nForce 500: nForce4 on Steroids?
by Gary Key & Wesley Fink on May 24, 2006 8:00 AM EST- Posted in
- CPUs
New Feature: LinkBoost
One of the feature sets unique to the nForce 590 SLI MCP - and highly touted by NVIDIA - is called LinkBoost. If a GeForce 7900 GTX is detected on the nForce5 system then LinkBoost will automatically increase the PCI Express (PCIe) and MCP HyperTransport (HT) bus speeds by 25%. This increases the bandwidth available to each PCIe and HT bus link from 8GB/s to 10GB/s. Since this technology increases the clock speed of the PCI Express bus by 25%, NVIDIA requires certification of the video card for this program to work automatically. In this case, the 7900GTX is the only compatible card currently, although you can manually set the bus speeds and achieve the same or better overclock depending upon your components.
In essence, NVIDIA is guaranteeing their chipset's PCIe and HT interconnect links are qualified to perform up to 125% of their default speeds without issue. While LinkBoost is an interesting idea, its actual implementation did not change our test scores outside the normal margin of error. The 25% increase in PCIe and HT yielded virtually the same performance as our system without LinkBoost enabled. The reason is that the performance boost is being applied in areas that have minimal impact on system performance.
LinkBoost is part of a package of easy-to-use auto overclocking features on the nForce5 designed for the OC newbie. If you fit in that category and you are excited about the 25% LinkBoost speed increase, you need to clearly understand that the 25% LinkBoost increase yielded little to no real performance increase. The true performance potential of this technology would have been realized on the AM2 CPU if the MCP's link to the CPU/Memory subsystem would have been dynamically increased from the base 8GB/s level, but NVIDIA does not control AMD CPU certification and thus left the CPU at stock speed.
The end result is that the Northbridge to CPU HyperTransport link remains at 8GB/s, and only the link between the MCP and SPP as well as the PEG slots get increased bandwidth. Having 8GB/s of bandwidth feed 10GB/s basically means you are still effectively limited to 8GB/s. It is possible that increasing the Northbridge to CPU bandwidth could improve performance slightly, but HyperTransport performance is rarely the bottleneck in current systems as you will see in our performance results.
New Feature: FirstPacket
As part of the overhaul of the networking features found in the NVIDIA nForce 500 Series, FirstPacket is a packet prioritization technology that allows latency-sensitive applications and games to effectively share the upstream bandwidth of their broadband connection. Essentially this technology allows the user to set network data packets for applications and games that are more latency sensitive with a higher queue priority for outbound traffic only.FirstPacket is embedded in the hardware and offers driver support that is specifically designed to reduce latency for networked games and other latency-sensitive traffic like Voice over IP (VoIP). When network traffic constrains a connection, latency is increased which in turn can result in dropped packets that would create a jitter and delay in VoIP connections or higher ping rates to the game server resulting in stutters and decreased game play abilities.
In the typical PC configuration, the operation system, network hardware, and driver software are unaware of latency issues and therefore are unable to reduce it. The standard interfaces that allow applications to send and receive data are basically identical to the OS in a typical system. This design results in latency-tolerant and large packet applications like FTP or Web browsers filling the outbound pipeline without regards to the needs of small packet and very latency-sensitive applications like games or VoIP applications.
FirstPacket operates by creating an additional transmit queue in the network driver. This queue is designed to provide expedited packet transmission for applications the user determines are latency-sensitive. The ability of the designated applications to get preferential access to the upstream bandwidth usually results in improved performance and lower ping rates. The FirstPacket setup and configuration is available through a new Windows based driver control panel that is very simple to use.
In our LAN testing, we witnessed ping rate performance improvements of 27% to 43% during the streaming of video from our media server while playing Serious Sam II across three machines attached to the network. We noticed ping rate performance improvements of 15% to 23% while uploading files via BitTorrent and playing Battlefield 2 on varying servers with VoIP conversations on Skype during game play. The drawback at this time is that only outbound packets are prioritized, so if you spend more time downloading than uploading the FirstPacket technology will have little impact for you. However, in NVIDIA's defense they cannot control the behavior or quality of service on other networked clients, so FirstPacket addresses the services NVIDIA can control - namely uploading.
Facebook
Twitter
RSS
64 Comments
View All Comments
nullpointerus - Wednesday, May 24, 2006 - link
TCP/IP CPU utilization scales with increased transfer rates. Online gaming is hardly designed to stress a 1Gb Ethernet. So this TCP/IP acceleration is primarily intended for LAN gaming. You don't really need a firewall on your LAN (unless perhaps you are hosting a LAN party on that machine?). It's acceptable to use your router's firewall if you really know how to configure your LAN properly:Modem
|
Router
|
+--A (game server)
+--B (game client)
+--C (game client)
+--D (game client)
In this situation, TCP/IP acceleration might be useful. Of course, if you leave yourself open to your LAN and one of the other computers on your LAN is compromised, you could very well be compromised, too.
I don't understand the comments about a third-party firewall. Perhaps only XP and Vista's firewalls will be supported initially?
Trisped - Wednesday, May 24, 2006 - link
A true router can configure a firewall on all ports, both incoming and out going. In this way you can set what ports can be sent and what ports can be received through a router, in the LAN or out side as the case may be. Of course, software firewalls are usually a little more powerful in that they check who sends what. I think hardware firewalls can do this too, but they don't on any of the Linksys, Netgear, or AirLink routers I have used lately.It is also important to note that the typical home router is not a true router. It is a 2 port router (1 port for the WAN and 1 for the LAN) and the rest of the ports are connected via a switch or hub. Switches can also have firewall, but most home solutions don't seem to.
mino - Wednesday, May 24, 2006 - link
The info WHO sends teh data is the most crucial. Enabling http(80) just for Opera and 993+25 for Thundebird is a huge difference to allow al kinds of malware/spyware go out on 80.And NO, no HW (or second machine be it linux router or cisco router) can detect (reliably) which application is sending the data.
Gary Key - Wednesday, May 24, 2006 - link
I will clarify this in the article. Windows XP firewall is not supported. Vista should be but that decision is not final at this time due to continuing changes from Microsoft.
Tanclearas - Wednesday, May 24, 2006 - link
So does this mean Nvidia has abandoned AA in NF4? I know that is my impression, based upon my experience with them (http://forums.anandtech.com/messageview.aspx?catid...">http://forums.anandtech.com/messageview...mp;threa..., but just wondering what the official word is.Honestly, the idea behind Nvidia's chipset-based firewall was a good one, especially for universities/colleges with literally thousands of computers on the inside of the corporate firewall. Protection at every level is worthwhile.
As for the Vista firewall working with NF5, I definitely would not count on Nvidia making it happen.
Gary Key - Wednesday, May 24, 2006 - link
Once the new 9.x platform drivers are released with support for chipsets prior to nF5, then yes, Active Armor firewall will be dropped. The new driver set will also have further changes in them to reduce the possibility of data corruption on the nF4, whether the CPU utilization goes up or down, I do not know at this point. I do not have a "stable" set of new platform drivers for the nF4 yet, cannot verify this myself. I will be speaking with the program manager tomorrow. ;-)
We did find out that NVIDIA expected Microsoft to accomplish a few tasks in the Windows Firewall program, it did not happen, so they are pulling support as Microsoft decided to go a different direction in Vista. I have beta 2 on the way so there will be some test runs with the new driver set to see what happens. I will have a full statement on this subject by the end of the week. Apparently, there were several promises made, not kept, or not communicated properly about 18 months ago between the two parties about the Firewall program, OS hooks, and its future.
mbf - Monday, June 12, 2006 - link
Is the nVidia firewall truly gone? I've just downloaded (but not installed) the 9.34 driver pack from the nVidia FTP site. The pack still contains a NAM setup (v60.16).What exactly constitutes the ActiveArmour part of the firewall? The hardware firewall was introduced with the nForce3 250gb chipset and the AA functionality was added in nForce4. I've never figured out what precisely AA does, though.
If it's true that nVidia has pulled the firewall from the feature set I'd say it's a truly bad choice. For me, the hardware firewall was one of *the* most important features on the nForce chipset. Well, not having to wonder losing this feature will make my switch back to an Intel chipset-based Core 2 rig much easier. It's a shame.
Stele - Wednesday, May 24, 2006 - link
I was actually wondering if XP and Vista's firewalls would even be supported, since they're arguably 'third-party' from nVidia's viewpoint. While the argument for using hardware-based solutions in routers etc is cogent, imho there is a case for protecting individual PCs against compromise from within the network. For example, there may be laptops in the LAN that are used on several other (and potentially unsafe) networks as well. Besides that, software firewalls have the advantage of being able to potentially alert users to and stop malware from making unauthorised outbound connections from an infected PC (or unauthorised inbound connections from an infected LAN peer). I wonder if it's possible to get around the issue in future versions of firewall software - then at least it's not a lost cause, just that we would have to wait for newer versions of the software to be released.As for the actual usefulness of the TCP/IP offload engine, perhaps the folks at Anandtech can design a specific test that would stress such an engine's capability to the limit, to provide clear and objective assessment of its effect on performance. After all, Anandtech developed a good custom test suite for server benchmarks that targets specific application types, so I figure this would be well within their considerable programming skills as well ;)
Meanwhile, I wonder how nVidia's SATA controllers have improved if at all over the generations, and it would be great to eventually see an ULi 1575/SB600/MCP comparison in the future.
nullpointerus - Wednesday, May 24, 2006 - link
Correction:You don't really need a firewall on your LAN clients...
Gigahertz19 - Wednesday, May 24, 2006 - link
Expect the new kitchen when Conroe is launched :)