❌

Normal view

Geekbench 7 chega com benchmark de GPU e testes multi-core reformulados

23 July 2026 at 20:47

A Primate Labs anunciou hoje o Geekbench 7, a geração mais recente de sua popular ferramenta de benchmark, que traz conjuntos de dados maiores e mais exigentes para o processamento de GPUs 1Graphics processing units, ou unidade de processamento gráfico., testes multi-core reformulados, entre outras novidades.

A Primate Labs afirma que o benchmark de GPU reformulado concentra-se em aprendizado de máquina e criação de conteúdo para refletir a forma como as GPUs são utilizadas atualmente. Ele inclui agora testes que rastreiam rostos e aplicam efeitos de filtro em tempo real a vídeos, realizam upscaling de imagens com aprendizado de máquina e desfocam fundos em transmissões de videoconferência.

Já os benchmarks multi-core foram projetados para refletir melhor o comportamento real dos softwares, visto que as novas cargas de trabalho só utilizam múltiplas threads quando as tarefas simuladas exigem.

Geekbench 7
Divulgação/Primate Labs

A Primate Labs adicionou ainda um teste expandido de edição de fotos, com um conjunto mais abrangente de edições. Ademais, a carga de trabalho de biblioteca de fotos oferece suporte à importação e ao processamento de formatos de imagem como JPEG XL e DNG. Além disso, os novos testes de áudio e vídeo agora incluem suporte a AV1 e Opus, além de simular cargas de trabalho de streaming e consumo de conteúdo.

No mais, ele tem a mesma aparência do Geekbench 6 — e a licença Pro continua custando US$100 (com desconto para US$80 até 6 de agosto). A versão Pro permite realizar benchmarks offline — sem envio de dados para o Geekbench Browser — e abrange todas as plataformas desktop compatíveis (macOS, Windows e Linux).

É possível baixar gratuitamente o Geekbench 7 para macOS e outros sistemas operacionais nessa página. A versão para iOS/iPadOS ainda não foi atualizada na App Store.

Notas de rodapé

  • 1
    Graphics processing units, ou unidade de processamento gráfico.

Nvidia RTX 5000 Super GPUs rumored to be ready — but they're on hold, and the reason why makes me nervous about pricing

17 July 2026 at 21:56
  • Nvidia's RTX 5000 Super models are being readied, and board makers have the chips, a new rumor reckons
  • However, these graphics cards are 'on hold' due to the high price of VRAM
  • The Super refreshes pack a lot of VRAM, and so that cost could have an unfortunate impact on MSRPs – and none of this sounds particularly comforting for would-be buyers

Nvidia's RTX 5000 Super refreshes are apparently being readied, but they're on hold for now – and the reason why makes me even more nervous about the pricing of these GPUs when (well, if) they finally emerge.

Of course, these RTX Super models are still rumors – albeit so persistent at this point, there must be at least some validity to them – and the latest assertion comes from VideoCardz.

The tech site has apparently spoken to a source at one of Nvidia's board-making partners who claims that their company has already received 'graphics cards', which presumably refers to GPU chips (which is what Nvidia supplies, along with VRAM, to partners that make the boards).

At any rate, while this board maker supposedly has at least some of the necessary chips, it has been told by Nvidia that the RTX 5000 Super refreshes are now 'on hold'.

Why? Apparently, the delay is down to the price of the video RAM, specifically the new 3GB GDDR7 modules that are required (as per previous rumors) for the refreshed graphics cards.

Compared to the 2GB GDDR7 modules currently in use with Nvidia RTX 5000 models, the 3GB memory chips are around three times more expensive right now. Ouch.

Analysis: Super stumble?

An Nvidia GeForce RTX 5080 sitting on its retail packaging

(Image credit: Future)

This presents an obvious stumbling block because, due to these Super variants being packed with VRAM, the total cost of materials gets expensive quite quickly.

Remember that the grapevine believes that the RTX 5080 Super and RTX 5070 Ti Super will bristle with 24GB of VRAM, and even the RTX 5070 Super will have a weighty 18GB on board. Gamers are going to love the latter as an upgrade to the lean 12GB of the vanilla version – but not if it pushes the price of the GPU through the roof.

Going by what VideoCardz tells us about pricing in US dollars as the market currently stands, in the case of the RTX 5070 Super, the greater quantity of RAM, and more expensive modules, would mean 18GB costs in the order of $360 (maybe a bit more), versus $120 for the current RTX 5070's 12GB loadout, whacking an extra $240 on the price (or the regional equivalent elsewhere).

These graphics cards were already expected to be pricey, but that kind of additional premium lumped on top, this could make them real wallet-worriers.

The good news for those hoping for new GPUs from Nvidia is that at least these RTX 5000 Super models do seem to be happening. It's just that we may have to wait some time for them yet – although another recent rumor did suggest they were due this year. The latter speculation also theorized that an RTX 5060 Super 12GB model was incoming, but that didn't make a lot of sense, and I note that it isn't mentioned in this latest rumor. As ever, don't go lightly with the seasoning on all this.

RTX Spark volta a unir Nvidia com a SEGA 30 anos depois do NV1

17 July 2026 at 12:36

Em 1995, a SEGA investiu 5 milhões de dólares na Nvidia e salvou a empresa da falência. Três décadas depois, Jensen Huang e o CEO da SEGA celebraram a parceria em Akihabara, anunciando que jogos clássicos e novos títulos como o Virtua Fighter Crossroads chegarão aos novos PCs RTX Spark.

The post RTX Spark volta a unir Nvidia com a SEGA 30 anos depois do NV1 appeared first on Tek Notícias.

RAM crisis prompts further RTX 3060 GPU resurrections as Palit launches new 12GB model — I just hope pricing makes more sense than it has done so far

16 July 2026 at 11:57
  • Palit has launched a new RTX 3060 model with 12GB of VRAM
  • This continues a recent theme of old GPUs being resurrected
  • These GPUs may pack 12GB of video RAM at a relatively affordable price, but they don't make sense compared to current-gen pricing

We might not have seen any new graphics cards so far from Nvidia this year — and no RTX 5000 Super refreshes are in sight (although there are rumors again) — but we're continuing to witness the resurrection of some old GeForce GPUs, with Palit officially revealing a new RTX 3060.

That popular workhorse of a gaming GPU has already made something of a comeback, returning to the shelves of retailers like Newegg as an alternative budget graphics card in a RAM-starved climate where prices of many of these current-gen boards are being hiked up (alongside other PC components, of course).

As flagged by VideoCardz, Palit has trumpeted the launch of its new GeForce RTX 3060 Infinity 2 OC graphics card as "the return of a classic", which is certainly one way of putting it.

Although to be fair, even today, the RTX 3060 remains a very popular GPU, ranking second on the Steam hardware survey for June 2026 (behind only the RTX 4060 laptop GPU in a rather bizarre twist I won't get into here).

Palit observes: "Combining a massive 12GB GDDR6 memory capacity with a clean, all-black dual-fan shroud, the Infinity 2 delivers honest, dependable performance without the premium price tag."

Analysis: let's hope for a Palit-able price tag, but so far, RTX 3060 pricing has been out of whack

Palit GeForce RTX 3060 Infinity 2 OC graphics card shown at an angle

(Image credit: Palit)

The key selling point for these RTX 3060 models is that they're loaded with 12GB of video RAM (VRAM), as Palit emphasizes with this new Infinity 2 OC model. Where this is side-stepping the RAM crisis (to some extent) is that these past-gen Nvidia cards use GDDR6 VRAM, a different memory supply to tap than the GDDR7 used in RTX 5000 graphics cards (except the RTX 5050, which goes with GDDR6, but that's the outlier here).

This means Nvidia can get its partners to produce graphics cards that have 12GB of GDDR6 VRAM (a loadout a good few gamers put a priority on securing), but sit at the cheaper end of the market. Due to cost issues, these days it's a lot less tenable to sell lower-end GPUs which stack up the GDDR7 VRAM (like the RTX 5060 Ti with 16GB).

It's sad that the RAM crisis has forced us to this point, but that's the reality of the GPU world at the moment. What's also problematic for me here is that the RTX 3060 might have 12GB of VRAM at a relatively affordable price point, but looking at current pricing in the US as an example, it's not meaningfully cheaper than the RTX 5060 with 8GB.

Yes, you're getting a lean RAM configuration with the latter current-gen graphics card, but it's still a much faster GPU — one that benefits from DLSS 4, too — and only just over 5% more expensive (at the time of writing). At this kind of price differential, the RTX 3060 12GB doesn't make a lot of sense.

If these RTX 3060 models were a lot cheaper, then there might be more of an argument. To be fair to Palit, we don't have a price tag on this new offering yet, but pricing needs to be a lot more compelling than what I'm seeing right now (especially in the UK, where the RTX 5060 8GB is actually cheaper than the 3060 at some retailers, although there isn't much stock of the latter).

Quel est le meilleur PC portable pour jouer en 2026 ?

16 June 2026 at 17:59
En 2026, le choix est tranché : pour la grande majorité des joueurs, un PC portable équipé d'une RTX 5070 entre 2 000 et 2 500 euros offre le meilleur rapport performance/prix du marché. Assez puissant pour encaisser les jeux les plus gourmands (AAA) en 1440p Ultra sans broncher, il laisse la...

Nvidia kills Windows XP-era Control Panel "after 20 years of dedicated service"

27 May 2026 at 16:06

Shiny new Nvidia apps like the GeForce Experience and the "Nvidia app" have come and gone, but the old Nvidia Control Panel and its rotating green Nvidia logo have existed as an option for managing basic settings since it was originally introduced in 2006.

That's ending with version 610.47 of Nvidia's Game Ready and Studio drivers for GeForce GPUs. Nvidia says the old Control Panel will no longer be installed by default, since "all actively supported Nvidia Control Panel features for GeForce users have been modernized and transitioned" to the new Nvidia app.

"The NVIDIA app contains all of the modern functionality of the NVIDIA Control Panel available for GeForce RTX GPUs, and much more, while running faster and more efficiently," writes Nvidia Technical Marketing Content Editor Andrew Burnes in the drivers' release notes.

Read full article

Comments

© Andrew Cunningham

❌