The main-stream market has been seeing a boom in performance over the last year or so. Cards like ATI's Radeon X700 Pro and nVidia's Geforce 6600GT have firmly replaced old, underwhelming cards like the 9600XT and 5700 Ultra and, thanks to a new 110nm manufacturing process which greatly reduces manufacturing costs, these new cards can be made available at performance levels equal to that of the old 9800XTs and 5950 Ultras while still staying under the $200 price target.

As you'd expect from Sapphire, they have taken ATI's X700 Pro and created a couple different cards based on it. The first is a 256MB X700 Pro called the Hybrid, and the second is a 128MB X700 Pro with a special cooling fan and kick ass bundle called the Toxic. In this review I'll be focusing on the more unique of the two, the Toxic X700 Pro.

While the Toxic line is not as established as some other vendor lines, it's still hard to mistake them for anything else. The bright orange fan housing and blue PCB, while not for everyone, won't ever be associated with anyone else. With the Toxic, Sapphire has taken a successful leap to differentiate themselves from the swell of red and green graphics cards out there and have a guaranteed advantage to get noticed at your favorite computer retailer.

Bundle

The Toxic includes full versions of two best selling UBI Soft games, Splinter Cell: Pandora Tomorrow (DVD) and Prince of Persia: The Sands of Time. These two gems are amongst the best games ever made and serve as a good base to show off the power of the Toxic X700 Pro graphics card. Also included on the software side are PowerDVD 5 and a driver CD that contains install manuals in PDF format, the drivers of course, and a little application called Automated Performance Enhancer, or APE for short. I'll have a little more about APE later.

On the hardware side of things Sapphire includes all the standard cables and adapters you need for taking advantage of the features the Toxic supports including a DVI to VGA adapter, S-Video cable, composite cable, and an S-Video to composite adapter. Also included in the box is something I've never seen included with a graphics card before; a blue 10cm Cold Cathode Fluorescent Lamp made by Sharkoon. I'll have a little more about that later too.

Here's a breakdown of what the Toxic bundle includes:

• Sapphire X700 Pro Toxic PCI-Express graphics card
• S-Video cable
• Composite cable
• S-Video to Composite adapter
• VGA to DVI adapter
• 10cm Sharkoon Cold Cathode Fluorescent Lamp
• Driver CD
• Splinter Cell: Pandora Tomorrow
• Prince of Persia: The Sands of Time
• PowerDVD 5
• Installation Manual

All X700 cards use ATI's codenamed RV410 graphics core. The RV410 is an approximately 120 million transistor chip with a die size of about 156mm² [13mm (w) x 12mm (h)] built with TSMC's 110nm manufacturing process. It has 8 pixel pipelines, 6 vertex shader units, and a 128bit memory data bus. It can support DDR, GDDR2, or GDDR3 memory in 128MB or 256MB memory configurations, but most commonly comes in the 256MB GDDR3 configuration.

It is DirextX 9.0 (Shader Model 2.0b) and OpenGL 2.0 compliant and supports ATI's own 3Dc normal map compression technology.

Videocards based on the RV410 core are currently only available using the PCI-Express x16 interface. AGP versions have been rumored, but so far there has not been anything officially announced.

The memory used on the Toxic X700 Pro is hidden under a set of blue anodized ramsinks. Carefully prying one of them off reveals Samsung ram, model K4J55323QF-GC20, which is 2.0ns memory rated for a maximum theoretical speed of 500MHz (1GHz DDR). This ram seems to be the most common type used on mid-stream cards such as the Toxic and is used on all the other graphics cards I am comparing with in this review.

Features

SMARTSHADER HD

• Support for Microsoft® DirectX® 9.0 programmable vertex and pixel shaders in hardware
• DirectX 9.0 Vertex Shaders
  • Vertex programs up to 65,280 instructions with flow control
  • Single cycle trigonometric operations (SIN & COS)
• Direct X 9.0 Extended Pixel Shaders
  • Up to 1,536 instructions and 16 textures per rendering pass
  • 2nd generation F-buffer technology accelerates multi-pass pixel shader programs with unlimited instructions
  • 32 temporary and constant registers
  • Facing register for two-sided lighting
  • 128-bit, 64-bit & 32-bit per pixel floating point color formats
  • Multiple Render Target (MRT) support
• Complete feature set also supported in OpenGL® via extensions

SMOOTHVISION HD

• 2x/4x/6x Anti-Aliasing modes
  • Sparse multi-sample algorithm with gamma correction, programmable sample patterns, and centroid sampling
  • Lossless Color Compression (up to 6:1)at all resolutions, including widescreen HDTV resolutions
  • Temporal Anti-Aliasing
• 2x/4x/8x/16x Anisotropic Filtering modes
  • Up to 128-tap texture filtering
  • Adaptive algorithm with bilinear (performance) and trilinear (quality) options

3Dc

• High quality 4:1 Normal Map Compression
• Works with any two-channel data format

HYPER Z HD

• 3-level Hierarchical Z-Buffer with Early Z Test
• Lossless Z-Buffer Compression (up to 48:1)
• Fast Z-Buffer Clear
• Z Cache optimized for real-time shadow rendering
• Optimized for performance at high display resolutions, including widescreen HDTV resolutions

VIDEOSHADER HD

• Seamless integration of pixel shaders with video in real time
• FULLSTREAM™ video de-blocking technology for Real, DivX, and WMV9 formats
• VIDEOSOAP™ noise removal filtering for captured video
• MPEG1/2/4 decode and encode acceleration
  • DXVA Support
  • Hardware Motion Compensation, iDCT, DCT and color space conversion
• All-format DTV/HDTV decoding
• YPrPb component output for direct drive of HDTV displays
• Adaptive Per-Pixel De-Interlacing and Frame Rate Conversion (temporal filtering)

Display Features

• Dual integrated display controllers
• Dual integrated 10 bit per channel 400 MHz DACs Integrated
• 165 MHz TMDS transmitter (DVI 1.0 / HDMI compliant and HDCP ready)
• Integrated TV Output support up to 1024x768 resolution

The Toxic cooler is manufactured by Arctic Cooling and sold in the retail channel as the ATI Silencer 1. The only difference is that the Toxic cooler is bright orange with a bluish fan while the original ATI Silencer is clear with a black fan. The Toxic cooler is also UV reactive, which explains the inclusion of the Cold Cathode Fluorescent Light in the bundle, and glows pretty nicely in the dark.

Toxic Cooler

Toxic Fan

Sharkoon 10cm CCFL

UV Reactive

The basic concept is the same as Arctic Cooling's classic VGA Silencer; that is the fan draws air from inside the case, passes it through a plastic housing over an array of aluminum cooling fins, and expels it out the back. Unlike the solid aluminum block of the classic VGA Silencer, the ATI Silencer consists of a piece of cooper about an eighth of an inch thick with 23 extremely thin aluminum cooling fins soldered onto it. The design has proven to be extremely effective and is quite a bit of overkill for a card like the X700 Pro that runs relatively cool even with the stock cooler.

Also on the Toxic are small blue anodized aluminum ramsinks that are held in place with a small piece of thermal tape. They aren't that hard to remove, so should the day come when you want to replace them you shouldn't have any problems.

Heatsink

Heatsink

Heatsink

Ramsinks

As for performance, there's really not much more that can be said that hasn't already be said about it. Arctic Cooling makes some of best VGA coolers on the market and they are all well respected for their excellent performance and silence. For a much more detailed look at the ATI Silencer 1 (along with the ATI Silencer 2 and 3), you can checkout my full review of the coolers here.

However, as good as the cooler is, Sapphire made one pretty big compromise that virtually eliminates the primary design advantage of the cooler. On the ATI Silencer 1 there is a plastic cover on the underside of the card which is there to help direct the hot air pulled off the heatsink out the back of the case. Apparently for component clearance reasons, Sapphire has removed that little cover. The result is that the heated air prematurely spills out of the cooling shroud back down onto the graphics card. Even though the core doesn't generate that much heat to make this a disastrous decision, the main design idea behind the cooler is still eliminated. It still looks good though..

Missing Plate

Duct Opening

Though the cooler is relatively narrow (except for bit where the fan is) it is fairly long and extends a good 2 inches past the end of the graphics card PCB. It's also a double-height cooler and takes up an extra expansion slot so, even though I didn't have any problems with the test system, there's a possibility that clearance issues might crop up on some motherboards.

To keep myself from benchmark induced insanity, I am not using the brand new Catalyst 5.2 drivers. Instead, I will keep the same test system and older Catalyst 5.1 drivers I used in my previous HIS X700 Pro IceQ Turbo review and reuse the benchmark numbers. I decided to use the HIS X700 Pro IceQ Turbo in iTurbo mode, which sets the card to 460MHz core and 480MHz (960MHz DDR) memory. I believe that the HIS card is a direct competitor to the Toxic, so I feel it's necessary to include the performance results achieved with it in this review. Besides, I already had the results from the review I just finished, so I might as well reuse them. Like the Toxic X700 Pro, I believe that the manufacturer overclocked settings are the ones that should be used when comparing these cards. The following is my standard test setup breakdown.

Anti-Aliasing and Anisotropy were applied in game where the options existed. For games that did not support those options natively, the graphics card control panel was used to force them on. V-Sync was also forced off for each test via the graphics card control panel. Otherwise all other graphics card control panel settings were left to their default values.

Batch files were used when possible for automated benchmarking. Manual benchmarking was done using FRAPS.

Benchmarking was done with Windows set to the "Adjust for best performance" profile, and all unnecessary Windows services and hardware devices were disabled. The latest drivers for each necessary hardware component were installed prior to testing and kept consistent throughout.

To avoid possible driver conflicts, testing of the ATI and nVidia graphics cards were done on two separate but identical Windows installations (using Norton Ghost images) on a pair of identical hard-drives.

Test System Specs

• CPU: Intel Pentium4 550, 3.4GHz Socket 775
• Motherboard: ABIT AA8 DuraMAX, Intel i925X chipset
• Videocards:
Specifications.
• Sapphire Toxic X700 Pro (APE), Catalyst 5.1
• Sapphire Toxic X700 Pro (Default), Catalyst 5.1
• HIS X700 Pro IceQ Turbo (iTurbo), Catalyst 5.1
• ATI Radeon X700 Pro, Catalyst 5.1
• Chaintech GeForce 6600GT, ForceWare Beta 67.66
• Memory: 1GB (2x512MB) OCZ DDR2 PC2-4200 Platinum Edition provided by the guys at Magnum-PC.com
• HDD: Seagate Barracuda V SATA, 7200rpm
• Sound: Onboard Realtek ALC880
• PSU: Antec NeoPower 480W
• Case: Coolermaster Stacker
• OS: Windows XP Pro SP2

Testing was done on a fully assembled PC.

Benchmark Software

• Half-Life 2, Source Engine 7 (DirectX)
• Far Cry, v1.3 (DirectX)
• Richard Burns Rally, v1.02 (DirectX)
• Doom3, v1.1 (OpenGL)
• Chronicles of Riddick: Escape from Butcher Bay (OpenGL)
• Call of Duty United Offensive, v1.41 (OpenGL)
• Pacific Fighters, v3.03 (OpenGL)
• 3DMark03, build 3.5.0 (DirectX)
• 3DMark05, build 1.1.0 (DirectX)
• ShaderMark 2.1, build 129 (DirectX)

Half-Life 2

Testing of Half-Life 2 was done using AnandTech's five recently released Source Engine 7 timedemos (which are available on this page). I ran each timedemo for each resolution and AA/AF setting used in the charts below, and then averaged the results to get the final score. Anti-aliasing and Anisotropy were set on the command line.

A batch file was used to automate testing; the command line is below for reference. This batch file was used for each card that was tested. The settings surrounded by < > change for each pass:

The Half-Life 2 results are below. Click the text links at the top of the chart to change settings.

Doom 3

Doom3 was tested using the built in timedemo, demo1.demo. I benchmarked the default "High Quality" mode, which sets Anisotropy to 8x, and then combinations of Anti-Aliasing and Anisotropy over the resolutions shown in the chart below. Anti-aliasing and Anisotropy were set on the command line.

Another batch file was used to automate Doom3 testing as well. The command line is below for reference. This batch file was used for each card that was tested. The settings surrounded by < > change for each pass:

The Doom3 results are below. Click the text links at the top of the chart to change settings.

Far Cry

Far Cry benchmarking was a little different. In the interest of time (Far Cry takes quite awhile to benchmark) I only tested anti-aliasing on the High setting and did not test anisotropy performance at all. I did however test over three different maps using custom timedemos. I averaged the results from each of those timedemo runs to get the final score. Anti-aliasing was set in the "Configure Far Cry" app.

The command line for the batch file I used to automate Far Cry benchmarking is below. This batch file was used for each card that was tested. The settings surrounded by < > change for each pass:

The Far Cry results are below. Click the text links at the top of the chart to change settings.

Chronicles of Riddick

Chronicles of Riddick was benchmarked using the built-in timedemo feature. No batch file was used (mainly because I didn't have time to make one). The timedemo used is a recording of the intro sequence. The Shader Model 2.0 path was used for all cards, including the 6600GT, which takes a huge performance hit using the Shader Model 3.0 path. Anti-aliasing and Anisotropy were set in the graphics card control panel.

The Chronicles of Riddick results are below. Click the text links at the top of the chart to change settings.

Richard Burn’s Rally

I used Fraps to benchmark Richard Burn's Rally using two custom demos recorded on the USA-Frasier Wells and Finland-Autiovaara stages. I started Fraps framerate logging as soon as the marshal launched the car and had it automatically stop logging after 90 seconds. I logged each pass two times, and then averaged the four results to get the final score for that setting and resolution. The default replay camera was used and sound was enabled. Anti-aliasing and Anisotropy were set via the graphics card control panel.

The Richard Burn's Rally results are below. Click the text links at the top of the chart to change settings.

Pacific Fighters

I benchmarked Pacific Fighters by loading the included "N1K1 vs BeauFighter.ntrk" track and logged framerates using Fraps from the beginning of the track for 90 seconds. All the in game details were set to their maximum levels, including "Landscape Detail" which was set to "Perfect", enabling Pixel Shaded water. Video was set to "Maximum Settings" for all graphics cards and sound was disabled in the Pacific Fighter Setup app. Anti-aliasing and Anisotropy were set via the graphics card control panel.

The Pacific Fighters results are below. Click the text links at the top of the chart to change settings.

Call of Duty United Offensive

I benchmarked Call of Duty using the built in timedemo and a batch file with a custom script to help automate the process. Sound was disabled and all graphical settings were set to their maximum. Anti-Aliasing and Anisotropy were set via the graphics card control panel.

The Call of Duty results are below. Click the text links at the top of the chart to change settings.

3DMark03

I benchmarked 3DMark03 using the default settings. I set it up to loop each test three times, it then automatically averaged the results and gave a final score. The results for each test are below.

Game Test 1 - Wings of Fury Game Test 2 - Battle of Proxycon Game Test 3 - Troll's Lair Game Test 4 - Mother Nature Fillrate - Single-Texturing Fillrate - Multi-Texturing Vertex Shader Pixel Shader 2.0 Overall Score

3DMark05

Like 3DMark03, I set 3DMark05 up to loop three times to get the final score. I tested 3DMark05 using the default detected settings. That means the VS 2_0 and PS 2_b paths are used for the X700 Pro's while the VS and PS 3_0 paths are used with the 6600GT. The 6600GT also makes use of the DST (Depth Stencil Texture) feature, which is enabled by default. DST helps the rendering of dynamic shadows in 3DMark05 and gives a fairly significant speed boost to graphics cards which support it.

Game Test 1 - Return to Proxycon Game Test 2 - Firefly Forest Game Test 3 - Canyon Flight Fillrate - Single-Texturing Fillrate - Multi-Texturing Pixel Shader Vertex Shader - Simple Vertex Shader - Complex Overall Score

ShaderMark 2.1

ShaderMark is a DirectX benchmark tool created to test DirectX 9.0 shader performance. The pixel and vertex shader code was written using Microsoft's High Level Shader Language. I left all the options to their default values for each card, and then ran the benchmarking mode. The results from each test are below (sorry about the mess, there's just too much data to represent graphically in the space we have available).

		Sapphire Toxic X700 Pro (APE)	Sapphire Toxic X700 Pro	IceQ Turbo (iTurbo)	ATI X700Pro	Chaintech 6600GT
1	pixel shader 1.1 full precision	s16e7	s16e7	s16e7	s16e7	s10e8
	pixel shader 1.4 full precision	s16e7	s16e7	s16e7	s16e7	s10e8
	pixel shader 2.0 full precision	s16e7	s16e7	s16e7	s16e7	s23e8
	pixel shader 2.0 partial precision	s16e7	s16e7	s16e7	s16e7	s10e5
	pixel shader 3.0 full precision	s0e0	s0e0	s0e0	s0e0	s23e8
	pixel shader 3.0 partial precision	s0e0	s0e0	s0e0	s0e0	s10e5
2	Per Pixel Diffuse Lighting	469	417	464	431	468
3	Per Pixel Directional Light Shader (Phong)	342	310	335	311	397
4	Per Pixel Point Light Shader (Phong)	342	310	335	311	397
5	Per Pixel Spot Light Shader (Phong)	282	255	276	256	364
6	Per Pixel Anisotropic Lighting	341	297	334	311	381
7	Per Pixel Fresnel Reflections	312	279	307	278	94
8	Per Pixel Car Surface Shader	194	94	183	167	85
9	Per Pixel Environment Mapping	596	533	596	544	120
10	Per Pixel Environment Bump Mapping	389	302	389	355	96
11	Per Pixel Bump Mapping	325	251	322	298	366
12	Per Pixel Shadowed Bump Mapping	116	38	114	103	238
13	Per Pixel Veined Marble Shader	163	41	160	135	204
14	Per Pixel Wood Shader	183	130	177	158	248
15	Per Pixel Tile Shader	133	113	130	121	189
16	Per Pixel Refraction and Reflection Shader witd Phong Lighting	154	121	153	134	46
17	Per Pixel BRDF-Phong/Anisotropic Lighting	202	176	200	178	84
18	Fur Shader Witd Anisotropic Lighting	19	9	19	17	31
19	Combination Effect	66	32	64	58	60
20	Dual Layer 8x8 PCF Shadow Mapping witdout Flow Control	21	19	21	19	29
21	Dual Layer 8x8 PCF Shadow Mapping witd Flow Control	-	-	-	-	42
22	High Dynamic Range Shader - low quality version witdout filtering	99	89	96	89	50
23	High Dynamic Range Shader - high quality witd fp filtering	-	-	-	-	54
24	High Dynamic Range Shader - high quality witdout fp filtering	70	62	68	63	38
25	Per Pixel Edge Detection And Hatching Shader using 1 RT and 2 Passes	56	49	55	50	51
26	Per Pixel Edge Detection And Hatching Shader using 2 RTs and 1 Pass	56	49	55	51	50

Getting an automatic overclock to 475/500 with APE is fantastic, especially when you consider it doesn't void your warranty. But we all know that factory overclocks always err on the side of caution, and a card like the Toxic is built for overclocking. One look at the massively oversized cooler and aluminum ramsinks will tell you that.

Since APE doesn't include any method to manually overclock, I used ATITool 0.0.23 to overclock the Toxic instead. Starting at the APE baseline of 275/500 I let ATITool automatically find the maximum core and the memory, I then went from that point and manually overclocked while checking for artifacts in 3DMark03. Ultimately I ended up at a very respectable 495MHz for the core speed and an absolutely unbelievable 570MHz (1,140MHz DDR) for the memory. That's 138MHz over stock and 70MHz over the rated limit for the ram modules. Whatever magic spell Sapphire cast on the card, it sure did work!

Below you'll find the performance results for the overclocked settings.

Half-Life 2 No Anti-Aliasing / No Anisotropy

1024x768   90.1  88.8  86.3  93.1  91.4  89.8  1280x960   76.6  74.2  69.8  82.7  79.0  79.2  1600x1200   48.4  46.5  43.4  69.1  58.7  58.0   0 55 110 Average Framerate Sapphire Toxic X700 Pro (495/570) Sapphire Toxic X700 Pro (APE) Sapphire Toxic X700 Pro X700 Pro IceQ Turbo (iTurbo) ATI X700 Pro Chaintech 6600GT

Doom 3 Default High Quality (No Anti-Aliasing / 8x Anisotropy)

800x600   77.0  73.1  67.6  72.4  67.6  82.6  1024x768   62.4  57.6  51.8  57.1  52.1  73.2  1280x1024   43.2  39.6  34.8  39.1  34.2  57.1  1600x1200   29.8  27.4  24.1  27.0  24.0  43.5   0 50 100 Average Framerate Sapphire Toxic X700 Pro (495/570) Sapphire Toxic X700 Pro (APE) Sapphire Toxic X700 Pro X700 Pro IceQ Turbo (iTurbo) ATI X700 Pro Chaintech 6600GT

The Splinter Cell: Pandora Tomorrow and Prince of Persia: Sands of Time bundle is virtually impossible to beat. The cooler is nice, and when you install the bundled UV cold cathode light, the orange and blue UV glow looks pretty sweet. The X700 Pro doesn't run very hot regardless of what cooler you use, but it's still too bad Sapphire removed that little plastic plate though; doing that cripples the main design feature of the cooler.

APE overclocking doesn't void your warranty and pushes the card to within a hairs breadth of X700XT speeds. While it's easy to use by just installing it and rebooting, I'm starting to wish that there was a little more control included with the app. Fan speed and temperature monitoring, maybe even controls for manual overclocking beyond APE's limits, would be pretty cool. Speaking of overclocking; 570MHz for the memory, fully 138MHz over stock is a feat in itself. What exactly Sapphire did to get the memory to run that fast is unknown, but I have a sneaking suspicion that it has something to do with only using 128MB of total memory.

The card performs very respectably in a lot of ways. Even with only half the amount of memory it manages to keep up to the other X700 Pros in most tests. It's really too bad though that there's only 128MB memory because we're at a point in time when the need for 256MB with some games is becoming very real, particularly when Anti-Aliasing is used and in cases such as with Half-Life 2 and Far Cry, even without Anti-Aliasing. There's just not enough room in there to hold all that data and performance suffers greatly because of it.

Unfortunately for Sapphire, 256MB X700 Pros from other manufacturers are available for pretty much the same price and they come with features very similar to that of the Toxic so even with the great bundle and excellent overclocking ability, it's difficult to recommend this card over some of those 256MB alternatives available out there.

---

---

Good

• Quiet cooling performance
• Excellent, almost unbelievable overclocking
• Outstanding bundle
• Kick ass looks

Bad

• Only 128MB of memory
• Crippled cooler
• Extra long cooler might cause some clearance problems