A loophole in the server backend of the German Corona warning app enabled remote code execution (RCE). The actual app was not affected. According to SAP, the loophole was not exploited. Personal data could not be accessed via the interface.
Although the contact recognition of the Corona warning app works decentrally on the smartphones, but the distribution of the random identifiers of infected people to the app runs via a central server.
Sylvester Tremmel has looked at the source code of the Corona warning app and explains the background to the gap
The gap was in the interface to Transmission of positive test results to the server. This is publicly available and does not require authentication. Only a TAN is required for the transmission of a positive result. The TAN is checked by an additional verification server, but only after it has been processed by the vulnerable code. So no positive corona test was necessary for using Lücke.
In the worst case, it would have been possible to write your own code the server and possibly smuggle in falsified results. In a blog post, SAP writes that the elimination of the vulnerability shows that “the open source and community process works perfectly and makes a decisive contribution to the security of the operation of the Corona warning app.”
Fund by GitHubs Security Lab The source code of the app and the of the server are publicly on GitHub. The vulnerability was found by chance by GitHubs Security Lab. Its researchers had looked for patterns for “Java Bean Validation” gaps in order to integrate the recognition patterns into the platform’s automatic code-scanning tools. During the search, they also found the hole in the code for the servers of the Corona warning app. There, the output of an error message was interpreted as a code.
After the discovery, the discoverers reported the vulnerability to SAP. Four days later it was closed for the time being and version 1.5.1 of the server was released. After tests by SAP and BSI, a second, more reliable fix was installed. The current version is 1.6.0 of the server.
A fork of the German Corona warning app is also being used in Belgium. However, the fork was created before the hole appeared in the code of the Corona app server. GitHub recommends that all countries that operate public or private forks of the server also apply the fix.
Cisco’s network management software DNA Spaces Connector, Integrated Management Controller (IMC) and IoT Field Network Director (FND) can be attacked via security holes classified as ” critical “. In addition, attackers could stealthily sneak into Webex meetings. Security updates are available.
Critical vulnerabilities Most dangerous is the vulnerability (CVE – 2020 – 3470) in IMC. Problems can arise here when processing HTTP requests, which results in memory errors. If this is the case, attackers could execute Schacode with root rights in the underlying operating system without authentication.
By successfully exploiting the vulnerability (CVE – 2020 – 3531) In FND, remote and unregistered attackers could access and change the back-end database due to insufficient authentication during REST API calls.
Since the management console of DNA Spaces Connector does not sufficiently check user input (CVE – 2020 – 3586 ), attackers could execute their own commands on vulnerable devices.
Webex spy Due a vulnerability (CVE – 2020 – 3419) attackers could be present at Webex meetings without appearing in the participant list. Hidden as a “ghost” from the other participants, attackers could eavesdrop on audio and video content, among other things. According to a warning from Cisco, this is only possible if attackers have access to meetings in the form of participation links and a password. Accordingly, the vulnerability is “only” classified with ” medium “.
Further vulnerabilities concern Expressway Software, Secure Web appliance and telepresence CE software. Here, attackers could, for example, gain unauthorized access to information or acquire higher user rights.
List sorted in descending order by threat level:
Integrated Management Controller Multiple Remote Code Execution IoT Field Network Director Unauthenticated REST API DNA Spaces Connector Command Injection IoT Field Network Director SOAP API Authorization Bypass IoT Field Network Director Missing API Authentication Webex Meetings and Cisco Webex Meetings Server Ghost Join Webex Meetings and Cisco Webex Meetings Server Unauthorized Audio Information Exposure Expressway Software Unauthorized Access Information Disclosure IoT Field Network REST API Insufficient Input Validation Webex Meetings API Cross-Site Scripting IoT Field Network Director Cross-Site Scripting Telepresence CE Software and RoomOS Software Unauthorized Token Generation Secure Web Appliance Privilege Escalation Webex Meetings and Cisco Webex Meetings Server Information Disclosure IoT Field Network Director Information Disclosure IoT Field Network Director Improper Access Control IoT Field Network Director File Overwrite IoT Field Network Director Improper Domain Access Control IoT Field Network Director Unprotected Storage of Credentials (of)
(Pocket-lint) – Call of Duty: Modern Warfare felt in many ways like a soft reboot of the COD franchise – it took some of the more iconic characters, time-shifted them, and told a more modern story that ultimately had all of the old bombast with a slightly more cynical air.
That pattern seems to have repeated itself in the form of Black Ops Cold War, doing the same for the Black Ops timeline with less revision of the continuity – but also far fewer changes to actual gameplay. This is a flashback Call of Duty, in effect, for better and worse.
Campaigns of old
Cold War picks up where the original Black Ops left off, broadly, before the move into futuristic battles of previous sequels. It’s set during the clandestine period of its title and sees some of the original game’s soldiers – the likes of Mason, Woods, and Hudson – joined by a new ringleader, Adler. It’s about as typical COD fare as you can get, with rogue agents and missing nuclear weapons once more at the heart of things.
Where Cold War makes an interesting change is in the addition of interludes in your team’s Berlin safe-house – where you can chat with your colleagues and get to know them for a refreshing change of pace.
Similarly, a couple of missions that dial back the fighting to let you stealthily and autonomously choose and complete objectives are also enjoyable. That these moments of calm don’t throw up meaningful choices that actually impact the narrative isn’t a great surprise, but still slightly disappointing.
Best upcoming PS5 games: PlayStation titles to anticipate
Best upcoming games for Xbox Series X: The top next-gen games
Other nice additions are a couple of optional side missions, locked behind codes that you can only decipher by finding (easily located) evidence in main missions. It’s a nice way to incentivise chasing collectibles, and it’s also smart that the codes rotate their answers across playthroughs to stop you from just looking up the solutions.
The rest of the time, this is the same old stuff: stop-and-shoot gameplay that can be hugely fun when you’re in the swing of things, and annoying when doltish enemy AI or poor checkpointing throws you off your stride.
Activision
The set-pieces are as explosive as ever, and things look pretty great throughout, with nice lighting choices in particular showing off what next-gen consoles can do if you’ve got one. The PS5 version’s use of DualSense’s adaptive triggers is also a promising sign for how nicely this immersive tech can work.
Still, there’s an undercurrent of political idiocy that rears its head repeatedly to remind you that America is the goodie, even if everything happening is its fault, which is eerily similar to the 2019 Modern Warfare’s message. This series’ fetish for the benefits of interventionist military action is looking more dated and amoral than ever.
Get your kicks online
Of course, the beating heart of COD is its online multiplayer. Treyarch, the game’s developer, has put a conscious rewind on Black Ops. The entire game is built on an older engine than Infinity Ward used for Modern Warfare, which means that some of that game’s introductions are gone.
Activision
That means there’s no mounting mechanic for recoil management, no tactical sprinting, and slightly less fluid movement overall. Whether these are missed is a matter of taste, but given that Modern Warfare felt like a big step forward for the franchise, their absence is perplexing. The differing engine also means you’d better get used to a new audio mix for footsteps and far fewer doors that you can open than your eyes might assume.
There’s a slightly meager range of maps at launch, with eight arenas for 6v6 modes – which is the core of the game – and these are slightly hit-or-miss in quality. Some work well in Team Deathmatch but are terrible in Domination or Hardpoint, while the lookout for Search and Destroy isn’t great at present.
At least the maps are fairly varied visually speaking – although there are still clearly loads of locations from the campaign that could be added down the line, and some more classic three-lane locations wouldn’t go amiss.
Another odd change this time is that lobbies are disbanded after every game. Ostensibly this helps the matchmaking engine keep you in pools of players close to your skill level, but it makes for a slightly stilted experience and robs you of a sense of community with your fellow players. It also means that performing well in a match will often throw you into a more punishing server next, a frustrating loop that can feel unrewarding. That’s made a bit more annoying by weapon unlocks that are supremely gradual, too.
Activision
Cold War’s multiplayer feels like a game of swings and roundabouts at launch. On the one hand, the gun balance doesn’t feel right, with the MP5 SMG absolutely dominating at silly ranges across all game modes. Then again, things are softened by the return of scorestreak rewards which don’t reset when you die, making it far easier for newer players to earn rewards like care packages and more.
For every hard edge there’s a soft cushion somewhere. Whether Black Ops can maintain players’ interest compared to the long-lived Modern Warfare will remain to be seen, especially once their progression systems both tie in with Warzone imminently.
Zombies are back
The last piece of the puzzle in a Treyarch COD is, of course, Zombies, and the mode returns to its roots in Cold War. For now, there’s one main map, Die Maschine, which is a reimagining of the original location from years ago.
Activision
It’s a frantic and fun mode, as always, played in squads of up to four or solo, and exploring the large facility takes plenty of time and skill, plus a little luck. Getting back into the rhythm of trailing zombies around until you’re ready to go to town on them doesn’t take too long, and there’s loads of gory fun to be had.
Zombies is actually the perfect tonic after or during a frustrating session in multiplayer, and we’d recommend exploring on your own initiative or under the guidance of a friend rather than looking up the best ways to find your way into the facility – keep some surprises back for yourself!
You’ll slowly uncover the full range of random weapon boxes, pack-a-punch upgrades, armor and ammo mods, and more to fight off the horde. We can only hope that Treyarch’s got another map in the pipeline at some point, to keep things fresh.
Activision
On PlayStation, there’s an additional time exclusive mode for a year, Onslaught, but it’s a simpler horde-style mode on multiplayer maps, so hardly a seismic loss to the Xbox and PC crowd.
Verdict
Zombies is probably the most unequivocal success in Cold War – and the mode in which a return to its roots is the most obvious win. With a slightly undercooked multiplayer component and a campaign that only does so much to rewrite the script, the value of the overall package will depend on where your personal focus lies.
Still, if you’re looking for a simple, explosive story to play through, and don’t mind some grinding as part of an undeniably addictive multiplayer component, there’s still absolutely loads of fun to be had here.
At the Supercomputing 2020 trade show, Intel revealed some additional details about its upcoming 3rd Generation Xeon Scalable ‘Ice Lake-SP’ processors due to be formally launched in Q1 2020. Among other things, Intel shared some preliminary performance numbers and said that a server running two 32-core Ice Lake CPUs is faster than a machine based on two of AMD’s 64-core EPYC processors.
In addition to disclosing some performance figures of Ice Lake-SP processors, Intel also confirmed that it is broadly sampling its next-generation code-named Sapphire Rapids CPUs that are made using its 10nm Enhanced SuperFin process technology.
(Image credit: Intel)
Intel’s upcoming 3rd Generation Xeon Scalable processor will feature up to 32 cores as well as eight memory channels that will support up to 6TB of DDR4-3200 memory/Intel Optane persistent memory. In addition, the new processor will support PCIe 4.0 connectivity as well as a number of new security technologies, including crypto accelerators.
(Image credit: Intel)
Intel says the new Sunny Cove microarchitecture coupled with special-purpose accelerators and extended memory support will allow the new processors to be significantly faster than predecessors which rely on the rather outdated Skylake microarchitecture. The new CPUs are made using Intel’s second-gen 10 nm process technology that is also used for client Ice Lake products.
According to Intel, a server running two third-gen 32-core Xeon Scalable ‘Ice Lake’ CPUs clocked at 2.20 GHz and paired with 256 GB of DDR4-3400 memory is 20% ~ 30% faster in select benchmarks (LAMMPS, NAMD STMV, Monte Carlo) when compared to a server powered by two 64-core AMD EPYC 7742 processors clocked at 2.25 GHz and equipped with 256 GB of DDR4-3200. That’s despite the fact that Intel’s system has half the number of cores.
(Image credit: Intel)
Intel yet has to formally introduce its 3rd Generation Xeon Scalable ‘Ice Lake’ processors, but a number of its customers, including Korea Meteorological Administration, The Max Planck Computing and Data Facility, The National Institute of Advanced Industrial Science and Technology (AIST), The University of Tokyo, Osaka University, and Oracle have already announced plans to deploy the new CPUs for their HPC needs.
In addition to disclosing some performance figures of Ice Lake-SP processors, Intel also confirmed that it is broadly sampling its next-generation code-named Sapphire Rapids CPUs that are made using its 10nm Enhanced SuperFin process technology.
In response to a new wave of attacks that have compromised standard approaches to Windows security, Microsoft announced its Pluton security processor that will reside inside of future consumer chips from both AMD and Intel – but it’s built using a technology that AMD pioneered with its custom processors for Xbox game consoles. It’s also based on a standard feature with AMD’s EPYC server processor chips. Now Intel will adopt the same approach to help secure PCs.
The new collaboration between Microsoft, AMD, Intel, and Qualcomm will enable more robust security that helps prevent physical attacks and encryption key theft while protecting against firmware attacks. Microsoft will also use the technology to streamline firmware updates via Windows Update.
The Pluton security processor comes as a result of recent new attack vectors that indirectly compromise the Trusted Platform Module (TPM), which has long been the preferred method of securing PCs from potential threats. The TPM, a small secondary chip inside the system that stores encryption keys for services like Bitlocker and Windows Hello, is still robust enough to protect encryption keys but malicious actors have learned how to penetrate the bus that connects the TPM to the CPU through physical attacks, thus compromising a system.
(Image credit: Microsoft )
Microsoft says that security must be built directly into the processor to prevent those attacks, hence the Pluton security processor. The new approach to securing the system isn’t really new at all, though – AMD pioneered the in-built security processor with the AMD Security Processor (ASP) in the Xbox game console back in 2013. This in-built 32-bit ARM Cortex-A5 processor is sandboxed from the rest of the processor, thus protecting it from attacks with exploits like Spectre, and provides secure encryption key generation and management to enable a hardware root of trust.
Image 1 of 5
Image 2 of 5
Image 3 of 5
Image 4 of 5
Image 5 of 5
AMD uses this same approach for its EPYC server chips and its commercial processors. For the Xbox, AMD’s secure processor runs Microsoft’s Pluton Security Processor firmware to enable tight integration between Microsoft’s software and AMD’s security hardware. AMD says it will be first to enable the same feature on all of its future client CPUs and APUs, though it hasn’t provided a specific timeline for the release.
Meanwhile, Intel says that it will continue to leverage its Hardware Shield feature in vPro, which only comes with specific commercial SKUs, but now also enable the Microsoft Pluton security processor to provide multiple root of trust options. Intel hasn’t provided a timeline for its adoption of the feature but says that it will support Pluton “as a choice.” It’s a bit unclear if Intel means that it hasn’t been forced to adopt the security processor approach or if the statement means the company will not enable the feature on all chips, thus providing customers with the ‘choice’ to purchase more expensive processors with the feature – much like with its vPro-enabled chips.
The Pluton processor, essentially Microsoft firmware/IP that runs on the sandboxed security processor, will emulate a TPM to maintain broad compatibility with APIs like BitLocker and System Guard. Pluton also uses a Secure Hardware Cryptography Key (SHACK) technology that prevents exposing cryptographic keys, even to the Pluton firmware itself, which will ultimately protect user information from physical attacks.
Finally, the Pluton processor secures the firmware updating process, streamlining the Windows Update process to provide a more unified and consistent method to update system firmware. This has become more important as a slew of security vulnerabilities have necessitated a rapid cadence of new firmwares to plug security holes like Meltdown and Spectre, but the current delivery system is fragmented. By building this functionality into the processor and using it to enable Windows Update to update firmware securely, Microsoft hopes that it, and the silicon vendors, can react to vulnerabilities quicker.
All Intel processors and the associated chipsets contain a so-called Management Engine (ME), which Intel now calls Converged Security and Management Engine (CSME). The latest version CSME 15. 0 debuted in the current mobile processors “Tiger Lake” of the eleventh Core-i generation.
In the 15 – page “CSME Security Whitepaper” Intel now explains many functions of the CSME as well as the innovations of the version 15. 0, which among other things improves the protection against attacks and manipulation. Accordingly, some security algorithms have been strengthened in order to make decryption with future quantum computers more difficult (Post-Quantum Cryptography, PQC). This applies to AES (now 256 bit), RSA key (3072 Bit), Elliptic-Curve Cryptography (ECC – 384) and SHA-2 digests (also 384 Bit).
ODCA instead of EPID Intel has also built in new functions to be able to react reliably with firmware updates in the event of an attack on essential cryptographic signatures and certificates. This is one of the reasons why Intel is replacing the Enhanced Privacy ID (EPID) previously used in the CSME with an On-Die Certificate Authority (ODCA). With the help of the ODCA, after a firmware update, the CSME can generate new security certificates for internal firmware functions without establishing a server connection.
The CSME generates and manages numerous cryptographic Key.
(Image: Intel)
The CSME 15. 0 is also the basis of the Control-Flow Enforcement Technology (CET) introduced with Tiger Lake to protect against attacks with Return-Oriented Programming (ROP) through Shadow Stack (SHSTK) and Indirect Branch Tracking (IBT). In addition, the CSME manages the keys for Total Memory Encryption (TME).
With CSME 15. 0, Intel also isolates different internal CSME functions from each other even more in order to limit the effects of a successful attack on one of these functions. As usual, according to its own information, Intel tries to keep the code scope of the basic security functions (Trusted Computing Base, TCB) as small as possible in order to reduce the probability of errors (minimum TCB). The CSME uses the Minix operating system with microkernel architecture.
This year only deliveries to important partners are to be made, in large quantities in the first quarter 2021. Zen 3 CPUs are coming first for supercomputers.
(Image: AMD)
AMD will be the server offshoot of Zen 3 – Epyc 7003 alias Milan – no longer this Introduce year. The delivery of the first processors to partners is said to have already started, but AMD only sells larger quantities from the beginning 2021. The chip manufacturer has scheduled the presentation for the first quarter 2021 – only then is with the specifications and reliable performance data.
So far AMD only presented desktop processors with Zen 3 architecture: The Ryzen – 5000 -CPUs (code name Vermeer) especially increase the single-threaded performance. The server versions with up to 64 should be like the 10 – Kerner Ryzen 9 5950 X benefit less from the improvements, since the thermal design power (TDP) is limited with the high core numbers. After all, all efficiency improvements come from the architecture – chip contract manufacturers continue to produce the CPU chiplets with 7 nanometer technology.
Chip shortage AMD apparently cannot keep up with the production of Zen-3 chip sets. The first four models Ryzen 9 7002 X, Ryzen 9 5900 X, Ryzen 7 5800 X and Ryzen 5 5600 X are hardly available worldwide. Delivery contracts with server partners are likely to have priority at the moment – among other things, the first parts of the mother-of-pearl supercomputer are to be included this year with Zen-3-Epycs go online. As a result, AMD will or can only deliver larger quantities to smaller server manufacturers from 2021 .
Most recently, Intel’s first 10 – nm server generation Ice Lake-SP, which competes with AMD’s Epyc 7002 (Rome) and Epyc 7003 (Milan) actually until the end 2020 should appear. According to the Taiwanese website Digitimes Intel is supposed to introduce the to the beginning 2021. (mma)
IBM is to operate the specialist service that stores the electronic prescriptions prescribed by the doctor on the server in encrypted form until they are redeemed by a pharmacy. This specialist service can then download the prescriptions into the insured person’s prescription app and into the pharmacy’s administration systems. The e-prescription is to be introduced nationwide in July 2021.
IBM has been awarded the contract by the Gematik project company. The source code of the specialist service is to be published before it is put into operation, as is the corresponding security report. The insured person’s app for the e-prescription will continue to be developed directly at Gematik in accordance with the requirements of the Patient Data Protection Act (PDSG), and Bundesdruckerei will presumably come into play as the identity provider.
IBM had applied for lot 1, with which an operator of the specialist service was sought. In a second lot, the identity provider is still being sought who will read the electronic prescriptions encrypted by the doctor and authenticate both the doctors and the pharmacies. Presumably this will be the Bundesdruckerei subsidiary D-Trust.
500 Millions of prescriptions In Germany over 500 millions of prescriptions are issued and processed each year, many of them follow-up prescriptions for which the patients actually do not Have to go to the practice. Larger general practitioners’ practices issue between 400 and 600 prescriptions per day. These are to be issued as e-prescriptions from July 2021, whereby the mandatory issue of simple recipes for 2022 it’s planned. Later, in level 2, narcotics prescriptions and in level 3 intensive care prescriptions and the green prescription are to be sent electronically to the specialist service.
The e-prescription is a QR code that is encrypted and stored on the prescription server. From there it will be picked up encrypted for transport by the pharmacy where the insured person would like to redeem the prescription. There is a smartphone app with a directory of all offline and online pharmacies.
In addition to deciding on a pharmacy, it should be possible to do so Sending an e-prescription to another person’s app, for example if you are not allowed to leave the house during quarantine. This shows the QR code in the pharmacy to a scanner, which then queries the server and picks up the e-prescription.
The AMD Instinct MI 100 is the first compute accelerator card with the new CDNA architecture and is produced in 7 nanometer technology at TSMC. AMD wants the PCIe 4.0 card and its 32 GByte HBM2 memory Nvidias A 100 compete and has not only revised the compute units known from the GCN architecture, but also built significantly more of them into the chip. The card should be available for system integrators from 6400 US dollars and undercut so Nvidias A 100 PCIe version clearly that from 10. 700 Euro is available.
A lot of flops , a lot of honor When it comes to technical specifications, AMD is cracking it again – probably also to win over competitor Nvidia with the A 100 – to limit accelerators in some areas.
The CDNA chip “Arcturus” has on the MI 100 120 active compute units (CUs) and even if AMD confirmed on request that this is the full expansion u act, recordings of the chip suggest that there are 8 CUs more. We asked AMD again for clarification and are currently waiting for the answer.
Block Diagram AMD Instinct MI 100
(Image: AMD)
The 120 As with the closely related graphics chips with GCN architecture (GraphicsCore Next), CUs each have 64 Shader cores, which results in ALUs for the entire chip 7680. Together with a maximum boost clock rate of 1480 MHz a throughput of , 07 TFlops with single precision (FP 32). As befits an HPC accelerator, the FP is 62 – rate at half, so 11, 5 TFlops – and not only above the 10 – TFlops brand but also round 19 Percent above the comparable value of Nvidias A 100 – Accelerator in the SXM4 -Format.
Arcturus-Die-Shot : CDNA accelerator with 108 active compute units and four HBM2 stacks
(Image: AMD)
The rake beast is fed as it was with the MI 50 / MI 60 of four HBM2 stacks. These hold 8 GB each and are marked with 1200 MHz clocked, what a transfer rate of 1, 228 TByte / s is good. An 8 MByte level 2 cache (6 TByte / s) is supposed to cushion the memory access. From the registers to the HBM2, everything is secured by ECC (SECDED).
In addition to PCI-Express 4.0, every MI 100 – Map with three infinity links à 92 GByte / s – together therefore 276 GByte / s. This means that there are now directly networked groups of four from MI 100 possible, which can form a coherent memory area.
Matrix Core Engines: A bit of Tensor The Compute Units of MI 100 are similar to those of the previous generation Graphics Core Next, but have been further upgraded by AMD for compute use. In order to achieve a higher throughput with matrix-matrix multiplications, AMD has expanded the circuits and register ports and calls the result the Matrix Core Engine.
AMD has a different approach than Nvidia with their tensor cores. The Core Matrix Engines work consistently with full FP 23-Accuracy. However, their maximum throughput is lower and they are not suitable for FP 60 – calculations. Therefore, it is difficult to compare the maximum throughput between the two approaches. Who consistently on full FP 23 – Accuracy is dependent on AMD, whoever also uses the alternative TF 32 or lower accuracy, the Nvidia accelerators promise more performance.
What both approaches have in common is that they use the BFloat format 16 (BF 16) support which with the value range of FP 32 (8-bit exponent) with the precision of FP 16 (7-bit mantissa, plus 1 sign bit) combined and is a de facto alternative to full FP 32 established in AI training Has. AMD gives in the CDNA white paper for BFloat 16 Indeed 10 Bit mantissa and 5 bit exponent to what actually FP 16 corresponds to.
Instinct MI 100 (PCIe) A 100 (SXM) Tesla V 100 Tesla P 100 Manufacturer AMD Nvidia Nvidia Nvidia GPU CDNA Arcturus A 100 (Ampere) GV 100 (Volta) GP 100 (Pascal) CUs / SMs 120 108 80 56 FP 32 Cores / SM 64 64 64 64 FP 32 Cores / GPU 7680 6912 5120 3584 FP 64 Cores / SM 32 32 32 32 FP 64 Cores / GPU 3840 3456 2560 1792 Matrix Multiply Engines / GPU
(Matrix Core Engine / Tensor Cores) 480 432 640 – GPU Boost Clock 1502 N / A 1455 MHz 1480 MHz Peak FP 32 / FP 64 TFlops 23, 07 / 10, 54 19, 5 / 9.7 15 / 7.5 10, 6 / 5.3 Peak Tensor Core TFlops – 156 (TF 32) / 312 (TF 32 Structural Sparsity) 120 (Mixed Precision) – Peak Matrix Core Engine TFlops 46, 1 (FP 32) – – – – Peak FP 16 / BF 16 TFlops 184, 6 / 92, 3 312 / 312 (624 / 624 Structural Sparsity) 125 / 125 21 ,1 / — Peak INT8 / INT4 TOps 184, 6 / 156, 6 624 / 1248 (1248 / 2496 Structural Sparsity) 62 / – 21,1 / — memory interface 4096 Bit HBM2 5120 Bit HBM2 4096 Bit HBM2 4096 Bit HBM2 Memory size 32 GByte 40 GByte 16 GByte 16 GByte Memory transfer rate 1 , 2 TByte / s 1,55 TByte / s 0.9 TByte / s 0, 73 TByte / s TDP 300 Watt 400 Watt (SXM) 300 Watt 300 Watt Transistors (billion) N / A 54 Billion 21, 1 billion 15, 3 billion GPU The Size n / a 826 mm² 815 mm² 610 mm² Manufacturing 7 nm 7 nm 12 nm FFN 16 nm FinFET + AMD Instinct MI 100 with complex IF connection and soldering points for up to three eight-pole connections.
(Image: AMD)
Without Radeon, without displays After Nvidia’s Tesla and Quadro waiver, AMD is now also changing the branding of the accelerator cards and removing the Radeon from the product name. The card is only called AMD Instinct MI 92 – whereby the number, unlike earlier Instinct cards, is no longer for the FP 16 – computing power is available.
In order to a lot of computing power in the TDP framework of 276 Watts, AMD has, according to its own information, omitted many hardwired functions that are necessary for a graphics card in the first CDNA chip “Arcturus”. This includes the rasterization units, tesselator hardware, special graphics buffers, the blending units in the raster output stages and the display engine. The MI 15 do not use it and Crysis does not run on it either.
Not removed However, AMD has the video engines, i.e. the specialized decoders and encoders. Reason: Machine learning is often used to analyze video streams or image recognition.
One of the first rack Server comes from Supermicro (Dell, HPE and Gigabyte also have similar products in their range). With the real cards it is noticeable that only an eight-pin connector is sufficient.
The macOS Big Sur update has been causing some older MacBook Pros to get stuck on a black screen during installation, with no way that users can find to fix it. The issue seems to be most common on late-2013 and mid-2014 models of the 13-inch MacBook Pro. Owners of other Macs have reported Big Sur installation issues, too — and some say they’ve been able to fix those issues — but those two models of MacBook Pro seem to be having the most severe problems for now.
The issues were spotted by MacRumors, which saw complaints of frozen laptops in its forums, on Reddit, and on Apple’s support site. There’s no clear indication of what could be causing the problem, and simple ways of fixing it — like resetting NVRAM or booting in safe mode — either can’t be accessed or don’t resolve the issue, users have said.
Big Sur has had an unusual number of launch issues. High-demand for the update seems to have slowed Apple’s servers to a crawl, which has led to a widespread slowdown of newer Macs due to a little-known feature that verifies with Apple that apps don’t have malware before each time they run. The server issues also made it impossible for some users to install the update, with the OS showing an installation error as people tried to launch the installer.
Given the black screen problems some users have been having, perhaps the download issues were for the best. If you’re thinking about updating, it may be wise to give it a few weeks while the bugs are worked out. The Verge has reached out to Apple for comment.
AMD announced its 7nm Instinct MI100 GPU today, along with a slew of design wins from the likes of Dell, HPE, and Supermicro. The Instinct MI100 marks the first iteration of AMD’s compute-focused CDNA architecture. The new architecture offers up to 11.5 TFLOPS of peak FP64 throughput, making the Instinct MI100 the first GPU to break 10 TFLOPS in FP64 and marking a 3X improvement over the previous-gen MI50. It also boasts a peak throughput of 23.1 TFLOPS in FP32 workloads, beating Nvidia’s beastly A100 GPU in both categories.
As expected from a data center GPU, the PCIe 4.0 card is designed for AI and HPC workloads and also supports AMD’s second-gen Infinity Fabric, which doubles the peer-to-peer (P2P) I/O bandwidth between cards. The Instinct MI100 also supports AMD’s new Matrix Core technology that boosts performance in single- and mixed-precision matrix operations, like FP32, FP16, bFloat 16, INT8, and INT4. That tech boosts FP32 performance up to 46.1 TFLOPS.
The cards come with 32GB of HBM2 memory, spread across four stacks, that provides up to 1.23 TB/s of bandwidth. AMD claims the cards offer up to 1.8x to 2.1X more peak performance per dollar compared to Nvidia’s A100 GPUs.
The cards boast up to 340 GB/s of aggregate throughput over three Infinity Fabric links and are designed to be deployed into quad-core hives (up to two per server), with each hive supporting up to 552 GB/s of P2P I/O bandwidth. AMD also announced that its open source ROCm 4.0 developer software now has an open source compiler and unified support for OpenMP 5.0, HIP, PyTorch, and Tensorflow.
Image 1 of 7
(Image credit: AMD)
Image 2 of 7
(Image credit: AMD)
Image 3 of 7
(Image credit: AMD)
Image 4 of 7
(Image credit: AMD)
Image 5 of 7
(Image credit: AMD)
Image 6 of 7
(Image credit: AMD)
Image 7 of 7
(Image credit: AMD)
The card has a 300W TDP and comes in the standard PCIe Add-In Card (AIC) form factor with two eight-pin connectors for power. Given the data center focus, the card lacks display outputs, and the passively-cooled card has a rear I/O shield with a large mesh for efficient airflow.
Peak Clock
Stream Processors
TDP
HBM2 Memory
Memory Bandwidth
PCIe Interface
FP64
FP32
Matrix FP32
Matrix FP16
INT4/INT8
bFloat16
7nm instinct MI100
1502 MHz
7680 (120 CU)
300W
32GB
1.23 TB/s
4.0
11.5 TFLOPs
23.1 TFLOPS
46.1 TFLOPS
184.6 TFLOPS
184.6
92.3
7nm Instinct MI50
1725 MHz
3840 (60 CU)
300W
32GB
1.024 TB/s
4.0
6.6 TFLOPS
13.3 TFLOPS
13.3 TFLOPS
26.5 TFLOPS
Nvidia A100 (PCIe)
1410 MHz
6912
250W
40GB
1.555 TB/s
4.0
9.7 TFLOPS
19.5 TFLOPS
156 TFLOPS (Tensor)
312 TFLOPS
624 / 1,248 (Tensor core)
624 / 1,248 (Tensor core)
Nvidia A1000 (HGX)
1410 MHz
6912
400W
40GB
1.555 TB/s
4.0
9.7 TFLOPS
19.5 TFLOPS
156 TFLOPS (Tensor)
312 TFLOPS
1,248 (Tensor core)
1,248 (Tensor core)
AMD dialed back the MI100’s peak clock rate to 1,502 MHz, down from 1,725 MHz with the previous-gen MI50, but doubled the number of compute units up to 120. The company also improved memory bandwidth to 1.23 TB/s. The net effect of the improvements to the CDNA architecture delivers a 1.74X gain in peak FP64 and FP32 throughput, and a whopping 3.46X improvement in matrix FP32 and 6.97X gain in matrix FP16, both due to AMD’s new Matrix Core technology that enhances the CUs with new Matrix Core Engines optimized for mixed data types.
AMD’s MI100 beats the Nvidia A100 in peak FP64 and FP32 throughput by ~15%, but Nvidia’s A100 still offers far superior throughput in matrix FP32, FP16 and INT4/INT8 and bFloat16 workloads.
AMD touts that the MI100 rivals the 6 Megawatt ASCI White, the world’s fastest supercomputer in 2000 that weighed 106 tons and provided 12.3 TFLOPS of performance. In contrast, the MI1000 brings power down to 300W, weighs only 2.56 pounds, and dishes out 11.5 TFLOPS of performance.
AMD Instinct MI100 CDNA Architecture
Image 1 of 7
(Image credit: AMD)
Image 2 of 7
(Image credit: AMD)
Image 3 of 7
(Image credit: AMD)
Image 4 of 7
(Image credit: AMD)
Image 5 of 7
(Image credit: AMD)
Image 6 of 7
(Image credit: AMD)
Image 7 of 7
(Image credit: AMD)
AMD split its architectures into the RDNA platform for graphics-focused work (gaming) and CDNA for compute workloads (HPC/AI workloads) so it could deliver targeted enhancements to each respective architecture. Naturally, that means the CDNA designs come without many of the traditional fixed-function blocks needed for graphical work, like rasterization, tesselation, graphics caches, blending, and the display engine. The CDNA architecture does retain some logic for HEVC, H.264, and VP9 decoding, which is important for machine learning workloads that focus on object detection.
The 7nm Instinct MI100 marks the first iteration of the CDNA architecture and comes with a PCIe 4.0 interface that supports a 16 GT/s link (32 GB/s bi-directional) to the CPU. AMD isn’t sharing the size of the 7nm die, which revision of 7nm the company uses, or the transistor count, but we do know the 120 enhanced CUs are split into four compute engines. Each CU features a Matrix Core Engine that boosts computational throughput for various numerical formats, which AMD describes as:
“The classic GCN compute cores contain a variety of pipelines optimized for scalar and vector instructions. In particular, each CU contains a scalar register file, a scalar execution unit, and a scalar data cache to handle instructions that are shared across the wavefront, such as common control logic or address calculations. Similarly, the CUs also contain four large vector register files, four vector execution units that are optimized for FP32, and a vector data cache. Generally, the vector pipelines are 16-wide and each 64-wide wavefront is executed over four cycles.”
“The AMD CDNA architecture builds on GCN’s foundation of scalars and vectors and adds matrices as a first class citizen while simultaneously adding support for new numerical formats for machine learning and preserving backwards compatibility for any software written for the GCN architecture. These Matrix Core Engines add a new family of wavefront-level instructions, the Matrix Fused Multiply-Add or MFMA. The MFMA family performs mixed-precision arithmetic and operates on KxN matrices using four different types of input data: 8-bit integers (INT8), 16-bit half-precision FP (FP16), 16-bit brain FP (bf16), and 32-bit single-precision (FP32). All MFMA instructions produce either 32-bit integer (INT32) or FP32 output, which reduces the likelihood of overflowing during the final accumulation stages of a matrix multiplication.”
The matrix execution unit handles MFMA instruction and reduces the number of register file reads because many matrix multiplication input values are re-used.
The shared 8MB L2 cache is physically partitioned into 32 slices (twice as much as MI50) and is 16-way set associative. Overall, the 32 slices deliver up to 6TB/s of aggregate throughput. The memory controllers support 4- or 8-high stacks of ECC HBM2 at 2.4 GT/s, with an aggregate theoretical throughput of 1.23 TB/s. That’s 20% faster than prior-gen models.
AMD Second-Gen Infinity Fabric
Image 1 of 2
(Image credit: AMD)
Image 2 of 2
(Image credit: AMD)
AMD’s CPU-to-GPU Infinity Fabric has proven to be a key advance that has helped the company win numerous exascale contracts. This technology enables shared memory/cache coherency between CPUs and GPUs to reduce latency, boost performance, and reduce power draw by reducing the amount of data movement inside the system.
The second-gen Infinity Fabric links operate at 23 GT/s and are 16-bit wide, just like with the previous-gen, but the latest revision supports a third link to enable quad-GPU configurations. This new design works best in quad-GPU hives, with a typical two-socket server supporting two hives – one per CPU.
These hives operate in a fully-connected topology, whereas the previous accelerators used a ring topology. The new topology boosts performance during all-reduce and scatter/gather operations, among others.
Overall, AMD’s second-gen Infinity Fabric dishes out twice the peer-to-peer (P2P) I/O bandwidth, with up two 340 GB/s of throughput per card (with three links). A quad-GPU hive provides up to 552 GB/s of P2P I/OP throughput, showing that it doesn’t scale linearly.
The fully-connected topology and shared address space is a key advantage for AMD over Nvidia and has led to several notable exascale supercomputing contracts. Notably, Nvidia has yet to announce an exascale supercomputer contract, but AMD’s accelerators have already enjoyed broad uptake in the supercomputing and HPC realms.
Image 1 of 7
(Image credit: AMD)
Image 2 of 7
(Image credit: AMD)
Image 3 of 7
(Image credit: AMD)
Image 4 of 7
(Image credit: AMD)
Image 5 of 7
(Image credit: AMD)
Image 6 of 7
(Image credit: AMD)
Image 7 of 7
(Image credit: AMD)
AMD also announced that fully-qualified OEM systems with the Instinct MI100 will be available from major OEMs, like Dell, Gigabyte, HPE, and Lenovo, by the end of the year.
Last week, a number of Mac users had trouble opening apps — a problem that seemed to be caused by an Apple security protocol responsible for checking that software comes from trusted sources. The slow-down prompted some to criticize Apple for collecting too much information about users’ activities; criticism which the company has now responded to with promises that it will change how these security protocols work in future.
Apple announced the changes via its support pages, adding a new “Privacy protections” section to a page entitled “Safely open apps on your Mac” (as spotted by iPhone in Canada). Apple says a service known as Gatekeeper “performs online checks to verify if an app contains known malware and whether the developer’s signing certificate is revoked.” It goes on to clarify how Apple currently uses the data, and outlines new safeguards that are being introduced over the next year.
Complaints about this verification process focused on a protocol known as the online certificate status protocol service, or OCSP. This security feature checks that an app’s developer certificate hasn’t been revoked before it’s allowed to launch. The outage lead to scrutiny of Apple’s practices, most notably by security researcher Jeffrey Paul.
In a blog post titled “Your Computer Isn’t Yours,” Paul claimed that this security process means Apple collects a hash of every program a Mac user runs, along with their IP address, over an unencrypted connection. The end result, wrote Paul, is that anyone use a modern version of macOS can’t do so without “a log of [their] activity being transmitted and stored.”
However, not everybody agreed with Paul’s analysis. One blog post by cybersecurity student Jacopo Jannone notes that the data sent to Apple’s OCSP server contains information relating to an app’s developer but not the app itself. It adds that Apple’s Gatekeeper service can send the hash of an executable, but that this is separate to OCSP and happens over an encrypted connection. Apple’s own support page notes that Gatekeeper uses “an encrypted connection that is resilient to server failures.”
In its updated support document, Apple makes clear that security checks it makes when authenticating software do not include a user’s Apple ID or device identity. The company also says it’s stopped logging IP addresses associated with the Developer ID certificate checks. “We have never combined data from these checks with information about Apple users or their devices,” writes the iPhone-maker. “We do not use data from these checks to learn what individual users are launching or running on their devices.”
However, something about these complaints do seem to have registered with Apple, as the company says it’s changing how it handles these checks in the future. Over the next year the company says it will roll out a new encrypted protocol for developer ID certificate checks while adding “strong protections against server failure” — that is, protections against the issues that stopped apps from opening last week. Finally, users will also be given the option of opting out of these security protections all together, a change that seems designed to appease complaints like Paul’s.
Apple first commented on the massive macOS problems last Thursday evening. In connection with the release of macOS 11 alias Big Sur, various servers at the company went on strike for hours, which led to a certificate check that was not possible. Macs that were on the Internet then refused to start numerous apps, and neither a restart nor other measures seemed to help. If the machine was disconnected from the network, the OCSP server request (Online Certificate Status Protocol) was automatically stopped and the apps ran as desired.
How gatekeeper works In a freshly supplemented support document, Apple now writes what the group is actually doing – and even praised improvement. macOS is “designed in such a way that it ensures user security, protects their data and at the same time respects their privacy.” The gatekeeper routine carries out an online check to determine whether an app contains malware and whether its developer signature has been withdrawn. Apple has “never combined data from these reviews with information about Apple users or their devices.” The data obtained during these checks are also not used to “find out what individual users start or run on their devices”.
IP addresses were saved The notarization checks whether an app contains known malware and uses “an encrypted connection that is” stable “against server errors. Why did the outages on Thursday nevertheless occur? Unfortunately, Apple does not explain. Furthermore, the company announces that these security checks “never” included the Apple ID of the user or the identity of the device. However, the company apparently saved the associated IP address until recently, but it now has “ended”. In the future, IP addresses in connection with developer ID certificate checks will no longer be logged, and they will also “ensure that all IP addresses collected are removed from the logs.” This is the first time that Apple is involved to this logging that poses a problem in terms of data protection, extremely.
Apple wants to revise the protocol and offer an opt-out Furthermore, the group announced that it will work in the future with a “new encryption protocol” to check developer ID certificates for their withdrawal – why this is planned – and whether the existing protocol has any weaknesses – says the company Not. In addition, Apple promises “strong protection against server failures” in the future, as well as a new settings menu with which users can completely switch off these security checks (opt-out). Apple did not specify the time for this.
The Supermicro C9Z490-PGW is a solid board to build your socket 1200 system around. The VRMs are more than capable, and once unlocked, performance landed in line with the competition. The appearance may not be for everyone, but many power users will like the 10GbE port and extra PCIe lanes thanks to a PLX chip.
For
Solid 90A power delivery
Server-type features (PXE boot, PLX chip)
Fast 10 GbE and integrated WiFi 6
USB 3.2 Gen2x2 Type-C port
Against
BIOS improved, but still feels clumsy
Long boot times
No integrated RGB lighting
Expensive
Features and Specifications
Supermicro is best known to most for server hardware. But over the last few years, the company has dipped its toes in the gaming realm. We’ve seen X99 and X299, as well as Z170 through Z390-based motherboards from the company, culminating with the C9Z490-PGW we have for review today. Like past iterations, the C9Z490-PGW board contains server-level features including PXE boot, four PCIe slots served by a PLX chip, in addition to parts and features we expect on a gaming motherboard.
With an MSRP of $395, this places the C9Z490-PGW squarely in the crosshairs of mid/upper mid-range Z490-based motherboards. Our PGW comes equipped with the full complement of six SATA ports, two M.2 sockets, Aquantia 10 GbE and Wi-Fi 6, a premium Realtek audio solution and more. There’s also the four PCIe slots that use a PLX chip for additional lanes. These are useful for those using multiple video cards, especially with compute workflows.
Performance on this board was on par with any board that follows intel specifications out of the box. To that end, other motherboards (that go beyond the spec) performed better in extended tests. However, you’ll see that the results weren’t that different overall. If you remove the stock turbo limits, then it is no different. Gaming performance was also on par with the other boards. The C9Z490-PGW happily handled our Intel Core i9-10900K CPU at 5.2 GHz without issue. RAM overclocking wasn’t quite set-it-and-go at DDR4 4000, but with a few tweaks, we ran our kit without issue.
Other features on the C9Z490-PGW include onboard power/reset buttons, five 4-pin fan headers, four SATA ports and two USB 3.2 Gen2 Type-C ports (one Gen2x2/20 Gbps). For those who would like to add RGB lighting to this board, you’re out of luck. There are no onboard headers for lighting, nor any integrated RGB.
Below is a complete list of specifications from Supermicro. Getting past that, we’ll tour around the board in detail.
Specifications – Supermicro C9Z490-PGW
Socket
LGA 1200
Chipset
Z490
Form Factor
ATX
Voltage Regulator
8 Phase (6+2) 90A MOSFETs
Video Ports
(1) HDMI (v2.0a)
(1) DisplayPort (v1.4)
USB Ports
(1) USB 3.2 Gen2x2 Type-C (20 Gbps)
(1) USB 3.2 Gen2 Type-C (10 Gbps)
(2) USB 3.2 Gen2 Type-A (10 Gbps)
(2) USB 3.2 Gen 1, Type-A (5 Gbps)
Network Jacks
(1) 1 GbE
(1) 10 GbE
Audio Jacks
(5) Analog + SPDIF
Legacy Ports/Jacks
✗
Other Ports/Jack
Wi-Fi Antenna
PCIe x16
(4) v3.0 (x16/x0/x16/x0, x8/x8/x8/x8)
PCIe x8
✗
PCIe x4
✗
PCIe x1
(1) v3.0 (x1)
CrossFire/SLI
AMD Quad and 2-Way CrossfireX
DIMM slots
(4) DDR4 4000(OC)
M.2 slots
(2) PCIe 3.0 x4 / SATA + PCIe (up to 110mm)
U.2 Ports
✗
SATA Ports
(4) SATA3 6 Gbps
USB Headers
(1) USB v3.2 Gen 2 (Type-C)
(1) USB 3.2 Gen1
(2) USB v2.0
Fan/Pump Headers
(5) 4-Pin (PWM/DC supported)
RGB Headers
✗
Legacy Interfaces
COM, TPM
Other Interfaces
FP-Audio
Diagnostics Panel
2-character and 4 LED
Internal Button/Switch
Power, Reset, Reset CMOS
SATA Controllers
✗
Ethernet Controllers
(1) Aquantia AQC107 (10 GbE)
(1) Intel I219-V (1 GbE)
Wi-Fi / Bluetooth
Intel AX200 Wi-Fi 6 (802.11ax, MU-MIMO) / BT 5.1
USB Controllers
✗
HD Audio Codec
Realtek ALC1220
DDL/DTS Connect
✗ / ✗
Warranty
3 Years
Starting with accessories, Supermicro includes what’s needed to get started, but not much else. Below is a list of what is included in the box along with the board.
Quick reference guide
Driver disk
(4) SATA cables
Labels for cables
IO plate
Wi-Fi antenna
Image 1 of 3
(Image credit: Tom’s Hardware)
Image 2 of 3
(Image credit: Tom’s Hardware)
Image 3 of 3
(Image credit: Tom’s Hardware)
Compared to most modern boards in this price range, the Supermicro C9Z940-PGW is far from flashy. A simple matte-black PCB sporting black heatsinks on all of the toasty parts. Contrasting this barrage of black are the silver/metal-colored reinforced DRAM and PCIe slots. The board’s appearance isn’t exactly offputting, but at this price point there are better-looking boards available, especially if shrouds and RGB lighting are your thing.
Speaking of RGB, the C9Z490-PGW doesn’t have any integrated on the board. Instead, it has a single white LED illuminating the SuperO name and a small strip above the rear IO. In fact, the board doesn’t have any RGB headers at all. Any RGB lighting will have to come from third-party ecosystems and controllers. That said, if you’re looking for a board like this in the first place, chances are you won’t miss the lack of lighting and basic styling.
(Image credit: Tom’s Hardware)
Looking closely at the top half of the board, we’ll start out on the left side. Here we see the black plastic The C9Z490-PGW uses well-apportioned heatsinks covering the VRMs, including other hot-running chips (like the PLX chip just above the top PCIe slot, for example).
Feeding power to the CPU is an 8-pin EPS plug located between the VRM heatsinks. The first of five 4-pin fan headers are located here, just next to the EPS connector. The second header is found just to the right of the DRAM slots. All fan headers support both PWM and DC control fans up to 1A/12W (the manual doesn’t list any power, so we assume the standard).
Next up are the four reinforced DRAM slots that alternate in grey and black (grey slots are the first to be populated). The locking mechanism is single-sided, which is helpful considering the proximity of the top PCIe slot. To the right of these are two more fan headers along with power and reset buttons. Sliding further down the right edge, we see the 24-pin ATX connector along with a front-panel USB 3.2 Gen2 Type-C header.
(Image credit: Tom’s Hardware)
The C9Z490-PPGW and its Wi-Fi free brother (C9Z490-PG, the W stands for Wi-Fi) come equipped with an 8+2 VRM setup. Power goes through an Infineon XDPE12284C controller (X+Y=8) that feeds six 90A Infineon TDA21490 MOSFETs for Vcore, along with 70A Infineon TDA21535 MOSFETs for the SOC. The phase count may not be as high as others. However, the premium Infineon ICs do a great job at handling our power-hungry Intel Core i9-10900K, even when overclocked to 5.2 GHz.
(Image credit: Tom’s Hardware)
Moving down to the bottom half of the board, we’ll look at the audio, PCIe slot breakdown and SATA storage. Starting on the left side, we see the Realtek ALC1220 IC, along with a few small audio capacitors. Just above this, under a black heatsink, is the Aquantia AQC107 10 GbE chip. There is some separation of audio bits to the rest of the board. But otherwise, there aren’t any covers on the IC to protect against EMI. That said, this implementation using the premium audio codec should be good enough for most users.
In the middle of the board, we see four full-length PCIe slots, each using reinforcement to prevent shearing and protect against EMI. The four slots work in a x16/x0/x16/x0 or x8/x8/x8/x8 configuration due to the PEX 8747 PCIe Gen3 switch that adds lanes (at the expense of some latency) to the CPU. This configuration is suitable for multi-GPU setups — in particular, compute type loads allowing users to stuff the slots full of cards. For gaming, Supermicro lists AMD CrossfireX as supported, but no mention of SLI. In addition to the four full-length slots is a single x1 size slot that is fed from the chipset.
Wedged between PCIe slots 1/2 and 3/4, are two M.2 sockets. The top socket supports drives up to 80mm, while the bottom up to 110mm. The top slot supports a maximum of PCIe 3.0 x4 speeds while the bottom slot is up to PCIe 3.0 x2. Both sockets only support PCIe-based modules as well as support RAID0 and 1. I would like to have seen at least one of these support SATA-based devices, especially considering there are only four SATA ports available.
As just mentioned, the C9Z490-PGW includes four SATA ports that support RAID0, 1, 5 and 10. All ports should remain enabled regardless of how the M.2 sockets are populated. For this type of board, I would like to have seen all six ports available, giving users more choices in saturating their storage options. In this same area is a 2-character debug LED that’s useful during POST to determine where the system may be hanging.
The bottom edge has a slew of headers including front panel audio, USB and more. In addition to the debug LED above, it also sports a 4 LED debug that stays lit if there is an issue during POST (CPU, DIMM, VGA and BOOT). Along with the debug LED, these can help pinpoint issues that occur before reaching Windows. Below is a list of all headers ordered from left to right.
Front panel audio
SMB to PCIe slots
USB 2.0 header
Standby power header
USB 3.2 Gen1 header
TPM header
Debug LEDs
USB 2.0 header
COM port
Front panel header
(Image credit: Tom’s Hardware)
The rear IO is full of inputs and outputs including Wi-Fi, dual Ethernet, audio, USB and video ports. On the video side is a DisplayPort and HDMI port for use with CPUs that have integrated graphics. There are a total of six USB ports out back – two USB 3.2 Gen1 (blue) and four USB 3.2 Gen 2 (one Type-C is Gen2x2, 20 Gbps). Our only concern here is the USB port count sits at six, which may not be enough for some users.
Above the USB ports are the Intel I219-V (left) and the Aquantia 10 GbE LAN ports. The C9Z490-PGW is the least -expensive motherboard that includes a 10 GbE port. The audio stack consists of five analog plugs and SPDIF. Last but not least is the small CMOS reset button.
If you buy something from a Verge link, Vox Media may earn a commission. See our ethics statement.
There is a plethora of streaming services available for anyone with a credit card handy. Hulu and Netflix are general entertainment providers offering licensed content and originals for subscribers. Others, like anime provider Crunchyroll and NFL Sunday Ticket, are targeted at specific interests. Streaming services like Disney Plus and Apple TV Plus are flashy newcomers that are trying to find their place in an overly crowded space.
It’s getting trickier to figure out where to spend your money every month. Netflix and Hulu used to be sufficient; subscribers could watch their favorite shows and popular originals like Stranger Thingsor The Handmaid’s Tale. Now, between live sports, movies, original TV shows, and even nightly news, no single service offers everything.
Take a look at the most popular shows streaming today: Friends and The Office were two of the most-watched shows on Netflix, but Friends has returned to its proper WarnerMedia home now that HBO Max has launched, and The Office will leave Netflix in 2021 and be a staple of NBCUniversal’s new streaming service, Peacock. New CW shows, which would normally arrive on Netflix after their seasons aired on TV, are now destined for either HBO Max or CBS All Access. ER and other NBC favoritesare on Hulu now, but they will eventually wind up on Peacock. Having trouble following along? So are we.
Some streaming services are better suited for your personal needs. Instead of listing every platform under the sun, we’ve listed a few key categories and the best streaming option in those sections. Note: some services may be duplicated depending on the category.
Photo by Robyn Beck / AFP via Getty Images
Amazon Prime Video
Pros: Different types of streaming options, incorporation of retail shopping, original series and popular licensed content
Who it’s for: People who use Amazon to shop regularly
Cost: $8.99 a month for Prime Video only / $12.99 a month or $120 a year for full Amazon Prime
Top original shows: TheMarvelous Mrs. Maisel, Transparent, Good Omens, Bosch
Compatible devices: Smart TVs, Blu-ray players, Roku, Google TV, TiVo, Nvidia Shield, Amazon Fire TV, Fire TV Stick, PlayStation 4, Xbox One, Android devices, iOS devices, Fire Phone, Fire tablets
Apple TV Plus
Pros: Free for anyone who purchased a new Apple product, low cost for standalone subscribers, star-studded shows
Who it’s for: People looking to stay up to date on the conversation
Cost: $4.99 a month, or free for one year for anyone who’s purchased a new or refurbished Apple product recently
Top original shows: Dickinson, The Elephant Queen, For All Mankind
Compatible devices: Amazon Fire TV, Apple TV, Browser, iOS, Roku
CBS All Access
Pros: Low cost, popular original series, big library of content
Who it’s for: Star Trek fans and people looking to binge past classics. CBS All Access is a general entertainment streaming service.
Cost: $5.99 a month for ad-supported and $9.99 a month for ad-free
Top original shows: The Good Fight, Star Trek: Discovery, Strange Angel
Compatible devices: Amazon Fire TV, Android, Android TV, Apple TV, Browser, Chromecast, iOS, Xbox, Roku, PlayStation
Disney Plus
Pros: Exclusive home for Disney TV shows and movies, low cost, entirely family-friendly, large library of shows and movies
Who it’s for: Families and Disney aficionados
Cost: $6.99 a month or $69.99 a year. It can also be bundled with ad-supported Hulu and sports streaming service ESPN Plus for $12.99 a month (see below).
Top original shows: The Mandalorian, Imagineering Story, High School Musical: The Musical: The Series
Compatible devices: Amazon Fire TV, Android, Android TV, Apple TV, Browser, Chromecast, iOS, Xbox, Roku, PlayStation
HBO Max
Pros: Includes HBO content, heavy focus on human curation, incredibly large movie and TV show library
Who it’s for: HBO Max is a general entertainment streaming service. It’s also a great option for anyone looking to upgrade their HBO Now accounts.
Cost: $14.99 a month. People with HBO Now, AT&T TV Now, or standard HBO cable subscriptions can get it for free.
Top original shows: The Wire, Succession, Game of Thrones, The Sopranos, Love Life
Compatible devices: Android, Android TV, Apple TV, Browser, Chromecast, iOS, Xbox, PlayStation
Hulu
Pros: Next-day programming for certain shows, large quantity of shows and movies, reality TV programming, variety of genres to choose from
Who it’s for: Everyone. Hulu is a general entertainment streaming service.
Cost: Ranges from $5.99 for an ad-supported plan to $60.99 a month for a subscription that includes live TV
Top original shows: The Handmaid’s Tale, Castle Rock
Compatible devices: Amazon Fire TV, Android, Android TV, Apple TV, Browser, Chromecast, iOS, Xbox, Roku, PlayStation
Netflix
Pros: Large quantity of movies and TV shows, easy-to-use interface, variety of genres to choose from, interactive programming
Who it’s for: Everyone. Netflix is a general entertainment streaming service.
Cost: Ranges from $8.99 for a basic plan to $15.99 for top-tier
Top original shows: Stranger Things, The Umbrella Academy, You, When They See Us
Compatible devices: Amazon Fire TV, Android, Android TV, Apple TV, Browser, Chromecast, iOS, Xbox, Roku, PlayStation
Peacock
Pros: Live news, some live sports, ad-supported cheaper tier, next-day programming
Who it’s for: Peacock is a general entertainment streaming service
Cost: A free ad-supported tier comes with a smaller portion of 7,500 hours of television and film content. Peacock Premium (the full 15,000-plus hours of content and access to everything) with ads will cost $4.99 a month, but it will be free to Comcast subscribers. Peacock Premium Ad-Free will be $9.99 a month or $4.99 a month for Comcast subscribers.
Top original shows: Brave New World, Intelligence
Compatible devices: Xfinity One, Xfinity Flex, Amazon Fire TV, Apple TV, Chromecast, Roku, iPhone, iPad, browsers, Android phone and tablet
Image: Disney
Disney Streaming Bundle (Disney Plus, Hulu, ESPN Plus)
Pros: Low cost, different types of streaming services, something for everyone
Who it’s for: With a sports streaming service, general entertainment platform, and family-friendly option, the Disney Streaming Bundle is made for a wide array of people
Cost: $12.99 a month
Compatible devices: Amazon Fire TV, Android, Android TV, Apple TV, Browser, Chromecast, iOS, Xbox, Roku, PlayStation
Disney Plus
Pros: Exclusive home for Disney TV shows and movies, low cost, entirely family-friendly, large library of shows and movies
Who it’s for: Families and Disney aficionados
Cost: $6.99 a month or $69.99 a year. It can also be bundled with ad-supported Hulu and sports streaming service ESPN Plus for $12.99 a year.
Top original shows: The Mandalorian, Imagineering Story, High School Musical: The Musical: The Series
Compatible devices: Amazon Fire TV, Android, Android TV, Apple TV, Browser, Chromecast, iOS, Xbox, Roku, PlayStation
Nick Hits
Pros: Catalog of classic Nickelodeon shows, full seasons available to stream, low cost
Cost: $7.99 a month or free with Amazon Prime Video Channels subscription
Top original shows: SpongeBob SquarePants, Hey Arnold!, iCarly
Compatible devices: Amazon Fire TV, Android, Android TV, Apple TV, Browser, Chromecast, iOS, Xbox, Roku, PlayStation
PBS Kids
Pros: Free, recognizable programming
Cost: Free
Top original shows: Sid’s Science Fair, CyberChase Shape Quest
Pros: Includes HBO, cloud DVR capabilities, multiple prestige channels
Who it’s for: Cord-cutters looking for HBO in their TV package
Cost: $55 per month for 45-plus channels; $80 per month for 60-plus channels
Compatible devices: Amazon Fire TV, Apple TV, Chromecast, Roku, Browsers, iPhone / iPad, Android Phone, Android Tablet, Samsung Smart TV, and Android TV
Hulu with Live TV
Pros: Comes with Hulu, on-demand playback, affordable pricing
Who it’s for: Cord-cutters looking to bundle a live service with streaming
Cost: $54.99 a month
Compatible devices: Apple TV, Chromecast, Roku, iPhone, iPad, Browsers, Android Phone Android Tablet, Android TV, Xbox, LG Smart TV, Samsung Smart TV, Vizio Smart TV, Amazon Fire TV
Sling TV
Pros: Cheaper than competitors, different package offerings, ability to rent new movie releases
Who it’s for: Cord-cutters not trying to break the bank
Cost: Sling’s three different packages come in at $30 a month (Sling Orange, 30-plus channels, or Sling Blue, 40-plus channels) and $45 a month (Sling Orange and Blue, 50-plus channels)
Compatible devices: Amazon Fire TV, Apple TV, Chromecast, Roku, iPhone, iPad, Browsers, Android Phone, Android Tablet, Android TV, Xbox, Samsung Smart TV, LG Smart TV, and Vizio Smart TV
YouTube TV
Pros: Easy-to-use interface, excellent DVR capabilities, local networks
Who it’s for: Cord-cutters looking for local broadcasts
Cost: $64.99 a month
Compatible devices: Apple TV, Chromecast, Roku, iPhone, iPad, Browsers, Android Phone Android Tablet, Android TV, Xbox, LG Smart TV, Samsung Smart TV, Vizio Smart TV, Amazon Fire TV
Photo by Stefan Etienne / The Verge
BritBox
Pros: Vast library of British television shows, top series like Doctor Who, classic British series
Who it’s for: Anyone looking to watch classic British television
Cost: $6.99 a month or $69.99 annually
Top original shows: Doctor Who, Absolutely Fabulous, Black Adder, Coronation Street
Compatible devices: Browsers, Roku, Chromecast, Apple TV, iPhone, iPad, Android devices, and Samsung Smart TVs
Boomerang
Pros: Classic cartoons all in one spot, family-friendly, new shows, wide variety of content
Who it’s for: Adults looking for a nostalgic kick and families
Cost: $4.99 a month or $39.99 annually
Top original shows: The best part of Boomerang is the collection of classics, like Scooby-Doo, Bugs Bunny, The Jetsons, Johnny Quest, Yogi Bear, Paddington Bear, and Tom and Jerry
Compatible devices: Android, iOS, Apple TV, Roku, Kindle, Browser, Amazon FireTV or Fire Stick
Crunchyroll
Pros: The biggest selection of anime titles, ability to read manga, streaming titles not long after airing in Japan
Who it’s for: Anime fans
Cost: $7.99 a month, $22.99 for three months, or $79.99 a year
Compatible devices: Apple TV, Amazon Fire TV, Roku, Xbox One, PlayStation 4, Chromecast, iOS, Android, Browsers
DC Universe
Pros: Wide array of DC TV shows, exclusive animated series, classic animated DC films, and digital comic books
Who it’s for: DC comics fans
Cost: $7.99 a month or $74.99 annually
Top original shows: Titans, Harley Quinn, Doom Patrol
Compatible devices: Amazon Fire TV, Android, Android TV, Apple TV, Browser, Chromecast, iOS, Xbox, Roku, PlayStation
Who it’s for: Anime fans looking for access to original and licensed shows as well as next-day offerings
Cost: There are three plans. The $5.99-a-month ($59.99 annually) Premium plan comes with ad-free streaming and two simultaneous streams. The $7.99-a-month ($79.99 annually) Premium Plus plan comes with ad-free streaming, offline viewing, and two simultaneous streams. The $99.99 annual Premium Plus Ultra plan comes with all of the above plus exclusive member perks.
Compatible devices: Apple TV, iOS, Android, PlayStation, Xbox, Amazon Fire TV, Chromecast, Roku, Amazon Kindle, LG and Samsung smart TVs
Plex
Pros: Ability to share content, good movie selection, host downloads
Who it’s for: If we’re being honest, people looking to share pirated content
Cost: A free tier includes playback of anything on your server, as well as podcasts, news, and web shows. There are also paid tiers at $4.99 a month, $39.99 a year, or $119.99 for a lifetime subscription, which gives people access to other subscription services and more apps.
Compatible devices: Browsers, iOS, Android, PlayStation 4, Xbox One, Amazon Alexa, Amazon Fire TV, Android TV, Apple TV, Chromecast, Roku, Sonos, TiVo
Shudder
Pros: Huge selection of titles, cost-effective, original movies and TV shows, podcasts, and live-streaming events
Who it’s for: Horror aficionados who want more than what’s on Netflix or Hulu
Cost: $5.99 a month or $56.99 annually
Top original shows: Creepshow, Deadwax
Compatible devices: Apple, Android, Roku, Fire TV, Browsers
Image: Criterion
Criterion Channel
Pros: Thematically programmed with special features, a wide array of titles, more than 1,000 classic and contemporary movies
Who it’s for: Cinephiles looking for classic films and additional material on beloved titles
Cost: $10.99 a month or $99.99 annually
Compatible devices: Amazon Fire TV, Android TV, Apple TV, Browsers, Chromecast, Roku
Kanopy
Pros: Free, large catalog of classic movies, wide array of independent titles
Who it’s for: University students and library users where Kanopy is available
Cost: Free
Compatible devices: Apple TV, Chromecast, browsers
Photo by Presley Ann / Getty Images for WarnerMedia
HBO Max
Pros: Includes HBO content, heavy focus on human curation, incredibly large movie and TV show library
Who it’s for: HBO Max is a general entertainment streaming service. Also a great option for anyone looking to upgrade their HBO Now accounts.
Cost: $14.99 a month. People with HBO Now, AT&T TV Now, or standard HBO cable subscriptions can get it for free.
Top original shows: The Wire, Succession, Game of Thrones, The Sopranos, Love Life
Compatible devices: Android, Android TV, Apple TV, Browser, Chromecast, iOS, Xbox, PlayStation
Showtime
Pros: Budget-friendly, good selection of new and older series, easy-to-use interface
Who it’s for: People who want prestige series on a budget
Cost: $8.99 a month
Top original shows: Billions, Shameless, The Affair
Compatible devices: Amazon Fire TV, Android TV, Apple TV, Chromecast, LG Smart TVs, Roku, Samsung Smart TVs, Xbox One
Top original shows: Outlander, Girlfriend Experiment, Power, Black Sails
Compatible devices: Kindle Fire Tablet, Amazon Fire TV and Stick, Android Phone, Android Tablets, Android TV, Nexus Player, Nvidia Shield, Razer Forge, Apple TVs, Browsers, iPhone, iPad, Roku, Xbox One
ESPN Plus
Pros: Carries big UFC fights, a number of other live sports options, low cost
Who it’s for: UFC fans and people looking for a cost-effective way to keep up with sports that air on ESPN
Cost: $4.99 a month or $49.99 annually
Compatible devices: Apple TV, Amazon Fire TV, Chromecast, Roku, Android TV, Android, iOS, Amazon Fire, Xbox One, PlayStation 4, browsers
MLB.TV
Pros: Plenty of games available to watch, different pricing tiers, single-team subscription, off-season plans
Who it’s for: Baseball fans trying to keep up with teams outside of their regional network
Cost: Currently, MLB.TV is not accepting new subscriptions. Previous prices were $24.99 a month (giving people the ability to cancel a little easier) or $116.99 a year for full-league coverage. A single team subscription was $89.99 a year.
Compatible devices: Roku, Apple TV, Chromecast, Amazon Fire Stick and Fire TV, Android TV, PlayStation 4, and Xbox One
NBA League Pass
Pros: Subscription packages for individual teams, ability to watch every game, international
Who it’s for: Basketball fans trying to keep up with their team or the NBA at large
Cost: There are four different options: NBA TV (live games, $6.99 a month or $49.99 annually), One Team ($59.99 annually gets you access to every single game from one team), All Teams ($99.99 annually, every single game), and All Teams, No Commercials ($124.99 annually, gives you every single game without ads)
Compatible devices: iPhone, iPad, Android Phone, Android Tablet, Kindle Fire, Roku 3, Apple TV, Android TV, Amazon Fire TV and Fire Stick, Xbox One, PlayStation
NFL Sunday Ticket
Pros: Every Sunday NFL game that’s not broadcasting in your city, Red Zone
Who it’s for: Football fans who don’t live in the same town as their favorite team and fantasy players who need to watch every single game
Cost: Currently, NFL Sunday Ticket is not accepting new subscriptions. Previous prices were $293.94 a season for the games, $395.94 a season for the games and Red Zone / Fantasy Zone channels
Compatible devices: Roku, Apple TV, Chromecast, Amazon Fire Stick and Fire TV, PlayStation 4, Xbox One, iOS and Android devices
NHL.TV
Pros: Three different subscription tiers, single team options
Who it’s for: Hockey fans looking to not miss out on a big game or favorite team
Cost: Currently, NHL.TV is not accepting new subscriptions. Previous prices started at $24.99 for one month of games, $119.99 a year for single team access, and $144.99 a year for the entire NHL season’s live games
Compatible devices: Roku, Apple TV, Chromecast, Amazon Fire Stick and Fire TV, PlayStation 4, Xbox One, iOS and Android devices
DAZN
Pros: Exclusive league games, available globally, fights are available to replay
Who it’s for: Boxing fans, MMA fans, people with interests in sports that aren’t the NFL, MLB, MLS, NHL, or NBA
Cost: $19.99 a month or $99.99 a year
Compatible devices: Android, Fire OS, iOS, Amazon Fire TV, Apple TV, Chromecast, Roku, PlayStation, Xbox, and browser
Update July 1st, 2020, 11:15AM ET: The monthly price for YouTube TV has been updated. Update July 16th, 2020, 11:40AM ET: Information about Peacock has been updated.
We use cookies on our website to give you the most relevant experience. By clicking “Accept”, you consent to the use of ALL the cookies.
This website uses cookies to improve your experience while you navigate through the website. Out of these, the cookies that are categorized as necessary are stored on your browser as they are essential for the working of basic functionalities of the website. We also use third-party cookies that help us analyze and understand how you use this website. These cookies will be stored in your browser only with your consent. You also have the option to opt-out of these cookies. But opting out of some of these cookies may affect your browsing experience.
Necessary cookies are absolutely essential for the website to function properly. This category only includes cookies that ensures basic functionalities and security features of the website. These cookies do not store any personal information.
Any cookies that may not be particularly necessary for the website to function and is used specifically to collect user personal data via analytics, ads, other embedded contents are termed as non-necessary cookies. It is mandatory to procure user consent prior to running these cookies on your website.
