PRICING & DISCOUNTS

We use GHZ as our biling unit ,which is not directly related to the processor frequency. GHz Billing Explanation GHZ is a value benchmarked against the PassMark score, where every 200 PassMark points equal 1 GHZ. This value is further calibrated using actual render times from multiple real-world scenes, enabling it to accurately represent the actual rendering performance of servers with different configurations. A higher GHZ value indicates faster rendering speed. For the same scene, the rendering cost is generally consistent across machines with different GHZ ratings. The numerical value of GHZ is typically larger than the core/thread count.

The platform's pricing unit is the render point. Render points can be purchased on the platform at an exchange rate of $1 USD = 5 render points. This rate may be periodically updated in the future due to fluctuations in currency exchange rates.

Separate billing models are applied to animation rendering and still image rendering

Animation rendering pricing

For animation rendering, billing is based on the actual render time consumed. When utilizing CPU-based rendering for certain applications, it is possible to select machines with high memory configurations. Generally, a larger memory capacity contributes to higher CPU rendering performance. An additional memory premium will be charged for using these high-memory machines.

Below is our animation pricing:

DCC	CPU			GPU
DCC	Base Price	RAM Configurations	RAM Fee	Price	VRAM/GPU Configurations
3ds Max	0.015/GHZ*hour	64,112,160,224,334	+4% Per 16GB Beyond 64GB	0.015/GHZ*hour	2080Ti 22GB
Maya	0.015/GHZ*hour	64,160,334	+4% Per 16GB Beyond 64GB	0.015/GHZ*hour	2080Ti 22GB;3090 24GB;4090 24GB
Houdini	0.015/GHZ*hour	64,160,334	+4% Per 16GB Beyond 64GB	0.015/GHZ*hour	2080Ti 22GB;3090 24GB;4090 24GB
Clarisse	0.015/GHZ*hour	64,160,334	+4% Per 16GB Beyond 64GB	0.015/GHZ*hour	2080Ti 22GB;3090 24GB;4090 24GB
Blender	0.015/GHZ*hour	x	x	0.015/GHZ*hour	2080Ti 22GB;3090 24GB;4090 24GB
Cinema 4D	0.015/GHZ*hour	x	x	0.015/GHZ*hour	2080Ti 22GB;3090 24GB;4090 24GB
Keyshot	0.015/GHZ*hour	x	x	x	x
Unreal Engine	x	x	x	0.024/GHZ*hour	4090 24GB
SketchUp	0.03/GHZ*hour	x	x	x	x
Rhino	0.03/GHZ*hour	x	x	x	x

Note:

For CPU rendering, some applications allow you to select machines with higher memory capacity, which will incur an additional fee.
For GPU rendering, the memory size cannot be selected and is fixed for each kind of GPU server. All GPU machine types are single-card configurations. The 2080Ti model refers to the upgraded version with 22GB of VRAM. GPU rendering charges are based solely on the GPU GHZ used; no fees are applied for CPU resources during GPU rendering.

Still rendering pricing

For still renderings, we employ a Linear Pricing model with a set billing cap. Once charges reach this threshold, rendering may continue (up to 7 days or 168 hours) without incurring additional fees. The rendering pricing structure is as follows:

DCC	Billing Approach	Configurations	Base Price	Linear Pricing	Billing Cap	RAM Configurations	GHZ Range
3ds Max	Time-based billing(Rendering time)	CPU Low-End	0.02/GHZ*hour	>25.00: A maximum price of 25.00	HiFi,Small GI Maps:20 Conservative Optimization:18 Medium Optimization:14 Highly Optimization:10	24,32,48,64	85~115
		CPU Economic	0.026/GHZ*hour	Single node: A maximum price of 25.00 Multi node: Distributed acceleration caps are calculated linearly based on the number of nodes (e.g., 4 nodes = 4 × 25 = 100 Render Points cap)	HiFi,Small GI Maps:25 Conservative Optimization:22 Medium Optimization:18 Highly Optimization:15	24,32,48,64,112	115~160
		CPU High-end	0.06/GHZ*hour	Single node: A maximum price of 25.00 Multi node: Distributed acceleration caps are calculated linearly based on the number of nodes (e.g., 4 nodes = 4 × 25 = 100 Render Points cap)	HiFi,Small GI Maps:25 Conservative Optimization:22 Medium Optimization:18 Highly Optimization:15	24,32,48,64,112,224	>=180
		GPU	0.02/GHZ*hour	>40.00: A maximum price of 40.00	40	x	x
	Resolution-based billing (Pre-render quote)	CPU Limited	0.5/Megapixel	x	x	x	x
	Resolution-based billing (Pre-render quote)	CPU Unlimited	1.0/Megapixel	x	x	x	x
SketchUp	Time-based billing	CPU	0.03/GHZ*hour	>25.00: A maximum price of 25.00	25		x
Rhino	Time-based billing	CPU	0.03/GHZ*hour	>25.00: A maximum price of 25.00	25		x
Keyshot	Time-based billing	CPU	0.02/GHZ*hour	>25.00: A maximum price of 25.00	25	x	x

Note:

For 3ds Max renderings in the table, you can choose one billing method from either Rendering Time or Pre-render Quote when submitting. Under the Rendering Time method, you can further select from four machine types: Low-end, Economic, Mainstream, and High-end. Each machine type also offers configurations with different memory sizes.
When billing by rendering time, you can also select rendering quality: Hi-Fi, Small GI Maps, Conservative Optimization, Medium Optimization, or Highly Optimized. Hi-Fi mode does not optimize parameters. The other four options progressively optimize more rendering parameters, reducing rendering time but resulting in increasingly noticeable differences in rendering quality compared to Hi-Fi mode. For Low-end and Economic machine types, selecting higher optimization levels results in progressively lower capped prices.

Acquiring Render Credits

Rendering on RenderWow consumes Render Points, where 1 point equals $0.20 USD. Render Points can be purchased through the Purchase page after logging into your account.

We also offer special promotions for animation users. When purchasing, you can opt to buy larger quantities at once to receive additional Render points.

Purchase Amount of Rendering Points	500-999	1000-1999	2000-2999	3000-3999	4000-4999	5000-99999999
Bonus Ratio	5%	10%	15%	20%	25%	30%

Server GHZ

The actual rendering performance gap between different servers' “single cores” can reach up to 10 times. Pricing at 1 Render Point/core*hour and 0.1 Render Point/core*hour may result in identical costs. How can pricing be fair and reasonable? GHz is the optimal approach.

1. Fundamental Principle. Servers with faster rendering have higher hourly rates, while slower rendering servers have lower hourly rates.

2. Determinants of Server Rendering Speed. A server's rendering speed is determined by its processor model, number of cores, and memory size, with processor performance being the primary factor. Key determinants of processor performance include architecture, manufacturing process, clock speed, core count, and thread count. Architecture and manufacturing process define the processor generation, e.g., 1st Gen i7, 2nd Gen i7, and 3rd Gen i7. When other metrics like clock speed, physical cores, and threads are equal, newer-generation processors offer faster speeds. Within the same architecture, higher clock speeds yield better performance, and more cores generally mean higher performance. However, comparing processors across different architectures cannot be done directly by counting cores or threads. This means it's possible for a server with significantly more cores to perform worse than one with fewer cores. Beyond making extremely “rough” judgments based on processor generation, clock speed, physical cores, threads, and memory size, a more accurate approach is to use internationally recognized software for direct server performance testing. This method enables fair comparisons between servers of different architectures and represents a more scientific evaluation method.

3. Methods for measuring server rendering speed. For CPU rendering, Renderwow utilizes PassMark software to evaluate server rendering performance. The official website for PassMark software is http://www.cpubenchmark.net/, which contains extensive publicly available PassMark scores for various server configurations. Renderwow measures server performance in GHZ units, where 1 GHZ ≈ 200 PassMark points. Simultaneously, Renderwow evaluates actual rendering performance based on real-world scenarios. It adjusts (downward) values derived from PassMark scores to ensure the calibrated GHz accurately reflects practical rendering capabilities. For GPU servers, since VRAM size significantly impacts rendering speed, we calibrate GPU performance by referencing OctaneBenchmark scores while considering cost.

4. Renderwow's pricing per hour of rendering on a single server. Renderwow measures rendering performance across different server configurations using GHZ metrics, and indirectly prices all configurations through GHZ-based pricing. For example, Renderwow's 3dsmax animation rendering base price is 0.015 Render Point/GHZ*hour. Thus, rendering for one hour on an E5-2680 V4 (28-core) server costs 85GHZ * 0.015 Render Point/GHZ*hour = 1.53 Render Point/hour.

The table below lists typical Renderwow server configurations and their corresponding GHZ values.

CPU Servers The model of the rendering server may change; this table may not be up-to-date or complete
CPU Model	Total thread cout	Frequency(GHz)	Renderwow GHZ
Intel Xeon E5-2670	32	2.60 GHz	85GHZ
Intel Xeon Cascade Lake 8255c	24	2.50 GHz	91GHZ
Intel Xeon E5-2680 v2	40	2.80 GHz	110GHZ
Intel Xeon E5-2680 v4	56	2.40 GHz	160GHZ
Intel Xeon Cascade Lake 8255c	48	2.50 GHz	160GHZ
Intel Ice Lake 83XX	64	2.80 GHz	291 GHZ
AMD EPYC 9xxxx	48	2.25 GHz	301 GHZ

GPU Servers The model of the rendering server may change; this table may not be up-to-date or complete
GPU Model	VRAM	Memory Size	Renderwow GHZ
2080Ti	22GB	128GB	300 GHZ
3090	24GB	114GB	530 GHZ
4090	24GB	114GB	750 GHZ