7 ms·
Lichee Console 4A – RISC-V mini laptop: Review, benchmarks and early issues
- BarsMonster 3y agoFirst tests of RISC-V laptop reveal a number of software and hardware issues, but show a path to freedom.
- dijit 3y agothis is my take-away too. RISC-V in a laptop was unthinkable just a few years ago, now it's a reality (at least a prototype) and performance is underwhelming, just like it has been for ARM before. For a true "freedom machine" I would be perfectly fine with a performance trade-off but it's also not a given that the gap will be so wide forever. This is an excellent step and to show that it's functioning is huge.
- LeonM 3y ago> a path to freedom Free how? Just because RISC-V is an open ISA doesn't mean that the silicon is open. None of the manufacturers release the full design of the SoC. Many RISC-V SoCs are still relying on blobs. Most (all?) current available RISC-V SoCs are designed and produced in China. So I wouldn't be so sure about a SoC designed by Alibaba to be free of blobs, license violations, patent infringement or even hardware backdoors.
- BarsMonster 3y agoIt's just a step. We are not there yet. But we have to continue one step at a time.
- Dylan16807 3y ago"a step" is not always progress in such a complicated landscape with so many dimensions to make progress in. I like having an open development process for an ISA, but it's one of the least important factors in total computer openness. Many other boards are probably more free despite using a less free ISA. The patents on basic x86_64 with SSE2 have all expired anyway.
- schemescape 3y ago> apt update fetches packages from Chinese server So is it not official Debian? If so, that’s concerning!
- planede 3y agoThe thing to be checked are the installed GPG keys that are trusted for package signing.
- sebtron 3y agoI interpreted it as "the default apt server is the official Chinese mirror" https://www.debian.org/mirror/list https://www.debian.org/mirror/list http://ftp.cn.debian.org/debian/ http://ftp.cn.debian.org/debian/
- jraph 3y agoYep. Or it could be official Debian + some specific packages, a bit like Mobian.
- BarsMonster 3y agohttps://mirror.iscas.ac.cn https://mirror.iscas.ac.cn
- rwmj 3y agoThere's a group as ISCAS who are rebuilding everything in Fedora for unclear reasons (Fedora has a perfectly good effort to build RISC-V packages at http://fedora.riscv.rocks http://fedora.riscv.rocks). Not sure if they're also rebuilding Debian too.
- aappleby 3y agoI have one of these as well - the distro I installed has a broken Chrome that crashes shortly after startup, hopefully that's fixed now. The case is prototype-quality but fine (the fact that it's metal helps a lot), the screen is nice, the keyboard is even more awkward than it needs to be for the size. I've been a bit disappointed by the performance in my tests (compiling various projects of mine with GCC), it seems closer to a Raspberry Pi 3 than a 4 or 5.
- m00x 3y agoI think it's mostly a case of software not being optimized for RISC-V. The sipeed website indicates that it's much faster if the benchmarks are built with optimized toolchains. I imagine it has x64 or arm7 emulator for most builds.
- FullyFunctional 3y agoCompletely false. This is compared to native versions on all platforms. Supposedly TH1520 gets a small uplift when compiled for its proprietary extensions, but I haven't seen a compiler using this and it would have no bearing on application from a standard distribution. I have both 7110 and TH1520, but I'm looking forward to SiFive P670 powered SG2380 which promises to be an exciting chip.
- adgjlsfhk1 3y agoit is a problem that the compilers for riscv are still pretty bad. it will take a few years (and more cheap risc-v dev boards) before gcc and clang actually know how to generate good code for risc-v (this is especially a problem for things like video payback and encryption that tend to be written with significant amounts of hand tuned assembly)
- brucehoult 3y agoThere is no problem with the compilers for RISC-V, what are you talking about?
- shrubble 3y agoMy question: is it bootable into Linux without the use of any binary blobs? That is, can it be booted essentially equivalent to a librebooted system?
- aappleby 3y agoProooobably? I've written bare-metal bootloaders for RISC-V chips (though running seL4 and not Linux) and there's a _lot_ less stuff in the startup process than on x86/x64.
- shrubble 3y agoThat is great to hear, actually! Thank you.
- drmpeg 3y agoIt's very likely that there's a blob for the SDRAM controller. Something like this: https://github.com/u-boot/u-boot/blob/master/arch/riscv/dts/fu740-hifive-unmatched-a00-ddr.dtsi https://github.com/u-boot/u-boot/blob/master/arch/riscv/dts/...
- FullyFunctional 3y agoAlso the GPU doesn't have a free driver [currently]
- spintin 3y agoThe GPU doesn't have a closed driver that works either. It's been a year for the JH7110 too, so I wouldn't hold my breath. On Vision Five 2 everything is still softpipe in the official Debian from Star Five! On the PineTab-V NOTHING involving the 3D part of the GPU works! YouTube works for consumption, but that's not interesting for production. To write/film a review of a SoC with GPU without running glgears is backwards. I think the only hope now is for the upstream to gain a working driver in a year or two.
- DeathArrow 3y agoDoes it use U-Boot or EFI?
- BarsMonster 3y agoIn kernel config I see: CONFIG_EFI_STUB=y CONFIG_EFI=y If that's what identifies it
- bradfa 3y agoThat just defines what the Linux kernel will be able to do, it doesn't necessarily imply that the hardware is using UEFI for firmware to launch another bootloader or OS.
- rwmj 3y agoIt can be both - uboot can emulate enough of EFI to start Linux.
- matthews2 3y agoIt uses U-Boot, but you can use U-Boot to launch an EFI executable. https://docs.u-boot.org/en/latest/board/thead/lpi4a.html https://docs.u-boot.org/en/latest/board/thead/lpi4a.html https://docs.u-boot.org/en/latest/usage/cmd/bootefi.html https://docs.u-boot.org/en/latest/usage/cmd/bootefi.html
- snvzz 3y agoAIUI Linux boot flow with this SoC goes internal rom -> u-boot SPL --> opensbi --> u-boot --> Linux
- snvzz 3y agoThis is TH1520, same SoC as the Lichee Pi 4A. A chip that's oddly faster than JH7110 (VisionFive2, Milk-V Mars) in microbenchmarks, but slower in practice (e.g. gcc). Presumably due to smaller cache. It is also less power efficient, and lacks the upstream support JH7110 enjoys[0]. I would look at Pinetab-V, a tablet-laptop based on that SoC, today. Better yet, wait for Milk-V Oasis[1] tba this June, as well as other boards based on SG2380, the first announced RISC-V SoC with serious performance: 16x SiFive P670 and 8x X280, all RVA22 compliant, plus vector 1.0 (standard) extension. 0. https://rvspace.org/en/project/JH7110_Upstream_Plan https://rvspace.org/en/project/JH7110_Upstream_Plan 1. https://community.milkv.io/t/introducing-the-milk-v-oasis-with-sg2380-a-revolutionary-risc-v-desktop-experience/780 https://community.milkv.io/t/introducing-the-milk-v-oasis-wi...
- BarsMonster 3y agoDamn, best products are always in the future :-) Oasis looks very capable and juicy, looking forward to get my hands on it.
- snvzz 3y agoSame. This seems to be the SoC and board I have been waiting a long time for. Do not forget to preorder, to get the unbelievably low price listed in their announcement. It should be fun to play with it this summer, and further down the line it will become my home server, replacing an old atom board that's considerably weaker. That's mostly enabled by the builtin I/O (pcie slots, sata ports...).
- codetrotter 3y agoI placed an order for 4 coupons :D
- davidlt 3y ago+1 on your initial comment. Exactly how I feel about the current situation. MilkV Oasis with SG2380 would be the end-game for majority of developers, but they are definitely loosing money if they keep starting price at 120 USD. They don't have it frozen (they changed the SoC specification some weeks ago) thus I wouldn't be surprised to see this slip into 2025. I wouldn't be surprised if this outperforms MilkV Pioneer.
- bpye 3y agoI could get over the perf but the power consumption is rough. It seems strange that the USB hub is using so much power. From a (leaked?) data sheet I don’t see any way to disable it - but it does have firmware which can be updated? There is also a 5V-3.3V LDO built in so I wonder if that may be what is dissipating so much power. The AMS1153 looks like a similar story…
- BarsMonster 3y agoUSB 3.0 can eat alot of power due to high speed, if it's always active...
- bradfa 3y agoUSB 3.0 host controllers and hubs do not need to consume large amounts of power nor do they need to always be active, if designed well. USB 3.0 has LPM (link power management) and Microsoft wrote up a nice overview: https://learn.microsoft.com/en-us/windows-hardware/drivers/usbcon/usb-3-0-lpm-mechanism- https://learn.microsoft.com/en-us/windows-hardware/drivers/u...
- happycube 3y agoIIRC pi4 also had VIA USB hub power issues early on.
- __michaelg 3y agoYup, something is strange there. There are 14in x86 laptops that consume 7W or 8W in idle (with the screen turned on but on minimal brightness).
- pat64 3y agoCouple of observations here; - RISC-V has miles to go, I worry that with the current pace of ARM it may never catch up. - The assertion you can’t buy UMPCs in 2024 is patently false. GPD and a whole host of manufacturers make em. - A non mobile optimised blog in this day and age? That was a chore to read on the go.
- G3rn0ti 3y ago> RISC-V has miles to go, I worry that with the current pace of ARM it may never catch up. Only assuming development progress happens linearly. In fact, technical progress goes exponentially. It took years to even get to the point to have RISC-V SoCs and now we are seeing the first laptop products. The pace will accelerate, especially with the promise of RISC-V being a cheaper platform to build hardware on than ARM and X86.
- sapiogram 3y ago> In fact, technical progress goes exponentially. Their point was that ARM is also progressing exponentially.
- rwmj 3y agoBut only a few companies (two? Arm and Apple?) develop new Arm designs, whereas anyone is free to develop new RISC-V designs.
- bigfishrunning 3y agoMost of those new Risc-V designs Are on the extreme low end; not a bad thing, but there's not a ton of high performance Risc-V stuff happening. It's more likely to fill in the ESP32 niche then the ARM one
- brucehoult 3y ago> but there's not a ton of high performance Risc-V stuff happening. Uhhh ... yes, there is. Several companies designing very high performance RISC-V were founded in 2021 or so. Most of them have top designers fresh from designing the latest Apple, Intel, AMD chips and they are aiming to match current x86 and Apple performance. You'll see their products in the market around 2026.
- tw1984 3y ago[flagged]
- BarsMonster 3y agoDo you imply that chip is fake and does not work? :-) Do you imply that there are no fake products on Amazon, which is also making chips (AWS Graviton)? Exactly the same issue: https://www.youtube.com/watch?v=y83BS_mK9GE https://www.youtube.com/watch?v=y83BS_mK9GE There is a reason why Aliexpress and Amazon suffer form the same curse.
- ulfbert_inc 3y agoIn Belarus where the author is from people don't have many options.
- ulfbert_inc 3y agoHis linkedin page says he's in Zurich, Switzerland, so I correct myself
- mkl 3y agoYou can edit comments less than 2 hours old by clicking their "edit" link.
- BarsMonster 3y agoI was originally from Belarus, but as you mention - I am indeed in Zürich for some years. Still, I can assure you, 99% of people in Belarus do work on x86, and cutting edge x86 hardware is being sold freely. Technological sanctions failed many years ago and at some point I will need to write an article about that...
- camel-cdr 3y agoSeems really fake tbh: https://github.com/T-head-Semi/openc910 https://github.com/T-head-Semi/openc910
- rchaud 3y ago
- f7alw8vc7lb58ih 3y agoHow does the Lichee Console 4A compare to a PinePhone (Pro) with attached Keyboard?
- dkjaudyeqooe 3y agoIs there a reason for a dearth of PCIe connectivity on these small SoCs? Is it difficult to implement? Perhaps a pincount issue? The big problem with these small and cheap systems is that you can't connect NVMe drives to them and you pay a huge performance penalty.
- dijit 3y agomost SOC CPUs don't really have a huge PCIe capability to begin with- possibly 1 or 2 lanes total; and you probably want USB, Ethernet etc; You wouldn't get very much performance anyway, as you only want many PCIe links in order to get bandwidth through them which you're unlikely to do on an anaemic CPU. to-wit: you wouldn't get good NVMe performance even if you had the pinouts and the CPU had 4x Lanes of PCIe anyway.
- adgjlsfhk1 3y agojust having 1 or 2 lanes of gen 3 with an m.2 would be really nice. 2 gen 3 lanes gives you 2gbps which isn't fast compared to the 4 gen 4 lanes on modern ssds, but it's fast enough that it won't slow down the rest of the system.
- nottorp 3y ago> you wouldn't get good NVMe performance even if you had the pinouts and the CPU had 4x Lanes of PCIe anyway. ... compared to eMMc or sata over usb? You won't get the full performance of the ssd, but i guarantee it will be much better.
- bradfa 3y agoGenerally high speed serdes IO blocks, which are needed for PCIe, in ASIC design are rather high power consumers. Even on Intel and AMD laptop spec CPUs there's generally many fewer PCIe lanes than on their desktop counterparts. The silicon space needed is also not small and sometimes the silicon process for making good low power but high performance serdes transceivers is not the best process for making the other things needed on the SOC, which can be solved with chiplet style designs but then you have another problem. Most inexpensive and small SOC just can't justify adding such costs and complexities given that it's very likely their target volume buyer doesn't need such features. I suspect this RISC-V SOC is priced in the $10-20 range, but I have no true understanding of the cost of this part. In that price range for SOCs almost no one has 4 PCIe lanes on offer, but some do have 1 or 2.
- cjs_ac 3y agoIs anyone making RISC-V SoCs outside of China? Trade restrictions have given China serious incentives to make RISC-V CPUs domesticly, unlike every other country that can just buy AMD64 and ARM chips. Given the geopolitics of the situation, while RISC-V is gaining marketshare in the microcontroller space, it looks like RISC-V will be the Chinese-bloc CPU ISA while the west sticks with what we already have.
- ecef9-8c0f-4374 3y agoWould not be the first time the cheaper technology becomes the market leader instead of the "better" one. Time will tell.
- jsheard 3y agoI've seen a few companies adopting RISC-V for embedded applications outside of China, like the management processor on Nvidia GPUs or Western Digitals SSD controllers, but those are places you'd probably use a small Cortex-M core rather than a big Cortex-A SoC if you were using ARM. Nvidia is still all-in on ARM for big processors.
- sylware 3y agoRISC-V is becoming a worldwide interop standard, and a real one, namely royalty free and which will be stable in time once we get all the "now we know" features, and all that at the ISA level (x86_64 and ARM are _NOT_ worldwide royalty free). And don't be mistaken, there is no perfect ISA (it is said "average"), it does not exist, only a set of trade-offs and compromises which will not fit all use-cases perfectly. RISC-V is a US Berkley initiative. Have a look at wikipedia where you'll find most of the answers to your questions. That said, to be a success, RISC-V will need _extremely_ performant implementations all across the board ("embedded", desktop, server), micro-archs and silicium process. It will have to survive its mistakes: for instance critical bugs in its major micro-archs or design flaws (you have to presume it will happen). And without access to the best silicium process, it _WON'T_ have performant implementations. Because there is no "enough" in the silicium industry, it wants always more transistors and less power consumption, and each new silicium process brings significant improvements on those metrics. This is where chinese chip designers are in trouble: Taiwan has the foundries with the best silicium process, and now you get US restrictions on EU EUV tools. Even though I wish intel and amd to switch to risc-v in the not to far away future, I would give attention at the people over there who would try to torpedo RISC-V. The other one, ARM, well, they will try everything to sabotage it, RISC-V is a death sentence for them... unless they clearly move to RISC-V ultra performant micro-arch design.
- michalf6 3y agoWhy a trackpoint without the middle click? How do you scroll?
- BarsMonster 3y agoPage-up/down keys... There are likely other ways )
- Havoc 3y agoHow difficult is working with risc v currently? I’m guessing back Linux works but most precompiled stuff not?
- saxonww 3y agoNot an expert, but I think anything compiled for the ISA implementation you have should work. If there's an issue with RISC-V it's that it's not one ISA with additional instructions accumulating across generations (which I think describes x86), it's a core ISA with a variety of extension blocks that might or might not be present in your implementation. I think this means that any random RISC-V binary stands a greater chance of not working on a chip you've bought than e.g. any random x86 binary with an x86 chip.
- adgjlsfhk1 3y agox86 has this problem too. Base x86 doesn't even support floating point. (and x87 which was the og way is now basically deprecated).
- BarsMonster 3y agoNearly all Linux packages you might want to install - work with nothing special required. Migration to ARM earlier this century made vast majority of open source software portable. Precompiled x86-only surely will not work.
- snvzz 3y agoDebian officially supports RISC-V and is about there[0]. Whereas qemu-user will run e.g. x86 and arm binaries, should you ever need to. This recent review of VisionFive2[1] has a little more info. There they even run popular games via x86 emulation. 0. https://buildd.debian.org/stats/ https://buildd.debian.org/stats/ 1. https://boilingsteam.com/vision-five-2-board-review/ https://boilingsteam.com/vision-five-2-board-review/
- windowsrookie 3y agoThe author mentions if you want a small laptop today, you need to go to eBay and buy an old UMPC. This is not correct, there are some modern mini laptops. I purchased a new 8" mini laptop with folding touchscreen and an intel N100 CPU from AliExpress. https://www.aliexpress.us/item/3256805658668421.html https://www.aliexpress.us/item/3256805658668421.html Specs: Intel N100 12GB RAM NVME SSD Slot 1280x800 Touchscreen (with pen support) USB-C port (supports charging and DisplayPort) HDMI Ethernet (1 gigabit) WiFi 6 Bluetooth 5.2 It's actually a pretty decent computer for being a no-name Chinese device. It has a backlit keyboard and my Microsoft Surface pen works perfectly on it. The 1280x800 LCD is low-res by todays standards, but at 8" is just fine. The 800 vertical resolution does mean that some modern apps get pushed behind the taskbar (on Windows 10). The cursor touchpad is difficult to be precise with, but you have the touchscreen with pen support if you need to select smaller items. Around Black Friday it was $245 + tax (Without an SSD, I preferred to install my own name-brand SSD) so keep an eye out for sales. The UEFI appears to be fully unlocked and has pages and pages of options. There are options to adjust CPU boost and voltage settings but I haven't been able to get the CPU to pull more than 6.5w. A downside of not buying from a real company: The drivers for the device are emailed to you in a .zip file from the seller. So that is pretty sketchy, but it's the same for any of these Aliexpress devices.
- kube-system 3y agoHere's another example, similar in form factor to old Sony VAIO UMPCs https://www.ayaneo.com/product/AYANEO-SLIDE.html https://www.ayaneo.com/product/AYANEO-SLIDE.html
- windowsrookie 3y agoA 6" display with Windows 11 sounds like an exercise in futility to me. That form factor probably works well for gaming, but In my experience it doesn't work well for anything else. I'm not a gamer so I prefer a more traditional laptop style device.
- alchemist1e9 3y agoAyaneo Slide vs GPD Win 4 seems to be an interesting debate. Based on my research so far, Win 4 has occulink and optional LTE module in it’s favor. Slide has tilt screen and supposedly slightly better keyboard. Is there anybody who is Linux focused and has laid eyes on both?
- nottorp 3y agoeMMc again :( Is it so expensive to add 1 lane of pciexpress or at least sata? Not via usb, that doesn't count.
- 999900000999 3y agoI'd suggest anyone who wants to play with RISC-V to buy a MangoPi instead. The software still isn't really in a working state. But I'd rather pay 30$ to figure that out over 800$.
- brucehoult 3y agoOf course a laptop with case, screen, battery etc costs more than a bare SBC. The Milk-V Mars, with very similar performance to this (see other comments for JH7110 vs TH1520), starts from $39 with 2 GB RAM (and 4 GB or 8 GB options) and of course quad core 1.5 GHz dual-issue CPU. That's around 8x the compute performance of the 1 GB MangoPi MQ-pro, for $10 more. Or, for more bare-bones, the Mars CM is $34.
- 999900000999 3y agoThanks for the suggestion. I'll admit I was really disappointed with the MangoPi. I still think I'm going to wait a few years before buying another device.
- stainablesteel 3y agowhy is the battery life is so bad?
- BarsMonster 3y agoUSB hub chip, USB<>SATA adapter - eating power for sure... SoC voltage scaling - need to be checked.
- sn9 3y agoAt a certain size, I'd rather just have a detachable split keyboard.
- Woshiwuja 3y ago420€. No thanks
- niutech 3y agoLichee Console 4A costs US$436.65 on AliExpress (https://www.aliexpress.com/item/1005006174268912.html https://www.aliexpress.com/item/1005006174268912.html), you can get much cheaper used netbook for a fraction of its price.