je_saist

They will not for two very good reasons.

• $499
• $699

Nvidia is only launching two cards, the GTX 470 at $499, and the GTX 480 at $699. According to the people who have already laid hands on the prototypes, the 470 is slower than the RadeonHD 5870, while costing a full $100 more MSRP.

The GTX 480, while faster than the existing HD 5870 gamers could buy last year, is only 5%-8% faster in "real" games. That's not exactly a huge performance delta... and certainly not worth an extra $300 dollars.

If you ARE into spending $700 on a graphics card, there's the RadeonHD 5970 4gb edition cards hitting the market: http://www.xbitlabs.com/news/video/d...cs_Boards.html

These dual-chip monstrosities on Crossfire supported games, don't just blow the GTX 480 out of the water. They'll do the equivalent of sending a ship parked in Pearl Harbor to Ohio State University by air mail.

From what we know, Nvidia's "mid-range" cards, the stuff to compete against the Radeon 57xx line-up, won't be landing until September or October at the earliest. Until, or unless, Nvidia gets their act together, price drops on existing RadeonHD 5000 series cards probably aren't in the cards. AMD and their vendors have no reason to lesson profits on what are the best (and only choices) for the gaming customer right now.

That being said, if AMD keeps on track for the RadeonHD 5000 series product refresh in June / July, we could see some price-cuts on the existing cards just to clear them out of the market.

Also, don't be too confident that OpenGL 3.2 / 3.3 / 4.0 support doesn't mean anything for CoH users. Television already has commented that they (the coders) are doing stuff in OpenGL that they apparently aren't ready to show yet. Given the timing of Television's comments and the release of the OpenGL 4.0 spec, there is the possibility that Paragon Studio's is planning to demonstrate Ultra Mode running against OpenGL 4.0 drivers from AMD for Going Rogue's July launch, which fits in with the expected date on AMD's RadeonHD 5000 product refresh.. and is enough time for AMD to get their OpenGL 4.0 drivers tested and ready for release.

Doh. Now it makes sense. BTX chassis. Yeah. That was Intel's attempt to "build a better mousetrap." Personally, Intel should just stay out of chassis development.

em. No. the next tier down from the HD 5850 is the HD 5830, then the HD 57xx series of cards, the 5750 and 5770.

The HD 4850 is a (generation) back.

Okay. the difference between the 4850 and the 5850 is an order of magnitude on performance, and differences in features. The RadeonHD 5000 series supports DirectX 11 / OpenGL 3.2 and OpenGL 3.3.

The 4850's performance comparative parts in the current AMD line-up are the HD 5750 and HD 5770... but each of these generally costs a bit more than the old 4850.

Now, I can tell you that you don't want the "single slot" Gigabyte 4850's. I put some together for a client and his computer is even louder at full tilt than my own Dual-slot Sapphire 4850 crossfire system.

Now, as to whether or not the 4850's will run ultra mode? Well. It is the "starting point" for Ultra mode. Key-words: starting point

Until the NDA is lifted, or I figure out how to give performance figures without giving performance figures, that's pretty much all I can say on how the 4850 performs in Ultra Mode.

interesting. I remembered the "monsters" using the same scale as GM's, but with a constant level con of one level lower than GM spawn.

I went into this in another post... here: http://boards.cityofheroes.com/showp...&postcount=371

Sadly, the HD 4670 is the most powerful AGP card that can currently be bought on the retail market. Now, while I can't tell you how the card does, or more preciously does not perform in Ultra Mode without torquing Public Relations off, it is quite powerful enough to run the non-Ultra Mode options at maximum quality, and in relatively high resolutions.

Ack! You mis-pronounced that! You've summoned Jack Emmert to be lead designer again!

I dunno. I've got one of the Agena 9600's, and I've never had a problem with L3 cache.

You can read about the TLB issue here: http://www.legitreviews.com/article/618/1/ :: http://www.techreport.com/discussions.x/13724

***



You might note that the processor isn't actually coupled against a Socket AM2+ motherboard though. It's coupled against the older (and slower) Socket AM2 Asus M2R32-MVP. The reason is actually pretty simple. Of the Socket AM2+ motherboards I've bought, they've all been bought for SLI setups and used Nvidia chipsets that will not run the Phenom 9600 in dual-channel mode. So far there hasn't been a Socket AM2+ Crossfire motherboard that's matched the price / capabilities of the M2R32-MVP. I got it for $120 back in '07, around 6 or so months after it's 2006 launch. It's only been recently that AM2+ boards like this MSI CrossFire capable board have been dipping into the ~$100 market with equivalent features. Granted, I wouldn't buy it because it's an MSI and every MSI board I've had comes with a BIOS that makes Intel's look good by comparison. Anyways, currently, of the boards listed on Newegg, none of the AMD chipset boards explicitly say they'll do x16/x16 in Crossfire, with most stating they'll do x8/x8. That's a step back compared to a board that's over 3 years old.

Of the boards Newegg has that return searches of x16/x16, on the Crossfire side there's a DFI, which is normally good. Only the last DFI I bought was the LANPARTY JR X58-T3H6... and it won't run I7's in Triple Channel mode.

Then there's the ASUS M4A79T-D. Great motherboard. I put together one for a client. Only.. come one. $180? No. Just. No. That's Intel motherboard pricing.

Anyways, because of my lack of willing to pay Intel-ish prices for AMD motherboards that have the feature set of a 3+ year old motherboard, might be a factor on why I haven't seen the TLB issues that were reported to be in the Agena 9600 B2 revision.

...

any chance you could take a picture and show us what you mean by this? I'm as confused as Father_Xmas.

Em. No. Not even close.

From what the developers said at Hero-Con, although it was subject to change, the existing Hero and Villain tutorials were not being changed. So far there has been no indication, either by a red-name post, nor by an off-the-record comment, that the existing tutorials are going to be changed or modified.

A completely NEW tutorial system was being created for Praetoria and Going Rogue, taking into account all of the mistakes the developers made with the Hero and Villain versions, as well as taking into account all of the gameplay changes since those tutorials were made. One of the case examples given was the IO system, which is "tutorialed" in the Universities of Cap and Steel Canyon. With Praetoria, this fundamental system will likely be tutorialed earlier, or will be a bit more extensive to better cover how IO's work.

The developers also flat out stated that the 10 base Hero and Villain archtypes would be able to choose from the new Praetoria tutorial, or the older Hero / Villain tutorials. The Hero and Epic Villain archtypes will have to go through the original Hero / Villain arcs, but will be able to swap morality alignments through the Praetorian alignment change.

The developers also flat out stated that those who chose to go through the Praetoria tutorial could elect to become a Hero without ever having a "Villain / Rogue" assignment, thus qualifying the 5 base Villain archtypes for the "I started a Hero and stayed a Hero" rewards. The opposite of the 5 base Hero Archtypes being able to go Villain without ever having had a Hero / Vigilante assingment was also clarified during the event.

What we don't know, either because the developers did not say, or it just hadn't been decided, is how the Praetoria Tutorial will interact with the Loyalist and Resistance Storylines. We do not know if there will be two separate Praetoria tutorials, with one pertaining to the Resistance and one pertaining to the Loyalists. We do not know if there is just one tutorial and players will have to select a starting Praetoria arc to work through. We don't know if Praetoria arcs will be based on the Hero City style where initial contacts are based on origin, and origin has an influence over what enemies you fight, or if the Praetoria arcs will be based on the Villain City style where you get the same contacts, the same missions, and the same experience over and over and over and over and over and over.

We do know that Praetoria, at minimum, is comprised of enough mission content to take a player from 1 to 20, and it was flat out stated by the developers that at level 20 the players would have to choose to be a Hero or a Villain, and go to the Rogue Isles or go to Paragon City.

This is what caused the confusion on the part of many players that Praetoria was just 1-20 content and that there was nothing else in the expansion, never minding the developers also explicitly stating there would be "morality" arcs in Praetoria to allow players to change sides. Such content that would obviously have to be post-20 in order for existing level 50 characters to swap.

Seriously, none of this stuff is exactly unkown. Almost all of it is found in this existing thread: http://boards.cityofheroes.com/showthread.php?t=196483

Yeah... I don't think I made this clear enough.

AMD Eyefinity, Nvidia 3D Vision Surround, and Matrox triple head are for 3 MONITOR SYSTEMS... not two monitor systems. This thread is about support for 3 or more monitors in an Eyefinity triple screen or Hex screen configuration.

If you do not understand what Eyefinity, 3D Vision Surround, or Triple Head do, please go watch this video on HardOCP: http://www.hardocp.com/article/2009/...hnology_review

***

Okay, with that clarification in place, here's why this thread is not about 2 monitors.

Bezels.

Those little black things that are on the border of your monitor. As I understand the centering of the camera the developers would have to break the rendering aspects in order to move your character off center from the camera position. This is not exactly an easy task to accomplish.

The advantage of triple screen wide, or hex screen (3 wide x 2 high), or enn screen (3 wide x 3high) is that the center focus point can still be positioned in a "middle" monitor. Hex, 3wide x2 high, can be a bit of a pain on FPS, but sometimes can be suited to RPG's where the character can be positioned in the middle of the bottom screen.

**

now, in theory, if Television and crew were to implement changes to the underlying engine to support triple, hex, and enn setups, it's possible that they could probably also implement offsets to support 2 wide monitors, 3 monitors wide with one monitor on top, 4monitors wide, or so on.

I really don't know.

I imagine though that supporting 3 screens wide would be an easier task out of the box, and doable this year.

The Walk animation speed is deliberately slow for RP purposes. It's supposed to be a Club Saunter, not a stroll through the park.

What bugs me about Ars's Article is that 3.2 is / was considered to be the "rough" feature parity equivalent of DX11. Most of the "big" DX11 features, such as Tessellation, made it into the 3.2 spec. From what I read on GLSL 4.0 and OpenGL 4.0, the changes to the spec go beyond what DX11 offers. Although, this wouldn't be the first time I've disagreed with Ars on their reporting as they've gotten basic history facts wrong.

****

Well, giving the timing of the 4.0 release and a certain developers Blue-Screen hijinx, it's possible we could see Apple, AMD/ATi, and Valve announcing Portal 2 as the first OpenGL 4.0 title at this years E3. Given Television's comments, part of me wonders if this game (our game) also has a 4.0 rendering mode in the works.

Presuming this is correct... We may see the first cards from ATi with "full" OpenGL 4.0 support in the already known RadeonHD 5000 series refresh later this year.

***

There's also another question about how much of OpenGL 4.0 is gamer-applicable. As I said in the first post, OpenGL 3.3 was an expected point release, but an OpenGL 3.x ES version was also expected. One of the strengths on which OpenGL has made a comeback is the OpenGL 2.0 ES release, which basically does what 3DFX did with Glide several years ago: limit the specifications to those that are applicable to one aspect of OpenGL.

It's very possible then that an OpenGL 4.0 ES specification might wind up being feature-equivalent with DX11.