Forums :: 6 Comments

Login with Forum Account

Don't have an account? Register today!
There's no reason a company like Qualcomm with its expertise in the sector and resources shouldn't be able to compete with Apples M1 but every reason to suspect they won't be able to due to years of lack of decent competition. Basically an Intel/AMD situation repeated in mobile SoCs.
Well there is competition - they are competing with the x86 chips. As for Apple you gotta remember that chip has 16 billion transistors built on a cutting edge 5nm process. That's about 3 times the number of transistors in a snapdragon 845 which was also built at 10nm. If you give Qualcomm the apple m1 transistor count and 5nm process they'd probably give it a run for it's money. For reference I think an AMD 5600X is 19 billion transistors on a 7nm process.
Dribble
Well there is competition - they are competing with the x86 chips. As for Apple you gotta remember that chip has 16 billion transistors built on a cutting edge 5nm process. That's about 3 times the number of transistors in a snapdragon 845 which was also built at 10nm. If you give Qualcomm the apple m1 transistor count and 5nm process they'd probably give it a run for it's money. For reference I think an AMD 5600X is 19 billion transistors on a 7nm process.

I actually don't think there's competition with x86 in that sense. I suspect we are going to see what has been going on since the late 1990s where general computing in a mobile form factor has been run on something super efficient but not particularly useful for anything outside simple tasks. Psion and Jornada 720 are excellent examples of this, the latter I think used a 200MHz ARM processor on a modified Windows CE OS. Now these systems are coming into their own with the advent of truly mobile working and internet access. I'd say there's now a market and available supporting technology to make this concept work properly and worth developing properly. It'll take a chunk of x86 sales, defintely, but I don't think there's competition so much as two different use cases.

If I want a macbook air, I'd not even consider x86 CPUs if there's a choice and mobility and light use is the aim. Mine has an i5. It's pointless and takes too much power for what it does (basically what my phone does with a screen and keyboard). But I can run a VM if needed. So I suspect it's going to be a use case division rather than competition.

I'm kind of thinking Intel might go down the hyper mobile route and AMD might go along the silly performance route. The former has a much bigger market share. If Intel decides to use ARM IP, I'd find that hilarious. AMD will be getting in on it by licencing GPU tech to Samsung for use on their mobile SoCs.
philehidiot
Dribble
Well there is competition - they are competing with the x86 chips. As for Apple you gotta remember that chip has 16 billion transistors built on a cutting edge 5nm process. That's about 3 times the number of transistors in a snapdragon 845 which was also built at 10nm. If you give Qualcomm the apple m1 transistor count and 5nm process they'd probably give it a run for it's money. For reference I think an AMD 5600X is 19 billion transistors on a 7nm process.

I actually don't think there's competition with x86 in that sense. I suspect we are going to see what has been going on since the late 1990s where general computing in a mobile form factor has been run on something super efficient but not particularly useful for anything outside simple tasks. Psion and Jornada 720 are excellent examples of this, the latter I think used a 200MHz ARM processor on a modified Windows CE OS. Now these systems are coming into their own with the advent of truly mobile working and internet access. I'd say there's now a market and available supporting technology to make this concept work properly and worth developing properly. It'll take a chunk of x86 sales, defintely, but I don't think there's competition so much as two different use cases.

If I want a macbook air, I'd not even consider x86 CPUs if there's a choice and mobility and light use is the aim. Mine has an i5. It's pointless and takes too much power for what it does (basically what my phone does with a screen and keyboard). But I can run a VM if needed. So I suspect it's going to be a use case division rather than competition.

I'm kind of thinking Intel might go down the hyper mobile route and AMD might go along the silly performance route. The former has a much bigger market share. If Intel decides to use ARM IP, I'd find that hilarious. AMD will be getting in on it by licencing GPU tech to Samsung for use on their mobile SoCs.

I agree a bit… I've had this debate with many people. The M1 isn't an astounding chip, it's a really decent chip. It uses a lot of transistors and has a large die area to accelerate x86 emulation and also does ye olde tied into dram for better performance trick that was really brought to the masses with ARM and the original Raspberry Pi. As said above IF Qualcomm threw billions of transistors at a chip they'd be able to match or surpass the M1 but Apple has a huge advantage with being so vertically integrated. I'm not sure Win 10, for example, would run as well because it won't be as integrated. Nor the rest of the hardware
Thin and light - well personally that side of things doesn't really bother me. My phone is good enough to do what I want without a laptop to carry around and if I really needed to take my current laptop then I could work around stuff. Most of my “work” is now based on livestreaming and working with people on livestreaming because gigs and stuff are off…
In short no SC8280 wont match the M1 but later designs probably will.

Thats the reason Qualcomm bought Nuvia because it was founded by Apple engineers responsible for many of the core features of Apple's Arm designs.

Apple must not have been able to patent these features so when Qualcomm could not match the M1 they bought Nuvia to close the gap.