Slashdot Mirror
AMD Stirs Athlon Into Geode Embedded Soup
An anonymous reader writes "AMD, which in recent months has gained ground against Intel in the battle for the desktop, today announced the addition of a line of high-performance, low-power embedded processors to its Geode embedded x86 processor family. The new processors will be known as the "Geode NX 1500@6W" and the "Geode NX 1750@14W," reflecting a new naming convention based on relative performance and power consumption. The Geode NX 1500@6W processor operates at 1GHz and the Geode NX 1750@14W operates at 1.4GHz. The two new embedded processors are essentially identical to AMD's Mobile Athlon processors, including packaging, but with tweaks to process technology and transistor selections that result in lower power consumption at reduced clock rates." If it meant better battery life, I could live with a processor this slow in a laptop, but according to the linked story, AMD doesn't see much of a market for that.
They were never consistant in the first place! I put the model number vs. MHz on a scatterplot a while ago and it wasn't linear, though it was close.
What if this signature were clever?
AMD's scoring isn't based on MHz, but speed.
At $65 and $55, they're a LOT cheaper than the Pentium M (I can't find one for under $195), although they are aimed at different markets.
My server
Yes, in August 2003. AMD has decided to include all of its low-power x86 processors in the "Geode" brand.
My server
Geode NX 1500@6W" and "Geode NX 1750@14W
Hello... whatever happened to marketing making things easy for consumers? Why not just go back to the K6, K7 route. Hey, you know the next K is better than the previous K.
For example, a 3000+ is not 3Ghz but an estimate Mhz comparison with Intel's processors.
Here the comparison is made against VIA processors. So a 1500 is a 1Ghz comparable with a 1500Mhz VIA processor.
It is better explained here http://www.theregister.co.uk/2004/05/24/amd_geode/
All the power "used" in a CPU is turned into heat. So yes, these Geodes give off less heat (hence the lack of a fan in Geode NX cooling solutions).
My server
Benchmarks used to determine model numbering.
My server
Yes... but one assumes that for the same family of processors the clock cycles per instruction remains the same so it should grow linearly.
I heard the same "you need hardly any proc power to browse the WWW, read e-mail, or do IM chat" type of comments back when 120 MHz chips were high-end and P75s were in the low-end computers. And that was true then, too, but it doesn't seem to stop the standard chips from becoming 25-or-so times faster.
For example, a 3000+ is not 3Ghz but an estimate Mhz comparison with Intel's processors.
Uhhh..no, the rating is not used to compare Intel processors it's suppose to compare to the Athlon T-bird.
An XP 3000+ is suppose to run like a T-Bird clocked at 3Ghz.
It just so happened that the XPs beat out the P4 at that same clockrating as well.
The change to the Northwood core, the change to a two-channel DDR400 memory subsystem with a 200MHz (QDR) bus are two big examples.
AMD had similar (but less significant) performance increases as well.
If they would have stupidly stuck with Intel's "Clockspeed is performance" mantra, the model numbers would have eventually become extremely misleading.
First generation Palomino Athlons do not perform as well as modern Thoroughbred Athlons anymore than Williamette Pentium IV's can compare to 800MHz FSB Northwoods.
If you plot your graph according to the average score of major benchmarks, you will find that up until about the AthlonXP 3200+ (possibly the 3000+), the rating system has been surprisignly accurate, and even a little conservative. The 3200+ rating is a bit overenthusiastic.
Athlon64's are now back to a conservative system of comparing performance.
Computer Science is no more about computers than astronomy is about telescopes. --E. W. Dijkstra
The reason that Intel is side-stepping the P4 for dual core is because dual core (and blades, while we're at it) has one purpose... When big computing powerhouses were building "10 year" datacenters 5 years ago they assumed that the speed, power consumption, and cooling needs of servers couldn't continue to grow at the rate that they were forever. They were wrong, they grew faster (look at Moore's law mapped out over the last decade or so, the last 2-3 years are beating the curve but the wall is coming).
Dual core is meant as a solution for companies who have maxed out their rack space, power, and cooling (virtually every company I know of and I work in a financial shop in NY with 6 digits worth of servers). Dual core is meant to stretch the life of that datacenter in all dimensions (space, power, and cooling) so that companies don't have to spend an (unbudgeted) small fortune to build new datacenters for a few more years.
I think they're going to be pretty popular. Now if AMD could get their foot in the door (the techies pulled for opteron but we're buying Itanics now) I'd be happy.
These Geodes are actually Mobile Athlon XP processors tweaked to run at low speeds and low power. I.e., they will spank the pants off of an old school Geode both in terms of Instructions per Clock and clock rate. It would be like putting a mouse up against a 3 headed fire breathing dragon in battle.
Of course, the Athlon based Geodes are using 5x the power of the old Geode, etc.
I think that AMD has just bought out the Geode name and is repositioning it slowly against the Pentium-M and Centrino now, and ignoring the markets that Geode used to sell in.
Power = C*V^2*F. C is a constant (capacitance switched), V is voltage, F is frequency. The 1GHz part runs at 1V, and is 6 watts typical:
6 Watts = C*1v*1v*1000000000hz
C = 6/1000000000
13.125 = C*1.25v*1.25v*1400000000Hz = C*1.56*1400000000Hz
Since they're the same core, the factor C is the same. The reason it isn't exactly 14 watts is most likely the static (leakage) power... even when nothing is switching, a small amount of current is flowing, just producing heat.
My server
Memory bandwidth doesn't tend to scale up with CPU speed, so while you can expect a linear speed increase for executing instructions in cache, most applications are going to be hitting system memory a lot and dragging performance down.
Check AMD's white paper on XP product numbering; you'll see they actually base their numbers on a wide range of benchmarks to try to give customers a number which actually reflects performance fairly well; that's important when, say, they increase the amount of on-die cache, as with the Barton; a 2500+ Tbred has a higher clockrate than a 2500+ Barton -- can you think of a clearer way of showing that their performance is largely the same?
Soekris uses them in some of their computers. Soekris's computers are primarily used as routers, firewalls, WAP, etc. Cool stuff.
Actually, it's 28% more amazing: According to my TiVo series 1 kernel log, the CPU is running at 54MHz.
According to Motorola, the 7457 running at 1GHz dissipates less than 10 watts (8.3 typical, 11.5 max). By contrast, your 400Mhz powerbook dissipates 3.29 watts typical, 7.43 max. Of course Apple runs their 7457 chips at 1.33 and 1.5GHz, because speed seems to sell more than battery life.
If you bought a newer battery for your powerbook, you would see it last a very long time -- they upped the wattage of the titanium powerbook batteries with each rev., and it got fairly high at the last rev compared to the first one (original ones were 50 watts, the last rev of the pbg4 batteries were around 65 watts).
- 36% display
- 21% digital circuitry (CPU, RAM, etc.)
- 18% Hard disks
- 18% Networking
- 7% Non-critical components
It definitely takes more than replacing the CPU to really save power. Amdahl's Law in action...For those exceptionally motivated with IEEE membership, search IEEE Xplore for "predictive shutdown," "dynamic voltage scaling," or "dynamic power management."
There are two types of people: those prepared for the zombie apocalypse and those who will be eaten.
Well, I'll answer you second question first because it's a quick and easy: Yes. Less power consumption == less heat. Your processor is not giving off any energy as noise and (hopefully) it's not glowing so it's not giving off any light energy. There might be a TINY fraction of energy given off as a signal on the bus, but for all practical purposes *ALL* energy consumed has to be given off as heat.
Now, to get back to the first question, the AthlonXP-M is available with power consumption of 25W, 45W or 62W TDP (Thermal Design Power), depending on the model you get.
These new products are rated at 9W TDP for the "1500@6W" model and somewhat higher for the 1750@14W model (the wattage listed in the chips model number is a "typical" wattage, while TDP is the maximum, hence the disparity in numbers).
Otherwise the processors are basically the same as the AthlonXP-M chips. Basically these are just "Low Voltage" and "Ultra-Low Voltage" versions of the AthlonXP-M chips, to use Intel's name for things. They are very directly competing with Intel's ULV Pentium-M chips that run at 1.0GHz and have a TDP of 7W. Performance and power consumption should be about the same, though the AMD chips sell for about 1/3 to 1/4 the price of what the Intel chips sell for.
As compared to desktop chips they consume quite a bit less power. Power consumption isn't quite such a big deal for desktop chips, so the numbers aren't pushed as much, but generally you're looking at 50-100W or more for a top-end desktop chip. AMD's AthlonXP 3200+ processor is rated for 76.8W.
Note that the definition of TDP (Thermal Design Power) varies from one chip to another. AMD defines the TDP of their desktop AthlonXP chips as the maximum power it will consume while running an absolute worst-case bit of code. They define their TDP for their mobile AthlonXP-M, mobile and desktop Athlon64 and workstation/server Opteron chips as being the maximum power consumption for any chip in that line (eg all of the "Low Power" AthlonXP-M chips are rated for 25W, regardless of clock speed, and all the "Mainstream" AthlonXP-M chips are rated for 45W, even though there is some overlap in clock speeds between the two lines).
Intel's TDP is defined slightly differently again. The Pentium-M and Celeron-M is defined much like AMD's desktop AthlonXP chips, ie absolute maximum power the chip can consume. The Mobile Pentium4-M is defined like AMD's mobile chips and their Athlon64/Opteron line, ie maximum for the line of processors. And then there's Intel's desktop P4 chips, which use a TDP that is kinda-sorta-almost the maximum power the chip will consume.
Ohh, and it just goes downhill from there. Don't even bother trying to figure out how Transmeta calculates the power consumption of their chips, because as best as I can tell they are just pulling numbers from a hat! Maybe there more info if you sign a bunch of NDAs as a developer, but from what I've seen on their website the actual power consumption of Crusoe and Efficeon chips seems to be firmly in the hands of the marketing department, not the actual specs.