Truth to be told AMD just shocked us in their blog. Now if you never read there – you are probably safe from the negative influence, but if you do, you could easily fall into doubt that those technical-marketing guys there don’t tell jokes… or at least not in the manner that will make you laugh. Read on to see how we decided to answer most of AMD questions (hey, everything has an end, we couldn’t possibly answer every answer they would throw, as they seem clueless about many things).
First of all – we are no fanboys at all. Well, one of us said he’d never buy AMD processor in a notebook, but he would, if they were good enough. In other words we like how AMD is filling a niche of lower-price-than-intel products, but we would never enter the oh-so-wrong comparison of their processors on a pure performance base. Because they will lose right now.
When it comes to mobile computing, there are few facts that everyone should know. For example, you never play games on battery, because in that case neither the CPU or the GPU will ever go into low power state, and those are the two most power hungry things in a laptop. Or you can run a game and make that notebook draw 80 or more watts (we take that you have decent graphics and processor) from you 6-cell battery which usually is set for like 50-55 W/hr. You can do the math on yourself.
Another fact is that people tend to forget there are different Power plans on Power options. Which always reflects battery life. They also forget their notebook on too much time, but wgo are we to judge them.
One more fact is that browsing too low power. Flash sites, hard to render objects, heavy scripting and big banners do draw current and the faster the processor, the worse. Wrong! The totally independent research of some forum guy playing with RMClock shows that if the processor finishes a job much faster than while it’s working at reduced speed, it will go faster into its power saving states (t hose C-states in the Advanced menu of RMClock; we advise you against using RMClock though) and require less power, compensating for the time it drew more better than if it was working at slower speed all the time without entering C-state. This research can be don with sole power meter by anyone, so it doesn’t need to be industrial grade research.
Now we can continue with facts, but let us tell you the reason we started this topic – AMD ignores our comments in their blog for more than 5 days (6th today). Now we know some of those posts there are old, but then, they are all part of a bigger questions AMD asks, and some of them have more comments than the number of our readers. And that’s after we decided to write there, which was initially out of question after we read that line:
Now the cynics and the wiser-than-thou will tell you that we are only doing this because our idle battery life is not as good or that we should use something other than 3DMark06. To them I say: Let’s engage in a discussion, and if you find a better test on “active” use than 3DMark06, we would love to discuss it with you! Let’s work together.
If you break that quote to more than one pats it starts to sound interesting. First of all, lets point the fact Nigel Dessau is senior vice president and chief marketing officer at AMD. Means he has the right to play bigger-than-you, but still doesn’t have to do in offensive manner as putting in one sentence “cynics” and “wiser-than-thou” right before he asks for civil discussion and even join forces with us (the cynics) and work together. First of all, dear Nigel, we are truly wiser-than the average reader you have there. Second to that is the fact we are cynical because of people working in marketing teams (like, for example, you) who tell fairytales to the not-so-wise in such persuasive manner that makes them sound right. Anyway, we decided to actually work with you and your colleague over AMD blogs, but our posts mentioning Intel don’t get approval. Some screenshots attached to prove it, and those are the more offensive ones (at least the last one):
The answers
We decided to invest some some time and answer in the best way we can to some of the questions AMD guys asked over the AMD blogs here. Don’t ask how we arranged those, it’s not chronological order but chaos :)
Objects in the Toolbar May Be Further Away Than They Seem
More to the point: does this number represent the PC’s battery life with the machine in use, or sitting idle?
When it comes to stated battery life of mobile phones there are two numbers – stand by time and talk time. Problems is that you don’t have “talk” mode on your notebook, you have average consumption and we already tried to tell you (but in a comment that still doesn’t have approval) the best way to describe average is to literally state the average notebook consumption in watts (which is the point between maximal and minimal consumption, really). because no one does that, we take most numbers represent the next stage of that math – you divide battery capacity to average consumption and state the number. You are right to ask, but nobody would answer.
For this reason, we propose that the industry needs another test to measure battery life, and we would like to propose adoption of the industry-standard 3DMark06 benchmark. The reason I like the sound of 3DMark06 is that it uses more graphics, it runs on Microsoft Windows XP and Windows Vista, and, most importantly, it runs the machine at a higher overall system utilization than other industry-standard benchmarks.
Take a look at the first fact about notebooks – you don’t play games while running on battery. Other than that we agree that 3DMark06 should be used (but as you will see, there are few caveats in that) to measure systems’ performance, as it represents to some extend how a system behaves in 3D and it could give you relative performance of the notebook compared to better known system or desktop.
We take that 3DMark06, as well as 2/3s of this graphic is not intended for use with laptops with integrated graphic cards like HD3200 or GMA 4500M. They are not meant to play games, and you actually cannot play games with those low end videocards. The middle comparison actually manages to tell us what you wanted to point out (the fact that 3DMark06′s consumption is much bigger than MM07′s, but it also compares 9600M GT to HD3450, which puts in question the performance per watt ration. Furthermore this is exactly where the aforementioned caveat shows – you can force the NVIDIA GPU to sit in its 2D speeds all the time while running on battery by simply using the Power saver windows mode – NVIDIA Powermizer’s levels are controlled that way, not by software switch ATI’s cards have in Catalyst Control Center. Well the result will clearly show such tactics to achieve more battery life and it’s not like someone would punish a notebook manufacturer that states result while running on AC and time the test loop ran on battery. Plus the 3Dmark06 method represents the more simplistic approach we offered above – state max consumption in watts and battery capacity in watts per hour, so anyone can derive the time by himself.
Because many manufacturers put their own power saving software add-ons or implement different techniques, we could use “lowest/max consumption”, which is the closest to the mobile metering we can get.
Shouldn’t we, as an industry, try and fix that?
You should and could, but it’s just not such a big deal you make it sound like.
All the Great Things are Simple
It’s about how we:
- Integrate technology to do new things
- Collaborate intimately with our customers and business partners to help solve their problems
- Impact the market
First bullet is OK, but let us tell you one secret – as customers we wanted drivers for our graphic card and when we didn’t get such, we blamed AMD, It doesn’t matter if it was some manufacturer fault we had to either stick to old ones or mod the desktop ones, it was AMD fault that we didn’t get drivers ready and waiting. This changed with Windows 7, but it’s prime example how you cannot rely on partners in some cases.
As for the third bullet – when it comes to notebooks, in order to impact the market, you have to have the better or at least a better platform, and “better” is not “cheaper”. You talk a lot about battery metering not because you expect that you will beat Intel on that field (meh, if they decide to release 0.9V Merom-3M cores at speeds close to the best AMD has, you’re done), but because you cannot make your processors to conserve more power, meaning AMD CPU’s power states aren’t as good as Intel’s.
For the typical consumer considering a notebook purchase, the first question is typically a usage question, maybe followed by a size and weight consideration.
Who told you that? As a long time notebook resellers we are we can guarantee you that outside US first question that sellers ask clients is about usage and more often than not people just don’t know or aren’t sure when it comes to general usage e.g. low-end systems (VISION level unknown). The first question the customer already asked himself is almost always one and the same and it’s about size. Weight is important, but most people consider that most notebooks have similar weight. Graphics is considered much too often by its memory size, which is plain wrong, but people don;t know this.
In testing we did, not once did the shopper ask for a processor brand as primary consideration.
That is, again, and not that we imply something that is not true, probably in US, plus it’s a seller’s job to tell him there is difference yet.
What mainstream consumers want is machines on which to do office work, to watch movies, to listen to music, to edit their photos and even edit their videos.
Then how about we talk about the important matter – screen resolution and color rendering capabilty, graphic card video acceleration and speakers quality? It’s a well known secret ALL notebooks today use TN panels (some better than others, yet they are all close to none to *VA/S-IPS, but much cheaper). We will accept that most graphic cards have similar DXVA level (although CUDA made NVIDIA better in that area), but speakers’level is low even compared to those radio spots back in seventies. There are good examples how a single-speaker system can sound much better and they don’t cost a fortune – look at XMI X-mini and similar sound systems.
Today AMD is introducing a new approach in retail we call VISION Technology from AMD.
And we already talked about it here.
We are also going to enable a VISION – Black edition for our technology partners who build the high-end, top of the line systems.
How would you do that without Core i7 or at least top of the line Core 2 Duo processor? What is more important… actually the most important thing – where is Phenom II Mobile? (AMD Pherion Ultra sounds nice, didn’t it?)
Test Driving AMD’s 2nd Generation Ultrathin Notebook Platform
Could it get better? Let me give you my first impressions on the AMD 2nd generation ultrathin platform, introduced today along with our new “Vision Technology from AMD” campaign
Problem is that you never mentioned battery life, when it’s key selling point of ultrathin notebooks.
I must caveat by saying that if you consider yourself a gamer, go for a system that has a higher end GPU like the higher end ATI Radeon HD 3000 or HD 4000 series. But if you play games but don’t consider yourself a gamer, no need to worry, you get a real ATI-branded graphics capability, not a generic brand in many systems.
It doesn’t really matter if you consider yourself a gamer. A notebook that cannot hold todays basic shooters and has to run even something as ancient as WoW on lower resolution in order to avoid stuttering (forget about Aion or Warhammer Online) is not to be considered even worthy of mentioning it amongst the word “gaming”.
To be continued…
Related articles:
Quick navi:
[...] See the original post: Twisted-Reviews.com – AMD asks.We answer. [...]