Framework Releases RISC-V Mainboard for Framework Laptop 13

[GFreeman]

We're happy to share that DeepComputing's DC-ROMA RISC-V Mainboard for Framework Laptop 13 is now in stock and shipping in the Framework Marketplace. This is very much a developer-focused board to help accelerate maturing the software ecosystem around RISC-V, so we recommend waiting for future RISC-V products if you're looking for a consumer-ready experience. We shared more detail on the Mainboard in an earlier blog post and video, but as a quick summary, this is powered by a StarFive JH7110 processor that uses the open source RISC-V ISA. The team at DeepComputing designed it to drop directly into a Framework Laptop 13 chassis or Cooler Master Mainboard Case. They have published setup guides for the hardware and for installing Ubuntu and Fedora.

To make it easier to jump into using a new partner-developed Mainboard or reusing an old one, we're also introducing the Framework Laptop 13 Shell today. This is a complete Framework Laptop 13 chassis with everything except for the Mainboard, memory, storage, and Wi-Fi. We're eager to continue making Framework products excellent platforms to extend on. Modularity and open source documentation is good for everyone!



We have a few other updates to share as well:

• We now have massive 8 TB WD_BLACK SN850X drives in stock, configurable with our DIY Edition laptops and available in the Marketplace. That means that on Framework Laptop 16 with a Dual M.2 Adapter, you can put in up to 26 TB of storage!
• We recently open sourced the mechanical design of the Framework Laptop 16 Graphics Module Shell, and we're already seeing new modules in development! We're also starting to see some really exciting Input Module designs come out of the Community, like an e-paper display module.
• We're continuing to make the entry pricing for Framework products more accessible. For the remaining quantity of 11th Gen Factory Seconds systems with German and British English keyboards, we've reduced pricing to a new low of €639 and £549. We also ended up with a limited quantity of Crucial-branded DDR5 modules through a purchasing mix-up, which we're selling at a discounted price through the Framework Outlet.
• We published a couple of new roles on our jobs page, and we'd love to have you or anyone you know help us deliver on this mission!

View at TechPowerUp Main Site | Source

 

[Chaitanya]

Is that a microSD slot or uSIM slot?

 

[Daven]

I get excited by less and less these days but the diversification of the CPU and GPU markets are one of the things I'm really into mainly because it could finally bring an end to fanboyism. I welcome any and all RISC-V client based products into the fray. Good luck!

 

[hsew]

The mad lads actually did it. And before releasing any ARM boards too! Just think, in 15-25 years, x86 will be dead, and ARM will only be found in Apple products.

 

[Daven]

The mad lads actually did it. And before releasing any ARM boards too! Just think, in 15-25 years, x86 will be dead, and ARM will only be found in Apple products.

A bold prediction indeed!

 

[TheLostSwede]

Is that a microSD slot or uSIM slot?

Please note that the board has 8GB of LPDDR4 soldered memory and uses MicroSD/eMMC storage.

microSD by the looks of it.

 

[Denver]

They're selling a highly outdated, sluggish core. For RISC-V to gain traction, even in a niche market, it needs a major company to drive a relevant design into mass production—giving developers confidence that their work won’t go to waste. Chips and Cheese demonstrated that even the latest SiFive cores have significant flaws;

(1) A RISC-V Progress Check: Benchmarking P550 and C910

 

[TheLostSwede]

The mad lads actually did it. And before releasing any ARM boards too! Just think, in 15-25 years, x86 will be dead, and ARM will only be found in Apple products.

Not a chance, RISC-V is too fragmented and to date, there isn't any serious player building notebook level chips on the architecture, even less a desktop level chip.
Until a much bigger player than these random xinese companies that are making RISC-V SoCs step in, not much is going to happen on the software side either, as the biggest issue right now is lack of support for so many things. Considering how far behind Arm is and how slowly/badly it's going, depending if you look at Linux or Windows, RISC-V isn't likely to win any ground in the PC market space any time soon.

On the other hand, RISC-V is starting to eat away at Arm's market share in the MCU end of the market, where big companies like WD has moved a lot of their controllers over to RISC-V from Arm. So for application specific MCUs, Arm will be fighting an uphill battle, but your time-line might apply to phones/tablets, but in no way the PC market, unless a company with a lot of money steps in. There is no indication of this happening today.

Even Framework is warning that this is not for most people.

This is very much a developer-focused board to help accelerate maturing the software ecosystem around RISC-V, so we recommend waiting for future RISC-V products if you’re looking for a consumer-ready experience.

 

[Daven]

They're selling a highly outdated, sluggish core. For RISC-V to gain traction, even in a niche market, it needs a major company to drive a relevant design into mass production—giving developers confidence that their work won’t go to waste. Chips and Cheese demonstrated that even the latest SiFive cores have significant flaws;

I'm thinking it took ARM about 20 years to get from a toaster and into a laptop so RISC-V development will take some time. Hopefully the timeline will be accelerated by learning from the past. That and it's getting harder and harder for Intel to block innovation through anti-competitive behavior. A weaker Intel alone allows other players to get investment funding and accelerate timelines.

 

[TheLostSwede]

They're selling a highly outdated, sluggish core. For RISC-V to gain traction, even in a niche market, it needs a major company to drive a relevant design into mass production—giving developers confidence that their work won’t go to waste. Chips and Cheese demonstrated that even the latest SiFive cores have significant flaws;

View attachment 383316

(1) A RISC-V Progress Check: Benchmarking P550 and C910

I can only agree, you just got your comment up before me.

Also note that webp attachments don't work in the forums.

I'm thinking it took ARM about 20 years to get from a toaster and into a laptop so RISC-V development will take some time. Hopefully the timeline will be accelerated by learning from the past.

That's not true though, as there were plenty of devices that used Arm chips that had a keyboard and a display back in the early 2000's, although many might not have been a full-fat laptop, but loads of them were running various light-weight operating systems. None of them really took off though

It seems like one of the first Arm based laptops launched in 1992.

Acorn Archimedes - Wikipedia

en.wikipedia.org

 

[Daven]

That's not true though, as there were plenty of devices that used Arm chips that had a keyboard and a display back in the early 2000's, although many might not have been a full-fat laptop, but loads of them were running various light-weight operating systems. None of them really took off though

It seems like one of the first Arm based laptops launched in 1992.

Acorn Archimedes - Wikipedia

en.wikipedia.org

Then I guess I limit my enthusiasm to just the end of Intel killing off the competition through anti-competitive behavior. Where any attempt by ARM to get into laptops in the mid 90s to the mid 2010s was probably crushed by Intel, RISC-V like ARM today have a better chance. I'm not saying you but many just don't understand how bad innovation and competition was killed by Intel in that past timeframe.

 

[TheLostSwede]

Then I guess I limit my enthusiasm to just the killing off of the competition by Intel. Where any attempt by ARM to get into laptops in the mid 90s to the mid 2010s was probably crushed by Intel, RISC-V like ARM today have a better chance. I'm not saying you but many just don't understand how bad innovation and competition was killed by Intel in that past timeframe.

Oh, Intel paid companies left, right and centre to get rid of any competition, so no argument from me there. Fortunately they did poorly in the phone/tablet space and that gave Arm a chance to grow. However, now it looks like Arm is trying to strong-arm (no pun intended) its way forward, using things out of Intel's playbook to prevent any real competition. Arm has already killed MIPS, even though that was some time ago, as well as several graphics IP companies that simply couldn't compete, as Arm made their graphics IP cheaper and worked to prevent its customers to integrate some third party graphics IP in their chips. This is also why, at least in part, why so many of the cheap xinese Arm chips have used current CPU cores, 2-3 generations old Arm graphics cores, as they didn't want to pay up for anything new, but in the end Arm forced them to move to newer solutions.

The entire market is messed up and there isn't much competition out there, regardless of what we may think as consumers. Sometimes I miss the way things were in the 90's and 00's, when there was some real competition and innovation.

 

[SOAREVERSOR]

Shit I'll take one.

 

[ScaLibBDP]

As a C/C++ Software Engineer I've been involved in RISC-V software development for almost 2 years and my optimism is significantly lower.

I don't think any positive statements from DeepComputing could be taken seriously. A presentation from a top person of DeepComputing company on 2024 RISC-V Summit North America I would rate as one of the worst. Even Google complained about a lack of Consumer grade ( Not DIY! ) RISC-V hardware with support of RVV v1.0 ( RVV stands for RISC-V Vector Extension ). Many companies continue to do most of the work using QEMU!

Canonical ( Ubuntu ), Fedora and Debian spent significant resources on porting these Linux OSs to RISC-V and I don't think they have good Return on Investment that would really justify all these efforts. Companies Must Make Money. Period. NVIDIA uses a lot of RISC-V cores in NVIDIA GPUs but it is a completely different thing! Microsoft, X, Amazon and Meta spent tens of billions dollars on NVIDIA GPUs rather than on RISC-V R&D and RISC-V hardware.

RISC-V International Lost Sense of Reality and continues releasing Absolutely Useless RISC-V Extensions.

RISC-V International paid some money for a 6-year forecast ( ! ) and I don't think any forecasts longer than 1 year should be made.

By June 2024 ( 2024 RISC-V Summit Europe ) more than 120 RISC-V Extensions ratified.

By October 2024 ( 2024 RISC-V Summit North America ) more than 150 RISC-V Extensions ratified.

NVIDIA does Not release so many new GPU ISA extensions and concentrated significant resources on software libraries!

I think a total number of all RISC-V instructions is Exceeding a total number of x86 & x86_64 instructions.

I didn't attend both RISC-V Summits since I consider it as a waste of financial resources, and time, but I've reviewed ALL (!) Youtube Videos released by RISC-V International for both RISC-V Summits. I don't think I will waste my time in 2025 since I didn't get too much useful knowledge from all these technical videos.

So, I'm Not negative and I'm rather pragmatic.

@Denver
>>...They're selling a highly outdated, sluggish core. For RISC-V to gain traction, even in a niche market, it needs a major company
>>to drive a relevant design into mass production-giving developers confidence that their work won't go to waste...

Correct. Once again, even Google fails to proceed from my point of view.

@TheLostSwede
>>...RISC-V is too fragmented and to date, there isn't any serious player building notebook level chips on the architecture,
>>even less a desktop level chip...

Correct. I think major companies will start loosing interest in RISC-V R&D and releasing RISC-V consumer grade products, for example a really good RISC-V Gaming Console or a good Not DIY RISC-V laptop.

@Daven
>>...I'm thinking it took ARM about 20 years to get from a toaster and into a laptop so RISC-V development will take some time. Hopefully the timeline will be accelerated by learning from the past...

Unfortunately, RISC-V International Lost Sense of Reality and continues releasing Absolutely Useless RISC-V Extensions.

 

[TheLostSwede]

@ScaLibBDP that's some interesting insights and from what I've seen, I'm not surprised.

What's with the bold company names though? No need for that here.

 

[Evrsr]

It will have the same issue ARM did, if they are only having really old designs used.

No one will be running modern software on these, let alone compiling any kind of code. These are much slower than the P550, IIRC, and that is Cortex-A53 fast.

 

[ScaLibBDP]

@ScaLibBDP that's some interesting insights and from what I've seen, I'm not surprised.

What's with the bold company names though? No need for that here.

>>...What's with the bold company names though? No need for that here...

I don't think your comment about Bold fonts is a right one.

Let's stay focused on The Topic and tolerate some little things. I do Not like that TPU spreads rumors but I tolerate it. I really like www.techpowerup.com website because you have the Best CPU and GPU databases!

It will have the same issue ARM did, if they are only having really old designs used.

No one will be running modern software on these, let alone compiling any kind of code. These are much slower than the P550, IIRC, and that is Cortex-A53 fast.

>>...It will have the same issue ARM did, if they are only having really old designs used...

Almost all RISC-V designs I would consider as New but the problem is that many companies are Not moving ahead with RISC-V processors tape outs.

They are expecting that another companies will buy a RISC-V design(s) as an IP(s) and do all the rest work with TSMC, Samsung, GlobalFoundries, etc, companies to bring all these designs to live.

In the RISC-V world only words RISC-V ISA are free, and for all the rest, especially hardware, customers need to pay money.