In a traditional news digest, Bloomberg product specialist journalist Mark Gurman shared his views on the company's range of new products, which will be equipped with different versions of M2 processor.Mac Pro workstations, in his opinion, will get M2 Ultra and M2 Extreme processors that will combine 24 to 48 processing cores and 76 to 152 graphics cores.Image source: ArleThe author believes that processors for the Apple Mac Pro will provide support for up to 256 GB of memory and will be two or four times more productive than the M2 Max.In fact, according to the source, Apple is already testing one of the Mac Pro configurations with the new processor.It offers 24 processing cores (16 performance and 8 economy), 76 graphics cores and 192 GB of memory.The prototype runs macOS Ventura 13.3.It should be recalled that the Ventura 13.0 version is due to debut this week.MacBook Pro notebooks with 14-inch and 16-inch displays, respectively, will feature the M2 Pro and M2 Max processors.The latter will feature 8 performance processing cores and 4 economy cores, as well as 38 graphics cores.Provision is made for the availability of 64 GB of memory.Similar characteristics will be available in versions of notebooks with both sizes of diagonal display.Mac mini, according to the source, will also be updated, it will get the M2 processor in the configuration already known from the MacBook Air and 13-inch MacBook Pro, offering 8 processing cores and 10 graphics.The company is also testing a version of the Mac mini with the M2 Pro processor, which will offer increased cores.The new generation Mac Pro, based on the M2 Ultra and M2 Extreme processors, won't go on sale until 2023, according to Mark Gurman, but the updated MacBook Pro and Mac mini could debut either in November, January, or next spring.At least, that was the announcement frequency for previous generations of Apple products.
NVIDIA will compensate manufacturing partners for the cost of packaging for the \"unannounced\" GeForce RTX 4080 12GB graphics cards
As reported by sources VideoCardz two partners working with NVIDIA, reported the company's intention to compensate them for the cancellation of RTX 4080 graphics cards with 12 GB of memory, at least - in terms of the cost of packaging for them.Source image: NVIDIAI'm talking about those partner companies that have already released boxes for the promised graphics cards, but never waited for deliveries.It is reported that the boxes will be collected and destroyed, as finished products are likely to come out under another name.According to sources, the manufacturer intends to simply forget about the intention to release the corresponding version of the GeForce RTX 4080 - the usual sticker with corrected data, pasted over the finished boxes already released sample, apparently, will not be enough.It is not known yet whether the company will compensate other costs, including, for example, work on packaging.As you know, NVIDIA unexpectedly \"de-announced\" the promised GeForce RTX 4080 12GB, it is not excluded that the card will still be released in a series of GeForce RTX 4070 Ti or 4070.In this case, the cost of packaging and preparation for shipment is very significant, in addition, the models will have to be removed from the former packaging, re-stickers with serial numbers and codes, and possibly even re-flash them, although the latter may be possible to avoid - just change the name in the driver, and PCI Device ID can remain the same.Canceling GeForce RTX 4080 12GB - not the first time that NVIDIA changed its plans at the last moment.For example, instead of GeForce RTX 3080 Ti with 20 GB of memory later it was decided to release a version with 12 GB.Nevertheless, some GeForce RTX 3080 Ti 20GB cards appeared later on Russian market, which indirectly confirms the fact that the cards had time to go into limited edition.However, in the case of GeForce RTX 4080 12GB technical problems may be much less, because there are no serious changes planned for the GeForce RTX 4080 12GB variant conversion with another name, if they are made at all.
AMD and NVIDIA will have to reduce prices of current generation graphics cards - GeForce will get cheaper than Radeon
Both NVIDIA and AMD find themselves in the situation when it is almost time to release new generation graphics cards, but the stock of current solutions is still too high.It is obvious that without measures to stimulate sales, there is no way to cope with surplus graphics processors, and the most effective step should be an official price reduction.Sources believe that here NVIDIA will surpass AMD already at the end of this month.Image source: NVIDIASeptember in former years in the PC market was associated with the beginning of the business season after the vacations and vacations.By the fall, AMD and NVIDIA have been traditionally preparing new graphics solutions in the gaming segment, and so far only AMD has openly confirmed its intentions not to betray these traditions, and about NVIDIA plans to release a family of GeForce RTX 40 (Ada Lovelace), many sources have recently said only unofficially and with an eye to the situation with inventory overstocking.Chinese media, citing its own sources in the supply chain, reported that at the end of August, AMD and NVIDIA will take steps to reduce the price of graphics cards of existing generations, with the latter to surpass the scale of generosity of the competitor.Already on sale you can find graphics cards GeForce RTX 3090 Ti at prices 50% lower than the recommended, and the plight of NVIDIA can be recognized by the rumors of intentions to use stocks GA102 chips to restart the release version GeForce RTX 3080 with 12 GB of memory.Note that recently NVIDIA has presented preliminary results of the last quarter.The company chose to warn of a 44% sequential and 33% year-over-year decline in revenue in the gaming segment.So a sale by NVIDIA is definitely needed.
The cooling system of the reference GeForce RTX 4090 Ti showed up in photos - like the GeForce RTX 3090 Ti, but even bigger
Photos of the cooling system allegedly belonging to the reference version of the GeForce RTX 4090 Ti graphics gas pedal appeared on the Chiphell Chinese forum.According to rumors, the GeForce RTX 40-series graphics cards will be based on the Ada Lovelace architecture.Image source: ChiphellAccording to the source of the photos, they show a huge radiator of the future NVIDIA flagship.Visually, it is very similar to the cooler reference model GeForce RTX 3090 Ti Founders Edition, where there is no customary shroud cooling system, and uses a large aluminum radiator with lots of ribs and a pair of fans located on both sides. The new generation video card radiator has a larger contact area for GPU and memory chip cooling.The GeForce RTX 3090 Ti has a similar one as shown in the photo below. GeForce RTX 3090 Ti Founders Edition cooling system.Image source: TechPowerUpIt looks like the reference versions of the GeForce RTX 40-series graphics cards will retain the V-shaped shape of one of the PCB faces used in the GeForce RTX 30-series models. Other photos show the metal heatsink frames of the reference version GeForce RTX 4090 Ti.Unfortunately, the quality of the provided photos does not allow us to determine their authenticity. It should be remembered that almost two years ago, before the official announcement of the GeForce RTX 3080, on the Chinese forum Chiphell there was a similar leaked photos of GeForce RTX 30-series graphics cards cooling system.According to the latest rumors, the announcement of GeForce RTX 4000 series graphics gas pedals is expected in mid-July this year.Graphics cards GeForce RTX 4090 class will get more than 16,000 cores CUDA, GDDR6X memory with data transfer rate of 21 Gbps, and their power consumption level will be 450 watts.
New leak of Navi 31, 32 and 33 GPU specs for upcoming AMD Radeon RX 7000 series
Rumors are surfacing again from Greymon55 informant, according to his data the flagship Navi 31 GPU, for RDNA3 architecture-based graphics cards includes not 15360 stream processors, but only 12288.It turns out that the statement about the power of 100 TFLOPs made a few days ago either turned out to be incorrect, or some details are still undisclosed.Navi 31 graphics processor with 6 shader cores, 12 shader arrays and 48 workgroup processors will be produced with 12288 stream processors, which is 20% fewer cores compared to the previously announced 15360 cores.The same is true for the Navi 32, which will be produced with 8192 stream processors instead of 10240 cores.For the Navi 31 this means 4096 cores instead of 5192, but it should be noted that Greymon55 previously mentioned fewer cores.With the updated specs of the Navi 31, the power will likely drop to 73 TFLOPs, and this at a GPU clock speed of 3 GHz.It remains to wait for information about Navi 32 and Navi 33 clock speeds.It should be mentioned that the release of the new third-generation Navi GPUs is still a few months away.And now there has been a pretty big leak about the specifications, which have been exchanged for over a year.So unless there are other reliable sources claiming otherwise, we can assume that these specs are closer to final.The specifications in the table above, represent only the complete GPU configuration, which cannot be guaranteed when released for the Radeon RX 7900 / 7700 / 7600 series.AMD may reduce the specs of each GPU to improve the performance of its Radeon RX 7000 graphics cards.
AMD to unveil Radeon RX 7000 graphics cards tonight
AMD will unveil Radeon RX 7000 graphics cards on the latest RDNA 3 architecture graphics processors tonight. The start of the presentation,...