ZOTAC GeForce RTX 3090 Trinity Review
Source: The Verge added 15th Oct 2020
Introduction
ZOTAC today launched its GeForce RTX 3090 Trinity Ampere graphics card. The card combines NVIDIA’s top-dog RTX 3090 GPU with 24 GB of GDDR6X memory, the latest IceStorm 2.0 cooling solution, and Spectra 2 ARGB illumination. ZOTAC also offers an industry-leading 5-year warranty subject to product registration. Unlike NVIDIA’s RTX 3090 Founders Edition with its Dual-Axial Flow-Through cooler, ZOTAC takes a more conventional approach with its IceStorm 2.0 thermal solution. All three fans are where you’d expect them. The cooler is longer than the PCB, so some of the airflow from the third fan flows through, out the backplate.
NVIDIA has taken an unconventional approach to the enthusiast segment with its GeForce RTX 30-series Ampere family. The $700 RTX 3080 launched last week has been labeled “flagship” by NVIDIA, and has been extensively shown beating not just its predecessor, the RTX 2080, by a high double-digit percent, but also the previous-gen flagship RTX 2080 Ti by a fair margin, while at least $500 cheaper. The new RTX 3090, on the other hand, is being launched as a “halo segment” product and extensively compared to the Turing-based TITAN RTX, which launched at $3,000.
What’s also unconventional about the GeForce Ampere series is NVIDIA’s use of a common silicon between the RTX 3080 and RTX 3090—the 8 nm GA102 graphics processor. With the previous generation, the RTX 2080 and its refresh, the RTX 2080 Super, were based on the smaller TU104 silicon, while the RTX 2080 Ti and TITAN RTX were built using the larger TU102 die. Between the RTX 3080 and RTX 3090, NVIDIA left itself plenty of headroom for future product segmentation.
With the RTX 3080 already capable of 4K UHD gaming with raytracing, the RTX 3090 has an interesting market position at its $1,499 starting price, which is about 50% higher than the launch price of the RTX 2080 Ti, but exactly 50% lower than the TITAN RTX. Besides enabling all but two streaming multiprocessors on the GA102 silicon, the RTX 3090 enjoys the full 384-bit wide memory interface of the die—no 352-bit business this time around. NVIDIA took things a notch further by arming the RTX 3090 with a staggering 24 GB of GDDR6X memory clocked at 19.5 Gbps—an astounding 940 GB/s memory bandwidth.
The comparisons to the TITAN RTX begin to explain the main application of the RTX 3090 to consumers as it offers the highest possible performance from the Ampere generation, with 4K UHD gameplay at higher refresh rates than the RTX 3080 can handle, 8K gameplay leveraging DLSS 8K, and “TITAN-class creator performance,” which probably underscores NVIDIA’s decision to give it 24 GB of memory.
NVIDIA carved the RTX 3090 out of the same GA102 silicon the RTX 3080 is based on by disabling just 1 of the 42 TPCs present on the silicon. With 41 TPCs (82 SM), the RTX 3090 enjoys a jaw-dropping 10,496 CUDA cores, 328 Tensor cores, 82 RT cores, 328 TMUs, and 112 ROPs. The GPU Boost frequency goes up to 1695 MHz. NVIDIA leveraged the new 8 nanometer 8FFN silicon fabrication node by Samsung to build the GA102. Ampere represents the 2nd generation of NVIDIA’s path-breaking RTX architecture that introduces real-time raytracing to the consumer segment by combining conventional raster 3D graphics with real-time raytraced elements, such as lighting, shadows, reflections, ambient-occlusion, and global illumination. The 2nd generation also introduces raytraced motion-blur and even has fixed-function hardware just to pull this otherwise difficult effect off. Find more details about the architecture in our NVIDIA Ampere Architecture article.
Unlike the TITAN RTX, which only comes in the reference-design Founders Edition version, the RTX 3090 can be built by partners, who have the freedom to implement their latest premium board designs with the chip. As we mentioned earlier, the Zotac RTX 3090 Trinity in this review comes with the company’s IceStorm 2 cooler that features a long series of aluminium fin-stack heatsinks held together by copper heat pipes, ventilated by three fans that each spin at a speed independent of the others. The card sticks to the reference 1695 MHz GPU Boost frequency, and its memory ticks at 19.5 Gbps (GDDR6X-effective). In this review, we take the card for a spin across our exhaustive list of game tests and compare it to our vast selection of high-end graphics cards to tell you if you should start saving for one.
| Price | Shader Units |
ROPs | Core Clock |
Boost Clock |
Memory Clock |
GPU | Transistors | Memory | |
|---|---|---|---|---|---|---|---|---|---|
| GTX 1080 Ti | $650 | 3584 | 88 | 1481 MHz | 1582 MHz | 1376 MHz | GP102 | 12000M | 11 GB, GDDR5X, 352-bit |
| RX 5700 XT | $370 | 2560 | 64 | 1605 MHz | 1755 MHz | 1750 MHz | Navi 10 | 10300M | 8 GB, GDDR6, 256-bit |
| RTX 2070 | $340 | 2304 | 64 | 1410 MHz | 1620 MHz | 1750 MHz | TU106 | 10800M | 8 GB, GDDR6, 256-bit |
| RTX 2070 Super | $450 | 2560 | 64 | 1605 MHz | 1770 MHz | 1750 MHz | TU104 | 13600M | 8 GB, GDDR6, 256-bit |
| Radeon VII | $680 | 3840 | 64 | 1802 MHz | N/A | 1000 MHz | Vega 20 | 13230M | 16 GB, HBM2, 4096-bit |
| RTX 2080 | $600 | 2944 | 64 | 1515 MHz | 1710 MHz | 1750 MHz | TU104 | 13600M | 8 GB, GDDR6, 256-bit |
| RTX 2080 Super | $690 | 3072 | 64 | 1650 MHz | 1815 MHz | 1940 MHz | TU104 | 13600M | 8 GB, GDDR6, 256-bit |
| RTX 2080 Ti | $1000 | 4352 | 88 | 1350 MHz | 1545 MHz | 1750 MHz | TU102 | 18600M | 11 GB, GDDR6, 352-bit |
| RTX 3070 | $500 | 5888 | 64 | 1500 MHz | 1725 MHz | 1750 MHz | GA104 | 17400M | 8 GB, GDDR6, 256-bit |
| RTX 3080 | $700 | 8704 | 96 | 1440 MHz | 1710 MHz | 1188 MHz | GA102 | 28000M | 10 GB, GDDR6X, 320-bit |
| RTX 3090 | $1500 | 10496 | 112 | 1395 MHz | 1695 MHz | 1219 MHz | GA102 | 28000M | 24 GB, GDDR6X, 384-bit |
| Zotac RTX 3090 Trinity | $1500 | 10496 | 112 | 1395 MHz | 1695 MHz | 1219 MHz | GA102 | 28000M | 24 GB, GDDR6X, 384-bit |