News Posts matching #GA107

Return to Keyword Browsing

ASUS RTX 3050 TUF Gaming its First TUF Gaming VGA with RGB Lighting

ASUS is giving final touches the to the TUF Gaming GeForce RTX 3050 graphics card. This is the company's first TUF Gaming product based on the RTX 3050, which was, until now, given the ASUS ROG Strix and DUAL OC custom treatment. What's more interesting is its cooling solution, which appears to be the first TUF Gaming graphics card with RGB LED lighting—a change from the functional, industrial look of TUF Gaming graphics cards.

This is also the smallest iteration of the TUF Gaming cooling solution the company debuted with the RTX 30-series "Ampere," and features a dual-fan setup (compared to triple-fan on every other TUF Gaming product from this generation). The card comes with a strip of RGB LEDs lined along the top of the card, at the edge of the backplate. It's being reported that this RTX 3050 card is based on the smaller "GA107" silicon, instead of "GA106," and hence comes with a lower typical board power than GA106-based RTX 3050 cards, with no difference in performance.

NVIDIA GA107-based GeForce RTX 3050 is Real, Comes with 11% Lower TDP, Same Specs

When NVIDIA launched the GeForce RTX 3050 "Ampere" based on the "GA106" silicon with specifications that could be fulfilled with the smaller "GA107," we knew that the company could eventually start making RTX 3050 boards with the smaller chip, and they did. Igor's Lab reports that RTX 3050 cards based on GA107 come with a typical board power of 115 W, which is about 11 percent lower than that of the GA106-based cards (130 W).

There's no difference in specifications between the two cards. Both feature 2,560 CUDA cores across 20 streaming multiprocessors, 80 Tensor cores, 20 RT cores, and a 128-bit wide GDDR6 memory interface, holding 8 GB of memory that ticks at 14 Gbps data-rate (224 GB/s bandwidth). The GA106 and GA107 ASICs share a common fiberglass substrate, and hence are pin-compatible for the convenience of board partners, with the latter having a smaller die, so any cooling solution designed for the launch-day RTX 3050 should work perfectly fine with those based on GA107.

NVIDIA GeForce RTX 2050 & MX550 Laptop Graphics Cards Benchmarked

The recently announced NVIDIA GeForce RTX 2050, MX570, and MX550 Ampere graphics cards have recently been benchmarked in 3DMark TimeSpy. The RTX 2050 and MX570 both feature the Ampere GA107 GPU with 2048 CUDA cores paired with 4 GB and 2 GB of 64-bit GDDR6 memory respectively. The MX550 uses the TU117 Turing GPU with 1024 CUDA cores running at 1320 MHz paired with 2 GB of 64-bit GDDR6 12 Gbps memory. The RTX 2050 and MX570 performed similarly in the 3DMark TimeSpy benchmark achieving a graphics score of 3369 while the MX550 scores 2510 points. These new laptop graphics cards will be officially launching in Spring 2022.

NVIDIA MX550 Rumored to Feature GA107 GPU with 2 GB of GDDR6 memory

The NVIDIA MX550 has allegedly surfaced as part of a new Lenovo laptop in a Geekbench listing paired with an Intel Core i7-1260P 12 core, 16 thread processor. The card is described as a "Graphics Device" in the Geekbench listing however according to ITHome this is actually the upcoming MX550 entry-level mobile graphics card. The card is supposedly based on the Ampere GA107 GPU with 16 Compute Units and 128 CUDA cores paired with 2 GB of GDDR6 memory. The MX550 is the successor to the MX450 launched in August 2020 and should offer a roughly 15% performance increase according to the Geekbench OpenCL score. We have limited information on the availability of the card or the remainder of the MX500 series except that NVIDIA may officially announce them sometime early next year.

NVIDIA Announces A2 Entry-Level Edge Tensor Core GPU Based on Ampere Architecture

NVIDIA has today launched the latest addition to its data-centric graphics card lineup, called A2 Tensor Core GPU. The company envisions the A2 Tensor Core GPU as an entry point of accelerators that power machine learning and artificial intelligence at the edge. It is based on Ampere architecture and built on Samsung's 8 nm node. The GPU has GA107 GPU with 1280 CUDA cores enabled, where the full GA107 has 3072 cores. This GPU has only 1280 CUDA cores because it is configured as a low-power, low-profile body with a configurable TBP (total board power) of 40-60 Watts, depending on the customer use case.

It directly replaced the previous generation T4 accelerator, an edge device, and promises to bring 20-30% higher performance in the same power envelope. As far as memory, it has 16 GB of GDDR6 memory running on a 128-bit bus and 200 GB/s memory bandwidth. The GPU core clocks at 1440 MHz base and 1770 MHz boost frequency, which outputs 4.5 TeraFLOPs of FP32 compute. A2 represents a significant upgrade over the previous T4 because it offers a 60% better price to performance ratio and 10% better power efficiency than the last generation. Overall, this release represents an excellent upgrade for customers that need basic AI processing at the edge, and no, gaming is not supported on these as video output is disabled.
Return to Keyword Browsing
May 19th, 2022 01:44 EDT change timezone

New Forum Posts

Popular Reviews

Controversial News Posts