Jump to content
The Inquirer-Home

Brit chip shop launches new processor

Smallest, greenest, efficientest
Monday, 23 February 2009, 11:48

CAMBRIDGE-BASED ARM HAS ANNOUNCED its news Cortex M0 32-bit processors with promises of all sorts of superlatives (OK so we made 'efficientest' up).

The plucky little British company which always fights well above its weight when it comes to making things tiny, reckons the new chippery is the smallest, most energy efficient ARM processor available and offers 32-bit power at an 8-bit price.

Armchip

The ultra low gate count of the Cortex-M0 processor enables it to be deployed in analog and mixed signal devices.

Drawing as little as 85 microwatts/Mhz, the new chips offer 32-bit power in the same footprint as most 16-bit chips and ARM expects them to be used in medical devices, lighting and gaming controllers as well as a host of other portable devices where cost, size and power consumption are top of the tick-list.

The chips are available for licensing from today.

L'Inq
ARM

 

Share this:

Comments
Netbook?

Would it be appropriate for a netbook? My N270-reference-design-in-slightly-differentiated-plastic seems to draw about 8 to 11 watts.

If the CPU were drawing micro- or milliwatts the battery would last a lot longer...

posted by : hoohoo, 23 February 2009 Complain about this comment
ARM isn't "Little"

I wouldn't call a company with a $2 Billion USD market cap "Little".

posted by : R, 23 February 2009 Complain about this comment
Units 101

You really have a problem with units, don't you?

µW/MHz

In HTML you write
µW/MHz
(BTW notice the capital H in HZ)

posted by : Metric, 23 February 2009 Complain about this comment
@Metric

µW/MHz. µ? A good portion of the readership mighn't know what µ refers to, not everyone has studied science/engineering after all. Micro - well they know it means bloody small.

And R - he may be describing it in terms of employee no. or assets. We just. don't.know.

posted by : Hdjn, 23 February 2009 Complain about this comment
@Hdjn - NOT

Sometimes I run across words that I do not know. Usually I look them up.

The target mental age for people who write for newspapers is said to be 7 years old - I hear this from communications people all the time. This explains why so much of newspaper journalism is shyte and perhaps why readership is falling.

re: units. If a reader is ignorant that is the reader's problem. I would not have articles dumbed down to accommodate the ignorant.

Metric questioning "85 microwatts/Mhz" is legit. Does the writer mean 85 uW per MHz clock speed - so a 1 GHz part would draw 85 mW? Does the writer not know what he is saying?

posted by : hoohoo, 23 February 2009 Complain about this comment
@Hdjn: micro?

You don't need to know or understand!

mini, micro nano, its all at www.apple.com ... all sexy and shiny!

Don't bother with this website - Apple for sure will use this chip (hint: cheap)

posted by : trendy, 24 February 2009 Complain about this comment
/ = per

It looks like the author means the chip consumes 85 microwatts per megahertz of clock frequency and the Cortex-MO runs at 50 megahertz. Atom punting technical shills can go home.

posted by : jamez, 24 February 2009 Complain about this comment
microwatts/MHz

Saying microwatts/MHz is a fairly standard way of saying microwatts used PER MHz, so yes, if the processor could scale to 1 GHz, it would use 85 milliwatts. Read a datasheet for any processor used in these embedded applications, they all use it, because often power usage (or thermal output) is more important than performance, so a designer will work out what frequency to run the processor based on the target thermal footprint. Having a go at the author shows your ignorance more than his.

posted by : geni, 24 February 2009 Complain about this comment
Another measure

Think of how you a velocity, you might write a velocity as 4.9 m/s. As the velocity is 4.9 metres per second, the metres quantity is 4.9 metres during a period of 1 second. So writing 85 microwatts/MHz is the same as 85 microwatts per 1 MHz of frequency. There is nothing incorrect about it. Its the same way a electrical signal can be measured as J/Hz, whereby each cycle contains so much energy.

posted by : geni, 24 February 2009 Complain about this comment
units are fine, not for netbooks

The M0 (swift) is for use in embedded micro-controller applications, and is not suited to running full operating systems, as you would expect a netbook to do. It would be fine with a real time OS (like eCos or threadx).

Those units - uW/MHz - are one of many benchmarked figures generated for CPUs. They are perfectly sound, since power scales with frequency.

posted by : john, 26 February 2009 Complain about this comment
Advertisement
Subscribe to the INQ Newsletter
Sign-up for the INQBot weekly newsletter
Click here to sign up Existing user
Advertisement
INQ Poll

Nvidia Fermi

Will graphics cards built with Nvidia's Fermi GPUs be a hit?