Logo   Banner   TopRight
Sapphire Toxic Radeon X1900 XTX
Date: Apr 10, 2006
Author: Marco Chiappetta
Intro, Specifications & Bundle

**Update: Sapphire re-branded this product the Toxic Radon X1900 XTX after our review was published.

Sapphire has been one of ATI's premier board partners for quite some time now. Their products tend to be very similar to those available directly from ATI themselves, but Sapphire does take additional steps to differentiate their offerings with unique accessory and software bundles, and occasionally unique cooling solutions. In the past, we've reviewed a couple of Sapphire's passively cooled "Ultimate Edition" cards and have also evaluated a few products in their "Toxic" family of products that feature oversized air coolers. Today though, we're going to take a look at a Sapphire's first liquid-cooled card, the Toxic Radeon X1900 XTX.

As its name implies, the Sapphire Toxic Radeon X1900 XTX is based upon ATI's latest flagship GPU, but instead of featuring a dual-slot heatsinks / fan assembly like ATI's cards, the Toxic is equipped with a slick, self-contained liquid-cooler that keeps the GPU cool without generating much noise. Sapphire also plans to ship the card pre-overclocked, which will give it a performance advantage over any standard Radeon X1900 XTX that strictly adheres to ATI's reference specifications. Check it out...

Sapphire Toxic Radeon X1900 XTX
Features & Specifications
Features - ATI Radeon X1900
. 380+ million transistors on a 90nm fabrication process
. Ultra-threaded architecture with fast dynamic branching
. Forty-Eight pixel shader processors
. Eight vertex shader processors
. 256-bit 8-channel GDDR3/GDDR4 memory interface
. Native PCI Express x16 bus interface
. Dynamic Voltage Control

Ring Bus Memory Controller
. 512-bit internal ring bus for memory reads
. Programmable intelligent arbitration logic
. Fully associative texture, color, and Z/stencil cache designs
. Hierarchical Z-buffer with Early Z test
. Lossless Z Compression (up to 48:1)
. Fast Z-Buffer Clear
. Z/stencil cache optimized for real-time shadow rendering
. Optimized for performance at high display resolutions, including widescreen HDTV resolutions

Ultra-Threaded Shader Engine
. Support for Microsoft DirectX 9.0 Shader Model 3.0 programmable vertex and pixel shaders in hardware
. Full speed 128-bit floating point processing for all shader operations
. Up to 512 simultaneous pixel threads
. Dedicated branch execution units for high performance dynamic branching and flow control
. Dedicated texture address units for improved efficiency
. 3Dc+ texture compression
_o High quality 4:1 compression for normal maps and two-channel data formats
_o High quality 2:1 compression for luminance maps and single-channel data formats
. Multiple Render Target (MRT) support
. Render to vertex buffer support
. Complete feature set also supported in OpenGL 2.0

Avivo Video and Display Engine
. High performance programmable video processor
_o Accelerated MPEG-2, MPEG-4, DivX, WMV9, VC-1, and H.264 decoding (including DVD/HD-DVD/Blu-ray playback), encoding & transcoding
_o DXVA support
_o De-blocking and noise reduction filtering
_o Motion compensation, IDCT, DCT and color space conversion
_o Vector adaptive per-pixel de-interlacing
_o 3:2 pulldown (frame rate conversion)
. Seamless integration of pixel shaders with video in real time
. HDR tone mapping acceleration
_o Maps any input format to 10 bit per channel output
. Flexible display support
_o Dual integrated dual-link DVI transmitters
_o DVI 1.0 / HDMI compliant and HDCP ready
_o Dual integrated 10 bit per channel 400 MHz DACs
_o 16 bit per channel floating point HDR and 10 bit per channel DVI output
_o Programmable piecewise linear gamma correction, color correction, and color space conversion (10 bits per color)
_o Complete, independent color controls and video overlays for each display
_o High quality pre- and post-scaling engines, with underscan support for all outputs
_o Content-adaptive de-flicker filtering for interlaced displays
_o XilleonTM TV encoder for high quality analog output
_o YPrPb component output for direct drive of HDTV displays
_o Spatial/temporal dithering enables 10-bit color quality on 8-bit and 6-bit displays
_o Fast, glitch-free mode switching
_o VGA mode support on all outputs
. Compatible with ATI TV/Video encoder products, including Theater 550
Advanced Image Quality Features
. 64-bit floating point HDR rendering supported throughout the pipeline
_o Includes support for blending and multi-sample anti-aliasing
. 32-bit integer HDR (10:10:10:2) format supported throughout the pipeline
_o Includes support for blending and multi-sample anti-aliasing
. 2x/4x/6x Anti-Aliasing modes
_o Multi-sample algorithm with gamma correction, programmable sparse sample patterns, and centroid sampling
_o New Adaptive Anti-Aliasing feature with Performance and Quality modes
_o Temporal Anti-Aliasing mode
_o Lossless Color Compression (up to 6:1) at all resolutions, including widescreen HDTV resolutions
. 2x/4x/8x/16x Anisotropic Filtering modes
_o Up to 128-tap texture filtering
_o Adaptive algorithm with Performance and Quality options
. High resolution texture support (up to 4k x 4k)

. Multi-GPU technology
. Four modes of operation:
_o Alternate Frame Rendering (maximum performance)
_o Supertiling (optimal load-balancing)
_o Scissor (compatibility)
_o Super AA 8x/10x/12x/14x (maximum image quality)
_o Program compliant

Top View

Side View

Front View



Sapphire's high-end cards have been shipping with a very comprehensive software and accessory bundle as of late. Included with the card itself, we found a basic user's manual, an addendum detailing CrossFire power requirements, a driver CD-ROM, and an assortment of other discs that included copies of Cyberlink's PowerDVD 6, Power Director 4, and Sapphire's "Software Select" CD.  Although it wasn't included in our sample, the Toxic will also ship with a copy of Sapphire's "TRIXX" overclocking utility.  The copy of Cyberlink's Power Director video editing software was included to compliment the card's video in / video out (ViVo) capabilities. In past reviews we've found an older version of Power Director to be quite useful and easy to use (see here).

The Software Select CD is a Sapphire exclusive that gives buyers a choice of their gaming bundle.  Two unlock codes are included with the disk, with which users can install two of four available games, Brothers in Arms (Road to Hill 30), Prince of Persia (Warrior Within), Richard Burns Rally, or Tony Hawk's Underground 2.

There was also a collection of cables and adapters included with Sapphire's Toxic Radeon X1900 XTX. There was a standard 6-ft S-Video cable included, a composite video cable, and two DVI-to-DB15 adapters along with a splitter for connecting the card to an HD-TV's components inputs, and another adapter equipped with S-Video and composite video inputs and outputs. Lastly, there was a molex-to-6-pin PCI Express power cable adapter thrown in as well.

The Card & Cooler

The Sapphire Toxic Radeon X1900 XTX is very similar to an ATI-built Radeon X1900 XTX on some levels. Both cards share the same PCB, GPU, memory and components. But because this card is equipped with a custom liquid-cooling apparatus, that cools only the GPU, the card's RAM chips are adorned with individual heatsinks to help keep temperatures in check.


Other than the differences in the cooling hardware, the Sapphire Toxic Radeon X1900 XTX differs from an ATI-built card only in its core and memory clock speeds. Sapphire plans to ships the Toxic at the end of this month with its Radeon X1900 XTX core clocked to 675MHz and its 512MB of memory clocked at 1.6GHz, increases of 25MHz and 50MHz (DDR) over stock reference specs. Should this change sometime between now and when the card ships, however, we'll update this review and let you know.



The real story in regard to the Sapphire Toxic Radeon X1900 XTX is the card's self-contained liquid cooling apparatus. The unit is built for Sapphire by Thermaltake. It is reminiscent of Thermaltake's Tide Water all-in-one VGA liquid-cooler, but Sapphires solution actually differs from the Tide Water in a couple of ways.  Sapphire's cooler is thinner and has a larger, lighted fan, and it generates a bit more noise than the Tide Water as well.

The cooler's water block and radiator are made from pure Copper for increased heat conductivity and dissipation, and the cooling fan can be run at one of two speeds thanks to a simple switch on the top of the cooler. At the lower setting, the fan spins at 2000RPM and generates 18db of noise. At the higher setting the fan spins at 2500RPM and generates 26dB. We found both modes to be very quiet during testing, and think the cooler will not be audible in a typical system with a closed case. The cooler requires almost no maintenance because it is closed loop, but should you have to clean the unit or ad water at some point in the future there is a cap on the reservoir and a fill port right on the side of the unit.

On top of its performance and quiet operation, we like the fact that there is enough slack in the tubing to install the cooler in a slot that's not adjacent to the graphics card itself.

Our Test System & 3DMark06


HOW WE CONFIGURED THE TEST SYSTEMS: We tested our NVIDIA cards on an Asus A8N32-SLI nForce 4 SLIX16 chipset based motherboard. However, the ATI powered cards were tested on an A8R32-MVP motherboard based on the CrossFire Xpress 3200 chipset. Both systems used the same AMD Athlon 64 FX-55 processor and 1GB of low-latency Corsair XMS RAM, though. The first thing we did when configuring these test systems was enter each BIOS and loaded the "High Performance Defaults."  The hard drive was then formatted, and Windows XP Professional with SP2 was installed. When the installation was complete, we installed the latest chipset drivers available, installed all of the other necessary drivers for the rest of our components, and removed Windows Messenger from the system. Auto-Updating and System Restore were also disabled, the hard drive was defragmented, and a 768MB permanent page file was created on the same partition as the Windows installation. Lastly, we set Windows XP's Visual Effects to "best performance," installed all of the benchmarking software, and ran the tests.

The HotHardware Test Systems
AMD Athlon 64 FX Powered

Processor -

Motherboard -

Video Cards -

Memory -

Audio -

Hard Driv
e -


Hardware Used:
AMD Athlon 64 FX-55 (2.6GHz)

Asus A8N32-SLI
nForce4 SLIX16 chipset

Asus A8R32-MVP
ATI CrossFire Xpress 3200

GeForce 7900 GTX

GeForce 7900 GT
512MB GeForce 7800 GTX
256MB GeForce 7800 GTX
Radeon X1900 XTX
Sapphire Toxic Radeon X1900 XTX

1024MB Corsair XMS PC3200 RAM

Integrated on board

Western Digital "Raptor"

36GB - 10,000RPM - SATA

Operating System -
Chipset Drivers -
DirectX -

Video Drivers

Synthetic (DX) -
DirectX -
DirectX -
DirectX -
DirectX -
OpenGL -
Relevant Software:
Windows XP Professional SP2
nForce Drivers v6.85
DirectX 9.0c (March Redist)

NVIDIA Forceware v84.17

ATI Catalyst v6.2

Benchmarks Used:
3DMark06 v1.0.2
Splinter Cell: Chaos Theory
FarCry v1.33*
Half Life 2*
Quake 4*

* - Custom Test (HH Exclusive demo)

Performance Comparisons with 3DMark06 v1.0.2
Details: http://www.futuremark.com/products/3dmark06/

Futuremark recently launched a brand-new version of their popular benchmark, 3DMark06. The new version of the benchmark is updated in a number of ways, and now includes not only Shader Model 2.0 tests, but Shader Model 3.0 and HDR tests as well. Some of the assets from 3DMark05 have been re-used, but the scenes are now rendered with much more geometric detail and the shader complexity is vastly increased as well. Max shader length in 3DMark05 was 96 instructions, while 3DMark06 ups the number of instructions to 512. 3DMark06 also employs much more lighting, and there is extensive use of soft shadows. With 3DMark06, Futuremark has also updated how the final score is tabulated. In this latest version of the benchmark, SM 2.0 and HDR / SM3.0 tests are weighted and the CPU score is factored into the final tally as well.

The Toxic's increased clock speeds give the card an obvious edge over a standard Radeon X1900 XTX in 3DMark06's default test, but as an added bonus they even gave the card enough of a boost to overtake NVIDIA's new GeForce 7900 GTX. The difference in performance is minimal though, as you can see.

If we break down the 3DMark06 results, we see that NVIDIA's current flagship is the clear victor in the Shader Model 2.0 tests, besting the Toxic by almost 150 points. The Toxic was again marginally faster than the reference Radeon X1900 XTX.

The Sapphire Toxic Radeon X1900 XTX was the dominant performer in 3DMark06's Shader Model 3.0 / HDR tests, which explains its overall victory in 3DMark06 as well. Sapphire's flagship video card outpaced the 7900 GTX by 220 points here, or about 9%.

Splinter Cell: Chaos Theory v1.05

Performance Comparisons with Splinter Cell: Chaos Theory v1.05
Details: http://www.splintercell3.com/us/

SC: Chaos Theory
Based on a heavily modified version of the Unreal Engine, enhanced with a slew of DX9 shaders, lighting and mapping effects, Splinter Cell: Chaos Theory is gorgeous with its very immersive, albeit dark, environment. The game engine has a shader model 3.0 code path that allows the GeForce 6 & 7 Series of cards and the X1000 family of cards to really shine, and a recent patch has implemented a shader model 2.0 path for ATI's X8x0 generation of graphics hardware. For these tests we enabled the SM 3.0 path on all of the cards we tested. However, High Dynamic Range rendering was disabled so that we could test the game with anti-aliasing enabled (a future patch should enable AA with HDR on the X1K family). We benchmarked the game at resolutions of 1,280 x 1024 and 1,600 x 1,200, both with and without anti-aliasing and anisotropic filtering.


Without any additional pixel processing used, the GeForce 7900 GTX held onto small leads at both resolutions, but the scales tip in favor of the Radeon X1900 XTXs, or more specifically the Toxic Radeon X1900 XTX, when AA and Anisotropic filtering are enabled. Once again, the Toxic's increased clock speeds give it enough of an edge to outpace the 7900 GTX at 1280x1024 4X/16X, where the reference X1900 XTX could not.

FarCry v1.33

Performance Comparisons with FarCry v1.33
Details: http://www.farcry.ubi.com/

If you've been on top of the gaming scene for some time, you probably know that FarCry was one of the most visually impressive games to be released on the PC last year. Courtesy of its proprietary engine, dubbed "CryEngine" by its developers, FarCry's game-play is enhanced by Polybump mapping, advanced environment physics, destructible terrain, dynamic lighting, motion-captured animation, and surround sound. Before titles such as Half-Life 2 and Doom 3 hit the scene, FarCry gave us a taste of what was to come in next-generation 3D gaming on the PC. We benchmarked the graphics cards in this article with a custom-recorded demo run taken in the "Catacombs" area checkpoint, at various resolutions without anti-aliasing or anisotropic filtering enabled, and then again with 4X AA and 16X aniso enabled concurrently.


Our custom FarCry benchmark told us essentially the same story as the Splinter Cell benchmark on the previous page. Without any anti-aliasing or anisotropic filtering used, the GeForce 7900 GTX was the fastest card of the bunch by a couple of frame per second. However, when anti-aliasing and anisotropic filtering were switched on, Sapphire's Toxic Radeon X1900 XTX comes screaming back to the pole position. The reference X1900 XTX though, fell short of the mark set by 7900 GTX at 1280x1024 when AA and aniso were enabled.

Half Life 2

Performance Comparisons with Half-Life 2
Details: http://www.half-life2.com/

Half Life 2
Thanks to the dedication of hardcore PC gamers and a huge mod-community, the original Half-Life became one of the most successful first person shooters of all time.  So, when Valve announced Half-Life 2 was close to completion in mid-2003, gamers the world over sat in eager anticipation. Unfortunately, thanks to a compromised internal network, the theft of a portion of the game's source code, and a tumultuous relationship with the game's distributor, Vivendi Universal, we all had to wait until November '04 to get our hands on this classic. We benchmarked Half-Life 2 with a long, custom-recorded timedemo in the "Canals" map, that takes us through both outdoor and indoor environments. These tests were run at resolutions of 1,280 x 1,024 and 1,600 x 1,200 without any anti-aliasing or anisotropic filtering and with 4X anti-aliasing and 16X anisotropic filtering enabled concurrently.


We're not going to spend too much time dissecting the results of our custom Half Life 2 benchmark, because all of the cards we tested posted excellent frame rates, and the test system was basically CPU bound with any of the 512MB cards installed. Technically speaking, the Sapphire Toxic Radeon X1900 XTX was the fastest overall, but a difference of one or two frames per second can hardly be considered a clear victory, when all of the cards are putting up 135+ frames per second.

F.E.A.R. v1.03

Performance Comparisons with F.E.A.R
More Info: http://www.whatisfear.com/us/

One of the most highly anticipated titles of 2005 was Monolith's paranormal thriller F.E.A.R. Taking a look at the minimum system requirements, we see that you will need at least a 1.7GHz Pentium 4 with 512MB of system memory and a 64MB graphics card, that is a Radeon 9000 or GeForce4 Ti-class or better, to adequately run the game. Using the full retail release of the game patched to v1.03, we put the graphics cards in this review through their paces to see how they fared with a popular title. Here, all graphics settings within the game were set to the maximum values, but with soft shadows disabled (Soft shadows and anti-aliasing do not work together currently). Benchmark runs were then completed at resolutions of 1152x864 and 1600x1200, with and without anti-aliasing and anisotropic filtering enabled.


A pattern similar to what we saw with FarCry and Splinter Cell: Chaos Theory emerged with the F.E.A.R. benchmark. Once again, without any additional pixel processing applied, the GeForce 7900 GTX was the top dog, besting the ATI powered cards by as much as 10%. Switch on 4X anti-aliasing and 16X anisotropic filtering though, and the Radeons come right back. The 7900 GTX remained slightly faster at the lower resolution, but at 1600x1200 with 4XAA and 16X Aniso enabled the Toxic is over 10% faster than NVIDIA's current flagship.

Quake 4 v1.0.5.2

Performance Comparisons with Quake 4
Details: http://www.quake4game.com/

Quake 4
id Software, in conjunction with developer Raven, recently released the latest addition to the wildly popular Quake franchise, Quake 4. Quake 4 is based upon an updated and slightly modified version of the Doom 3 engine, and as such performance characteristics between the two titles are very similar.  Like Doom 3, Quake 4 is also an OpenGL game that uses extremely high-detailed textures and a ton of dynamic lighting and shadows, but unlike Doom3, Quake 4 features some outdoor environments as well. We ran this these Quake 4 benchmarks using a custom demo with the game set to its "High-Quality" mode, at resolutions of 1,280 x 1,024 and 1,600 x 1,200 without anti-aliasing enabled and then again with 4X AA and 8X aniso enabled simultaneously.


The GeForce 7900 GTX, and the 512MB GeForce 7800 GTX for that matter, clearly outpaced the Radeon X1900 XTXs in our custom Quake 4 benchmark. The 7900 GTX was about 18% to 25% faster than Toxic Radeon X1900 XTX in the default tests. When anti-aliasing and anisotropic filtering were enabled simultaneously though, the GeForces' leads get much smaller, but it was again able to outpace the Radeon by approximately 10% to 13%.

Overclocking the Toxic X1900XTX

As we neared the end of our testing, we spent a little time overclocking the Sapphire Toxic Radeon X1900 XTX using the clock frequency slider available within ATI's Catalyst driver suite, under the Overdrive tab. To find the card's peak core and memory frequencies, we slowly raised their respective sliders until we began to see visual artifacts on-screen while running a game or benchmark, or the test system became unstable.

Overclocking the Sapphire Toxic Radeon X1900 XTX
(Fast 3D Video Card) + Overclocking = Even Faster Card


Due to the fact that the Sapphire Toxic Radeon X1900 XTX is scheduled to ship at higher clock speeds than a standard Radeon X1900 XTX, and we tested the card at its higher clocks, it was obviously faster than a "stock" card. We were hoping that the Toxic's liquid-cooling apparatus would allow for even higher frequencies though, and we weren't disappointed. When we let ATI's driver's detect the Toxic's optimal frequency, Overdrive reported peak core and memory clock speeds of 689MHz and 801MHz for the core and memory clocks, respectively. However, we took things a step further and were actually able to max out the sliders available on the overdrive tab for final overclocked speeds of 690MHz and 801MHz. We also tried overclocking the card using a third-party tool, but at 700MHz / 810MHz we could not complete a loop of 3DMark, so we backed things down to 690MHz / 801MHz for additional testing. With the card configured at these clocks, we re-ran a couple of tests and saw marginal performance increases in both, as you can see in the graphs above.

While we had the card overclocked, we also checked the GPU core temperature after running a handful of benchmarks and found that the core ran at about 60oC after an extended period of use and it peaked at about 88oC. That's not too much lower than we've seen with the stock Radeon cooler, but the Toxic's liquid-cooler does its job much quieter than the stock cooler.

Our Summary & Conclusion

Performance Summary: The Sapphire Toxic Radeon X1900 XTX was an exceptional performer throughout all of our testing. Because the card is due to ship in a "pre-overclocked" state, it was faster than a stock Radeon X1900 XTX in every benchmark we ran. And the Toxic Radeon's higher core and memory clocks also gave it an edge over the GeForce 7900 GTX in a few tests (3DMark06, FarCry, Splinter Cell), where a stock Radeon X1900 XTX fell short of the mark set by NVIDIA's current flagship.

Sapphire's Toxic Radeon X1900 XTX is appealing on a few different levels. First off, the card ships with a very complete accessory bundle that includes video editing and DVD playback software, two full-version games, and an assortment of cables that give users the ability to connect the card to a multitude of different display types. The Toxic's liquid cooling apparatus is also another plus in our book. The card's liquid cooler does a better job at keeping the GPU cool than a stock heatsink assembly, and it does so while generating much less noise overall. We also like the fact that the cooler can be relocated into a different slot position within a system, and that it doesn't have to be right up against the video card. The Toxic Radeon X1900 XTX also proved to be a very competent overclocker, and because it's powered by ATI's flagship Radeon X1900 XTX GPU and 512MB of RAM it's obviously a heck of a performer too.

With an MSRP of $569, the Sapphire Toxic Radeon X1900 XTX is not cheap by any means. But considering that "stock" X1900 XTXs are currently selling for about $530 to $580, and Sapphire will be offering a $30 mail in rebate bringing the final price down to $539, we think the Toxic Radeon X1900 XTX's price-premium is well justified. The Toxic card's quieter liquid-cooler and complete bundle easily make up for the increased cost. Based on the information we currently have from Sapphire, we think the Toxic Radeon X1900 XTX is one of the more attractive high-end video cards we've tested here on HotHardware. We're giving the Toxic Radeon X1900 XTX a strong 9 on the Heat Meter, and a rare HotHardware Editor's Choice award.

  • Near Silent Operation
  • Liquid-Cooling
  • Good Overclocker
  • AVIVO Support
  • Complete Bundle
  • ViVo
  • Great Performer
  • Premium Price
  • Not Shipping Yet
  • Smallish RAM sinks

Get into HotHardware's PC Hardware Forum Right Now!

Content Property of HotHardware.com