Original Link: http://www.anandtech.com/show/1030



Things have been quiet on the NVIDIA side as of late. If we look back, one can see that the GeForce4 Ti 4200 AGP 8x and the GeForce4 MX 460 AGP 8x parts (NV28 and NV18 respectively) were the last NVIDIA products to be released. Launching one month ago these two chips were fairly un-revolutionary on the whole, offering NV25 and NV17 performance in an AGP 8x certified package. In fact, one has to go back over 10 months to recall NVIDIA's last major consumer desktop product; the NV25 based GeForce4 Ti 4600. A glance at NVIDIA's roadmap from earlier this year quickly suggests that this was not the plan. The company's roadmap proved to be quite aggressive, in fact too aggressive for even NVIDIA to keep up with. As a result, today we are left NV30-less.

While NVIDIA was attempting to balance their production woes with their launch cycle, ATI was taking advantage of the situation at hand. ATI aggressively pressed ahead with new core designs. The past three months have found a number of ATI products come to market based on the RV250 and R300 core, including the Radeon 9700 Pro, the Radeon 9000 Pro, and the Mobility Radeon 9000. Users and analysts alike began to wonder how NVIDIA would attempt to curb ATI's progress.

On the desktop side, we have yet to see NVIDIA's plan of attack. The company's response to the R300 core, the NV30, has yet to make a public appearance. The RV250 competitor, the NV18, has been gaining a bit of momentum but its lack of DX8 shaders has been dissuading customers. While it is up in the air how NVIDIA will respond to ATI's threat on the desktop side, today we see how the company plans to but a damper on ATI's mobile graphics processor party.

Today NVIDIA responds to the ATI Mobility Radeon 9000, a chip launched at the end of August that has already found its way into twenty shipping notebook machines. NVIDIA's answer is the NV28M based GeForce4 4200 Go, a mobile version of the NV28 core. The GeForce4 4200 Go does more than just include the full feature set of the desktop NV28 core used in the GeForce4 Ti 4200 AGP 8x; it also incorporates the power saving technology that is so important for notebook machines. Follow us as we take a look at the first revolutionary product from NVIDIA in just about nine months and see if the GeForce4 4200 Go has what it takes to bring mobile gaming to the next level.



The NV28M

What exactly is the NV28M chip that powers the GeForce4 4200 Go? Well, to find the answer one need not look further than the desktop GeForce4 Ti 4200 AGP 8x, the first desktop product from NVIDIA to use the updated NV28 core.

The mobile NV28M chip is based on the desktop NV28 chip which is based on the desktop NV25 chip. Confused yet? We thought so. Let's go through the steps that produced the NV28M.

The first chip in the evolution which created the NV28M was the NV25. Launched nine months ago, the NV25 core is the chip that powers NVIDIA's enthusiast level GeForce4 Ti series chips. At the time, the NV25 core provided a marked improvement over the aged NV20 (GeForce3) core. On top of using a 0.15 micon architecture, the NV25 implemented a number of new features that further pushed the power of 3D graphics accelerators. For one, the NV25 improved upon the NV20's DirectX 8.1 functionality by implementing dual vertex shader units, an addition which helps games that use vertex shading operations run more smoothly.

Except for the addition of the second vertex shader unit, the 3D pipeline of the NV25 remained very similar to that of the older NV20. The NV25 contained the same single pixel shader unit and 4 pixel pipelines capable of processing two textures per pixel. The other hardware upgrade was made not to the 3D pipeline but rather to the memory architecture. The NV25 was the first NVIDIA chip to implement what is known as Lightspeed Memory Architecture II. You can read up on the details of this technology in our initial GeForce4 Ti review, but in summary LMA II effectively increases memory bandwidth offered by managing the data traveling over the memory bus more efficiently.

So that is a brief summary of what went into the NV25, the chip that is found in all GeForce4 Ti 4200, GeForce4 Ti 4400, and GeForce4 Ti 4600 cards out on the market today. In what ways is this chip, the NV25, different from the NV28 used in the GeForce4 Ti 4200 AGP 8x (and perhaps future NVIDIA products)? Not many.

Hardware wise, the NV25 differs from the NV28 in only one way: the NV28 includes an AGP 8x host interface. The NV28 maintains the same 3D pipeline, the same memory controller, and the 2D components. It is easy to see why many were disappointed with the release of the NV28. On the hardware side, there is very little difference between the new NV28 and the existing NV25. On the performance side, the NV28 gains no tangible performance by moving from an AGP 4x design to an AGP 8x one. So does that leave the NV28 without promise. Not exactly.

First off, NVIDIA has only released the GeForce4 Ti 4200 in AGP 8x form. There is no question that both the GeForce4 Ti 4400 and GeForce4 Ti 4600 will receive this chip eventually. Many hope that these incarnations of the NV28 will prove to be more attractive than the current GeForce4 Ti 4200 AGP 8x. But NVIDIA did not see the potential for the NV28 limited to the desktop market. Enter the NV28M.



The NV28M is the mobile variant to the NV28, just as the NV17M (GeForce4 440 Go) was the mobile variant to the NV17 (GeForce4 MX). What does being a mobile variant of the chip entail? In the case of the NV28M it means that the chip maintains 100% of the desktop NV28's hardware but incorporates power saving technology. The same 3D pipeline with dual vertex shaders; the same LMA II; the same AGP 8x support. Added to the core is PowerMizer and the ability to run at lower clock speeds and save power when possible, which we will discuss more in a minute. For now keep in mind that the GeForce4 4200 Go (NV28M) is based on the current generation highest performing desktop offering from a company, tge first time this has happened in the mobile graphics industry. Historically, mobile chips have been based off of mainstream parts as shown by the GeForce4 4x0 Go's base in the GeForce4 MX series and the ATI Mobility Radeon 9000's base in the Radeon 9000 cards. The GeForce4 4200 Go changes that trend and makes a huge statement as to where mobile graphics is going. There are some problems that arising from doing this, especially in the case of the NV28 specifically. Lets talk about some of the limitations of turning the NV28 into a mobile graphics processor.

Unlike the NV17 core found in the GeForce4 MX cards, the NV28 core was not designed specifically with a mobile variant in mind. As a result, the NV28M does not have the dedicated laptop functionality that the NV17 offered. The two big things that the NV28M is missing that the NV17M has are an integrated TV encoder (for TV-output) and an integrated LVDS (for driving the laptop display). This means that notebook manufacturers have to turn to at least an external LVDS and possibly an external TV encoder, making designing notebooks that use the GeForce4 4200 Go a bit more difficult.

The lack of integrated display hardware is not the only stumbling block in the GeForce4 4200 Go's way. As a desktop chip, emphasis on designing the NV28 did not lie in reducing heat and power consumption. Power saving hardware, such as PowerMizer, has the ability to reduce the bottom line power consumption of a chip but the peak power characteristics of a mobile variant remain fairly similar to that of the desktop one. This results in very high peak power consumption and thermal characteristics in the NV28M under situations when the chip is stressed, such as when rendering a 3D scene.

NVIDIA would not disclose the maximum power draw of the NV28M but they did mention that it was high. They were able to get the voltage down to 1.25 volts when at the maximum performance setting (the same as the Mobility Radeon 9000) but apparently the power draw of the chip is still desktop-like.

Another compromise that NVIDIA had to make when turning the desktop NV28 into the NV28M has to do with the chip packaging. The NV28 is composed of many more transistors than the NV17. As a result, both the chip and package of the NV28 is larger than that of the NV17. As a result, the GeForce4 440 Go and the GeForce4 4200 Go are different sizes. One characteristic of the entire GeForce4 4x0 Go line (with exception to the GeForce4 440 Go) is that all the chips are pin compatible. This means that all a manufacturer has to do to go from producing a notebook with a the high-end GeForce4 460 Go to the lower-end GeForce4 420 Go would be to pop out the old chip and pop in the new one. NVIDIA planed to expand the 31x31 BGA package size to future mobile NVIDIA products but obviously the NV28M marks a deviation from this path. Not a big deal for now (let's hope future mobile NVIDIA parts revert to this design), but it does make getting the GeForce4 4200 Go into a laptop a bit more difficult.

Likely due in large part to the GeForce4 4200 Go's thermal requirements, the NV28M will not come in a MAP package. You may recall from our GeForce4 4x0 Go review that the previous Go chips were available in discrete packages (which use external memory) as well as what NVIDIA called mobile AGP packages, or MAP packages (which include the GPU core on the same package). The way this was done in the past was by mounting the GPU core and the memory packages on the same substrate, thus requiring less physical space for the full video setup.

The lack of a MAP GeForce4 4200 Go solution turns out to actually be one of the more minor thorns in the chip's side. As it turns out, not many MAP solutions have made their way into market (this is true for ATI as well) as OEMs have been opting for cheaper external memory. In fact, the most appealing market for a GPU and memory combination chip is the thin-and-light and ultra-portable market, where thermal and power requirements keep the highest performing chips out.

Obviously, as a mobile part, the GeForce4 4200 Go runs at lower core and memory clock speeds compared to the desktop GeForce4 Ti 4200 AGP 8x. The desktop version of the chip (the NV28 in a 64MB configuration like the configuration of the GeForce4 4200 Go) runs at a 250MHz core and a 500MHz (250MHz DDR) memory bus. In contrast, the GeForce4 4200 Go has the ability to run at a maximum speed of 200MHz core and 400MHz (200MHz DDR) memory. Expect to find chips performing at this pace only in desknote type systems. The majority of notebooks, including the notebook version that we were given to test on, will be running at around the 175MHz core mark and a 400MHz (200MHz DDR) memory bus (ours ran at 175MHz/200MHz DDR). The reason for this: the GeForce4 4200 Go produces a lot of heat and draws a lot of power running at 200/200MHz DDR.



PowerMizer: The same but different

We mentioned previously that the only thing added to the NV28 core to make the NV28M core used in the GeForce4 4200 Go is PowerMizer power management hardware. The PowerMizer hardware used in the NV28M is borrowed directly from the NV17M, meaning that the two possess almost the same feature sets. You may recall that the key features of PowerMizer (detailed here) combine to produce a chip that is much more sensitive to power requirements. Unlike in a desktop, a mobile graphics processor can not afford to draw as much power and create as much heat as it wants. PowerMizer holds back the power consumption of the GeForce4 4200 Go by implementing a number of power saving features such as dynamic clock gating and dynamic clock and voltage scaling. On the hardware side these features are comparatively very similar to those found in ATI's POWERPLAY power saving hardware (with the exception of the automatic color depth reduction which ATI offers but NVIDIA does not).

Again, NVIDIA did not provide us with the total wattage that the GeForce4 4200 Go draws nor did we have enough time with our test system to measure the power consumption ourselves (look for this in a future GeForce4 4200 Go review). What we do know is that the NV28M core in the GeForce4 4200 Go can range in voltage from 1.05 volts (idle) to 1.25 volts (full performance) and can clock scale down to 1/8th the full clock speed of the mobile core or 25MHz when the chip is running at 200MHz. We were also told the memory can scale from the full 200MHz clock speed to 100MHz while idle. The dynamic voltage scaling has been redesigned in the NV28M to reduce leakage in the chip, but again that is something that will have to wait for more in-depth testing to verify. Note that the core clock speed that the NV28M is able to go down to in a 2D static state is much lower than the lowest state of the Mobility Radeon 9000 core which can only underclock itself to an 80MHz (then again, the Mobility Radeon 9000 draws a good deal less power clock for clock than the GeForce4 4200 Go). On the flip side of the coin, the memory speed on the Mobility Radeon 9000 can dip lower than the GeForce4 4200 Go's memory can, as the ATI solution is able to go down to 80MHz.

The largest thing updated for the GeForce4 4200 Go's PowerMizer release is in the software end. PowerMizer has become more friendly in this release of the software, as one is now able to set performance of the chip to one of three settings while on AC power and while on battery power. The maximum performance mode leaves the GeForce4 4200 Go running at full voltage and core and memory speed. The maximum power saving mode forces the chip to run at the lowest mode desired, i.e. at 1.05 volts and a 25MHz core speed and a 100MHz memory speed. The balanced mode sets performance between these two and can be set to OEM preferences. Again, thanks to the limited time we had with the test unit, we were unable to test the system in anything but the maximum performance mode but we will be sure to take a look at this in future AnandTech reviews.

The functionality of PowerMizer's software side is still noticeably less powerful than that of ATI's POWERPLAY. You may recall that the Mobility Radeon 9000 implemented a new interface for the POWERPLAY functions that allowed for much smarter power management. ATI is still a step ahead of NVIDIA on this front and although the new PowerMizer front end is better than the previous version, it still leaves some to be desired (for example multiple profiles for various battery states).

Before we get to the test, here is a table comparison of the GeForce4 460 Go, the GeForce4 4200 Go, and the Mobility Radeon 9000.

Mobile GPU Comparison
 
NVIDIA GeForce4 460 Go
NVIDIA GeForce4 4200 Go
ATI Mobility Radeon 9000
Core Clock - Max
250MHz
200MHz
250MHz
Core Clock - Min
?
25MHz
80MHz
Fill Rate
1000 MTexels/s
1600 MTexels/s
1000 MTexels/s
Memory Bus
128-bit DDR
128-bit DDR
128-bit DDR
Memory Clock- Max
500MHz
400MHz
460MHz
Memory Clock - Min
?
100MHz
80MHz
Memory Bandwidth
8000 MB/s
6400 MB/s
7360 MB/s
Memory Supported
64MB
128MB
64MB
Power Management
PowerMizer
PowerMizer
POWERPLAY


The Test

As we mentioned before, our time with the test GeForce4 4200 Go notebook was limited and as a result today we are only able to bring you performance numbers from the chip running at its full speed. Keep on the lookout for a review that details not only performance of the GeForce4 4200 Go in various power saving modes but also goes in depth about the power consumption of the chip.

Also due to testing constraints we are unable to say what kind of notebook we tested the GeForce4 4200 Go on. The only information we can provide you with is in the table below but do note that the Mobility Radeon 9000 system and the GeForce4 4200 Go system tested were identical.

Windows XP Professional Test System

 

Hardware
 
Notebook Test System
Desktop System
CPU(s)
Mobile Pentium 4-M 2.0GHz
Pentium 4 2.0GHz
Motherboard(s)
845 DDR
Memory
512MB PC2100
512MB PC2100
Hard Drive
40.0GB
IBM Deskstar DPTA-372050 20.5GB 7200 RPM
CDROM
DVD/CDRW
Philips 48X
Video Card(s)
ATI Mobility Radeon 9000 64MB 128-bit DDR
NVIDIA GeForce4 4200 Go 64MB 128-bit DDR
NVIDIA GeForce4 MX 440 64MB
NVIDIA GeForce4 Ti 4200 64MB
Ethernet
Onboard Ethernet Adapter
Linksys LNE100TX 100Mbit PCI Ethernet Adapter

 

Software

Operating System

Windows XP Professional SP1

Video Drivers
ATI 7.75
NVIDIA 40.72 Beta - GeForce4 4200 Go
NVIDIA 40.72 Beta

 

Benchmarking Applications
 

Croteam Serious Sam 2: The Second Encounter - Extreme Settings
Epic Unreal Tournament 2003 - High Quality Settings
Lucas Arts Entertainment Jedi Knight 2 - High Quality Settings



Performance - Unreal Tournament 2003 Flyby

Unreal Tournament 2003 Flyby
800x600x32
Desktop GeForce4 Ti 4200

GeForce4 4200 Go

Mobility Radeon 9000

Desktop GeForce4 MX 440

147.6

137.0

106.4

104.8

|
0
|
30
|
59
|
89
|
118
|
148
|
177

Right off the bat we can see that the GeForce4 4200 Go brings mobile gaming to places it has never been before. If we recall the incremental gains that the Mobility Radeon 9000 offered over the GeForce4 440 Go (top dog when the Mobility Radeon 9000 was released), we can see that the performance that the GeForce4 4200 Go offers is nothing less than stunning. The GeForce4 4200 Go outperforms the Mobility Radeon 9000, the prior king of the hill, by an impressive 29%. In fact, at 800x600x32 the performance of the mobile GeForce4 440 Go is only 8% off the performance of the high performing desktop GeForce4 Ti 4200 (which should be identical in speed to the GeForce4 Ti 4200 AGP 8x).

Unreal Tournament 2003 Flyby
1024x768x32
Desktop GeForce4 Ti 4200

GeForce4 4200 Go

Desktop GeForce4 MX 440

Mobility Radeon 9000

130.3

106.6

73.2

71.9

|
0
|
26
|
52
|
78
|
104
|
130
|
156

Bumping the resolution up to 1024x768x32 in Unreal Tournament 2003 still shows the GeForce4 4200 Go at the head of the mobile pack. The lead rises to a massive 46% boost over the Mobility Radeon 9000. The performance of the GeForce4 4200 Go drops relative to the desktop GeForce4 Ti 4200 as this time around the cards are separated by 22%. The increasing performance delta between the desktop and mobile part can be attributed to the additional stress placed on the GPU at higher resolutions. Whereas at 800x600x32 both the GeForce4 Ti 4200 and the GeForce4 4200 Go were somewhat CPU bound, they become video card bound at these higher resolutions.

Unreal Tournament 2003 Flyby
1280x960x32
Desktop GeForce4 Ti 4200

GeForce4 4200 Go

Desktop GeForce4 MX 440

Mobility Radeon 9000

96.9

69.8

50.6

46.4

|
0
|
19
|
39
|
58
|
78
|
97
|
12

Boosting the resolution of Unreal Tournament 2003 up to 1280x960x32 still paints the GeForce4 4200 Go in good light. The chip, running at its full clock speed, is able to outperform the Mobility Radeon 9000 in its top configuration by 38%. Again the gap between the mobile chip and the desktop one running at a higher memory and core speed grows, this time amounting to 39%. Like in other resolutions, the GeForce4 4200 Go also has no problem what so ever beating out the desktop GeForce4 440.

Unreal Tournament 2003 Flyby
1600x1200x32
Desktop GeForce4 Ti 4200

GeForce4 4200 Go

Desktop GeForce4 MX 440

Mobility Radeon 9000

65.1

49.3

34.0

33.2

|
0
|
13
|
26
|
39
|
52
|
65
|
78

The GeForce4 4200 Go certainly makes a statement in Unreal Tournament 2003 when at 1600x1200x32. Even in today's most stressful gaming environment the chip is able to run at reasonable speeds at this high resolution. The flagship NVIDIA mobile product outperforms the flagship ATI one by 45%. With the GeForce4 4200 Go long gone are the days where subsequent mobile products only increases performance marginally. The speed that the GeForce4 4200 Go offers over the Mobility Radeon 9000 in Unreal Tournament 2003 is not trivial in any sense of the word.



Performance - Serious Sam: The Second Encounter

Serious Sam 2 - Little Trouble
Extreme Quality 800x600x32 (aniso off)
Desktop GeForce4 Ti 4200

Desktop GeForce4 MX 440

GeForce4 4200 Go

Mobility Radeon 9000

117.7

78.7

59.0

55.6

|
0
|
24
|
47
|
71
|
94
|
118
|
141

Serious Sam: The Second Encounter paints a different picture of the GeForce4 4200 Go. Although the chip still performs better than the Mobility Radeon 9000, the performance gains the solution offers are not as impressive as those demonstrated in Unreal Tournament 2003. In addition, the desktop chips have no problem flexing their muscle and outperforming all the mobile solutions. We can attribute this to the nature of the Serious Sam 2 demo: it appears to be much more CPU and platform dependent than video card dependent. At 800x600x32, the GeForce4 4200 Go runs 6% faster than the Mobility Radeon 9000 but falls behind the desktop GeForce4 Ti 4200 by a whopping 99%.

Serious Sam 2 - Little Trouble
Extreme Quality 1024x768x32 (aniso off)
Desktop GeForce4 Ti 4200

Desktop GeForce4 MX 440

GeForce4 4200 Go

Mobility Radeon 9000

98.8

52.7

46.5

43.2

|
0
|
20
|
40
|
59
|
79
|
99
|
119

At 1024x768x32, the GeForce4 4200 Go runs 8% faster than ATI's offering but both chips still fall a good deal behind the desktop chips.

Serious Sam 2 - Little Trouble
Extreme Quality 1280x1024x32 (aniso off)
Desktop GeForce4 Ti 4200

Desktop GeForce4 MX 440

GeForce4 4200 Go

Mobility Radeon 9000

68.2

34.1

31.2

29.9

|
0
|
14
|
27
|
41
|
55
|
68
|
82

As the desktop GeForce4 MX 440 system begins to become video card limited, the performance difference between the GeForce4 4200 Go and the desktop MX card shrinks. The GeForce4 4200 Go still remains 4% faster than the Mobility Radeon 9000.

Serious Sam 2 - Little Trouble
Extreme Quality 1600x1200x32 (aniso off)
Desktop GeForce4 Ti 4200

Desktop GeForce4 MX 440

GeForce4 4200 Go

Mobility Radeon 9000

48.5

23.9

22.7

21.9

|
0
|
10
|
19
|
29
|
39
|
49
|
58

The extreme quality setting of Serious Sam 2 brings the GeForce4 4200 Go to its knees at 1600x1200x32. Performance is more than half that of the desktop GeForce4 Ti 4200 but still remains 4% faster than the Mobility Radeon 9000.



Performance - Jedi Knight 2

Jedi Knight 2 v1.03 - JK2FFA
High Quality 800x600x32 (aniso off)
Desktop GeForce4 Ti 4200

Desktop GeForce4 MX 440

GeForce4 4200 Go

Mobility Radeon 9000

123.4

118.7

109.0

107.2

|
0
|
25
|
49
|
74
|
99
|
123
|
148

Another CPU and platform dependent benchmark, Jedi Knight 2 at 800x600x32 shows all the solutions we tested performing fairly close together. Although CPU and platform limited in this instance, the GeForce4 4200 Go squeaks by the Mobility Radeon 9000 by offering a 2% performance advantage.

Jedi Knight 2 v1.03 - JK2FFA
High Quality 1024x768x32 (aniso off)
Desktop GeForce4 Ti 4200

Desktop GeForce4 MX 440

Mobility Radeon 9000

GeForce4 4200 Go

122.1

109.8

100.6

98.1

|
0
|
24
|
49
|
73
|
98
|
122
|
147

Jedi Knight 2 at 1024x768x32 proved to be the only situation where the GeForce4 4200 Go fell to the Mobility Radeon 9000. Again, the benchmark is CPU bound at this resolution so the performance numbers are less of an indication of graphics performance and more an indication of CPU performance. The lead offered by the Mobility Radeon 9000, at 3%, is negligible at best.

Jedi Knight 2 v1.03 - JK2FFA
High Quality 1280x1024x32 (aniso off)
Desktop GeForce4 Ti 4200

Desktop GeForce4 MX 440

GeForce4 4200 Go

Mobility Radeon 9000

119.9

82.6

81.8

73.8

|
0
|
24
|
48
|
72
|
96
|
120
|
144

As the resolution is pushed up we see the GeForce4 4200 Go distance itself from the Mobility Radeon 9000. The lead offered at 1280x1024x32 in Jedi Knight 2 grows to 11% over the ATI product. Performance is well off the desktop GeForce4 Ti 4200, however, performing 47% slower.

Jedi Knight 2 v1.03 - JK2FFA
High Quality 1600x1200x32 (aniso off)
Desktop GeForce4 Ti 4200

GeForce4 4200 Go

Desktop GeForce4 MX 440

Mobility Radeon 9000

106.7

77.2

60.8

51.8

|
0
|
21
|
43
|
64
|
85
|
107
|
128

1600x1200x32 is the resolution that the GeForce4 4200 Go is able to shine at in Jedi Knight 2. As the Mobility Radeon 9000 system becomes more and more video card bound, the GeForce4 4200 Go system takes off. Now the NVIDIA part runs 49% faster than the ATI one since the video system acts as a much smaller bottleneck. The GeForce4 4200 Go also dominates the desktop GeForce4 MX 440 in this situation, running 27% faster.



Anti-alised Performance - Unreal Tournament 2003

Unreal Tournament 2003 Flyby
800x600x32 2x Anti-Aliasing Enabled
Desktop GeForce4 Ti 4200

GeForce4 4200 Go

Mobility Radeon 9000

Desktop GeForce4 MX 440

132.1

105.4

71.2

62.2

|
0
|
26
|
53
|
79
|
106
|
132
|
159

Match the GeForce4 4200 Go up with Unreal Tournament 2003 and turn on anti-aliasing and you will see some pretty exciting numbers, as we quickly found out. We ran the Unreal Tournament 2003 Flyby with 2x anti-aliasing enabled and were extremely pleased with the results. At 800x600x32, the GeForce4 4200 Go offered 48% greater performance than the Mobility Radeon 9000. The chip was also able to easily beat the desktop GeForce4 MX 440 and fall only 25% behind the desktop GeForce4 Ti 4200.

Unreal Tournament 2003 Flyby
1024x768x32 2x Anti-Aliasing Enabled
Desktop GeForce4 Ti 4200

GeForce4 4200 Go

Mobility Radeon 9000

Desktop GeForce4 MX 440

97.8

69.7

46.2

40.0

|
0
|
20
|
39
|
59
|
78
|
98
|
117

The numbers are equally impressive at 1024x768x32. The GeForce4 4200 Go outperforms the Mobility Radeon 9000 by 51%, a number that is truly astounding.

Unreal Tournament 2003 Flyby
1280x960x32 2x Anti-Aliasing Enabled
Desktop GeForce4 Ti 4200

GeForce4 4200 Go

Mobility Radeon 9000

Desktop GeForce4 MX 440

64.9

43.6

29.0

26.1

|
0
|
13
|
26
|
39
|
52
|
65
|
78

Not surprisingly, the GeForce4 4200 Go does well with anti-aliasing enabled at 1280x960x32. Although frame rate numbers are beginning to drop, the GeForce4 4200 Go still beats the Mobility Radeon 9000 by 50%. It seems that the GeForce4 4200 Go is made to run with anti-aliasing enabled.



Conclusion

As the performance numbers show, the GeForce4 4200 Go and the NV28M core bring mobile gaming to a whole new level. NVIDIA may have not been first to market with DirectX 8 support in a laptop (ATI holds that prize), but they are the first to announce a mobile graphics chip based completely on the current high-end performance part. The NV28M looks promising, does it not?

The answer to that is not so easy actually. There is no question that the performance of the GeForce4 4200 Go leaves little to be desired. The chip regularly outperforms the Mobility Radeon 9000 and by no small amount; in today's most stressful gaming situation, Unreal Tournament 2003, the GeForce4 4200 Go performed, on average, 39.5% better than the Mobility Radeon 9000 in the resolutions we tested at. Turn on 2x anti-aliasing and that average jumps to 49.7%. We complained a bit in our Mobility Radeon 9000 review that performance was not drastically better than other mobile offerings out there. We certainly do not have the same complaint for the GeForce4 4200 Go but we do have some complaints.

First off is the heat and power consumption that stems off the fact that the GeForce4 4200 Go is based off the NV28 core. The fact that NVIDIA would not comment on the maximum power that the chip draws suggests that the chip draws an insane amount of power while in full performance mode. It does not matter much while the system is plugged in, but using the chip to its full potential away from a power outlet will probably drain your battery pretty darn fast (something that we will certainly look into in future GeForce4 4200 Go reviews). The GeForce4 4200 Go will be most at home in a desknote or desktop replacement notebook.

Finding a home for the GeForce4 4200 Go is actually the second, and perhaps largest, problem we have with the chip. ATI set a new mobile standard for time to market: notebooks based on the ATI Mobility Radeon 9000 were shipping seven days after the product was announced. Unfortunately for consumers NVIDIA will not be following the same path. We were told to expect the GeForce4 4200 Go to be in laptops towards the end of this year and the start of next year. That is quite a while to wait, especially with the holiday season about to go into full swing. We are also somewhat concerned with what OEMs will opt for a GeForce4 4200 Go solution being that the power and thermal requirements of the chip are very demanding

There is no question that the GeForce4 4200 Go holds the performance crown in the mobile world today, however it only holds this crown on paper. If NVIDIA is able to bring the product to market by the end of this year, they will find themselves in a very comfortable situation; the chip will be the undoubtedly the mobile GPU of choice. However, if OEMs are slow to adopt the GeForce4 4200 Go or if NVIDIA is not able to bring the chip to market on time, they will be faced with some competition. ATI is not ready to roll over and let NVIDIA steal the performance crown from them one more time. There are certainly forthcoming mobile ATI products on the horizon that set to not only perform on par with the GeForce4 4200 Go but offer better power management and thermal characteristics. From what we have seen today, it will be difficult to match or beat the GeForce4 4200 Go performance-wise any time in the near future but ATI remains optimistic.

With performance that can be described as nothing short of breath taking, the GeForce4 4200 Go sets a new standard in mobile gaming. No longer must performance be sacrificed in order to play games on a system that can be moved from point a to point b without hassle. All we have left to do now is hope that GeForce4 4200 Go systems make their way to market soon, as the chip is every mobile gamers dream. With the GeForce4 4200 Go, NVIDIA sets the mobile performance bar higher not by inches but by feet.

Log in

Don't have an account? Sign up now