View Full Version : Please don't shoot the messenger...PowerPlay

03-15-00, 12:38 PM
This will be long, even for me! (LOL)

I'm Cisco Certified, and I work for an ISP. I think most of you know that, but I state it again for those who don't, who might think I'm talking completely out of my derrier.

I'm starting to wonder if the Cisco/Valve backed "PowerPlay" should be relabled "MarketingPloy". Here's what I'm going to do. As opposed to blindly assuming that everything Cisco does is good and light and made of fairy dust, I'm going to decypher some of their babble speak for you and explain what they're talking about in people speak.

First of all, the best explanation of who the players are and a good interview is here:


I'm going to C&P the relevant sections, and then give you my take on them. As always, your mileage may vary, but I talk to Cisco on average 3 times a week, every week. I am not an enemy of Cisco, I'm a Systems Integration Engineer, I'm one of the people who will have to make this work, and I'm starting to have serious reservations abotu it.

First off, I'm going to more or less ignore the gaming companies. For two reasons. One, they don't write or deal with network standards, and two, they're very much the backseat partner in this.

Okay..let's get on with this.

"1) What is PowerPlay?

PowerPlay is a set of standards and protocols for improving games and entertainment on the Internet. "

Standards defined by who? What guarantee do I have that a Cisco defined "standard' is going to work on ANY other router type? None. This is an important point, boys and girls, There are half-a-dozen major manufacturers of routers, if this "standard" doesn't run on any one of them, it's going to create areas of the net where powerplay won't work.

"2) Who is involved in PowerPlay?

Initially PowerPlay has been defined by Cisco, the leading supplier of Internet infrastructure, a leading US ISP, and Valve, creators of Half-Life. Over time, PowerPlay will become an open industry standard."

Over time? How much time? The problem here, is that a standard has to be open from the get go. You don't have a proprietary standard and THEN release it to the IETF, IEEE etc, It has to come FROM them to BE a standard. They need about 3 years or so to look at it, tweak it, and change it and fight, bitch and scream at each other over it until they hammer something out and release it. THAT, is a standard. Not Cisco giving them a prepackaged Cisco router only solution.

"3) Why did these companies get together?

When the Internet was originally conceived, it wasn't designed to deliver high quality consumer entertainment. As a result, developers have focused on mitigating the worst rather than optimizing for the best. In order to bring about the necessary changes in the Internet to allow for great consumer experiences, it became clear to the three companies that a coordinated solution between content developers, service providers, and infrastructure manufacturers was necessary."

This could be shortened to:
"To make money. Duh."

"4) How does it work?

The initial focus of PowerPlay is on Quality of Service (QOS). There are a wide variety of protocol and deployment issues related to improving QOS. Valve brought experience and technology related to client applications design, our ISP partner addressed deployment issues, and Cisco addressed router and access concentrator issues."

Oh boy! QoS! Well! That will just make everything peachy keen! Umm..not.

QoS is a primary Cisco marketing tool. Which will work on certain Cisco routers, everyone else can go blow. QoS comes out of Voice over IP, which is very INTOLERANT of delay, , packets arriving out of order, also known as latency, or ...lag. Sound familiar?

QOS can be thought of like this. Traffic A, in any Queue, will ALWAYS be sent first, in order to maintain QoS, which might be based on the type of traffic (like voice), the priority level of the customer (someone who paid to be first in line), application, or other traffic shaping considerations.

"5) What's the result?

Users who have a PowerPlay enabled system will have a dramatically better entertainment experience. Problems with lag, packet loss, jerky play and so on will dramatically be reduced. Basically PowerPlay brings the quality of LAN play to the Internet."

Let's change that to:
People using an end-to-end Cisco only router solution will experience less latency or lag, because their gaming packet will be sent first out of the Queue, and other types of traffic will experience increased delay.

"6) How will PowerPlay be rolled out?"

Official Answer skipped because it's not really important and is a really long statement about deployment. Short version:
Modems first, everyone else later.

"7) Is this just for games?

Games are probably the most demanding application, but other applications such as voice or streaming video will benefit from the issues addressed by PowerPlay. For example, bandwidth reservation, which will be addressed in PowerPlay 2.0, is applicable to a wide variety of Internet communications and entertainment applications."

That folks, is QoS in a nutshell. Call it "PowerPlay", or "CiscoOnlyPlay", or "ProprietaryPlay" if you want. It's Cisco Marketing. It's a Cisco tool, running on some Cisco routers, (not all support it) and it's traffic shaping via QoS. Period.

"8) Why is the Internet important for entertainment?"

Because it's closer than the movie theater and cheaper too. NEXT!

Here's the quote from Cisco marketing:

"As the market leader in remote access equipment-the point of contact where a gamer first touches the Internet-Cisco has tremendous opportunity to improve a gamer's dial-up connection. But to achieve maximum improvement, the game itself, the network it is deployed on and the dial-up connection have to be optimized. Gamers get the best possible Internet gaming experience with PowerPlay because it combines Cisco's key enhancements with those of Valve and our service provider customers."

Mathew Lodge, Mgr, Product Marketing

Buy Cisco end to end, have a bunch of people configure a number of routers, (with all that entails in terms of misconfigurations, forgeting to save the config changes etc, etc...), and implement QoS and you can have faster gaming. Of course, you're gonna start to hear screaming from all the people trying to watch porn videos and run voice over IP, but hey, they can always buy a higher QoS than you..in which case, you're back to where you started from.

NOW, mind you, IF the software companies can optimize the netcode to run on say, an ATM format, with fixed length cells, then that might show some improvement across the board for everyone regardless of their system. And you betcha ATM is very suited to low latency applications. OTOH, just to make things more annoying, ATM isn't as suited to downloading, because of it's fixed byte length. It's a good side / bad side issue.

In the end, I want to say very clearly that I'm not against Cisco, or any solution taht speeds up the gaming experience, however, I do not want people to pin their hopes for a free boost to their speeds on this. Because the way *I* see it shaking out, and this is just a guess, more or less, is that some services WILL switch over, and start using this as a marketing point. HOWEVER, your main line corporate backbones (upon which the majority of the internet runs, like it or not), are NOT going to invest in new gear just to help the gaming speed of someone thousands of miles away. IF, OTOH, it's a STANDARD, then they're more likely to become compatible, for their own marketing reasons.

Cisco is trying to use it's market presence and power to create an exploitable niche so that it can move more Cisco product. That's really all that's going on.

As always, your mileage may vary, and if anyone has any RFC's or white papers on this I'd love to take a look at them.


"Yeah Baby, YEAH!!!"

03-15-00, 01:24 PM
Hehehehe... well said. I can see those modem users rubbing thier hands together just waiting for this technology to come out: "Our day has come fellow lagsters. Those high bandwitch turds are going to rue the day they called us HPB's!!" Or something.


03-15-00, 03:40 PM
I see your point Bouncer. Powerplay is a very good "theory" but it has too many holes to fall into...On the other hand if they do get it to work it will not only change the online gaming experence but Valve and Cisco and the other co. will be very rich... http://www.speedguide.net/ubb/biggrin.gif