– Next-Gen-graphics card: Unknown Nvidia GPU showed up in the Geekbench

0


Both Geekbench results put the day in focus, because they have been created by the not-yet-released Nvidia graphics card. They have now come to the end of 2019 at the latest, but for the time being at the end of the week noticed. In the case of GPUs, this can be the place of work) for Next-Gen Generation of the music industry.

These are the facts

The two entries in the Geekbench come from the two different graphics cards or settings. The fastest model, it is equipped in accordance with the Reference point 118 processing., the frequency is specified with a 1.11 GHz, and the memory configuration of 24 GB of memory. In the model, the most to be slow, it depends on the 108 Units of processing, and the work of the 1.01-GHz, and 4Seven GB, you can use a large memory.

Geekbench result with 118 Units in the processing of
Geekbench results for the 118-Units of the processing to the image: Geekbench)
Geekbench results for the 108 Units of the processing
Geekbench results for the 108 Units of the rendering (image: Geekbench)

The faster the card is OpenCL Score 184.096 points, and the slowest in the 141.654 points. To put that in perspective, The GeForce RTX 2080 It comes to 129.019 point, the Titan’s V-132.601 and professional, from one place to Another V100 on-Line Version of 153.741 points. The fastest variant, which truncates words, 42 per cent better than the nvidia GeForce RTX 2080 tokyo It is 20% better than the Tesla V100, slower and slower, I have to give you the current one But the Top model, however, it has failed.

What to look for in the entries?

These are the facts but how they are interpreted? First of all, an important question is whether the results are real, to do just that. With the security can not be answered. For example, the odd configuration of the memory of 47 gb may be a bug in the limit of the Tools. As soon as it is possible, however, that Nvidia has made the experiences, for the purpose of, for example, the ALUs and the memory of it has been turned off. So, now, you can check that the values are consistent. This would be the case with a healthy dose of scepticism is seen.

When Nvidia announced a new GPU-Generation, the proportion of the ALUs through the “Unit-of-Computing, and Nvidia’s GeForce products, the “Streaming Multiprocessor”, it has not been changed, as well as in 118 Pa at 7.552 ALUs, 108 Pa at 6.912 ALUs (each CU has 64 ALUs in the Back, and the Turing machine). The GeForce RTX in 2080 It can be to rely on the 4.352 ALUs, the Tesla V100 to 5.120. Plus it would be a little less than 50 per cent, in comparison with a professional solution.

The specified clock rates are very low

The specified clock rate of 1.11 GHz and 1.01 GHz are quite low. Geekbench reports, usually in the middle of the Round, in the case of the Tesla-V100 1.455 MHz. Since it seems unlikely that Nvidia needs to lower the stroke is so strong that it would eat up the extra performance is gross of close to it, it is not the most likely of the models to work with for the final blow or it will just read incorrectly. It would have been set too low, it would have to be in the performance, however, is able to be improved.

The storage capacity should be the fastest model, the 23.8 GB, the slow-46.8 MB. It may be just a little too close to it, because, especially on the experimental status of the new Hardware, it’s a whole lot that can be played around. 47 it should be, in the end, however, 48. Both of these abilities can be used with GDDR6, as well as HBM2E to understand and, right now, Nvidia is offering the Quadro RTX 8000 for a video card that has 48 gigabytes (GB).

At the end of the first quarter, the Situation couldn’t be more clear

At the end of the matter, the results of the Consumer and the professional, and the products come as of yet. The Board is currently of the professional, from one place to Another the products in the store, the points alone expansion to up to 48 GB. Due to the expansion of the memory, there seems to be a self, to a potential close to the Titan that is way too high. In addition, the new employee will be Generating in the near future, you will be presented with the in-house exhibition on the global tobacco epidemic at the end of the first quarter.

The patterns of Consumption of the GeForce series, it seems certain, however, is likely just a few months later, according to the value of the Cross-entries from October and November would be too soon. Add to that, the unrealistic large storage capacity. If it is, in fact, the Pro-Generator of the current, even though the final Name is still uncertain), the question arises, to what extent do the Consumers of the products or technical adjustments, which can be obtained. Why Nvidia fits in with the products for your HP/KI and the market is becoming more and more strong, on their requirements, games, the, and, for example, the processing of the programs may have different requirements, and the need to achieve the best results, as well as other pieces of Hardware.