Chinese MIPS compatible

L

lightp2

Gawd
Joined
Oct 28, 2009
Messages
672
I read about the processor over at Semiaccurate

1. This is a very possible potential.
2. I heard they already pay royalty to MIPS so processor import should not be a problem.
3. If they can produce 8 or 16-core setup
4. Then licensed, adapt and tuned the old IRIX to run on new generation hardware.
5. There could be a market potential for such configuration

The processor itself is not faster than x86, so the potential more likely in ,multiprocess-multi-threading environment. To compete in Linux market also makes little sense due to x86 orientation from all the top tiers. However, the old SGI IRIX nostalgic maybe could see a resurrection in combination with ATI/nVidia GPU.
 
Hi, yep I think you are right but I read the stuff in the semiaccurate forum.

I am not a technical expert so cannot offer any processor comment. What I do see is

1. Short-term it seems none are able to match the x86 fab race, performance and cost. (meaning you have a lot of those low-cost 2C4T/QC processors floating around with reasonable price, abundance of x86 software).
2. So it is pointless to go and argue that market prospect, except different reasons.
3. The current low-power is struggle between ARM and the x86 push from Intel/AMD.

4. Linux/FreeBSD are OK but then why you need to do this MIPS setup when you can get cheap and good x86 industry solution? That perhaps is a problem. There are also a lot of those who are trying to sell servers with many ATOM processors.

5. So it seems the only reasonable ground, if China can deliver a cheap multicore setup, to tap the old SGI IRIX base. I understand the old Irix is MIPS-based, but somehow MIPS is now only focus on different market, not desktop performance.

6. Since China looks intend on putting resources into developing this, maybe combining it with IRIX can create a segment that attract old Irix customers. As SGI-(now Rackable I think) still holds the Irix OS, might as well make it worthwhile with something.

7. You probably need to add ATI/nVidia GPU into the mix since short-term wise none can do performance GPU over them in sustainable scale.
 
Last edited:
As seeing more info come out from elsewhere

1. It is now quite obvious irrespective of up or down, short to midterm-wise, China intends to put resources into bringing up successive designs and implementations with internal market size as basis for outlook.

Before continue, I like to divert a bit to elaborate a point.

1. Since [H] has majority US-based readers, I think this is appropriate example

2. Whereas other parts of the world play soccer (alias Football), America has American Football.

3. If you follow American Football, just like any team sports, there tends to be average, plus some very outstanding star players.

4. Some teams have a very solid above average team players across the board. Managed and play properly, with luck, they can dominate the conferences for extend period. There's nothing wrong with this scenario. But it does mean, after a while, there could only be a few very strong teams left. Again, nothing wrong. The western scientific theorists and practitioners like to say something about this particular scene. I leave that to the wise.

5. However, it does have some issues where some are not going to accept this for long due to various reasons. Therefore in American Football, I think every year when new crop of outstanding players make themselves available for selection into Professional Football, the lower ranks teams may have priority. (The main purpose is to balance the talent pool, and hope better results will come,hope,not guarantee) I do not know the detail and exactly how it works. I am just using this as example for next point.

6. So you know in any team sports, there tends to be weakness and also strong points.

7. Enter the Coach. His main duty, in addition to recruiting/admin/marketing functions, is to identify the situation and make the best of it when fielding team on field.

8. Now switch to NCAA Football, When facing an opposing national title contender, the Coach probably knows his stuff on several levels. If you have the rigth stuff, your planning and strategy will react accordingly. If you do not have the necessary balance line, you try to figure out how to protect or negate the disadvantage, if possible. If you flatout don't have the stuff to do it right, you at least put on a show to please the paying fans.

9. If you faced level-rank teams, obviously the oppurtunity is much fancier, and the team motivation is very likely much higher, and that gives them breath room for brigther future ahead of them.

10. Unless the team is made of steel, you cannot schedule 10 opponents of Conference Champions potential in a row (unlikely due to conference games, just a point)

-- OK back to technology --

2. So sometimes we can find the analogy is not necessary correct in technology field.

3. You can find some companies hitting hard at opponent strongest points. In appears that if an opponents use passing-attack for quick score, they will reply with grander passing attach so both can go into high-scoring affair (100% likely to please the fans :) ) In the mean time, defence is left to wander on their fate.

4. And you can find the debates in sports, and also in tech.

---------------------

OK. I am not 100% sure how to correlate American Football, US MBA and Tech industry.

I feel this post is incomplete, but I do not have the wisdom to decipher the oracle. It is remotely related in macro view, but not 100% obvious how this relates to the OP.
 
I don't think this is going to be used commercially in the US in any shape or form.

Maybe in nettop form or smaller somehow.
 
lightp2 said:
6. Since China looks intend on putting resources into developing this, maybe combining it with IRIX can create a segment that attract old Irix customers. As SGI-(now Rackable I think) still holds the Irix OS, might as well make it worthwhile with something.
Click to expand...

I'm in!!
 
This is interesting, so far in the US the only single remaining major tech industry we have is in CPUs. The USA seems to be very good at innovating a field, developing new tech then getting lazy and losing the industry. CPUs are the one last stand the US has. So far the Japanese and others who have tried have been unable to topple the US based companies in CPU design and production. They have however managed to topple almost every other industry that was developed for computer use in the USA.

If China gets into this I could see this being the final blow to the US tech industry. The reason is China throws massive government help behind any industry it wishes to take over. Eventually through massive government investment, heavy protectionism and subsidies they kill the entire industry out of most other countries.

One of the reasons they like to do this is so they do not need to pay royalties to foriegn countries. Also a person should note that now days free market capitalism is pretty much a pipe dream every country uses their government to forge their market position. Without government invovlement it seems it is impossible for any company to compete with companies in other countries that are being supported.

It will not be an overnight thing but over the course of 10 - 15 years I can definitely see China converting their home market over to using these chips and in doing so since the produce pretty much all of the worlds motherboards they would be able to tip all the scales in the favor of the home grown processor.

This is exactly the way intel creates hell for chipset companies, might be a sort of ironic turn of events to see intel eventually eroded by the same practices they used to stifle competition.
 
Last edited:
My personal thinking is that China is working with western (American) tech partners not just for the economic benefits. Eventually, China will have the ability and know how to adapt (steal) western technology and manufacture it as their own. I believe China is playing the long game, and that they are using their current power to get them in a position where they don't need technology partners. Companies may have to partner with China to get in the door, but ultimately that partner company is the Chinese communist government.

The Longsoon might not be all that relevant in the short term based just on its capabilities. In the next decade I think China will be able to design and manufacture their own technology in order to supplant the current western companies. As more and more companies rush to China for its emerging markets and its cheap manufacturing, more companies will be in a compromised position in regards to doing business with China.

China is the Soviet Union with more brains, more people, more money, and more patience. Imagine if the USSR was able to partner with western companies to develop its technology instead of starting from scratch. Ultimately the fall of communist Russia was do to economics. China won't have that problem.
 
IRIX on Loongson? Quite impossible. Loongsons don't have SGI's ARC PROM. And why would China need to be any more indebted to an American company? IRIX is a dead end, and everybody knows it. I've tried using it as a desktop, and I ran straight for a Sun Blade 100 (which was even worse actually) and a Powermac G5.
 
Not sure what they'd do with Irix, though ;) The market was lost to Windows, OSX, and Linux in the graphics front. HPC spread to the various UNIX vendors and Linux.

Though if they turned it open source, I'm sure they'd end up with a following.
 
ckryan said:
China is the Soviet Union with more brains, more people, more money, and more patience. Imagine if the USSR was able to partner with western companies to develop its technology instead of starting from scratch. Ultimately the fall of communist Russia was do to economics. China won't have that problem.
Click to expand...

China is probably more capitalist than the United States... They have an unhampered market and less welfarism, that's why they're growing so fast. You cannot say something is similar to the soviet union when it has primarily privatized production.
 
lightp2 said:
MIPS enters Android Honeycomb tablet race

http://www.theregister.co.uk/2011/05/31/mips_partners_with_ingenics_to_port_honeycomb/
Click to expand...
It's doomed as a competitive processor on the Android platform. :p

The MIPS 74K core seems to perform very favorably (even faster at 1GHz) vs a single core Cortex A9. But... there's no competitive high performance OpenGL on MIPS SoCs, MIPS is not a primary development platform for Android (even the heroic x86 effort is still lagging, MIPS Android will be much worse off) and just because it's not ARM, all the optimization efforts put into ARM versions of Android will have to be redone, plus native ARM apps won't run of course.

The Reg article points out where these MIPS processors will have the best chance of success: in China where it has a small market share already.
 
Just an update note so that it is overall consistent. [I understand it is one-year-later.]

http://www.engadget.com/2012/01/12/ainovo-79-novo7-paladin-ice-cream-sandwich-tablet-hands-on/
http://www.anandtech.com/show/5438/mips-technologies-targets-mobile-push

User Jayllo and Verge - MIPS-based products are now available, Android tablets from China, even the endorsement from Google itself. To others, it appears MIPS-based Android Tablets have certain interest from generic users, even users from the [H] Forum.

Secondly, from anandtech article even MIPS technology (the company) itself seems favorable to the idea so it is not just a Chinese entry-level tablet thing.

Third, to address pxc concern, Novo tablet is the first generic Google Android ICS tablet, while ICS is still pending for many other brands and models.

Fourth, at least currently, those MIPS tablets are not doing head-on push against Apple. It seems to come from different angle, so it appears there are still sensible market needs.

Summary : Maybe MIPS is going to keep the CPU world in balance, in addition to x86 and ARM.
 
lightp2 said:
Third, to address pxc concern, Novo tablet is the first generic Google Android ICS tablet, while ICS is still pending for many other brands and models.
Click to expand...
Being first doesn't matter. It won't run many of the "good"/popular Android apps which run native on ARM due to specific hardware reliance or native ARM code necessary for perfornance. No Netflix, no Chainfire3D, no alternative browsers, no Skype, etc.

Some of the problems with a MIPS-based Android tablet are explained here in a short review last month: http://tabletrepublic.com/forum/ainol-novo-7-basic/ainol-novo-7-basic-my-impressions-so-far-232.html Also note the performance problems even in the games that "work".

It will likely never attract much interest until the Android ecosystem starts catering to MIPS SoCs. At the low price Novo sells for, it may be acceptable as a limited, niche product.

I'm happy more options are available, but MIPS Android devices are not anywhere equal to modern ARM Android offerings. The reduced functionality is priced into the MIPS models.
 
You must log in or register to reply here.
Back
Top