VIA Quits Motherboard Chipset Business
arcticstoat writes "Following the media hit that was VIA's Nano processor, VIA says that it's now quitting the motherboard chipset business that used to be its bread and butter product for years. VIA's vice president of corporate marketing in Taiwan, Richard Brown, explained that: 'Intel provides the vast majority of chipsets for its processors and, following its purchase of ATI, AMD is also moving very quickly in the same direction.' VIA will still be developing chipsets for integrated motherboards featuring the Nano CPU, but will no longer produce chipsets for Intel and AMD CPUs. Was this the right decision, and where does this leave other third-party chipset manufacturers such as SiS?" Seems like this is a tough business to stick around in.
Seems like this is a tough business to stick around in.
Considering Nvidia reject the reports of its exit from the chipset market out of hand and demanded a retraction from the original source (Digitimes), I don't think that story is worth linking to...
The ultraportables is a fast growing market, and if, as I suspect, VIA focuses on cheap low-consumption CPU + chipset, they are in a great position to capitalize from this market.
"The agriculture ministry is not in charge of Gundam" - Japanese ministry official.
They are getting out of this market, because it's not going to exist for much longer.
I am TheRaven on Soylent News
...I'm not sure how they'll do without, but look at what's happening with the latest processors. The memory controller and more and more other things are moving into one and the same chip. it won't be long before laptops are essentially one chip with traces going out to all the accessories = much simpler than today because almost all the heavy lifting is inside the chip.except memory and the only reason I don't see that going in is because none of the players have taken any interest in that.
Live today, because you never know what tomorrow brings
In this case, you have 3 companies making binary compatible "platforms". This is direct competition with the added benefit of less hardware quirks and incompatibilities from trying to support everyone else. This is the very reason why Apple is hesitant to vary their hardware.
If you fear one will dominate than the others, Intel won that fight with it's partner Microsoft in the 90s.
I think my only question in this is where nVidia will fit in? All three companies (AMD/Intel/VIA) make their own integrated everything at this point, including video accelerators. These meet or exceed most people's needs except the die hard gamers. But if most people get what they need out of the box, why would they even consider a video card upgrade? Why would they (or a vendor) buy a motherboard with an outsider chipset?
I use a VIA C3 intgerated system as a file server, the only thing I had to stick in it was the RAM and hard drive. My next desktop upgrade will probably be an AMD "Spider". I like this direction.
Average Joe doesn't buy a video card upgrade anyway, so nvidia's market there shouldn't be too badly affected. Of course, if AMD/ATi decide to introduce incompatibilities into their chipset that make it hard for other video cards to work, that's another matter. Also don't nvidia do integrated graphics? They might have a problem there.
Perhaps we'll see nvidia entering the CPU business some time soon... Maybe they'll be the new AMD, who knows?
Nobody else has this sig.
"lapped" ... it's about 20% faster clock for clock, and the top clocks are about 20% higher, so that's about 44% faster at the most, and certainly not if you start scaling to multiple CPUs where AMD is still leading (check out the 4P 16C benchmarks for AMD against Intel).
AMD now have an in-house chipset maker who are making some very well received and functional chipsets (AMD 790GX for example), have improved Linux support incredibly (Day 1 Linux Support for HD4000 series graphics cards, drivers were on the shipped CD).
I think you are seeing the natural integration difficulties in 2007 and this year as a long-term issue, whereas it is clearly a short-term issue. Barcelona was flawed even before the acquisition, R600 was an underperformer before it as well. RV770 and the fixed Phenoms are good options now, and there are good vibes for the coming year as well.
The end came when AMD's acquisition of ATI put Via in the same position they were in with Intel. To be fair, nVidia got stabbed in the back the same way. Both Via and nVidia had their turn as the de facto standard AMD chipset manufacturer, and the switch between them happened natrually; AMD buying ATI took it away from both of them by force. AMD's betrayal of their third-party chipset makers was galling. Not only is Via quitting, but there are rumors of nVidia doing the same thing.
AMD didn't betray anyone. Via hasn't released a chipset with any innovative features in years, the only reason they had any products were to cover the legacy (AGP) and low-end markets. Their changing market focus has been obvious. nVidia has released a number of products with very high-profile defects, such as chipsets with severe data corruption bugs, and GPUs that fail prematurely due to packaging issues. nVidia chose to gamble that keeping SLI proprietary wouldn't piss Intel off enough to deny them a Nehalem bus license, and they lost. nVidia makes chipsets for extreme gamers who want SLI, and those consumers will buy Nehalem platforms because they are the fastest. If all nVidia has left is the AMD market, they really have no reason to keep making chipsets. The fact that their chipsets have a reputation for running hot and having issues doesn't really help at all.
Okay, so PC Hardware is never broken? Cheap-ass power supplies that don't actually deliver 12V, caps with bad fluid, or shoddy connectors don't happen? I mean, at a basic point, you're essentially asserting that PC hardware doesn't break. Clearly, because Linux, FreeBSD, etc. are capable of running on "any system," even one where voltage levels are below the required specs or PCI buses don't actually work according to spec.
There are three levels to flawed hardware. One is the "fundamental flaws," and you're right, they're not as common as some would have you believe. I can only think of a few off the top of my head: Via had an issue with PCI buses on their systems that basically dropped data when under heavy load. ALi had a problem with early AGP implementations where the voltage was actually well below the maximum required by the AGP spec, thereby causing crashes with some video cards. I've seen mainboards where USB sound "cards" were drawing power on the same bus that was supposed to be connected to front USB ports - if more than two bus-powered devices were connected, the sound card didn't work correctly. That strikes me as a fundamental flaw, although admittedly one caused by the mainboard manufacturer and not the chipset manufacturer. I'm surprised that given the number of design flaws in any field you believe that the PC mainboard industry is exempt from this issue.
The second level to flawed hardware is the driver problems. You've got new systems running older versions of Linux. Congrats. I'd first wonder if you were trying to use anything more advanced than network support on those systems, but I'll leave that for another discussion. Can I recompile the kernel with whatever slap-dash, shoddy binary blob driver I want and expect your systems to run correctly? That's the problem Via continually plagued the world with on its chipsets - I don't doubt that most of them, hardware-wise, were as decent as any other randomly selected chipset. Their drivers were simply awful, though. If you caught them a year and a half after release, yeah, they were probably okay. When you first got a KTxxx, you could end up with something that worked fine or you could end up cursing the day you were born. You don't need OS-level workaround to fix a problem, just a halfway decent driver. On-board sound drivers: garbage. Conflicts with sound cards: guaranteed. Issues when PCI slots four or five were filled: common. None of them were likely hardware-level problems, but they were just as bad. At a certain level, they were fundamental flaws in that chipset drivers are as much a part of the system as the chipset itself. I wouldn't buy Via stuff at all after NVidia filled the hole in the AMD market (and Intel always had pretty solid stuff), but it doesn't mean I didn't have to troubleshoot it for others. If you've got older versions of Slack that aren't attempting to use ATA-133 (or hell, ATA-33 at that age), sound at any level other than SB emulation, or USB anything, I'm not surprised you aren't seeing problems (note: I'm assuming here that you're using whatever version of the kernel shipped with Slack 3.3). Those of us who bought or worked on Via boards and tried to use such features, though, were greeted with the Sisyphean task of hoping that Via's latest 4in1 driver would fix whatever bizarre problem they'd introduced last month.
The final level to flawed hardware is the cheap components used by manufacturers who sourced Via or Sis or ALi chipsets. It did no good to save a couple of bucks on the chipset if you didn't slash costs elsewhere to compensate. You ended up with sub-par equipment all around. Perhaps Via products would have a better rep if they weren't constantly relegated to the $60 mainboard (although the early days of the Athlon and the boards from quality manufacturers say otherwise). As it was, though, cheap chipsets went hand-in-hand with lousy mainboard build quality and quality control. Not Via's fault, necessarily, but it wasn't making me any happier with their product.
I'd close by asking what Windows update is for if Microsoft refuses to acknowledge Windows has any bugs, ever.