ATI’s Radeon X1950 Pro graphics card

SINCE LATE AUGUST, we’ve been first-person witnesses to the fall parade of video cards. Personally, I’ve reviewed so many video cards that I’m having trouble separating a GeForce from a Radeon, let alone a GT from a GS or a Pro from, er, an amateur.

Fortunately, the new product ATI is unveiling today brings with it good news in several forms: the Radeon X1950 Pro is a strong new contender at the value-oriented $199 price point, and it’s based on a brand-new mid-range GPU. ATI hasn’t had the best of luck with mid-range graphics processors, but it looks like that’s about to change. What’s more, this new graphics processor at long last incorporates CrossFire capability directly into the GPU. Gone are the external dongles and proprietary CrossFire Edition graphics cards, replaced by simple, SLI-like bridge connectors between the cards.

Sounds nice, doesn’t it? But is the Radeon X1950 Pro’s formula sufficient to challenge the excellent GeForce 7900 GS at $199? Let’s take a look.

ATI’s new middle manager: the RV570
I’m going to be a huge geek and start off by talking about the new GPU. I could kick things off with a discussion of the product itself, but that would totally appeal to a broader audience and spoil the challenge for our ad sales guy. Can’t have that.


The RV570 GPU

The GPU that powers the Radeon X1950 Pro is code-named RV570. The RV570 is a true mid-range part, with quite a bit more power than the RV530 GPU that drove the ill-fated Radeon X1600 XT. The RV570 shares its technological DNA with the rest of ATI’s R500-series GPUs, but it has its own mix of on-chip resources, including eight vertex shader units, 36 pixel shader processors, and 12 texture address units. Given the design of the X1000-series GPUs, that means it also has 12 Z-compare units, 12 render back-ends, and can manage a maximum of 384 concurrent threads. There will be a quiz on these numbers after class.

TSMC manufactures this chip using its 80nm fab process—a “half-node” process that’s an incremental improvement over the more common 90nm node. This slight process shrink ought to help the RV570 to run slightly cooler and perhaps hit higher clock speeds than its siblings in the Radeon X1000 family, all of which are 90nm chips.

ATI estimates the RV570’s transistor count at 330 million. The chip’s dimensions are roughly 16.7 mm by 13.8 mm, for a total die area of about 230 mm2. Compare that to the Nvidia G71 GPU on the GeForce 7900 GS. Although methods of counting transistors vary, Nvidia says the G71 has 278 million transistors, and at 90nm, the G71 is about 196 mm2. Of course, the G71 is Nvidia’s high-end GPU, hobbled slightly for use in the 7900 GS, while the RV570 is brand-new silicon aimed at the middle of the market.

In spite of its middle-class pedigree, the RV570 still gets membership in the Radeon X1950 club. That’s a departure from the past, when the GPU powering the video card determined its model number. ATI says performance, not silicon, now determines its naming scheme. That means it’s possible we might see a future Radeon card in, say, the X1650 series based on this exact same GPU.

The Radeon X1950 Pro
Whatever you call it, the Radeon X1950 looks pretty suave, with the same red transparent cooler motif present on the rest of the X1950 lineup, only this time in a sleek, quiet, single-slot cooler.


A pair of Radeon X1950 Pro cards

Onboard this card you’ll find an RV570 GPU clocked at 575MHz and 256MB of GDDR3 memory running at 690MHz (1380MHz effective, thanks to DDR black magic.) Sticking out of the PCI slot cover is a TV-out port and a pair of dual-link DVI ports. Those DVI ports have full support for HDCP in order to protect you from the movie industry.

Err, wait. Other way ’round.

Anyhow, as you’d expect, the X1950 Pro drops into a PCI Express x16 slot and has a six-pin auxiliary power connector to keep the GPU cranking.

ATI says these puppies should be on the virtual shelves at online stores starting today for $199, and unlike with some past mid-range Radeon products, ATI will not be relying solely on its partners to get the boards out there. There will be “built by ATI” versions of the X1950 Pro available, as well.

For those of you tracking GTs versus XTXs versus Pros at home, the Radeon X1950 Pro is indeed a replacement for the Radeon X1900 GT. The X1900 GT is based on an R580 GPU with portions deactivated, but its basic capabilities work out to almost exactly the same as the X1950 Pro: 36 pixel shaders at 575MHz, albeit with slightly slower memory. One of the X1900 GT’s weaknesses was the lack of a matching CrossFire Edition card for it. The GT could pass CrossFire data via a PCIe link, but at the risk of reduced performance. The X1950 Pro solves that problem, and the X1900 GT will be phased out as the X1950 Pro takes its place.

 

CrossFire internalizes, goes native
Nearly a year since the debut of its CrossFire multi-GPU scheme, the red team has finally integrated CrossFire’s image transfer and compositing capabilities directly into a graphics processor. Before now, the high-performance implementations of CrossFire required the use of a specialized CrossFire Edition video card that came with an FPGA chip onboard to handle image compositing. Getting data from the other Radeon card to the CrossFire Edition required the use of an external dongle cable that hung out of the back of the PC like a hemorrhoid. Thanks to Preparation RV570, that’s no longer necessary.

Any Radeon X1950 Pro can talk to a peer via the two “golden fingers” connectors on the top of the card.


Each Radeon X1950 Pro has two CrossFire connector, uh, connectors

Look familiar? These are similar to Nvidia’s SLI connections, but they’re neither physically nor electrically compatible with an SLI bridge. (The golden fingers thingy is wider, for one.)


CrossFire connector (left) and SLI bridge (right)

Here’s a look at SLI and CrossFire bridges side by side. Right now, SLI bridges come with SLI-ready motherboards, but ATI has a different plan. They will provide one CrossFire bridge with each Radeon X1950 Pro, so users with older motherboards won’t be stuck without one.

Oddly enough, native CrossFire requires two connections between cards in order to work properly. Each of these connections is a 12-bit link, and native CrossFire will scale up to 2650×1600 at 60Hz via a dual-link arrangement. (I’m fairly certain SLI scales to that same resolution using just a single bridge connector.) ATI says CrossFire could work with just one of these two connectors between cards, but that its graphics drivers currently enforce a dual-link config. In fact, when first setting up our Radeon X1950 Pro CrossFire test config, I attached just one connector, and the system refused to go into CrossFire mode.


Gotta have two links to make it go

The presence of two CrossFire bridges naturally raises some suspicions in the curious mind. This is an all-new thing for ATI, and surely they’ve put some thought into it. Why complicate a dual-card setup unnecessarily by requiring the installation of two physical bridges? Could the excess capacity be there for future three- and four-card configurations? The two, staggered connectors per card could work well in a sort of daisy-chain of graphics cards, if need be.

Hmmmm.

Can’t wait to see what they do with this one. Potentially, any model of Radeon with native CrossFire could be put into a serially connected team of three or four cards by interleaving the connectors used. With DirectX 9’s three buffer limit (which I explain in my really cool Quad SLI article,) a triple-card CrossFire rig might be the sweet spot for performance. Mobos with three PCIe x16 slots are already beginning to show up in various quarters.

 

Test notes
We did run into a few snags in our testing, although none of them affected the Radeon X1950 Pro. Most notably, when we tried to run a pair of GeForce 7600 GT cards in SLI, we encountered some odd image artifacts that we couldn’t make go away. The image artifacts didn’t appear to affect performance, so we’ve included results for the GeForce 7600 GT in SLI. If we find a resolution for the problem and performance changes, we’ll update the scores in this article.

Also, the 3DMark06 test results for the Radeon X1950 XTX CrossFire system were obtained using an Asus P5W DH motherboard, for reasons explained here. Otherwise, we used the test systems as described below.

Our testing methods
As ever, we did our best to deliver clean benchmark numbers. Tests were run at least three times, and the results were averaged.

Our test systems were configured like so:

Processor Core 2 Extreme X6800 2.93GHz Core 2 Extreme X6800 2.93GHz Core 2 Extreme X6800 2.93GHz
System bus 1066MHz (266MHz quad-pumped) 1066MHz (266MHz quad-pumped) 1066MHz (266MHz quad-pumped)
Motherboard Asus P5N32-SLI Deluxe Intel D975XBX Asus P5W DH
BIOS revision 0204 BX97510J.86A.1073.2006.0427.1210 0801
0305
North bridge nForce4 SLI X16 Intel Edition 975X MCH 975X MCH
South bridge nForce4 MCP ICH7R ICH7R
Chipset drivers ForceWare 6.86 INF Update 7.2.2.1007
Intel Matrix Storage Manager 5.5.0.1035
INF Update 7.2.2.1007
Intel Matrix Storage Manager 5.5.0.1035
Memory size 2GB (2 DIMMs) 2GB (2 DIMMs) 2GB (2 DIMMs)
Memory type Corsair TWIN2X2048-8500C5 DDR2 SDRAM at 800MHz Corsair TWIN2X2048-8500C5 DDR2 SDRAM at 800MHz Corsair TWIN2X2048-8500C5 DDR2 SDRAM at 800MHz
CAS latency (CL) 4 4 4
RAS to CAS delay (tRCD) 4 4 4
RAS precharge (tRP) 4 4 4
Cycle time (tRAS) 15 15 15
Hard drive Maxtor DiamondMax 10 250GB SATA 150 Maxtor DiamondMax 10 250GB SATA 150 Maxtor DiamondMax 10 250GB SATA 150
Audio Integrated nForce4/ALC850 with Realtek 5.10.0.6150 drivers Integrated ICH7R/STAC9221D5 with SigmaTel 5.10.5143.0 drivers Integrated ICH7R/ALC882M with Realtek 5.10.00.5247 drivers
Graphics Radeon X1800 GTO 256MB PCIe
with Catalyst 8.282-060802a-035722C-ATI drivers
Radeon X1900 XTX 512MB PCIe + Radeon X1900 CrossFire
with Catalyst 8.282-060802a-035515C-ATI drivers
Radeon X1900 XT 256MB PCIe + Radeon X1900 CrossFire
with Catalyst 8.282-060802a-035515C-ATI drivers
Radeon X1900 GT 256MB PCIe
with Catalyst 8.282-060802a-035722C-ATI drivers
Radeon X1950 XTX 512MB PCIe + Radeon X1950 CrossFire
with Catalyst 8.282-060802a-03584E-ATI drivers
Dual Radeon X1950 Pro 256MB PCIe
with Catalyst 8.291-060822a-036024C-ATI drivers
 
Radeon X1900 XT 256MB PCIe
with Catalyst 8.282-060802a-03584E-ATI drivers
   
Radeon X1950 Pro 256MB PCIe
with Catalyst 8.291-060822a-036024C-ATI drivers
   
Radeon X1900 XTX 512MB PCIe
with Catalyst 8.282-060802a-03584E-ATI drivers
Radeon X1950 XTX 512MB PCIe
with Catalyst 8.282-060802a-03584E-ATI drivers
   
BFG GeForce 7600 GT OC 256MB PCIe
with ForceWare 91.47 drivers
   
Dual BFG GeForce 7600 GT OC 256MB PCIe
with ForceWare 91.47 drivers
   
XFX GeForce 7900 GS 480M Extreme 256MB PCIe
with ForceWare 91.47 drivers
   
Dual XFX GeForce 7900 GS 480M Extreme 256MB PCIe
with ForceWare 91.47 drivers
   
GeForce 7900 GT 256MB PCIe
with ForceWare 91.31 drivers
   
Dual GeForce 7900 GT 256MB PCIe
with ForceWare 91.31 drivers
   
XFX GeForce 7950 GT 570M Extreme 512MB PCIe
with ForceWare 91.47 drivers
   
Dual XFX GeForce 7950 GT 570M Extreme 512MB PCIe
with ForceWare 91.47 drivers
   
GeForce 7900 GTX 512MB PCIe
with ForceWare 91.31 drivers
   
Dual GeForce 7900 GTX 512MB PCIe
with ForceWare 91.31 drivers
   
GeForce 7950 GX2 1GB PCIe
with ForceWare 91.31 drivers
   
Dual GeForce 7950 GX2 1GB PCIe
with ForceWare 91.47 drivers
OS Windows XP Professional (32-bit)
OS updates Service Pack 2, DirectX 9.0c update (August 2006)

Thanks to Corsair for providing us with memory for our testing. Their quality, service, and support are easily superior to no-name DIMMs.

Our test systems were powered by OCZ GameXStream 700W power supply units. Thanks to OCZ for providing these units for our use in testing.

Unless otherwise specified, image quality settings for the graphics cards were left at the control panel defaults.

The test systems’ Windows desktops were set at 1280×960 in 32-bit color at an 85Hz screen refresh rate. Vertical refresh sync (vsync) was disabled for all tests.

We used the following versions of our test applications:

The tests and methods we employ are generally publicly available and reproducible. If you have questions about our methods, hit our forums to talk with us about them.

 

Pixel-filling power
The table below shows a few key specs for the Radeon X1950 Pro and a number of current graphics cards. By no means is this table a complete representation of a graphics card’s capabilities, though. It’s simply a look at the theoretical peak pixel painting and texturing power of a card—simple stuff that doesn’t include pixel shader math—alongside memory bandwidth.

  Core
clock
(MHz)
Pixels/
clock
Peak
fill rate
(Mpixels/s)
Textures/
clock
Peak
fill rate
(Mtexels/s)
Effective
memory
clock (MHz)
Memory
bus width
(bits)
Peak memory
bandwidth
(GB/s)
Radeon X1650 Pro 600 4 2400 4 2400 1400 128 22.4
GeForce 7600 GT 560 8 4480 12 6720 1400 128 22.4
All-In-Wonder X1900 500 16 8000 16 8000 960 256 30.7
Radeon X1800 GTO 500 12 6000 12 6000 1000 256 32.0
GeForce 7800 GT 400 16 6400 20 8000 1000 256 32.0
Radeon X1800 XL 500 16 8000 16 8000 1000 256 32.0
GeForce 7800 GTX 430 16 6880 24 10320 1200 256 38.4
Radeon X1900 GT 575 12 6900 12 6900 1200 256 38.4
GeForce 7900 GS 450 16 7200 20 9000 1320 256 42.2
GeForce 7900 GT 450 16 7200 24 10800 1320 256 42.2
Radeon X1950 Pro 575 12 6900 12 6900 1380 256 44.2
XFX GeForce 7900 GS 480M 480 16 7680 20 9600 1400 256 44.8
GeForce 7950 GT 550 16 8800 24 13200 1400 256 44.8
Radeon X1900 XT 625 16 10000 16 10000 1450 256 46.4
XFX GeForce 7950 GT 570M 570 16 9120 24 13680 1460 256 46.7
Radeon X1800 XT 625 16 10000 16 10000 1500 256 48.0
Radeon X1900 XTX 650 16 10400 16 10400 1550 256 49.6
GeForce 7900 GTX 650 16 10400 24 15600 1600 256 51.2
GeForce 7800 GTX 512 550 16 8800 24 13200 1700 256 54.4
Radeon X1950 XTX 650 16 10400 16 10400 2000 256 64.0
GeForce 7950 GX2 2 * 500 32 16000 48 24000 1200 2 * 256 76.8

When it built the Radeon X1000 series, ATI decided to concentrate on pixel shader power rather than raw fill rate or texturing capability—hence the 12 texture units and 36 pixel shader processors on the RV570. As a result, the X1950 Pro’s peak pixel and texel fill rates are lower than its most direct competitor, the GeForce 7900 GS. The X1950 Pro is very competitive in terms of memory bandwidth, though, and that probably matters most.

The quick synthetic fill rate benchmarks below will test how closely the theoretical peaks match up with what the cards can deliver.

The X1950 Pro performs about as expected here. These numbers won’t tell us too much about what kind of performance to expect in today’s shader-laden games, though, so let’s move on to the games themselves.

 

Quake 4
In order to make sure we pushed the video cards as hard as possible, we enabled Quake 4’s multiprocessor support before testing.

Quake 4 is kind of rough territory for the Radeon X1950 Pro. Its additional memory bandwidth doesn’t help it outperform its predecessor, the Radeon X1900 GT, and the GeForce 7900 GS is well ahead of both of them. The X1950 Pro does scale up fairly well to two cards in CrossFire, though, for what it’s worth.

Notice something else here, too. The Radeon X1900 XT 256MB performs quite a bit better than the X1950 Pro, and it’s selling for only 60 bucks or so more than the X1950 Pro’s list price at some vendors. You may want to keep an eye on how the X1900 XT 256MB compares to the X1950 Pro in the following pages.

 

F.E.A.R.
We’ve used FRAPS to play through a sequence in F.E.A.R. in the past, but this time around, we’re using the game’s built-in “test settings” benchmark for a quick, repeatable comparison.

Things look pretty good for the Radeon X1950 Pro at first glance here, but upon further inspection, the picture isn’t so rosy. Although the X1950 Pro’s average frame rates are higher than those of the GeForce 7900 GS, the median low frame rates of the GeForce card are higher—which generally translates into smoother gameplay. And CrossFire is a disaster here for all of the Radeons.

 

Half-Life 2: Episode One
The Source game engine uses an integer data format for its high-dynamic-range rendering, which allows all of these cards to combine HDR rendering with 4X antialiasing.

What Quake 4 giveth, Half-Life 2 taketh away. The GeForce 7900 GS ends up being slower here by margins similar to those it won by in Quake 4.

 
The Elder Scrolls IV: Oblivion
We tested Oblivion by manually playing through a specific point in the game five times while recording frame rates using the FRAPS utility. Each gameplay sequence lasted 60 seconds. This method has the advantage of simulating real gameplay quite closely, but it comes at the expense of precise repeatability. We believe five sample sessions are sufficient to get reasonably consistent and trustworthy results. In addition to average frame rates, we’ve included the low frames rates, because those tend to reflect the user experience in performance-critical situations. In order to diminish the effect of outliers, we’ve reported the median of the five low frame rates we encountered.

We set Oblivion’s graphical quality settings to “Ultra High.” The screen resolution was set to 1600×1200 resolution, with HDR lighting enabled. 16X anisotropic filtering was forced on via the cards’ driver control panels.

The Radeon X1950 Pro just edges out the GeForce 7900 GS here, but these aren’t really meaningful performance differences between the cards. I do think the Radeon does a better job of playing Oblivion for two reasons, though. First, ATI’s default texture filtering is simply superior to Nvidia’s right now, and it shows up in the cobblestone streets of the Imperial City. Second, when running through an area of the city for the first time, the GeForce cards tended to slow down intermittently in order to load up new texture or the like. The ATI cards were much smoother in the same circumstance.

Ghost Recon Advanced Warfighter
We tested GRAW with FRAPS, as well. We cranked up all of the quality settings for this game, with the exception of antialiasing. However, GRAW doesn’t allow cards with 256MB of memory to run with its highest texture quality setting, so those cards were all running at the game’s “Medium” texture quality.

Again, the X1950 Pro comes out ahead, but it’s pretty close. Native CrossFire looks to scale just as well as the dongle-bound version, though, and as well as SLI.

 

3DMark06

The Radeon X1950 Pro seems to have an edge on the GeForce 7900 GS in Direct3D games generally, and 3DMark06 bears that out. That edge is exaggerated when we move to dual-GPU configs, where the performance gaps are multiplied.

We find nothing unexpected in 3DMark’s additional synthetic tests. Since both have 36 pixel shaders and eight vertex shaders at 575MHz, the Radeons X1900 GT and X1950 Pro perform almost identically here, even though they are based on two different GPUs.

 

Power consumption
We measured total system power consumption at the wall socket using an Extech power analyzer model 380803. The monitor was plugged into a separate outlet, so its power draw was not part of our measurement. We tested all of the video cards using the Asus P5N32-SLI SE Deluxe motherboard, save for the CrossFire system, which required a different chipset. For that system, we used an Intel D975XBX motherboard.

The idle measurements were taken at the Windows desktop. The cards were tested under load running Oblivion using the game’s Ultra Quality setting at 1600×1200 resolution with 16X anisotropic filtering.

With this smaller 80nm GPU, ATI has trimmed peak power use considerably. The Radeon X1950 Pro draws about 20 fewer Watts when running a game than the Radeon X1900 GT, despite performing slightly better. The GeForce 7900 GS still uses less power than the X1950 Pro, but only by about 15W under load. Also, thanks to the lower power draw of the Intel 975X chipset, the Radeon X1950 Pro CrossFire system matches the power use of the GeForce 7900 GS SLI rig.

Noise levels and cooling
We measured noise levels on our test systems, sitting on an open test bench, using an Extech model 407727 digital sound level meter. The meter was mounted on a tripod approximately 14″ from the test system at a height even with the top of the video card. The meter was aimed at the very center of the test systems’ motherboards, so that no airflow from the CPU or video card coolers passed directly over the meter’s microphone. We used the OSHA-standard weighting and speed for these measurements.

You can think of these noise level measurements much like our system power consumption tests, because the entire systems’ noise levels were measured, including CPU and chipset fans. We had temperature-based fan speed controls enabled on the motherboard, just as we would in a working system. We think that’s a fair method of measuring, since (to give one example) running a pair of cards in SLI may cause the motherboard’s coolers to work harder. The motherboard we used for all single-card and SLI configurations was the Asus P5N32-SLI SE Deluxe, which on our open test bench required an auxiliary chipset cooler. The Asus P5W DH Deluxe motherboard we used for CrossFire testing didn’t require a chipset cooler, so those systems were inherently a little bit quieter. In all cases, we used a Zalman CNPS9500 LED to cool the CPU.

Of course, noise levels will vary greatly in the real world along with the acoustic properties of the PC enclosure used, whether the enclosure provides adequate cooling to avoid a card’s highest fan speeds, placement of the enclosure in the room, and a whole range of other variables. These results should give a reasonably good picture of comparative fan noise, though.

We measured the coolers at idle on the Windows desktop and under load while playing back our Quake 4 nettimedemo. The cards were given plenty of opportunity to heat up while playing back the demo multiple times. Still, in some cases, the coolers did not ramp up to their very highest speeds under load. The Radeon X1800 GTO and Radeon X1900 cards, for instance, could have been louder had they needed to crank up their blowers to top speed. Fortunately, that wasn’t necessary in this case, even after running a game for an extended period of time.

You’ll see two sets of numbers for the GeForce 7950 GT below, one for the XFX cards with their passive cooling and another for the BFG Tech cards, which use the stock Nvidia active cooler. I measured them both for an obvious reason: they were going to produce very different results.

The Radeon X1950 Pro seems nice and quiet to my ears. The cooler is about as beefy as it gets for a single-slot design, with lots of copper fins, and that translates into very low noise levels under load. There’s no great change in fan speeds on the X1950 Pro between idle and fully loaded, though, and that means it’s a little louder than some cards at idle. Frankly, I don’t think that’s something one would notice inside of a PC case.

Overclocking
We try to use the built-in overclocking utilities in the graphics vendors’ drivers to test overclocking most of the time, but that method hasn’t always proven foolproof, especially with ATI cards. The tool that finds the optimal frequency in ATI’s drivers was wildly optimistic with our Radeon X1950 Pro cards, which led to screen corruption and lock-ups during our testing. I manually dialed the clock speeds down until the X1950 Pro was stable at a 594MHz core clock and 776MHz memory. The memory slider in the Catalyst Control Center insisted on 773MHz memory for the CrossFire config, so I tested it with the video RAM running 3MHz slower.

Your mileage may vary, but we didn’t gain much by overclocking the X1950 Pro.

 
Conclusions
Just a month ago, when we reviewed the Radeon X1900 GT, I said that the GT’s performance was “virtually on par with the GeForce 7900 GS” and that “You’d be hard pressed to tell the difference between the two in day-to-day use.” I thought the ATI card had better image quality but suffered from poor CrossFire options. The X1900 GT also drew more power than the GeForce 7900 GS and cost a little more. Now, with the Radeon X1950 Pro, we have the same basic performance and image quality as the Radeon X1900 GT, but it comes with lower power draw, lower noise levels while gaming, and a much improved CrossFire scheme that works almost exactly like SLI. And the Radeon X1950 Pro now matches the GeForce 7900 GS’s $199 price tag.

If forced to choose between the Radeon X1950 Pro and the GeForce 7900 GS, I’d distract you by pointing to the Radeon X1900 XT 256MB selling at Newegg for $259. Forking over the additional 59 bucks is probably worth it, given that card’s outstanding performance. You’ve got to be willing to give up native CrossFire and an adjacent slot in your system, but for most of us, that’s probably worth it.

Pinned down and forced to choose strictly between the Radeon X1950 Pro and the GeForce 7900 GS, I’d most likely pick the Radeon X1950 Pro for use in my own system. Nvidia’s iffy texture filtering becomes really bothersome in games like Oblivion and Guild Wars, and since the X1950 Pro only pulls about 15W more under load than the 7900 GS, why not grab it instead? Also, we’ve been down this road half a dozen times in the past month, but it bears repeating that the Radeon X1000 series has some feature advantages that translate into better image quality than what Nvidia’s G71 can offer, including smarter, more flexible antialiasing and angle-independent anisotropic filtering. The possibility of running two cards in SLI or CrossFire serves to focus more attention on those image quality issues, too, because a dual-GPU config leaves plentiful headroom for advanced filtering and AA.

Still, the choice between the GeForce 7900 GS and the Radeon X1950 Pro is a tough one. It may come down to brand preference or some minor feature difference for many folks. Take your pick of either card, and you won’t go too wrong—unless, of course, you consider buying a DirectX 9 card just weeks or months before the DirectX 10 cards come out “going wrong.”

More generally, I’m glad to see ATI get a mid-range GPU right again. After the still-born Radeon X700 XT and the incredible sojourn of the underperforming Radeon X1600 XT from $249 to under $99 in less than a year, I began to get the impression that the red team didn’t take the meaty part of the enthusiast market seriously. The RV570 is proof that’s not so. Native CrossFire is also a nice step forward, and I’m curious to see what new possibilities it opens up down the road. 

Comments closed
    • Prospero424
    • 13 years ago

    I’ve been down on ATi for a couple of years now, but these Crossfire changes might be just what I’ve waited for to see multi-card setups as useful to someone who doesn’t want to spend over $500 at a time for their video card setup.

    I bought a 7800GT (which I’m very happy with) with the intention of perhaps purchasing a second one 6 months later (this was a year ago next month) for around $200-250 to roughly double my high-res performance. The problem with this plan is that Nvidia ceased production on the 7800GT so fast that it wasn’t and isn’t available in any volume in a time frame that would have let that work. Sure, they brought out the 7900GT, but that won’t work with a 7800GT in SLI, so the prospect of a multi-card setup for me was made very, very unattractive by the high prices and low availability of the 7800GT.

    It looks like Crossfire won’t suffer from this problem, since I would guess, say, an X1950 will be able to work with whatever their next mid-range card will be using the bridge connectors. Sure, you might have some pipelines disabled when doing this with a newer card, but I’m willing to bet you’ll still see a substantial boost over either one by themselves.

    Now, if ATi would just get to serious work on their Linux drivers, they would really, really have something going again. I hope AMD will have a positive impact on their software development in this regard.

      • BaldApe
      • 13 years ago

      I thought the recent updates to nvidias drivers allow for mixed sli configurations…..

    • Severus
    • 13 years ago

    Those DVI ports have full support for HDCP in order to protect you from the movie industry.

    Err, wait. Other way ’round.

    Hehehe.

    • kilkennycat
    • 13 years ago

    Yet another paper-release from ATi. Following the pattern securely established with the X1800 series. No sign of the X1950 Pro on Newegg, even among their “Out-of-Stock ” list, where it would appear if Newegg had already exhausted their first shipment. Just the X1950XTX and the old X1950 “Crossfire Edition” are present in the X1950 category.

    BTW, on a related subject, beware of buying the X1900GT. ATi has allegedly changed the clock speed from 575MHz to 512MHz and has attempted to compensate by increasing the memory speed — without changing the Product Nomenclature! Before the X1900GT and other products using the R580 GPU are obsoleted, ATi seems to be trying to quietly use up a stock of R580 chips that could not be binned at 575MHz — a profit -maximizing move by ATi, no doubt. The R580 is a very large die and expensive to fabricate and test. For more details on the subject and comments on this move by ATi, see page 2 of today’s Anandtech (otherwise very favorable) report on the X1950 Pro:-

    §[<http://www.anandtech.com/video/showdoc.aspx?i=2858<]§

    • DrDillyBar
    • 13 years ago

    So when will we see a comparison of X19xx folding skill? Points in 24-48 hours or something as an additional benchmark?

    • Flying Fox
    • 13 years ago

    I wonder which partner is going to bold enough to come out with a passively cooled version? Or is that even possible? 🙄

      • madgun
      • 13 years ago

      you might expect that from ASUS..

        • Flying Fox
        • 13 years ago

        That means I need to give up a kidney for that. 🙁

        • CampinCarl
        • 13 years ago

        Or, probably more likely, Gigabyte.

      • Flying Fox
      • 13 years ago

      Damn, just saw the releases from Asus and Gigabyte. Gigabyte slaps a VF700 on while Asus has its own cooler too. I wonder if they will release “silencer” editions later…

      Or, I could just get a reference and slap an aftermarket solution that uses standard fans?

    • nightmorph
    • 13 years ago

    Hmm, looks like the hardwarezone.com article is a little better informed about some of the technology in this card, and what it’s used for:

    §[<http://www.hardwarezone.com/articles/view.php?cid=3&id=2090<]§ See the section on the bridge connection in particular. I was suprised; normally, TR reviews don't have so many guesses in them -- TR is always known for having the definitive answers, reviews, etc. But I think I liked the hardwarezone review a little better -- interesting that Powercolor's version uses a quiet aftermarket cooler from the get-go (AC Accelero 2), I'd like to see this pitted against the quiet single-slot version here in TR's review.

    • marvelous
    • 13 years ago

    ATI changed the specs on us again?

    1950 was to be clocked higher than 1900gt.. Now the core clock is same and just memory clocked higher and doesn’t overclock worth a damn even on a smaller die… 🙁

    Nothing a raised voltage on the core couldn’t cure…

    I think this is little 2 late… If it came out like year ago this would have been a good buy… Not anymore..

      • BaldApe
      • 13 years ago

      I’m surprised they didnt up the core voltage for the review. Given the lower power consumtion, I would think they backed of the core voltage slightly. Even if they didnt, I would still love to see what the new 80nm core can do with some extra juice.

      About to order a new gpu. I’m leaning toward the x1900xt, but if the x1950 pro overclocks well with some extra voltage, I will probebly purchase it instead.

    • Austin
    • 13 years ago

    y[<:oO<]y 7900GS tend to be super o/cable and that should be factored in too; commonly 450mhz to a whopping 600mhz or more is a big increase (> 30%). The X1950Pro seems to be already clocked near its maximum potential, not surprising since the X1900XTX & X1950XT offer only 50mhz more clockspeed @650mhz. This is only a good thing if you don't like to o/c in order to keep your warranty. r[<;o)<]r The X1900XT_256MB gives a fair bit more perf for a little more cash but so do the 7900GTO and 7950GT too. I don't think DX10 is a huge factor as early cards will surely be $500+ monsters and it's likely to be March-April 2007 (ie about 6 months) before we see any $200 DX10 cards.

    • MadManOriginal
    • 13 years ago

    Who needs specific 7900GTO benchmarks? At stock we know it’s faster than a 7950GT, and depending on oc can be identical to a 7900GTX. It’s not all that hard to mentally place it on the charts.

    • MadManOriginal
    • 13 years ago

    I was looking forward to reading about this card, a good mid-range ATi card with HDCP – for future use in an HTCP when it’s no longer decent enough for games – would have been perfect. But I have to say I’m a bit underwhelmed, maybe my expectations got me too hyped up. Yes it’s on par with the 7900gs although it’s ‘only’ on par with it, but the x1900xt 256MB option crops up not to mention the spectacular enthusiast bargain 7900GTO anomoly. But there’s the catch for me – I’ve seen only one HDCP-enabled x1900xt 256MB, an Asus model which automatically means expensive anyway and for some stupidass reason it has a max resolution of only 2048×1536. I haven’t seen any HDCP 7900GTOs. So it looks like my hopes for this card have been pretty well dashed.

    Is it possible ATi’s wierd pipeline:shader mixture doesn’t downscale well? Because honestly just being ‘on par’ with the 7900gs doesn’t cut it, they really need to beat it to gain market and mind-share.

    I’ll tell you what ATi ought to do to get back the hearts of enthusiasts – make their cards unlockable or BIOS-flashable to enable more pipelines like the x800-series days. I understand why they can’t do this on certain models that are made with GPU cores that don’t have the sub-units in the first place but the x1900xt 256MB proves it’s not purely an issue of the production cost of the GPU either. This would drive tremendous mid-range sales to enthusiasts and not hurt the average BestBuy-type market.

    Thanks for the review, *sigh* too bad it’s only ‘just competitive.’

    • jobodaho
    • 13 years ago

    Conclusion page:

    If forced r[

    • toxent
    • 13 years ago

    Nice review Scott. I’d read these things just for the comments you make!

    Keep up the good work!

    • Damage
    • 13 years ago
      • indeego
      • 13 years ago

      Duggity Duggg{<.<}g

      • Proesterchen
      • 13 years ago

      From the review:

      q[

      • DrDillyBar
      • 13 years ago

      Diggity.

    • BoBzeBuilder
    • 13 years ago

    The 7900GTO may be the best deal to ever hit the market, it would be nice to see one of those cards in your review.

    • Fighterpilot
    • 13 years ago

    The new coolers look pretty groovy.Nice performance and IQ.
    Sure looks a good thing at that price…..they’ll sell heaps I expect.

    • flip-mode
    • 13 years ago

    Nice review. Nice balanced conclusion and recommendations. It’s quite a relief to see ATI do something proper at the midrange.

    Those cards look pretty. It’s just too bad one has to turn one’s case upside down to appreciate it.

    And as you already mentioned – too bad this only comes to us now, just before the release of DX10.

    Sadly, with a 2 year old and a 1 month old, a thesis to finish and a full-time job, I’m thinking I might never get to buy another gaming card again. I guesss that’s OK by me though.

    • Ardrid
    • 13 years ago

    Excellent as always. Though if you had to choose btw an X1950 Pro and a 7900 GS, only to be distracted by an X1900 XT 256MB, I would further distract you by mentioning the 7900 GTO 😀

    • Proesterchen
    • 13 years ago

    Nice review, Scott!

    If it wasn’t for the display limitations of Crossfire (it’s like SLi in that regard, single display only, right?) and the impending G80 release, at price of only ~170 Euros according to some shops, I might just have considered changing my current X1800 XT for a dual-RV570 Crossfire rig.

    • Samlind
    • 13 years ago

    And the F@H results are…..????

    j/k

    • CampinCarl
    • 13 years ago

    Good review…

    but

    Anyhow, as you’d expect, *[

Pin It on Pinterest

Share This