technique

2023-11-17 10:00:00
#Nvidias #gaming #flagship #worse #speculated #respect #Giant #dimensions #TechFeed.cz

2023-11-17 07:00:00
#RedMagic #Pro #gaming #cannon #adheres #eccentric #futuristic #design #TechFeed.cz

As we informed you on Thursday shortly after the official unveiling in of this articleValve suddenly came out with an upgraded version of the popular handheld console Steam Deck. However, this isn’t anything along the lines of Steam Deck 2.0, but rather a minor, albeit significant, update that brings OLED displayimproved cooling and also a slightly more advanced processor.

Valve boasted in press materials 30-50% battery extension, and as it is well known, the efficiency of the used hardware tends to be the most important part of advances of this type, apart from the battery itself. Not only does the display consume less, but it also uses energy significantly better slightly modified APU.

The manufacturer of the chip for the new Steam Deck is, of course, again AMD. Codenamed SoC Sephiroth it reuses the processor architecture Zen 2 and architecture RDNA 2 for integrated graphics, but what has changed is the manufacturing process.

Source: NvidiaRTX 4080 SUPER disappointing: 20GB memory nowhere and only minimal improvement. How are the other models doing?

While the original “Van Gogh” APU was manufactured with TSMC’s 7nm technology, the new Steam Deck uses OLED more modern 6nm process. Therefore, although the TDP value remains in the same range from 4 to 15 W, the actual consumption during gaming should be lower.

Moving on to the specifications, the chip offers four CPU cores and eight threads. The maximum frequency is 3.45 GHz. The GPU relies on 512 stream processors and a clock speed of 1.6 GHz. The new Steam Deck also boasts a little bit more faster operating memory – 16GB RAM works with 6,400 MT/s instead of the original 5,500 MT/s. Again, we are talking about LPDDR5 technology.

The new Steam Deck OLED will available from November 16 this year. It will be sold in configurations with 512GB SSD and 1TB SSD, while the price is 569 euros and 679 euros (14,000 CZK and 16,650 CZK in conversion).

2023-11-11 10:00:00
#Steam #Deck #OLED #modern #processor #parameters #TechFeed.cz

The last few weeks have been discussing new additions to the family of desktop graphics cards Nvidia Ada Lovelace. The green team should officially rebuild them early next year, during a presentation that is part of the popular trade show CES 2024. The broadcast will be available to watch on January 8, as we informed you here on Thursday.

Specifically, there was talk of the fact that Nvidia will release three new models, i.e. SUPER variants of already existing graphics cards from the RTX 40 series. Specifically, we are talking about RTX 4080 SUPER, RTX 4070 The SUPER a RTX 4070 SUPERwhich will focus on adding CUDA cores, or graphics memory.

But as far as the specifications are concerned, the situation was confused from the beginning. The statements of individual leakers or news portals often differed, so it was not really known what to expect. Trusted foreign server Wccftech however, he has now obviously gotten to all the key information we rank among not only the parameters, but also the details regarding the approximate performance.

RTX 4080 SUPER – 20GB MEMORY NOWHERE AND PERFORMANCE INCREASE ONLY MINIMAL

Let’s start with the best model. RTX 4080 SUPER will bring 10 240 CUDA jader, through the larger AD103 graphics chip in the AD103-400 configuration. 320 TMU and 112 ROP should be available. The cache capacity is then 64 MB. Regarding graphics memory, previously discussed Unfortunately, we will not get 20 GB, and the wider bus with the alleged 320-bit interface will not be present either. So you can count on 16GB of VRAM of the GDDR6X type and a 256-bit bus. The TGP value remains at the same 320 W as in the case of the standard RTX 4080.

If we then focus specifically on performance, the RTX 4080 SUPER is said to be average 3-5% faster than the classic RTX 4080. However, Wccftech notes that some models from Nvidia partners will be able to provide up to a 10% increase in performance.

RTX 4070 TI SUPER – STRANGE NAME BUT HIGHER VRAM

Now let’s look at the parameters of the strangest graphics card of the whole lineup, and perhaps also the strangest effort from Nvidia in recent times – RTX 4070 The SUPER. The confidential designation, which at first even the insiders themselves did not want to believe, was already confirmed in a recent leak, and the closer we get to CES 2024, the more likely the release of this model seems.

The RTX 4070 Ti SUPER will reportedly get it 8 448 CUDA jader, either with the AD103-275 chip or with the AD102-175 chip. Both variants have 48MB L2 cache. It will be a pleasant change 16GB of GDDR6X memory, which will use a bus with a 256-bit interface. Compared to the classic RTX 4070 Ti, owners of the SUPER variant will therefore be able to use 4 GB more VRAM (and with higher throughput). The RTX 4070 Ti SUPER will reportedly deliver 15% increase in performance, which definitely sounds better than the RTX 4080 SUPER. TGP will remain at 285W.

Source: AMDRadeon RX 7900M performed better than the mobile RTX 4080 in the first 3DMark tests. Price also plays in AMD’s favor

RTX 4070 SUPER – NEW POWER CONNECTOR AND LOTS OF CORE

And finally the RTX 4070 SUPER. This card boasts by a high increase in CUDA cores, which will get 7,168 (instead of the original 5,888). As for the GPU specifically, either the AD104-350 or the AD103-175 will be available. This time too, the card will offer 48MB of L2 cache. Unfortunately, neither the memory capacity nor the bus width changes. However, due to the addition of a high number of CUDA cores consumption increases, from 200 W to 220 W.

Despite the fact that the RTX 4070 SUPER will offer more than 20% more CUDA cores than the RTX 4070, the increase performance by only 14% on average.

Wccftech also mentions that it will get the RTX 4070 SUPER new power connector, which will be CEM 5.1 Gen5. Previous RTX 4070 models used either eight-pin or twelve-pin (depending on the AIB approach).

HOW IS THE PRICE ON IT?

Many of you would certainly also be interested in the price (and of course us too), but unfortunately, Nvidia seems to have not yet agreed on specific amounts. Previously, we could only hear that the RTX 4080 SUPER will receive the same price as the classic RTX 4080 at its original release. In the Czech Republic, it could therefore cost something around 36,000 crowns, which will nicely fill the gap between RTX 4080 and RTX 4090.

2023-11-10 16:57:13
#RTX #SUPER #disappoints #20GB #memory #improvement #minimal #models #TechFeed.cz